Praise for The Logic Model

Guidebook

“A very thorough treatment of the practice of logic modeling with concrete

guidance based on actual programs and projects. The Guidebook is useful
both as a teaching tool and as a more general way of familiarizing
practitioners with logic models. I am impressed with the comprehensiveness
and detail …”

—Richard Elmore, PhD, Graduate School

of Education, Harvard University

“Better thinking and planning through logic models can contribute to stronger
results. The Guidebook supports rigor and quality. It’s a great tool for the
important work of creating sustainable social change.”

—Joseph M. Stewart, Trustee, W. K. Kellogg Foundation;

CEO & Chairman, Stewart Industries

“This book should be in the hands of anyone with intentions of leading

change. It is a much-needed and practical guide based on years of real-world
experience. The advice about quality is essential to improving the social
sector.”

—David Ray, Head of Policy & Advocacy, CARE

“A holistic roadmap for design, plans, and evaluation. This text offers sage
advice on metacognition and easy, clear steps to improve effectiveness.”

—Wendy Puriefoy, Chief Executive Officer,

Public Education Network

“This book should be in the library of every individual involved in program

development and evaluation. It is a powerful tool for practitioners and
students.”

—Sylvie Taylor, PhD, Antioch University Los Angeles

“The Guidebook is an essential tool for practitioners looking to improve

organizational performance and maximize impact in a resource-constrained
world. Here, the capable authors provide much-needed practical direction for
program evaluation and planning.”

—Matthew Knott, Chief Operating Officer, Feeding America

“The material in this book has enduring value. It is a ‘keeper’ for students
and me.”

—Simon Fass, PhD, School of Economic,

Political and Policy Sciences, The University of Texas at Dallas.

“Regardless of sector, logic models are valuable tools to design systems and
improve strategy.”

—Kori Reed, Vice President Cause & Foundation, ConAgra Foods

“The Guidebook is an important resource. It provides savvy counsel, is

accessible and focuses on results. The authors’ attention to quality is an
essential contribution to how we design and execute work.”

—William Rudnick, Co-Managing Partner, Chicago Office, DLA Piper

“The Guidebook fills a niche in the skills and knowledge needed by nonprofit
managers to be successful in their work. It leads the field in providing both
the theory and practice of using logic models as a critical management tool.”

—Kathryn Agard, PhD, (retired) Executive Director,

Dorothy A. Johnson Center for Philanthropy &
Nonprofit Leadership, Grand Valley State University

“The Guidebook is a tremendous resource for the novice to expert. It offers

explicit counsel on the steps to ensure quality in design, plans and evaluation.
I recommend it for anyone in philanthropy and social change.”

—Jill Wohlford, Vice President for Learning & Strategy,

Completion by Design

“It is the only text I am aware of that focuses specifically on logic modeling.
The links from theory to practice are important. It contains many practical
illustrations of innovative and diverse logic models. The Guidebook also
offers support to more experienced professionals by providing a range of
approaches and raising important considerations in model development.”

—Gary Miron, PhD, Professor, Evaluation, Measurement & Research,

Western Michigan University

“The Guidebook is easy to read and understand. I like how logic models
make assumptions visible. This makes it more likely to choose effective
strategies and secure desired results.”

—Faye Richardson-Green, Director Global Learning

& Development, Steelcase, Inc.

“I especially liked the learning aids, the clear writing style, the many figures
and examples and the listings of important points within each chapter. This is
all good teaching methodology.… Logic models are an important tool in
planning and evaluation. Both planners and evaluators should know how to
use them.”

—James Sanders, PhD, Professor Emeritus, Western Michigan University

 2
Edition

The
Logic Model Guidebook
 For Taylor, my earth angel.
I know you will soar.
For Tim, with profound gratitude, admiration, and respect.

For Courtney and Nicholas, my greatest lessons learned. You have

taught me much.
 2
Edition

The
Logic Model Guidebook
Better Strategies for Great Results

Lisa Wyatt Knowlton Cynthia C. Phillips

Phillips Wyatt Knowlton, Inc
FOR INFORMATION:

SAGE Publications, Inc.

2455 Teller Road
Thousand Oaks, California 91320
E-mail: [email protected]

SAGE Publications Ltd.

1 Oliver’s Yard
55 City Road
London EC1Y 1SP
United Kingdom

SAGE Publications India Pvt. Ltd.

B 1/I 1 Mohan Cooperative Industrial Area
Mathura Road, New Delhi 110 044
India

SAGE Publications Asia-Pacific Pte. Ltd.

3 Church Street
#10-04 Samsung Hub
Singapore 049483

Acquisitions Editor: Helen Salmon

Senior Associate Editor: Lauren Habib
Editorial Assistant: Kaitlin Perry
Production Editor: Libby Larson
Copy Editor: Kim Husband
Typesetter: C&M Digitals (P) Ltd.
Proofreader: Wendy Jo Dymond
Indexer: Marilyn Augst
Cover Designer: Bryan Fishman
Marketing Manager: Nicole Elliott
Permissions Editor: Adele Hutchinson

Copyright © 2013 by SAGE Publications, Inc.

All rights reserved. No part of this book may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying, recording, or by any information storage and
retrieval system, without permission in writing from the publisher.

Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

Wyatt Knowlton, Lisa.

The logic model guidebook : better strategies for great results / Lisa Wyatt Knowlton, Cynthia C.
Phillips. — 2nd ed.

p. cm.
Includes bibliographical references and index.

ISBN 978-1-4522-1675-1 (pbk.)

1. Proposal writing for grants. 2. Logic. I. Phillips, Cynthia C. II. Title.

HG177.K56 2013
658.15′224—dc23 2012016268

This book is printed on acid-free paper.

12 13 14 15 16 10 9 8 7 6 5 4 3 2 1
 Brief Contents

Preface
Acknowledgments
About the Authors

PART I: CONSTRUCTION
1. Introducing Logic Models
2. Building and Improving Theory of Change Logic Models
3. Creating Program Logic Models
4. Modeling: Improving Program Logic Models

PART II: APPLICATIONS

5. Logic Models for Evaluation
6. Display and Meaning
7. Exploring Archetypes
8. Action Profiles
Name Index
Subject Index
 Detailed Contents

Preface
Acknowledgments
About the Authors

PART 1: CONSTRUCTION
1. Introducing Logic Models
Basic Concepts
Models and Modeling
Logic Model Benefits
Logic Models Defined
Logic Model Uses
Two Types: One Logic
Historical Background
Examples
Theory of Change Model Example
Program Logic Model Example
Program Logic Model and Evaluation Design
Limitations of Logic Models and Modeling
Models Begin With Results
Logic Models and Effectiveness
In Summary
Learning Resources
2. Building and Improving Theory of Change Logic Models
Building a Theory of Change Model
Getting Started
Preferences and Styles
Evidence Based and Plausible
 The Big Picture
Multiple Strategies and Results
Realistic Models
Knowledge and Assumptions
Action Steps: Creating a Theory of Change Logic Model
Improving Theory of Change Models
Multiple Perspectives
“Unpack” and Share Assumptions
Toggling
Promising Practices and Benchmarking
Group Process
Nonlinear Theory of Change Models
Doing the “Right Work”
Tough Questions
In Summary
Learning Resources
3. Creating Program Logic Models
From Theory of Change to Program Models
Assumptions Matter
Key Elements of Program Logic Models
Nonlinear Program Logic Models
Hidden Assumptions and Dose
Building a Program Logic Model
Program Logic Model Example
From Strategy to Activities
Action Steps for a Program Logic Model
Creating Your Program Logic Model
Guiding Group Process
In Summary
Learning Resources
4. Modeling: Improving Program Logic Models
Modeling and Effectiveness
Context Challenges
 Common Pitfalls: Blind Spots and Myths
Logic, Scale, and Specificity
Politics, Persuasion, and Perception
A Learning Culture and External Review
Quality Techniques
Modeling
Testing Model Quality: SMART and FIT
A “Mark Up”
Quality Questions
A Quality Model
“Better” Decisions
In Summary
Learning Resources
PART II: APPLICATIONS
5. Logic Models for Evaluation
Getting More Out of Evaluation
Connecting Management With Measurement
Evaluation for Effectiveness
Evaluation Design Basics
Where Consumers Add Value
Where Logic Models Add Value
A Design Example
Two Kinds of Learning
Key Evaluation Questions
Indicators
Indicators and Alignment
Results Require Choices
Performance Standards
Quality Evaluation Designs
A Quality Framework
In Summary
Learning Resources
6. Display and Meaning
 Variation and Learning
Graphic Display
Complexity and Meaning
Content, Uses, and Creation
Model Benefits
Alternative Approaches
Selected Examples
Example 1. Eco Hub
Example 2. Wayne Food Initiative
Example 3. Promoting Preschool Change
Example 4. Collaborative Learning, Inquiry, and Practice
Example 5. New York Healthy Weight Model
Example 6. Evaluation System Development
In Summary
Learning Resources
7. Exploring Archetypes
The Blank Page Challenge
Archetypes and Learning
Recipes for Change
Value of Archetypes
More Critical Thinking
Selected Archetype Examples
Example 1. Federal Block Grants
Example 2. Education Readiness and Success
Example 3. Communications
Example 4. School Improvement
Example 5. Public Health Research
In Summary
Learning Resources
8. Action Profiles
 Strategy, Evaluation, and Learning
Profile 1. Building Civic Engagement
Profile 2. Better Corporate Giving
Profile 3. Kyrgyzstan Decent Work Country Programme
Profile 4. Alabama Tackles Asthma
Profile 5. Resilient Communities
Profile 6. Sheltering Families
Profile 7. Environmental Leadership
In Summary
Learning Resources
Name Index
Subject Index
 Preface

R
esponding to and creating change is demanding. Every day, people
in nongovernmental organizations, the private sector, universities,
and community-based organizations are responding to or creating
change. Models can help us see what is and what we want to create.
They can be powerful tools that support learning and performance. They can
help us with metacognition: thinking about our thinking.
Logic models are used in a huge range of topical content and functions
worldwide. They can easily explicate the influence of actions on results. If
our aim is coping with change and generating it, a critical review of “do” and
“get” is a vital action. As we face complex challenges like climate change,
education quality, poverty, homelessness, water distribution, healthcare
inequities, aging, and hunger, we need potent ways to communicate the
current situation and the desired one. As we consider ways to innovate,
transfer, and market knowledge—we need powerful approaches to new
contexts. As we deliberate a sustainable planet—we need to be able to co-
create options with shared meaning. Logic models are tools that help these
examples of important work.
We wrote the Guidebook because we care about results. We know people
need better skills, knowledge, and tools to have influence. While logic
models are never perfect, they do offer a partial remedy for better decisions,
plans, and adaptation. They can contribute to effectiveness and are consistent
with Palchinsky’s Principles to

seek out new ideas and try new things;

when trying something new, do it on a scale at which failure is
survivable; and
seek out feedback and learn from mistakes as you go along.

This second edition of the Guidebook provides the reader with a basic
understanding of how to create and use logic models. This is important for
people who work in the nonprofit, government, and private sectors with
responsibilities to lead and manage. Evidence-based models can be
particularly helpful to create programs, plan, communicate, and evaluate.
Logic models can provide important help that guides better thinking and
focused inquiry. Logic modeling is a process that contributes to clarity about
the sequence and the content of interactive relationships. Logic models
display relationships of many kinds: between resources, activities, outcomes,
and impact. They can also articulate the interaction of environmental barriers
and facilitators. The physical display models provide allows a chance to
critically review the relational logic among the “pieces” and context. And
they can be a platform to prompt important questions about assumptions and
choices. Logic models can significantly aid strategy development if we use
them to consider what’s plausible, feasible, and optimal vis-á-vis intended
results.
All logic models should be considered drafts. Every model example in the
Guidebook has flaws. Because models represent perception and reflect
choices, they have consequent limitations. Any individual has “blind spots,”
so people and groups that author models include those. Regardless, models
and modeling offer a potent alternative to lengthy narrative because visual
display is such a powerful, common way to create shared understanding and
test quality.
There are no perfect models, but the quality of models certainly can range
from simply “cockamamie” to highly strategic. Quality is a vital matter in
creating models. The best standard we can offer to ensure the potential of its
intended outcomes is prior evidence. However, when generating innovation,
it’s important to simply acknowledge rationale and “see” the prototype on
paper. This can ensure fidelity of implementation and focus evaluation or at
least document the initial approach in contrast to what actually is executed.
Modeling can be an exciting process. It includes a cycle of display,
review, analysis, critique, and revision to develop a model. These action
steps, best done with colleagues or stakeholders, can contribute significantly
to more informed models and are more likely contribute to results. Using
logic models in a systemic and disciplined approach to design, planning,
communication, and evaluation can contribute to individual and
organizational learning.
The Guidebook is a practical text for students and field practitioners. It is
organized with the assumption the reader has no knowledge or prior
experience. We hope it supports your changes in awareness, knowledge, and
skill relative to models and modeling.

New to the Second Edition

Each chapter in the second edition retains some of the prior “classic”
resources and includes many contemporary additions. We have added a large
number of model examples, associated descriptions, and an entire new
chapter with seven profiles that show the reader how models are used in the
field.
We’ve retained the initial organization of the text in two parts:
construction and application. Model construction is covered in Chapters 1
through 4. The application of models is in Chapters 5 through 8. The
construction chapters introduce readers to models and their creation and
improvement. The application chapters offer a more thorough review of use
and include many new examples of models in context.
Every chapter still includes an overview, learner objectives, questions,
and exercises along with learner resources. Chapter 1 introduces models,
their benefits, and some caveats. Chapters 2 and 3 detail two types of models:
theory of change and program logic models. Chapter 4 describes improving
model quality. This is a vital contribution because we suggest that the model
quality correlates to quality in planned strategy and tactics. Ultimately, these
are important influences for implementation, evaluation, and intended results.
Chapter 5 focuses on how models can be of significant use to monitoring
and evaluation. Logic models are an important tool in many aspects of
evaluation design, planning, and execution. They are also very useful to those
who are evaluation consumers. Chapter 6 provides examples of the range of
display for models. We have included several new ones. Likewise, in Chapter
7 we identify some reliable evidence-based examples that can be archetypes
for your work. These show how it’s possible to “borrow brilliance” and build
on the great work of others. Chapter 8 is all new content. It profiles some
wonderful work using models as a central tool and process. These profiles
show how logic models contribute to a range of functions and disciplines.
We hope this text is read and used in ways that support better thinking,
strategies, and models. If so, we’re confident you’ll secure great results!
 Acknowledgments

O
ur work is valuable because of amazing people, our clients, who
care about change and results. Our first and warm thanks go to
them.
This edition of the Guidebook benefited from many new
contributors and more than a dozen new models. We appreciate the time and
effort these colleagues made to enrich the text. Some of the models that
appeared in the first edition have been retained. In all, contributors include
the following:

Chapter 6
Example 1: Eco Hub—Adrian Jones, Integration and Application
Network, University of Maryland Center for Environmental Science

Example 2: Wayne Food Initiative—Tes Thraves, North Carolina State

University

Example 3: Promoting Preschool Change—Gale Berkowitz, DrPH,

(former) Director of Evaluation; Kathleen Reich, MPP, Program
Officer, Leader, Preschool Grantmaking; Lois Salisbury, JD, Director,
Children, Families and Communities Program, The David and Lucile
Packard Foundation. Julia Coffman provided the Kingdon models.

Example 4: Collaborative Learning, Inquiry, and Practice—Beverly A.

Parsons, Ph.D., Executive Director, InSites

Example 6: Independent Sector—Sherry Rockey, (former) Vice

President Independent Sector
Chapter 7
Example 2: Pathways Mapping—Lisbeth Schorr, PhD, and Vicky
Marchand

Example 4: Center on School, Family, and Community Partnerships,

Johns Hopkins University—Joyce Epstein, PhD.

Example 5: National Center for Injury Prevention & Control (CDCP)

—Sue Lin Yee, MA, MPH, and Howard Kress, PhD.

Chapter 8
Profile 1: Civic Engagement—Seattle Works—Tara Smith and Dawn
Smart, MA Clegg Associates

Profile 2: Better Corporate Giving—ConAgra Foods Foundation—

Kori Reed, Vice President Cause & Foundation

Profile 3: Kyrgyzstan Decent Work Programme—International Labour

Organization, Craig Russon, PhD, and Alexey Kuzmin, PhD Process
Consulting

Profile 4: Alabama Tackles Asthma—Alabama Department of Public

Health—Debra Hodges, PhD

Profile 5: Resilient Communities—Post Carbon Institute—Johanna

Morariu, MA, Innovation Network

Profile 6: Sheltering Families—Haven House—Angela Mayeaux,

Executive Director

Profile 7: Environmental Leadership—Paint Product Stewardship

Initiative—Matt Keene and Chris Metzner
Our thanks to those who graciously submitted example models for this
edition. Given limitations in space, most don’t appear here in print. We hope
there will be other ways to make your efforts more visible. The skills and
considerable knowledge of the crew at SAGE were essential to many aspects
of this book.
Several other exceptionally capable professionals contributed to new
understandings in model development and applications. They include Sal
Alaimo, PhD; Johnny Morell, PhD; Rosalie Torres, PhD; and Rodney
Hopson, PhD.
Past and recent readers offered valuable critique and constructive
feedback on all chapters. They include Kathryn Agard, EdD; Tom Chapel,
MA, MBA; Richard Elmore, EdD; Simon Fass, PhD; Nancy Horn, PhD;
Gary Miron, PhD; Janice Molnar, PhD; Lois-ellin Datta, PhD; David Osborn,
DMin; Faye Richardson-Green; Craig Russon, PhD; Jim Sanders, PhD;
Sylvie Taylor, PhD; and Rob Walsh, PhD Their comments and insights were
helpful.
We deeply appreciate the generous and thoughtful comments on our text
by Matt Knott, Wendy Puriefoy, David Ray, Kori Reed, Bill Rudnick, Joe
Stewart, and Jill Wohlford, as well as several of the readers noted above. We
applaud your leadership and the vital work you do in behalf of those most
vulnerable—across the globe.
Even if it’s the second time around, family provides important support in
the many challenges of creating a book. They were bystanders to long hours
at the computer and witnessed worry about the details. We are deeply
grateful. Lisa applauds, with love, Timothy, Taylor, and Meg. Cynthia offers
the very same to Courtney and Nick.
 About the Authors

Lisa Wyatt Knowlton, EdD, is a cycling enthusiast, Lake Michigan fan,

adoption advocate, and voracious reader. She holds a BA in international
relations from Michigan State University, an MPA. from Western Michigan
University, and an EdD in management and policy from Western Michigan
University. Her work history includes senior roles in programming and
management of private, community, as well as corporate philanthropy. She
has managed change projects for the W. K. Kellogg Foundation as well as the
Aspen Institute, the Bill and Melinda Gates Foundation, the Independent
Sector, and the Ball, Nokomis, and Kauffman Foundations. Lisa is a W. K.
Kellogg National Leadership Fellow with experience in Central America,
Asia, and Europe. Lisa authors a blog called tinker. She is a contributor to
Leadership in Nonprofit Organizations (Sage, 2011). Her areas of
specialization include strategy, organization development, leadership, change
management, and systems thinking. She is chief strategy officer, management
guru, and learning coach with Phillips Wyatt Knowlton, Inc. Lisa has cross-
sector experience and speaks Spanish. You can reach her via e-mail at:
[email protected].
Cynthia C. Phillips, PhD, is a birder, recovering aerobics instructor with 30
million meters rowed, and cyber-sleuth. She received a BS in biology and
chemistry from Indiana University, an MA in educational leadership from
Western Michigan University, and a PhD in measurement, research, and
evaluation from Western Michigan University. Her experience includes
consultation with the W. K. Kellogg Foundation, the Ball Foundation, the
Bill and Melinda Gates Foundation, Nokomis, Kauffman, and the David and
Lucile Packard Foundations in the design and implementation of evaluation,
evaluation training, and knowledge management projects. She is the author of
the W. K. Kellogg Foundation Logic Model Development Guide. Cynthia is a
sought-after presenter on logic models and knowledge management. Her
areas of specialization and expertise include evaluation/measurement;
knowledge management; organizational learning; logic models, quantitative
methods, and qualitative methods; and electronic data collection and
dissemination. Cynthia offers a user-friendly approach to evaluation capacity
building. She is chief idea engineer and measurement expert with Phillips
Wyatt Knowlton, Inc. You can reach her via e-mail at
[email protected].

Phillips Wyatt Knowlton, Inc. (PWK) is a measurement and management

resource with cross-sector and international experience. They focus on
systems performance and social change with clients worldwide. PWK
provides expert counsel in strategy, organization effectiveness, research, and
evaluation. For more information, see www.pwkinc.com.
 PART I

Construction
 1

Introducing Logic Models

T
his chapter introduces logic models. There are two types: theory of
change and program. This chapter describes model benefits and uses
and explains the role of modeling in both program and organizational
effectiveness. The process of modeling begins with results.
Regardless of type, quality models are evidence based.
 LEARNER OBJECTIVES
Explain the difference between models and modeling
Recognize the benefits and uses of logic models
Demonstrate how to “read” a logic model
Recognize types of models and their characteristics
Describe the ways that models can support effectiveness

Work in any sector, whether private, charitable or government, requires

design, planning, monitoring, and evaluation. Each of these functions solves
problems, and evidenced-based models are a great aid in any context. Perhaps
you have been asked to design a new program, lead a change project, create a
marketing strategy, or plan an evaluation. Did drafting a narrative to circulate
among colleagues feel fragmented or inadequate? Did you think, “Where do I
begin?” Logic models and modeling can be a potent option to resolve your
dilemma. The Guidebook provides the practical support you need to create
and use models. It will also enhance your understanding of the relationships
between actions and results. Step by step, we describe how logic modeling
can be used as both a tool and a process that resonate with learning and
performance management.

Basic Concepts

Models and Modeling

Logic models support design, planning, communication, evaluation, and
learning. They are often used when explaining an idea, resolving a challenge,
or assessing progress. They can untangle and clarify complex relationships
among elements or parts.
Logic models are a graphic way to organize information and display
thinking. They are a visual approach to the implicit maps we all carry in our
minds about how the world does or should work. Logic models are tools that
convey a scheme, program, or project in a brief, visual format. Logic models
describe planned action and its expected results. A model is a snapshot of an
individual’s or group’s current thinking about how their idea or program
might work.
Modeling is a technique. The process of modeling encourages iterative
development of an idea, program, or project. It can create a safe space to start
a debate, generate ideas, support deliberations, and allow one to think more
clearly about specific relationships. A single, coherent logic reflects a
consistent thread that connects design, plans, execution, and evaluation. This
thread of evidence-based logic is critical to program and organizational
effectiveness.
Modeling allows careful consideration of the relationship between
activities and results. When tackled by a team or small group of stakeholders,
models can be improved by engaging the knowledge and experience of
others. We think modeling is significantly underutilized as a valuable process
with real benefits. We believe the best models are socially constructed in a
shared experience that is facilitated. The shared understanding and meaning
they produce among colleagues are valuable and enable success in
subsequent steps of implementation and assessment.

Logic Model Benefits

In addition to extraordinary execution, organizational effectiveness ultimately
requires design, planning, monitoring, and success measures. Logic models
can contribute to the quality of all of these. In Chapters 1 through 4, we
address models from the design and planning perspective. In Chapter 5, we
offer more detail about their use with monitoring and evaluation. Logic
models also

Develop common language among stakeholders.

Offer highly participatory learning opportunities.
Document and emphasize explicit outcomes.
Clarify knowledge about what works and why.
Identify important variables to measure and enable more effective use of
evaluation resources.
Provide a credible reporting framework.
 Lead to improved design, planning, and management.

When logic models and modeling are used as a standard technique, they
can influence an organization’s effectiveness. Logic models offer the
strategic means to critically review and improve thinking. And better thinking
always yields better results. Modeling can happen well before resources are
committed or final decisions get made. This offers a way to pretest quality
and limit risk.
Effectiveness is not limited to—but certainly depends on—a clear vision,
capable implementation, and the means to monitor both processes and results.
Logic models can be tremendous supports for creating and communicating a
common understanding of challenges, resources, and intended success.
Moreover, models can also be used to calibrate alignment between the “big
picture” and component parts. They can illustrate parts of or whole systems.
Choosing a perspective can influence the level of detail. When modeling, this
specifies boundaries as well as the breadth or depth of display. For example,
a logic model can show the learning objectives for an elementary Spanish
curriculum, what a school district will do to secure student achievement, or
what the federal government will provide in educational resources for
second-language learning.

Logic Models Defined

Logic models are a visual method of presenting an idea. They offer a way to
describe and share an understanding of relationships (or connections) among
elements necessary to operate a program or change effort. Logic models
describe a bounded project or initiative: both what is planned (the doing) and
what results are expected (the getting). They provide a clear road map to a
specified end. The development of models (or the modeling process) provides
an opportunity to review the strength of connection between activities and
outcomes. Through the experience of critical review and development,
models can display participants’ learning about what works under what
conditions.
Models are the product of modeling—which we believe is best done in
small groups of stakeholders with the aid of intentional facilitation. They
complement systems thinking as a tool and technique for achieving valid but
simplified representations of real-world complexities. Common synonyms for
logic models include idea maps, frameworks, rich pictures, action, results or
strategy maps, and mental models. Although logical frameworks (logframes)
and causal loop diagrams (systems dynamics) are used for purposes similar to
logic models, they are fundamentally different but complementary tools.

Logic Model Uses

While often used in the nonprofit sector among large nongovernmental
organizations and foundations, logic models are of increasing interest among
community-based organizations and the private sector, too. Because models
enhance learning through the iterative exchange of information and
experience, they offer important features to organizations that value evidence,
diversity, dialogue, feedback, inquiry, great planning, and teams. Models can
be used in program design, planning, implementation, and evaluation. For
example, logic models can be used to design a marketing program, display a
purchasing process, describe a school district’s education improvement plan,
create a community leadership program, or establish the best ways to resolve
conflict.

Two Types: One Logic

We describe two types of models: theory of change and program. They differ
by level of detail and use but represent the same logic. A theory of change
model is simply a general representation of how you believe change will
occur. A program logic model details resources, planned activities, and their
outputs and outcomes over time that reflect intended results.
These two model types are different in their appearance and use. The
level of detail and features distinguish theory of change and program logic
models. Program logic models include more features than theory of change
models. This concept of “view” is important and is discussed further in
Chapter 4 because it influences the quality and utility of models. Theory of
change models are conceptual, and program logic models are operational.
Model types and their relative features are indicated in Table 1.1.
Relative to time frame, level of detail, volume of elements, display, and
focus, the model types contrast. They are alike because they share the same
research, theory, practice, and/or literature. Essentially, the types are different
views of the same evidence-based logic that have a shared origin.
Model use differs in purpose(s). Theory of change models display an idea
or program in its simplest form using limited information. These models offer
a chance to test plausibility. They are the “elevator speech” or cocktail napkin
outline of an idea or project. Program logic models vary in detail but offer
additional information that assists design, planning, strategy development,
monitoring, and evaluation. Program models support a display that can be
tested for feasibility. They are the proposal version of an idea or project
because they have fleshed out far more detail that often includes activities,
resources, outputs, and other elements of interest to those creating and/or
using the model. The relationships between elements, both the relative
interaction and sequence, are valuable for understanding intended work and
causal connections. They can be a huge help in creating action plans.

Historical Background
Use of theory of change and program logic models began in the 1970s. Carol
Weiss (1995) and Michael Fullan (2001) and Huey Chen (2005) are among
the pioneers and champions for the use of program theory in program design
and evaluation. U.S. Agency for International Development’s logical
framework approach (Practical Concepts, Inc, 1971) and Claude Bennett’s
(1976) hierarchy of program effectiveness were among the earliest uses of the
types of visual displays that have evolved into the program logic models we
know today.

Table 1.1 Features of Model Types

 Logic models did not receive much recognition, however, until after the
United Way of America came out with its publication Measuring Program
Outcomes in 1996. This publication promoted the structures and vocabulary
of logic models. The W. K. Kellogg Foundation also was instrumental in
spreading the use of logic models with its Logic Model Development Guide
(2001). For those readers interested in more detail on the historical evolution
of logic models, see the references provided at the end of this chapter.
Thinking about thinking, or metacognition, is present in many new
management and leadership texts. Because our thinking affects our actions,
it’s an area that’s well worth understanding better.

Examples
In the examples that follow, we briefly explain the general concepts and
terms related to a theory of change and to a program logic model. Chapters 2
and 3 provide more depth. Although we show one of each type of model, it is
important to keep in mind that these are but two examples from a much
broader continuum of possibilities. There are many ways to express or
display the ideas and level of detail.

Theory of Change Model Example

Figure 1.1 shows a simple theory of change model for leadership
development. Read from left to right, it suggests that some strategies, for
example, curriculum and experiences, can positively influence people so they
can more effectively tackle community challenges. This theory relies on the
assumptions that training, experiential learning, and community orientation
will have a substantial influence on individuals’ skills and ultimately result in
community development. It also relies on a particular framing of the
“problem(s).”
Chapter 2 focuses on creating theory of change logic models. They are
the critical foundation for any change effort. Often, these models exist as part
of an internal mental framework that is “dormant” or undisclosed. They can
also imply considerable knowledge, experience, research, and practice. The
evidence base for theory of change models typically is not made explicit.

Program Logic Model Example

Program logic models inventory, from start to finish, a specified program
effort. For example, a program logic model for a community leadership
program (based on the theory of change) would include the specified
resources/inputs, activities, outputs, outcomes, and impact. Resources or
inputs are what are needed to ensure the program can operate. Activities are
the tactical actions (e.g., events, services, publications) that occur to fulfill the
promise of each strategy. Together, activities make up the program design.
Outputs are descriptive indicators of what the specific activities generate.
Outcomes are changes in awareness, knowledge, skill, or behavior. The
impact reflects changes over a longer period. Figure 1.2 displays a simple
program model for the same community leadership program shown as a
theory of change model in Figure 1.1.

Figure 1.1 Community Leadership Academy Theory of Change

This program model suggests desired results include more and better
leaders and community development. It implies the leadership development
agenda is about resolution of community challenges and that, if resolved, it
contributes to community development.
To “read” this model, first note the intended impact (ultimate aim) of the
program: community development. Then, move to the far left-hand side,
where resources or inputs essential to the program are listed. Logic models
employ an “if–then” sequence among their elements. When applied to the
elements in each column, it reads, “If we have these resources, then we can
provide these activities. If we pursue these activities, then we can produce
these outputs. If we have these outputs, then we will secure these outcomes,”
and so on.
This model is just one very simple representation of how a program might
be designed and planned for implementation. Many variations on this
example could represent program design and planning for community
leadership development that meets standards of logic and plausibility. We
know that Figure 1.2, in fact, represents a program with some definite flaws.
More discussion about how the program could be improved through a “mark
up” (or critical review) that tests the program design is described in Chapter
4.

Figure 1.2 Community Leadership Academy (CLA) Program Logic

Model
Program Logic Model and Evaluation
Design
This guidebook also offers some support for using logic models to assist in
evaluation design. This book will address only the framing of broad inquiry.
At this level, evaluation questions are the foundation for evaluation design
and planning. If we apply this to the community leadership program example,
it is appropriate to focus on the program’s intended results. The summative
evaluation question is, What difference did the program make in the
community’s development? Perhaps a place to begin is in determining the
contribution made by the program to the development of more and better
community leaders. A clear, coherent program logic model provides great
assistance during evaluation design. A model points out the key features and
shows the relationships that need assessment.
In this example, an evaluation could consider both changes in the
awareness, knowledge, skills, and behavior of participants as well as the
community development impact. Stakeholders (funders, participants, and
other influentials) might also want to know about the content selection and
quality of training. They might be curious about implementation fidelity and
adaptation, too. Figure 1.3 demonstrates a program logic model with typical
evaluation questions.
This program logic model is serving evaluation. The five key evaluation
questions are applied at specific locations on the illustrated program model.
Key questions for the Community Leadership Academy (CLA) displayed
include

1. Is the CLA doing the right things?

2. Is the CLA doing things right?
3. What difference has the CLA made among participants?
4. What difference has the CLA made across the community?
5. What are the ways community needs can and should be addressed by the
CLA?

Positioning questions on the program model identifies where evaluative

evidence might be found to address inquiry. Labeling on the model also helps
to establish the relationship between program, implementation (processes),
results, and evaluation. Question 1 “tests” the logic constructed during
evidence-based planning. This question requires thoughtful connections be
drawn across activity accomplishment, implementation fidelity, and the
attainment of desired outcomes/impact. It addresses the overall effectiveness
of the selected strategies and the related action in achieving the desired
results. Question 2 examines implementation fidelity/variance as well as the
scope, sequence, penetration, and quality of activities. Questions 3 and 4
focus on the extent to which outcomes and impact have been achieved.
Question 5, like Question 1, should span the whole model to surface program
improvement needs. Questions 1 and 5 are more reflective but are essential to
improved effectiveness.

Figure 1.3 Community Leadership Academy (CLA) Program

Evaluation Model

These evaluation questions can be very helpful in the initial design and
development of the program, as they help to aim the program intervention.
The next step is establishing indicators. Models also help in guiding the
conversation and exploration needed to determine indicators or the measures
of progress for an effort. These issues are addressed in greater detail in
Chapter 5.

Limitations of Logic Models and Modeling

It is important to note that the proper reference, “logic model,” is no
guarantee of logic. While many models do demonstrate some modicum of
logic, a logical representation does not equal plausibility, feasibility, or
success. There is some danger in seeing a graphic display on paper and
considering it “true.” This notion of omnipotence stems from limited domain
knowledge, vested interest, and lack of perspective. Typically, models do not
take unintended consequences into account, although every program has side
effects. The modeling process usually does not include program critics, and
most stakeholders are not likely to be grounded in the research literature.
Realistically, even when program theory and logic are constructed and
build on the insights of broad representative stakeholder groups, can anyone
be sure who is right? Every model should be considered a draft. They are
deterministic, incomplete approximations of what usually are more open
systems. They provide the simple illustration that makes evaluation and
program improvement more accessible to individuals and groups. The mere
existence of a model does not mean that the model or the plans it represents
are ready for implementation or that it will readily deliver the intended
results!
Chapters 2 and 4 tackle model improvement and development in greater
detail. It is essential to note that a model is a graphic display of the program
captured at one point in time. Models, we believe, should change to reflect
best thinking and current evidence as these evolve. Creating and displaying
variations of a model are experiences that can develop thinking about
strategies/activities and results. This development is a critical process in
model quality and, ultimately, in the feasibility of the efforts described.
We believe the greatest value of logic models is their use in an iterative,
intentional process aimed at improving the thinking they illustrate. This is
best done through a facilitated sequence with selected stakeholders.
Obviously, logic models do not ensure plan implementation fidelity or
quality. Nor do they remedy any of the many concerns about organizational
structure and culture that can deeply affect program and organizational
effectiveness. Important action steps associated with quality include
identification of both assumptions and evidence used in models.

Models Begin With Results

Determining the results you desire is the first step in effectiveness, because
knowing where you are headed is critical to picking the best route to use. In
our experience, models begin with results. Results consist of outcomes and
impact; each appears in a sequence over time. While impact is the ultimate
end sought, sometimes synonymous with vision, outcomes are earlier
indications of progress toward results. We think results are the place to begin
when you are struggling with choices about strategies (with a theory of
change) or activities (with a program logic model). It is important to avoid
moving prematurely to specify what you want to do. In any change work,
program design, or problem solving, specifying those outcomes most likely
to occur soon and then those that will take more time to emerge helps
determine what route (action path) might be best to use.
People commonly complain their work is both activity focused and
frantic. Considerable time and effort are spent on a flurry of tasks that
frequently lack a clear relationship to intended results. Logic models can
assist in sorting priorities because they both rely on and help build a visual
literacy that makes action and expected consequences clear. Through the
models and modeling, stakeholders can identify potent strategies/activities
likely to contribute to the results sought. And those with less (relative) value
can be sidelined or discarded.

Logic Models and Effectiveness

In the workplace (and in life), almost everyone is interested in effectiveness.
To that end, we provoke important thinking when we ask these questions:

Are you doing the right work?

Can you make better decisions?
Are you getting superior results?
 All of these questions apply in any context—whether it is in government
or in the private or the nonprofit sector. They are among the most critical
questions for managers and leaders because they focus on key levers that
influence performance. We know from practical experience and assessment
that doing the right work along with great decisions secures superior results.
Logic models can help with the design that ensures the right work, the plans
and implementation that reflect better decisions, and the evaluation that tests
both pathways and progress toward success. For these reasons, they are an
exciting tool and process for anyone interested in more effective programs,
projects, and organizations.
Figure 1.4 demonstrates key points of the design, planning,
implementation, and evaluation that the two types of models can support.
Theory of change models are most helpful during the design of a program or
project. As plans or evaluation require greater detail, program logic models
can make a substantial contribution to these later stages of work. The types of
models and their uses form a continuous loop that can provide feedback
about a program throughout its life cycle.
Logic models as both a tool and a strategic process offer considerable
value to programs and, subsequently, organization effectiveness. They can be
used for different purposes at different times in the life cycle of an idea
(program, project, or change effort). Theory of change models can
dramatically influence program planning because they rely on knowledge to
offer choices about doing the right work. In this stage, the selection of
strategies relative to results occurs. Program logic models help with more
precise decisions about which activities in a given strategy are most effective.
Program logic models can also be used to support evaluation design. They
can assist in pointing to optimal areas of inquiry and help to determine
whether progress is being made and what difference has occurred relative to
results.

Figure 1.4 Effectiveness and Logic Models

 Some organizations use logic models routinely. They can become a
standard tool that promotes alignment and synergy. For example, evaluation
can be designed and implemented more easily when a clear theory of change
and program logic model have already been determined. These tools and
related processes can also assist learning and dissemination in significant
ways. Logic models and modeling can be vital elements in performance
management because they rely on evidence, support informed decisions
about strategy, and assist with assessment. Performance management seeks
predetermined results and adapts actions to obtain them.

IN SUMMARY

Logic models are simply a visual display of the pathways from actions to
results. They are a great way to review and improve thinking, find common
understandings, document plans, and communicate and explicate what works
under what conditions. We think theory of change models are distinct from
program logic models in several important ways. Theory of change models
present a very high-level and simple explanation of “do and get.” Program
logic models offer a detailed map that can be implemented when
supplemented with work plans. In this chapter, we also distinguished between
models as tools and modeling as a process. A quality feature of logic models
is that they are evidence based. Logic models can be used for learning,
improving, and greater effectiveness.

LEARNING RESOURCES

Reflection

1. In what circumstances can you use logic models in your work or field of
study?
2. What benefits does each type of model provide? And to whom?
3. What do logic models display? And what is missing?
4. How are theory of change models and program models alike? Different?
5. What kind of logic models have you seen before? Which are most
commonly used?
6. What current models/processes are commonly used for program design
in your organization? What work cultures are best suited for logic
models?

Application
Select and draw one of the following: promotion of a new brand of ketchup, a
driver’s training program, or a domestic violence awareness campaign. Have
others independently draw the same project you select. What do all the
drawings have in common? What areas are different? Why? When and how
do these differences become reconciled? How did the levels of detail differ
among the drawings? What can these drawings tell us about mental maps?

References and Supplemental Readings

 Texts
Chen, H. (2005). Practical program evaluation: Assessing and improving
planning, implementation and effectiveness. Thousand Oaks, CA: Sage.
Collins, J., & Hansen, M. T. (2011). Great by choice: Uncertainty, chaos and
luck—why some thrive despite them all. New York: HarperCollins.
Fullan, M. (2001). The new meaning of educational change (3rd ed.). New
York: Teachers College Press.
Funnell, S. C., & Rogers, P. J. (2011). Purposeful program theory: Effective
use of theories of change and logic models. Hoboken, NJ: Wiley & Sons.
Kane, M., & Trochim, W. M. K. (2006). Concept mapping for planning and
evaluation. Thousand Oaks, CA: Sage.
Murray, D. K. (2009). Borrowing brilliance: The six steps to business
innovation by building on the ideas of others. New York: Gotham Books.
Murray, D. K. (2011). Plan B: How to hatch a second plan that’s always
better than your first. New York: Free Press.
Practical Concepts, Inc. (1971). The logical framework. Approach and
training materials developed for the U.S. Agency for International
Development. Washington, DC. Unpublished manuscript.
United Way of America. (1996). Measuring program outcomes: A practical
approach. Alexandria, VA: Author.
W. K. Kellogg Foundation. (2001). The logic model development guide.
Battle Creek, MI: Author. Retrieved December 1, 2011, from
http://www.wkkf.org/knowledge-center/resources/2006/02/WK-Kellogg-
Foundation-Logic-Model-Development-Guide.aspx
Weiss, C. H. (1995). Evaluation (2nd ed.). Upper Saddle River, NJ: Prentice
Hall.

Journal Articles
Bennett, C. (1976). Analyzing impacts of extension programs, ESC-575.
Washington, DC: U.S. Department of Agriculture, Extension Service.
Dwyer, J. (1996). Applying program logic model in program planning and
evaluation. Public Health and Epidemiology Report Ontario, 7(2), 38–46.
Fitzpatrick, J. (2002). A conversation with Leonard Bickman. American
Journal of Evaluation, 23(3), 69–80.
Funnell, S. (1997). Program logic: An adaptable tool for designing and
evaluating programs. Evaluation News and Comment, 6(1), 5–17.
Julian, D. (1997). The utilization of the logic model as a system level
planning and evaluation device. Evaluation and Program Planning,
20(3), 251–257.
Julian, D. A., Jones, A., et al. (1995). Open systems evaluation and the logic
model: Program planning and evaluation tools. Evaluation and Program
Planning, 18, 333–341.
Sartorius, R. (1991). The logical framework approach to project design and
management. Evaluation Practice, 12(2), 139–147.

Internet Resources
Davies, R. (2008). The logical framework: A list of useful documents.
Retrieved December 1, 2011, from http://mande.co.uk/2008/lists/
the-logical-framework-a-list-of-useful-documents/
Duigan, P. (n.d.). Brief introduction to program logic models (outcome
models). Retrieved December 1, 2011, from
http://outcomescentral.org/programlogicmodels2.html
Evaluation logic model bibliography. (n.d.). Madison, WI: University of
Wisconsin Extension Service. Retrieved December 1, 2011, from
http://www.uwex.edu/ces/pdande/evaluation/evallogicbiblio.html
Israel, G. D. (2001). Using logic models for program development. Retrieved
December 1, 2011, from http://edis.ifas.ufl.edu/wc041
List, D. (2006). Program logic: An introduction. Retrieved December 1,
2011, from http://www.audiencedialogue.net/proglog.html
McCawley, P. F. (1997). The logic model for planning and evaluation.
Retrieved December 1, 2011, from
http://www.uiweb.uidaho.edu/extension/LogicModel.pdf
 2

Building and Improving Theory of

Change Logic Models

T
his chapter identifies the basic elements of a theory of change logic
model. They are evidence based and plausible. This chapter describes
the steps to create and improve a theory of change model. It also
names criteria for a “good” model.
 LEARNER OBJECTIVES
Identify basic elements of a theory of change model
Identify the contributions a theory of change model lends to a change effort
Create a simple theory of change model
Apply critical review for theory of change model plausibility

Logic models offer an exciting way to combine narrative and graphics to

display the mental maps people hold about a specific program or change
initiative. These mental constructs are also sometimes called “idea maps.”
While the process of creating a model can be solitary, there are significant
benefits when models are generated in a small group among stakeholders
with a shared agenda. Logic models can be used over the entire life of a
change effort—their boundaries should be consciously determined by the
participants who create the model.

Building a Theory of Change Model

Getting Started
While logic models can be used for many purposes, there are two basic types:
theory of change and program models. Understanding these types is
important to their development and use. The choice of which to use reflects
whether the model needs to describe broad and general concepts about
change or more detailed operational elements essential to design, plans, and
management. It is possible to begin with either a program logic model or
theory of change model.
We believe it is important that a program model always accompany a
theory of change because the assumptions held in the theory of change have
fundamental value for program operations and success. These assumptions
should be consistent and anchor choices made in the development and
selection of strategies to fulfill intended results. When assumptions are
evidence based, then a single coherent logic and alignment can occur that
enables success. Relying on knowledge, whether theory, research, practice,
and/or literature, is essential to a good model.

Preferences and Styles

People vary considerably in what level of detail they prefer to describe their
mental maps. This is an important consideration for those who lead the
modeling process. Invariably, any small group will include people with a
strong preference to start at a broad, high level and those who feel far more
comfortable beginning with detail. Both approaches have value because the
best program or change effort design eventually should display models with
these features. Accommodating differences in how any individual approaches
the display task is a common tension in the creation process.
Our aim is to guide you consciously from big ideas to finer points. For
this reason, we provide a theory of change model description and example
first, then a parallel representation of a program logic model in the next
chapter. So that content matter does not confound the process, we have
chosen to use community leadership development and health improvement
for all model content in Chapters 1 through 5. A broad range of subject
content is offered in the models found in Chapters 6 and 7.

Evidence Based and Plausible

Theories of change can be grounded either in an established claim with proof
or in a hypothesis. Programs based on proofs are a replication of something
that has worked. Hypotheses are rationales based on research literature that
show promise of working and are therefore something worth trying.
Programs based on hypotheses are innovations. If the theory of change is
supported by a body of evidence, there is a stronger chance that the strategies
chosen will secure the desired results. Frequently, however, this “standard” is
overlooked. In the urgent fever to get to implementation, the design and plan
quality can be shortchanged and rely, instead, on faulty assumptions, old
practice, or little or no evidence.
We suggest practitioners construct the theory of change model with
grounding it in literature, experience, or other evidence that promotes
plausibility. Most theories of change will exhibit some degree of logic.
Plausibility, however, is a more appropriate litmus test for work that has
inherent opportunity cost. Later, in Chapter 3, we suggest that the program
model must also be feasible if there are authentic intentions of securing
results. During the construction of a change model, it is important to explore
or discover what works under what conditions. This is about the choices
made in selection of strategies relative to anticipated and therefore planned
results. When constructing a program logic model, the realistic criterion of
limited resources is also in play. Any program, project, initiative, or
organization has some limits on time, talent, and financial resources. In the
migration from theory of change to program logic model, users can shift their
thinking from what “could work” to what “should work.”

The Big Picture

A theory of change logic model offers the big picture of strategies that could
generate your intended results (or impact). This construct is illustrated by
Figure 2.1.
A basic theory of change model contains just two elements: strategies and
results. The intent is to illustrate the connection between what you will do
with what you hope to get.
Strategies reflect a choice of optimal actions (via activities or tactics) to
secure intended results. They represent an allocation of resources focused on
a clearly defined objective. Marketing, training, political advocacy, and fund
development are examples of common strategies. This element describes
your actions or what you plan to do.
Strategy is the overall plan that gives coherence and purpose to the
specific actions that organizations undertake. For some nonprofits, however,
the meaning can be murkier and framed as an ambiguous aspiration. When
we use the term strategy, we rely on an implicit but evidence-based assertion
that connects means and ends. Wherever the word strategy appears in our
illustrations and narrative, we assume that the “case” for selection is sound—
in other words, that it has strong potential to secure impact. Later, in Chapter
3, we also use the term as an umbrella for nested clusters of activities (or
tactics) that aim at specific single outcomes or clusters of outcomes.
Results reflect the long-term effect of strategies. They are the “get” from
what we are doing. Results are ultimately secured through the change(s)
generated by the preceding strategies. They can reflect a single outcome or
multiple outcomes over time.

Multiple Strategies and Results

In reality, many programs are more complex than is shown in Figure 2.1.
Most often, several strategies combined (over time) in a particular sequence
yield results. And we generally both create and experience results as the net
yield of several strategies working together. When displaying theory of
change models, this can be challenging both to conceptualize and to
represent. For example, to become proficient in a new language, it is most
likely the combination of instruction, practice, and cultural immersion that
generates proficiency. Likewise, to be a profitable cereal company may
require a high level of competency in research/development, marketing,
production, and distribution strategies. In addition, great health outcomes for
neurosurgery may rely on expertise in diagnostics, surgical techniques, pre-
and postsurgical care, and rehabilitation therapies. Because multiple
strategies often contribute to results, a more complex representation of a
theory of change might look like Figure 2.2. We call the path from each
strategy to result a “strand.”

Figure 2.1 Basic Theory of Change

Recall the theory of change model for the Community Leadership

Academy (Figure 1.1). In that example, the outcome desired was more and
better community leaders. This model described two simple strategies as
essential contributions to the recipe: curriculum and experiences. An applied
example of a multi-strategy model for securing improved health is displayed
in Figure 2.3. Read from left to right, the theory of change suggests if we
provide exercise, nutrition, stress reduction, and some other key strategies,
then we will secure improved health for participants who follow the program.
It is also important to note that strategies may interact (although not shown
here). This theory of change represents a generalized construct for many
health-improvement programs.
Figure 2.3 simply provides a gross summary of strategies and intended
results for a health-improvement program. A theory of change model displays
some of the underlying assumptions about change and is a view at 65,000
feet. This view is how a farm looks from an airplane window in contrast to
the view from a tractor seat. It simply shows the specific strategies that the
designer believes will achieve a desired result. Theory of change models are
distinct in that they include assumptions (either explicitly or implicitly), offer
the big picture of the bounded mental map, and name impact. Theory of
change models do not provide the detail essential to action planning,
implementation, or evaluation. They simply state what you plan to do and
what you expect to get. It is easier to explore ideas and manipulate them at
this stage. The why and how of these models are embedded in assumptions
and eventually reveal themselves in the strategies selected. For example, in
Figure 2.3, some of the underlying assumptions might include the following:

Figure 2.2 Multiple-Strategy Basic Theory of Change

Figure 2.3 Health-Improvement Theory of Change

 Increased exercise and improved nutrition are known to contribute to
improved health.
Only those who participate in the program will achieve results.
Participants need to be recruited.
Stress may be a contributing factor to poor health.
Participants’ fidelity to the program is critical to achieve results. They
will need parallel increases in awareness, knowledge, and skill in order
to change behaviors that most impact health.

Realistic Models
Theory of change models should demonstrate plausibility. This means they
“could work.” Given the realities of limited time, as well as human and social
resources, logic alone is inadequate. In fact, the logic displayed in a model
can be uninformed or misinformed. For example, world peace is a tangible
and clear desired result, but a theory of change that relies solely on
communication (e.g., newsletters and websites) is not plausible in securing
world peace. Or consider the desired result of hiring more mid-level scientists
at your research institute. Are outreach strategies with local math and science
teachers and students logical action steps? Yes, but meetings with those
targets can be helpful only in a pipeline that can tolerate a decade of delay. It
is not a best strategy given urgent human resource needs this week and next
month.
Knowledge and Assumptions
So far, we have described a basic theory of change model for improved
health that is specifically composed of doing (strategies) and getting (results).
Each of us brings along some other contributions to our theory of change that
are more closely held. While not often named, we commonly bring what we
believe (our assumptions) to theories of change, too. The most viable
assumptions used to select strategies are rooted in knowledge, and that
knowledge generally includes research, practice, and theory. Figure 2.4
illustrates the knowledge base for beliefs that precedes assumptions and
strategies in a theory of change.
It is critical to recognize the role of beliefs. They are important
determinants in choices about strategies for both creating and improving a
theory of change model. Figure 2.4 illustrates how knowledge and beliefs
contribute to a program’s underlying or driving assumptions. Assumptions
are often informed by knowledge, which can include research, practice, and
theory. We find that making assumptions explicit can improve our chances
for program success. Sometimes assumptions are informed by experiences,
habits, or values that do not also reflect knowledge. Mediating or moderating
factors such as program context are useful to consider as barriers or
facilitators to program success at this stage. Dogma, misinformation,
ignorance, and wishful thinking are hazards here. Often, assumptions can
differ significantly among and between both stakeholders who create and
those who execute. They can also dramatically affect how problems are
identified and framed. For model utility, it’s important to cite what
problem(s) we’re trying to solve and find a way to frame a problem so that it
is meaningful to others.
Modeling can help surface vital differences among stakeholders and offer
a disciplined process for resolution based first on plausibility, then on
feasibility during subsequent versions. This is why, in part, modeling offers
considerable value beyond the construction of models alone. It’s important to
note that dialogue is critical to exploration of knowledge and assumptions
that are embedded in models. Engaging multiple stakeholders is critical to
quality as well as meaning.
However, modeling can be an uncomfortable process because it nearly
always raises differences among participations’ perceptions, experience,
knowledge, training, and other factors. Identifying and negotiating these can
be challenging. This navigation is most easily done with external assistance.
If not, then it can be useful to explicate the criterion for decisions about
model content and display. Simply who participates in modeling can be
loaded with politics, since it will very likely influence the model content.

Action Steps: Creating a Theory of Change

Logic Model
We recommend that people begin building a theory of change model by
specifying their intended results. Most often, it is easiest to be clear first
about the intended results. Our experience with clients is they know what they
want to accomplish. They often label this as desired results.

Figure 2.4 Informing a Theory of Change

Once results are named, we suggest identifying the strategies required to

achieve the results you seek. Strategies are about how intended changes will
occur. And assumptions are the preceding knowledge: the research, practice,
and theory that inform choices about strategies. They significantly influence
which strategies are chosen as pathways to your intended result.
So the steps to generate a theory of change logic model are ordered in this
way:

1. Identify results desired.

2. Name the strategies that will deliver your intended results.
3. Define the assumptions (see Figure 2.4) that support your specified
strategies.

Figure 2.5 displays these actions in a three-step sequence. Most theory of

change models generally do not display underlying beliefs or assumptions.
Nevertheless, these are important elements to explore consciously when
creating a theory of change. We suggest those assumptions are named in
association with the theory of change. Assumptions can simply be a bulleted
list on the same page or reverse side. Remember, a theory of change model is
simply one representation of the “truth,” not a substitute for it. The model
draft becomes a place for starting discussion and testing meaning, coherence,
assumptions, and plausibility. Engaging others in modeling offers the
opportunity for critical review and improvement over time through the
generation of versions. By starting with a theory of change model, it is easier
to arrive at shared understanding of what your program will do and can
achieve.

Figure 2.5 Steps in Creating a Theory of Change Model

 Improving Theory of Change Models
We offer several common processes to consider as you explore iterations of
your theory of change model. While improvement is definitely not limited to
these suggestions, an application of these four will likely contribute to the
development of any attempt:

Engage multiple stakeholders.

Share explicit assumptions.
“Toggle” or test alternative content in model versions.
Explore promising practices and consider benchmarking.

Multiple Perspectives
People hold and operationalize theories of change in both their work and
personal lives. Most experienced parents, for example, have a recipe that
contains the primary strategies they believe are vital to parenting a “good
kid.” Parents can vary considerably, however, in what they mean by a good
kid. Likewise, even if we agree on what a good kid might know and be able
to do, it is highly likely that from one parent to the next, there will be many
variations on parenting strategies to ensure the “good kid” result. This
example suggests the considerable importance of ensuring that all
stakeholders in your program or change effort are specifying results and the
strategies needed to get there with the same meaning and level of specificity.
Developing and improving the theory of change for your program is one way
to begin the conversations needed to reach shared understanding.
In the health example we started this chapter with (see Figure 2.3), we
identify improved health as the result sought. It is important to ensure that
everyone has a highly consistent understanding of what “improved health”
means. To one participant, it may be weight loss. Another could interpret it as
normal blood pressure. Others may feel improved health is a combination of
several positive outcomes. If you ask a half-dozen people what improved
health means to them, it is quite likely there will be variation in their
individual answers.
Specifying what the results mean, such as improved health in this
example, becomes critical for your program design as well as essential for
measuring progress toward and determination of results. If the meaning and
measures of results are shared and understood similarly, then it is more likely
strategy choices will align with your intended impact. It is more likely
indicators of progress will be appropriate, too.

“Unpack” and Share Assumptions

The most significant opportunity to improve theory of change models lies in
unpacking the knowledge and beliefs employed in assumptions. This means,
in practice, that any theory of change for a program or social change effort
should be grounded in knowledge. If results are connected to strategies that
reflect research, practice, theory, and experience, there are far greater chances
for success than with strategies that lack this grounding. Figure 2.6 displays a
combination of the elements found in knowledge. A combination of little or
no practice, experience, research, and theory in your model means the effort
it represents is an idea that may be highly innovative but is not likely to
succeed. A combination of practice, experience, research, and theory in your
model suggests the effort it represents is more likely to succeed.
The best theory of change model deliberately pursues alignment among
research, theory, practice, and experience. The stronger models build on the
knowledge and good work that precede them. These substantiated models and
their associated programs or social change efforts gather and then use
codified knowledge from prior efforts to inform effective program design.
Figure 2.6 describes the geography of choices and emphasizes that relative
success relies on a depth of practice, experience, research, and theory.

Figure 2.6 Success in Theory of Change Models

Toggling
Another practical way to improve models is what we refer to as “toggling.”
We define toggling as finding the optimal fit between a selected set of
strategies and plausible results. For example, options to improve school
nutrition could include planting a garden, removing vending machines, or
changing the lunch menu. Toggling “tries on” options and makes a best
choice. In this critical review, the model builder is experimenting with the
best combination of strategies to secure the results sought. Inviting others to
join in this iterative tactic in real time can be very productive. Sometimes it is
best to refine or focus the specified results. For example, a program or social
change effort could specify one of these results: “end childhood obesity” or
“create schools with improved nutrition choices for children.” During
toggling, it may become apparent the result is not plausible. Often, ambitions
are greater than what is feasible. It is important to guard against grand
ambitions. They are possible hazards that can result in flawed models. Figure
2.7 demonstrates the interactions between strategies and results as choices are
made in the final determination.
This figure displays the testing that occurs in an effort to determine the
best combination of strategies to secure your intended results. Once a
preliminary theory of change model is drawn, the modeling process begins.
The model is tested though iterative cycles of inquiry. The basic questions
addressed are “Are the results focused and narrow enough to discern
optimum strategies?” and “Is the connection between the strategies named
and the results desired as strong and direct as needed to be effective with the
population of interest?”
 Toggling can also involve a review of both the duration and sequence for
chosen strategies. The objective is to specify a model that is plausible. The
specifics of what is feasible are developed in the program logic model (and
are discussed in following chapters). People sometimes mention confusion
when they talk about the “fog of war.” In our experience, there is
considerable “fog” or ambiguity in program design and planning. A clear and
plausible theory of change is the foundation of intended work and requires
considerable attention and scrutiny. Just as with the Cheshire Cat in the story
Alice in Wonderland, if you do not know where you are going, any road will
do. Fuzzy, ambiguous theories of change rarely net the success intended.

Figure 2.7 “Toggle”: Improving Theory of Change Models

Promising Practices and Benchmarking

It is valuable to explore strategies and results of programs (or social change
efforts) similar to yours. A better understanding of the rationale for their
strategies and related results can deeply inform your design choices. In the
private sector, this is often referred to as “benchmarking,” a systematic
discovery and comparison process that can be a simple way of both clarifying
and improving your design early on. In benchmarking, one simply looks
around at promising practices to inform and then make good choices about
your own. Benchmarking may include a review of documents, a survey, and
discovery with peers/competitors. It establishes the status of other efforts,
programs, or organizations on specified features or issues. The big questions
benchmarking can help answer are what others are doing and why? Chapter 7
provides some examples of archetypes, general recipes that are evidence
based, that can also help in your early efforts to construct a model.
 School-improvement efforts, for example, often include quality
instructional and assessment practices among their many strategies aimed at
teachers to, ultimately, improve student academic achievement. There is
evidence these strategies can positively influence student academic
achievement. It follows that a school-improvement effort, then, might be
more successful if these strategies (or some adapted version) are included in
the program plan. The converse is also true. A school without these strategies
as standard operating procedure or as part of a new plan is less likely to
secure improved student academic achievement. A theory of change can
show what you are and are not thinking about. The accompanying modeling
process provides the opportunity to reflect and improve on the underlying
logic for your program.

Group Process
Consider involving others in co-creating a theory of change model. Let’s
build on the improved health example from earlier in this chapter and aim at
obesity prevention. How could you guide a group in exploring a countywide
program design intended to maintain healthy weight and prevent obesity? In
tackling this question, it’s important to anticipate the need for data prior to
the convening. Gathering and sharing information about research, practice,
and theory makes for a much smarter dialogue. It’s also possible to include
experts who bring data and field experience literally to the table. In general, a
guided group process could follow these action steps in a daylong work
session or over a series of meetings.
Remind participants, again, of the intention of the work to establish a
theory of change that articulates a single relationship between results and
strategies. The assignment is to identify strategies most likely to get the
planned results given the context, target audiences, and other factors. So it’s
important, first, to secure a shared understanding of the results intended. Ask
all the participants, on their own, to identify the result they want the program
to achieve in the next 3 years. It’s vital to specify a period to bound the
program effort. Have participants post directly (or transfer) their intended
results for public sharing. This first posting will likely display a range of
expectations and assumptions about what results are desired. Reconcile those
that are similar and do discovery on what’s “underneath” the postings.
Through dialogue, find the result that the group believes is most feasible
given the context. Features of context might include historical and current
rates of obesity and overweight, definitions of those terms, an inventory of
physical fitness options and their physical proximity, socioeconomic data for
the county population, and access to healthcare and weight loss resources,
along with aspects of prevailing culture. Create a list of resources, including
specific funds that could be designated for the program. Your participants can
probably name many other features of context. These are the influences as
well as data that help to inform the current reality. It may help to post facts
and features of context so they are present to dialogue. This portion of the
process should rely on facts as well as perception.
Then, consider your target audience(s). Will your program effort be
designed to influence males, females, teens, young adults, all residents
between 10 and 50 years of age? Or some combination of these
characteristics? Employ learning from the context discussion to inform your
choices. Be aware the selection you make may require you to adjust the
group’s intended result. The effort to name and understand the results is well
worth the effort because it frames subsequent action steps.
Last, ask participants to name strategies that the program should include.
Post them. Often, people will name tactics or specific activities. Getting to
the same level of detail just requires some modification. This is another great
opportunity to insert more information. For example, identify independent
research, practice, and theory shown to influence weight management. Share
some benchmarking information from effective programs that have already
tackled this same challenge and those that failed to make progress. Be sure to
include their costs and related organizational resources.
Ultimately, the group should determine a clear list of strategies and
specified results that are not simply feasible but optimal—that is, highly
likely to secure the impact. This may require some “toggling.” Use the
Guiding Questions (below) to critically review the work of the group. Look
forward to Chapter 6 and review the New York state Healthy Weight
Partnership. It offers some great ideas about strategies and results (defined by
their mission and vision). The NY model cites target sectors/settings to
segment their program plans since the work is focused on all state residents.
As you construct, then review a theory of change, the following questions
may be useful:
 Guiding Questions for Reviewing a Theory of Change
Model

1. Are the results specified with shared meaning among all

stakeholders?
2. Did we uncover our assumptions and carefully examine research,
practice, and theory as the grounding for our choices in strategies?
3. Did we “toggle” between strategies and results to ensure plausibility
given our assets and limitations?
4. Have we carefully reviewed similar programs to learn what strategies
worked under what conditions to secure results?
5. Does the model clearly show the relationship of strategies to results?

Nonlinear Theory of Change Models

Theory of change logic models are not always displayed in a linear fashion
(as they have been in the text so far). Realistically, few theories, programs, or
change efforts occur in a precisely linear sequence. The world is much more
complicated and integrated than the simple, step-by-step actions and reactions
as drawn here. Most change occurs iteratively, or in cycles, and with multiple
interactions among many features. Any change is also connected to a much
larger system than illustrated by the theory of change model. In our
experience, systems and holism can be difficult to manage and even harder to
evaluate or communicate. In generating a theory of change, it is important to
represent how change occurs as an aspect of a far more comprehensive and
vast geography. When using a systems view, it is important to consider the
key leverage points or strategies that are most influential given time,
expertise, and resources. A simple example of a nonlinear theory of change
logic model is shown in Figure 2.8.
In Figure 2.8, the four strategies that contribute to results occur in a
specific sequence over and over again. Their interaction contributes to the
center target: results, an aggregate of progress over time. The intersection and
influence of external issues and/or the environment can be illustrated as well.

Figure 2.8 Nonlinear Theory of Change Model

In the nonlinear theory of change model, Figure 2.9 represents an

alternative view of the Community Leadership Academy theory of change
model shown earlier. Through two strategies identified here—curriculum and
experiences—this change effort expects to produce more and better
community leaders and, ultimately, community development. The
curriculum, the experiences, and the participants interact. Chapters 6 and 7
also provide some additional examples of nonlinear models.

Doing the “Right Work”

In Chapter 1, we mentioned effectiveness and three critical questions:
 Are you doing the right work?
Can you make better decisions?
Are you getting superior results?

The first question was about the right work. This is about attending to
making the strongest, most direct and plausible connection between your
strategies and results. It is about the focus of time, energy, talents, and
resources in relation to your specified success. Eventually, right work is also
about detailing those specific activities that are subsumed by each strategy
that is chosen for display in the program logic model. Giving conscious
attention to the criterion used in selecting strategies at this stage, and again
later, will identify how implementation can make a big difference in the
likelihood that your program will secure results. The right work is clarified
and confirmed if there is a shared understanding of the problem you plan to
resolve and there is agreement on how it can be accomplished. Specificity
here, on the front end, contributes to the results you and your colleagues
intend to secure. Ambiguity can doom the best-intentioned efforts to failure.

Figure 2.9 Community Leadership Academy Nonlinear Theory of

Change Model
 If your end result in a construction project is a great house, then cooking
and sewing probably are not the most relevant strategies. However, planning
with well-detailed blueprints as well as appropriate purchasing (e.g., quality
lumber) and contracting should be ripe for your attention. It is surprisingly
easy to spend time on the wrong work. It can be an unconscious or conscious
choice. Theory of change models should display planned results and specify
the most relevant and influential strategies to secure the results. The strategies
are determined from a universe of possibilities. Often, people include
strategies (and later, activities) they have always done or are most familiar
with doing. If replication is intentional, then repeating what has been done
before might be appropriate. As time passes and knowledge changes,
however, results may require we use what has been learned about new or
different strategies (and activities) to be more efficient and effective.
Remember, a theory is only as good as its last test.

Tough Questions
Of course, there are many ways to secure a named and intended result.
Discarding strategies/activities that are peripheral, modest contributors or less
than optimal in potency can focus limited resources. Models and their
iterations can develop a disciplined way of thinking that contributes to new
understandings about what will generate progress toward results. Once results
are specified, the discovery and discussion that should be encouraged during
your modeling attends to these two big questions:

What are the many ways we could resolve this challenge?

Then, what are the most effective and efficient ways to secure results?

Subsequently, as model versions are explored, it is important to inquire

further. For example, are we doing something that has a reasonable chance of
working? Are we doing something that should be done? Are we clear enough
about the work that we have shared expectations for what it includes and can
yield? How does our model rely on research, practice, theory, and/or
literature? We know that the politics of power and dynamics of resources
often preclude these conversations. A predisposition to activities and
busyness can overwhelm a disciplined and interactive process, too. However,
the benefits and relative value of getting things right at the start are
considerable. The opportunity cost or waste in missing this step is huge.

IN SUMMARY

Logic models display mental maps people hold about cause and effect.
Combined, theory of change coupled with program logic models are the most
potent design prescription. Theory of change models specify and link
strategies with results. Most change efforts require multiple strategies.
Knowledge is a critical input for models and can include research, practice,
and theory. What people believe affects the content and format of models.
Improving theory of change models requires multiple perspectives,
unpacking assumptions, shared language, toggling, and the exploration of
promising practices.
 LEARNING RESOURCES

Reflection

1. What role do assumptions and beliefs play in a theory of change model?

2. How can you test a theory of change model for plausibility? Why bother
with this step?
3. Are there blind spots in the modeling process? If so, what are they?
4. What are the implications of a change model that relies on a hypothesis
versus one based on a claim with proof?
5. What are some ways that theory of change models can be improved
and/or developed?
6. What challenges do complex and highly interactive systems present in a
theory of change model? Where and how do you bound the presentation
of a theory of change model?

Application
1. Have a conversation:
A. Ask colleagues to share their beliefs about parenting (or their mothers’
or fathers’ beliefs) to ensure a happy, confident, successful young
adult. From this conversation, draw a theory of change. What are their
most important strategies? Can you identify their beliefs, values,
assumptions? Do they cite any evidence for their choices? Is research,
practice, or theory part of their explanation? How are their views
similar to or different from yours? Do they have a shared understanding
and agreement about parenting with their spouse (or among their
parents)? How does your response to these questions influence the
model?
B. Ask a friend or colleague to share a recipe for marketing a new car
model. What are the most important strategies for ensuring profit?
What evidence supports their choice of strategies? How do assumptions
inform their theory of profitability? How does your response to these
 questions influence the model?
2. Ask several people to list the many ways that “improved health” might be
described. Why does this outcome/result have different meanings? Could
these differences influence modeling?
3. Find a news article that describes a change effort (in a government,
nonprofit, or private sector). Draw it. Can you detect the efforts
underlying theory of change? How was it informed: based on a claim or a
hypothesis?
4. Considering the drawings from Questions 1 and 3, how do choices of
strategies influence the likelihood of achieving your intended results?
What changes, if any, could be made to improve the plausibility of these
models?

References and Supplemental Readings

Texts
Bickman, L. (1987). The functions of program theory. In L. Bickman (Ed.),
Using program theory in evaluation. New directions for program
evaluation, 33, 5–18. San Francisco: Jossey-Bass.
Bono, E. (1999). Six thinking hats (2nd ed.). New York: Little, Brown &
Company.
Chen, H. T. (1994). Theory-driven evaluations. Thousand Oaks, CA: Sage.
Donaldson, S. I. (2007). Program theory-driven evaluation science:
Strategies and applications. Mahwah, NJ: Lawrence Erlbaum.
Porter, M. E. (1995). Competitive advantage: Creating and sustaining
superior performance. New York: Free Press.
Reisman, J., & Gienapp, A. (2004). Theory of change: A practical tool for
action, results and learning. Baltimore: Annie E. Casey Foundation.
Retrieved December 7, 2011, from
http://www.aecf.org/upload/PublicationFiles/CC2977K440.pdf
Scheirer, M. A. (1987). Program theory and implementation theory:
Implications for evaluators. In L. Bickman (Ed.), Using program theory
in evaluation. New directions for program evaluation, 33, 59–76. San
 Francisco: Jossey-Bass.
Weiss, C. H. (1995). Evaluation (2nd ed.). Upper Saddle River, NJ: Prentice
Hall.
Weiss, C. H. (1995). Nothing as practical as a good theory. In J. P. Connell,
A. C. Kubisch, L. B. Schorr, & C. H. Weiss. (Eds.). New approaches to
evaluating community initiatives: Concepts, methods and contexts (pp.
65–92). Washington, DC: Aspen Institute.
Wholey, J. S. (1987). Evaluability assessment: Developing program theory.
In L. Bickman (Ed.), Using program theory in evaluation. New directions
for program evaluation, No. 33 (pp. 77–92). San Francisco: Jossey-Bass.

Journal Articles
Birckmayer, J. D., & Weiss, C. H. (2000). Theory-based evaluation in
practice: What do we learn? Evaluation Review, 24(8), 40–43.
Bolduc, K., Buteau, E., Laughlin, G., Ragin, R., & Ross, J. A. (n.d.). Beyond
the rhetoric: Foundation strategy. Cambridge, MA: Center for Effective
Philanthropy. Retrieved December 7, 2011, from
http://www.effectivephilanthropy.org/assets/pdfs/CEP_BeyondTheRhetoric.pdf
Chen, H. T., & Rossi, P. (1983). Evaluating with sense: the theory-driven
approach. Evaluation Review, 7(3), 283–302.
Connell, J. P., & Kubisch, A. (1998). Applying a theory of change approach
to the evaluation of comprehensive community initiatives: Progress,
prospects and problems. Washington, DC: Aspen Institute. Retrieved
December 7, 2011 from
http://www.dochas.ie/Shared/Files/4/TOC_fac_guide.pdf
Donaldson, S. I. (2001). Mediator and moderator analysis in program
development. In S. Sussman (Ed.), Handbook of program development
for health behavior research and practice (pp. 470–496). Thousand Oaks,
CA: Sage.
Fullan, M. (2006). Change theory: A force for school improvement (Seminar
Series Paper No. 157). Victoria, Australia: Centre for Strategic Education.
Retrieved December 7, 2011 from
http://www.michaelfullan.ca/Articles_06/06_change_theory.pdf
Kramer, M. R. (2001, May/June). Strategic confusion. Foundation News and
 Commentary, 42(3), 40-46.
Monroe, M. C., Fleming, M. L., Bowman, R. A., Zimmer, J. F.,
Marcincowski, T., Washburn, J., & Mitchell, N. J. (2005). Evaluators as
educators: Articulating program theory and building evaluation capacity.
New Directions for Evaluation, 108, 57–71.
Rogers, P. J., Petrosino, A., Huebner, T. A., & Hacsi, T. A. (2000). Program
theory evaluation: Practice, promise, and problems. New Directions for
Evaluation, 87, 5–13.
Sridharan, S., Campbell, B., & Zinzow, H. (2006). Developing a stakeholder-
driven anticipated timeline of impact for evaluation of social programs.
American Journal of Evaluation, 27(6), 148–162.

Internet Resources
Sharpe, M. (2009) Change theory. Muncie, IN: Ball State University.
Retrieved December 7, 2011 from http://www.instituteforpr.org/wp-
content/uploads/ChangeTheory-1.pdf

For a comprehensive website and online suite of tools sponsored by

ActKnowledge and the Aspen Institute on Community Change, see the
following:
Anderson, A. (n.d.). A community builder’s approach to theory of change: A
practical guide to theory development. New York: Aspen Roundtable on
Community Change. Retrieved December 7, 2011, from
http://www.dochas.ie/Shared/Files/4/TOC_fac_guide.pdf
Theory of change. (n.d.). Retrieved December 7, 2011, from
http://www.theoryofchange.org
 3

Creating Program Logic Models

T
his chapter identifies the basic elements of a program logic model.
Generally, these models have enough detail to support design,
planning, management, or evaluation. This chapter describes a
program logic model example and the action steps to create a model
with a small group.
 LEARNER OBJECTIVES
Describe the relationship between theory of change and program logic models
Identify basic elements for a program logic model
Create a simple model
Recognize limitations of display

From Theory of Change to Program Models

Theory of change logic models are literally the foundation for program logic
models. When well developed, they can ensure intellectual rigor for program
logic models. Figure 3.1 illustrates the relationship of a theory of change
model (composed of strategies and results) to the primary elements of a
program logic model.
Strategies reflect the resources, activities, and outputs needed to achieve
results. Results reflect the sequence of outcomes over time through impact.
Outcomes (for individuals) are generally progress in changes in awareness,
knowledge, skill, or behavior among targeted audiences. There are also
outcomes for organizations and systems. Although a plausible and evidence-
based connection can be established, impact is often well beyond the scope
(or feasibility) for the program being modeled. Together, outcomes, which
are closer to the effort, of multiple strategies plus impact (further away) make
up results. While program logic models are often built on a theory of change,
it is also possible to infer a theory of change from a program logic model.

Figure 3.1 Relationship of Program and Theory of Change Models

Assumptions Matter
It is important to be aware that specific assumptions are not illustrated in
Figure 3.1. Recall that assumptions are informed by beliefs and knowledge.
Too often, program models are built without the benefit of explicitly naming
the assumptions and underlying theory of change. This omission can help
explain why tremendous conflict, even chaos, can erupt during program
development, planning, implementation, or assessment. In the absence of
explicitly named assumptions, either a clear theory of change does not exist
or people hold multiple and conflicting variations that reflect their deeply
held views about what should/could work and why. This can lead to diffuse
or dilute programs that lack the focus and intensity needed to produce
intended results. Because of these implications, omitting this “foundation” for
your idea, program, or social change effort undermines its potential for
success.
As noted previously, conceptualization and learning styles differ from
person to person. Organizational culture can also affect how design, planning,
monitoring, and measuring occur. Given these practical issues, we strongly
suggest that both theory of change and program logic models eventually be
created to form the foundation of shared meaning for all aspects of the
program. The sequence in which they are developed certainly should and will
reflect the stakeholders’ preferences.

Key Elements of Program Logic Models

Program logic models display what an existing idea, new program, or focused
change effort might contain from start to finish. The elements in a program
logic model consist of the recipe for a bounded investment of financial and
social capital for a specified result. The level of detail increases so that the
relationships shown by the model illustrate essential linkages needed to make
a plan fully operational for each of the strategy strands identified in the
theory of change. The primary elements for each strand of a program logic
model include resources, activities, outputs, outcomes, and impact. Figure 3.2
is a template of the elements for most program logic models.

Figure 3.2 A Basic Program Logic Model

These program logic model elements are defined as follows:

Resources are essential for activities to occur. They can include human,
financial, organizational, community, or systems resources in any
combination. They are used to accomplish named activities. Sometimes
resources are called inputs.
Activities are the specific actions that make up the program. They reflect
tools, processes, events, technology, and other devices that are intentional in
the program. Activities are synonymous with interventions deployed to
secure the desired changes or results.
Outputs are what specific activities will produce or create. They can
include descriptions of types, levels, and audiences or targets delivered by the
program. Outputs are often quantified and qualified in some way. They
simply characterize the application of activities with selected audiences.
Outcomes are about changes, often in program participants or
organizations, as a result of the program. They often include specific changes
in awareness, knowledge, skill, and behavior. Outcomes are dependent on
preceding resources, activities, and outputs. Sometimes outcomes are parsed
by time increments into short, intermediate, and long term. Time spans for
outcomes are relative and should be specified for the idea or project
described. However, short term is often 1 through 3 years, intermediate-term
outcomes 4 through 6 years. Long-term outcomes might be achieved in 7
through 10 years. The intervals specified for any given model would depend
on the size and scope of the effort.
For example, a small-scale project such as an adult education typing class
in one location might produce knowledge and skill outcomes in 6 weeks,
where behavioral changes such as use or changes in employment might take
somewhat longer. Alternatively, a program targeting changes in global water
quality might specify changes in the awareness and knowledge of
international policymakers within 1 to 3 years; actual environmental
improvements might not occur within decades. Typically, dividing the project
duration into thirds works pretty well as a starting point. Relying on a
literature or other evidence base can help inform what is feasible.
Being clear about timing and expected results is important. The time span
for outcomes is project specific. Time is one of several important
considerations. The logical sequencing of any given outcome chain matters,
too. Think about what will happen first, then what is likely to happen next.
Also keep in mind that the sequence may or may not be lockstep. Under some
conditions, there may be different points of entry into a sequence. The
important thing is to explore the interconnections and dependencies that do
exist among the outcomes and impact you specify.
Impact is the ultimate intended change in an organization, community, or
other system. It carries an implication about time. It varies in its relative
timing to the actual program or change effort. Sometimes impact occurs at
the end of the program, but more frequently, the impact sought is much more
distant. For some efforts, this may mean impact can be cited in 7 through 10
years or more. This can have important implications, as it is well beyond the
funding cycle for many typical grant-funded programs or the patience of
many managers or politicians. The logic model is one way to show how the
work you can do within these constraints may contribute to a larger, grander
impact.
The “planned work” of a program logic model includes resources,
activities, and outputs. These are the essential elements that are used to secure
results or make change happen. The “intended results” include what the
program produces: outcomes and impact.

Nonlinear Program Logic Models

Just as in theory of change models, very few ideas, programs, or projects
actually occur in a linear progression. Purposely, to aid learning, we
simplified the display of elements as a straight sequence. Reality suggests
cycles, iterations (additional attempts), and interactions are very common.
This more organic development is shown in Figure 3.3.
In this circular display, there is no specific starting point. Although the
logic model elements are constant, the work of design, planning, managing,
or evaluating might begin with any element. In addition, this view shows how
cycles of the same activity might occur over time. Keep in mind that the
illustration groups activities together. A more detailed view could be
staggering to portray. Sometimes, capturing reality in a display impedes
communication.

Hidden Assumptions and Dose

A program logic model displays the elements that are most critical to
establishing and operating a program or social change effort. It specifies the
activities and their often interdependent relationship as well as what they are
expected to generate. Program logic models do not necessarily include
assumptions, but they rely on them. They offer a view of the map that can
inform action planning and, later, implementation. Program logic models can
also define the “dose” (e.g., number, type, and duration of activities),
quantify and describe the effects and benefits of the program for a given dose
and the ultimate change expected. Dose is an important concept in
effectiveness. A dilute dose can have the same impact as none at all. For
example, if your intended result is a large voter turnout in an election, a
classified ad is not an adequate communication strategy. A comprehensive
media plan coupled with free transportation to the voting booths has greater
chances of success. So, it is important to design a program with enough of the
right activities to secure the outcome you intend.

Figure 3.3 Nonlinear Logic Model

Building a Program Logic Model

Program Logic Model Example

An example of a simple program logic model for securing improved health is
displayed in Figure 3.4. Read from left to right, this program model suggests
that if we recruit and retain participants and provide exercise, nutrition, and
stress reduction, then we will secure improved health. Note the development
of detail connecting strategies to results in this model compared to the theory
of change (see Figure 2.3). The program logic model provides detail for the
theory of change by explicating the elements from a basic logic model for
each strategy strand. In a program model, the details relative to resources,
activities, and other elements are named.
Although still an overview and incomplete, this illustration provides a
more detailed view of what this health improvement program wants to do,
plans to measure, and hopes to achieve. Beginning on the left with resources,
this model includes funds, facility, faculty, and coaches, as well as eligible
and willing participants, among its requisite inputs. To keep it simple, the
strategies contain implied clusters of activities in this illustration. The
specific activities that contribute to outputs are not named. Outputs from the
intervention strategies and associated activities (exercise, nutrition, stress
reduction) could be numerous. For this illustration, we show only the
overarching categories of information that could be considered. Each
category would be repeated for each of the strands. These would include
details about the scope, sequence, and quality of the curriculum; staffing
qualifications; and information about participants and their participation.
Activities “inside” these strategy strands contribute to changes in knowledge,
skill, and adherence. Eventually, they can contribute to increases in strength,
endurance, nutrients, flexibility, and relaxation. Concurrently, over time,
these same strategies also yield reduced fat/calories. The retention and
recruitment strategy strand also generates some outputs and outcomes.
Aggregated, activities within this strategy secure and keep participants in the
program. Note that this model uses arrows to show relationships. Sometimes
they reflect a cluster (indicating synergies) rather than just one:one
relationships.

Figure 3.4 Improved-Health Program Logic Model

 As is typical of many programs, several strategies may be shown as
contributing collectively to outcomes rather than each strategy making its
individual contribution to distinct outcomes in isolation. Collectively, the
long-term outcomes generate improved health, which could be measured in a
variety of ways (e.g., blood pressure, blood lipid and sugar profiles, weight).
In contrast to the big-picture view that theory of change models offer,
program logic models provide a closer, more detailed picture of operations.
This view of the program provides adequate detail to then create work plans.
Program models can provide a reliable outline for work plans that are used to
implement and manage a program or larger change effort. Just like theory of
change models, program models are often logical—but here, feasibility,
given limited time and resources, is the appropriate standard for assessing
their value. A common question about program logic models focuses on their
level of detail. Essentially, the level of detail in program logic models should
be determined by their intended use and users. Although somewhat
situational, program logic models build out strategies to activities. Sometimes
they can even get to the fine detail of tasks, although more often that is
described in an operations or action plan.

From Strategy to Activities

Some program logic models can be extremely complex, but the steps to
create them are generally the same as for more simple efforts (see Figure
3.2). Large-scale programs or multiyear change efforts (sometimes called
“initiatives”) often are composed of many strategies aimed at target audiences
across many sites over considerable time.
Program logic models usually do not display underlying beliefs or
assumptions. They are nevertheless important elements in the conscious
exploration of multiple target audiences. Sometimes programs or change
efforts are implemented in a cascade with some overlap in time, which
requires a particular sequence of strategies and associated activities. When
this is the circumstance, it can be helpful to focus on a function, a given
strategy, or one partner’s designated work. The task is often simplified by
thinking about a single aspect and then connecting it back to the whole with
some of the inherent complexity reduced. Ultimately, program execution
relies on integrated action—but the work that precedes it may require focused
developmental attention on smaller parts.
Using the health improvement program example, Figure 3.5 provides an
orientation to how the exercise strategy strand might be reduced to activities.
It breaks the strategy into greater detail for the purposes of selection and
design.
In Figure 3.5, it becomes evident that exercise as a strategy is made up of
several key activities. They include physical exercise (strength and
endurance), education, and assessment. Together, all of these activities
represent a comprehensive strategy, exercise, that is just one means to
improved health. Recall that the whole theory of change for this example also
includes stress reduction, nutrition, and retention. It is the combination of
strategies reflected in the whole program that is most likely to secure results.
Each strand of a comprehensive program logic model needs to illustrate the
contribution of each strategy as well as the interdependence.
As you specify the activities content of your strategy, you are naming
more precisely what makes up the given strategy. Later, the whole model is
tested for feasibility—both practically before implementation and literally
when the program is evaluated.
In Figure 3.6, we provide a view with greater detail for only the exercise
strategy. In this illustration, we show the detail of activities within the
exercise strategy. It also suggests the many decisions hidden in program
design and planning. In choosing activities, it is critical that the relationship
among strategies and activities is intentional. The strategies and the cluster of
appropriate activities should also be chosen with reference to a target
audience. Remember, logic models use if–then sequences from left to right in
the columns and among the features as you read from left to right.

Figure 3.5 One Strategy With Multiple Activities

Figure 3.6 Exercise Strategy Detail

Action Steps for a Program Logic Model
The practical construction of a program logic model often begins with one or
more information sources (e.g., research, interviews, documents). We
recommend that people begin both theory of change and program logic
models with the named ends. People are most clear about their intended
results (outcomes and impact). Our experience is that you do know what you
want to accomplish. The results sought reflect both the impact intended and
the outcomes over time. Next, name the changes or outcomes that will be part
of your progress toward impact. Unpacking this sequence is important
because it makes it easier to see the strength of the connection between what
you do and what you can get. We suggest tackling the activities required to
achieve the outcomes you have specified in your third step. Activities are
about how intended changes will occur. Fourth, resources/inputs become the
essential ingredients of activities. And finally, outputs reflect the information
needed to verify that activities named earlier in the process reach the right
audiences and are of the quality and quantity needed to produce results. So,
the steps to draft a program logic model are ordered in this way:

1. Identify the results that one or more strategies will ultimately generate.
2. Describe the stepwise series of outcomes (or changes) that will show
progress toward impact.
3. Name all the activities needed to generate the outcomes (for each
strategy).
4. Define the resources/inputs that link directly to and will “supply” the
activities.
5. Identify the outputs that reflect the accomplishment of activities.

Figure 3.7 illustrates these action steps and their sequence.

Creating Your Program Logic Model

The format of a logic model helps organize information in a useful way.
Think of an idea, project, or program you manage now or want to create and
its results. For each strategy, brainstorm elements that might be cited in short-
term outcomes first but are clearly linked to your intended results. Do the
same for resources, activities, and outputs. It is important to make choices
about the outcomes that are feasible with your limited resources. This is
discussed in greater detail in the next chapter.
With some experience, you will begin to recognize commonly used
strategies that reflect knowledge from your field or discipline. For example,
marketing/communications, recruitment, retention, professional development
or education, advocacy, and policy are strategies often found in program
models. Examples of activities under a marketing/communications strategy
could include preparing a database of target markets, generating news
releases, creating and sending a newsletter, establishing a website, and
distributing public service announcements. We suggest you tackle one
strategy at a time. Aim to define the same level of detail for each strategy.
Selected examples of archetypes, or tested recipes, are described in Chapter
7.

Figure 3.7 Steps in Creating a Program Logic Model

Guiding Group Process

We think the best method for generating a program logic model engages a
small group, especially if the members are stakeholders in the idea or
program it represents. Stakeholders are situational, but they generally are
those who have an interest in or are likely to benefit from the program. Logic
modeling often includes funders, program staff, and program participants.
Intentionally including stakeholders supports best contributions as well as
some subsequent benefits relative to implementation. The facilitation of
modeling requires some advance planning and a commitment to both
discipline and quality during the process.
If you’ve already constructed a theory of change, use it to catalyze the
creation of the program logic model. If not, defining shared understanding for
specified results gets your group process effort started. It is important to note
that models may need to be updated to respond to the dynamics of an external
environment (context). They also reflect living systems that are not
mechanistic but changing. For these two reasons (and others), it is necessary
to expect models will be revised. In association with some public
specification of time, impact and outcomes can be explored and selected.
This can be accomplished a number of ways.
 We have had success in using the action steps noted, particularly when
each participant contributed to brainstorming the model elements by
nominating contributions on sticky notes. This quickly generates a large
number of possibilities for each element. Redundancies should be noted and
celebrated as commonly held. Then, the group can sort them: those that must
be kept, that could be kept, and those that are not relevant. Once the results
are named, then it is possible to compose content for the other elements. In
this disciplined sequence, each stakeholder contributes to the whole, and each
contribution has the benefit of an internal test relative to design.
There are several variations on this approach. From a group, you could
invite individuals or pairs to generate models in the sequence shown
previously and then reconcile the variations. This approach helps avoid
“groupthink” but requires strong process facilitation with content knowledge.
A generic model or template for a given program may be available. With
some advance planning, it’s possible to identify one of these archetypes and
introduce it to your group. Then, the content adaptations can focus on
improving it so that the content is relevant to your purposes, conditions, and
planned results. Chapter 7 includes examples of models that can be used in
this way. Sometimes a call for proposals or funding opportunity will
articulate the range of acceptable content in a model.
Regardless of the process, strategic decisions about activities and the
relationships between elements should be made from among all the content
generated. It is important to consider criteria for choices that reflect context,
target audience(s), research, practice, literature, and program benchmarking,
as well as resource parameters. It can be very helpful to have draft models
critically reviewed in a “mark up.” This is described in the next chapter as a
quality assurance process.
We often use Microsoft Visio to construct our models, but many other
applications such as Word and PowerPoint have drawing options. These as
well as Inspiration software are all readily available. The Supplemental
Readings list at the end of the chapter identifies some examples of other free
and commercial software applications. Take care in using technology for
model creation, because it can exclude valuable participation.

IN SUMMARY
High-quality program logic models depend on the evidence base found in
their parallel but simpler theory of change models. Program logic models
display several important elements: resources; activities; outputs; short-,
intermediate-, and long-term outcomes; and impact. To create a program
logic model, start with the intended results: outcomes and impact. Then,
activities (which are consistent with strategies in the theory of change model)
are selected. Next, resources and outputs are cited. We believe creating
models with deep participation of stakeholders improves their quality and
encourages their use.

LEARNING RESOURCES

Reflection
1. What are the implications of a program logic model built without a
specific theory of change?
2. Think of a successful business and its product or service. What is the
underlying program logic that shows the explanations for profitability?
3. Feasibility relies on several aspects. Can you name some?
4. What are strengths and limitations of a linear or a nonlinear display?
Would individuals from different fields (and their relevant cultures)
answer similarly or differently? Why?
5. Why is being specific about results important?

Application
Specify the result of a shared program, project, or idea. Draw a theory of
change model for the program, project, or idea. Then, attempt a program
logic model. Using sticky notes or pieces of paper, brainstorm the outcomes
that need to happen to secure the result. Organize them into short,
intermediate, and long term. Pick one short-term outcome. Brainstorm what
activities are critical to that outcome. Organize the activities relative to a
single or multiple strategies. For given strategies and their activities, name
the resources needed. From the activities, cite what outputs are possible.
Organize these elements as one model.

References and Supplemental Readings

Texts
Frechtling, J. (2007). Logic modeling methods in program evaluation. San
Francisco: Jossey-Bass.
Green, E. L. (2005). Reinventing logic models: A stakeholder-driven group
approach. Unpublished doctoral dissertation, University of Cincinnati,
OH.
Mayeske, G. W. (1994). Life cycle program management and evaluation: An
heuristic approach. Washington, DC: United States Department of
Agriculture.
Stringer, E. T. (2007). Action research (3rd ed.). Thousand Oaks, CA: Sage.
United Way of America. (1996). Measuring program outcomes: A practical
approach. Alexandria, VA: Author.
Westley, F., Zimmerman, B., & Patton, M. Q. (2007). Getting to maybe: How
the world is changed. Toronto: Vintage Canada.
Wong-Rieger, D., & David, L. (1996). A hands-on guide to planning and
evaluation. Ottawa: Canadian Hemophilia Society.

Journal Articles
Cooksy, L. J., Gill, P., & Kelly, P. A. (2001). The program logic model as an
integrative framework for a multi-method evaluation. Evaluation and
Program Planning, 24(2), 119–128.
McLaughlin, J. A. (1999). Logic models: A tool for telling your program’s
performance story. Evaluation and Program Planning, 22(1), 65–72.
Millar, A., Simeone, R. S., & Carnevale, J. T. (2001). Logic models: A
systems tool for performance management. Evaluation and Program
Planning, 24(1), 73–81.
Porteous, N. C., Sheldrick, B. J., & Stewart, P. J. (2002). Introducing
program teams to logic models: Facilitating the learning process.
 Canadian Journal of Evaluation, 17(3), 113–141.
Renger, R., & Titcomb, A. (2002). A three-step approach to teaching logic
modeling. American Journal of Evaluation, 23(4), 493–503.
Rush, B., & Ogborne, A. (1991). Program logic models: Expanding their role
and structures for program planning and evaluation. Canadian Journal of
Program Evaluation, 6, 95–106.

Internet Resources
For comprehensive bibliographies and links to additional resources, see
Logic model resources. (n.d.). Atlanta, GA: The Evaluation Working Group
of the Centers for Disease Control and Prevention. Retrieved December
8, 2011, from http://www.cdc.gov/eval/resources/index.htm#logicmodels
Jung, B. C. (1999–2012). Evaluation resources on the Internet. Retrieved
December 8, 2011, from http://www.bettycjung.net/Evaluation.htm
University of Wisconsin Extension. (n.d.). Logic model bibliography.
Retrieved December 8, 2011, from
http://www.uwex.edu/ces/pdande/evaluation/evallogicbiblio.html

For logic model development toolkits, see

Administration for Children and Families. (2006). Instructional guide—
Creating and using the logic model for performance management.
Retrieved December 7, 2011, from
http://www.acf.hhs.gov/programs/cse/grants/resources/logic_model/
Bureau of Justice Assistance. (n.d.). Developing and working with logic
models. Washington, DC: Author.
http://www.ojp.usdoj.gov/BJA/evaluation/guide/pe4.htm
Centre for Development Innovation (2010). Logical framework approach.
Wageningen, the Netherlands: Author. Retrieved April 18, 2012 from
http://portals.wdi.wur.nl/ppme/index.php?Logical_Framework_Approach
Community toolbox: Developing a theory of change or logic model. (n.d.).
Lawrence: University of Kansas. Retrieved December 8 2011, from
http://ctb.ku.edu/en/dothework/tools_ tk_4.aspx
Enhancing program performance with logic models. (n.d.). Madison:
 University of Wisconsin Cooperative Extension Service. Retrieved
December 8, 2011, from http://www.uwex.edu/ces/lmcourse/
Graig, E. (n.d.). Logic models: A tutorial. Useable Knowledge, LLC.
Retrieved December 8, 2011, from
http://www.usablellc.net/resources/logic-model-tutorial
Jackson, B. (1997). Designing projects and project evaluations using the
logical framework approach. Geneva: The World Conservation Union.
Retrieved April 18, 2012, from
http://cmsdata.iucn.org/downloads/logframepaper3.pdf
Point K: Practical tools for planning, evaluation and action—Logic model
builder. (n.d.). Washington, DC: Innovation Network. Retrieved
December 8, 2011, from http://www.innonet.org/index.php?
section_id=64&content_id=185

For logic model development software, see

Duigan, P. (2011). Do View 3.0: Visual results and outcome planning.
Retrieved December 8, 2011, from http://doview.com/
Milstein, B., & Chapel, T. (2011). Tool: Peer discussion of software
programs to help you model. Retrieved December 8, 2011, from
http://ctb.ku.edu/en/tablecontents/sub_section_tools_1877.aspx
ImpactMap. (2011). ImpactMap: Navigate social change. Retrieved
December 8, 2011, from http://www.impactmap.org/
Results Technologies, Inc. (n.d.) Results online: Outcome tracking software
and outcomes-based applications. Retrieved December 8, 2011, from
http://www.resultstechnologies.net/index.php?process=solutions
 4

Modeling

Improving Program Logic Models

T
his chapter focuses on improving models through simple processes
that test feasibility. With careful and deliberate review, models for an
idea, program, or project can change and mature in their quality.
Logic models that are accurate and realistic representations of what
you do and will get can increase the likelihood of effectiveness.
 LEARNER OBJECTIVES
Apply simple review and improvement steps to models
Identify common errors in program logic models
Recognize the value of multiple versions of models
Recognize contributors to model quality

The process of modeling supports better thinking about a given idea or

effort. It can establish routines wherein alternative possibilities are considered
and explored. Modeling may be an important antidote to snap judgments. We
see modeling as an essential step that has tremendous potential to position a
project, program, or initiative for greater effectiveness. It generally involves
several versions or attempts at models that result from a critical review of the
information displayed.

Modeling and Effectiveness

Once a shared model is created, then those who created the model and others
who are external to the program should spend time on its critique and
revision. Great plans and programs come from great models. Great models
require several rounds of revision. Modeling is the most important content in
this book because critical thinking is what refines the content of the model.
Modeling is a means to illustrate where you are headed, make better
decisions about getting there, and get closer to the right work. We think it is
important to test both plausibility and feasibility through modeling. In
Chapter 2, we stated that plausibility is the most important criterion for a
theory of change model. The process of toggling between combinations of
strategies and results to secure an evidence-based model tests and verifies
plausibility. We provide techniques that address the aspect of feasibility once
plausibility has been established or confirmed. These criteria reflect the
widely used Program Evaluation Standards (The Joint Committee on
Standards for Educational Evaluation, 1994) that include utility, feasibility,
propriety, and accuracy. Given the strong relationship between programming
(strategy) and evaluation (results), we draw upon what constitutes sound
evaluation practice to inform logic model development.
This chapter identifies some of the common context challenges that
influence models and mitigate their potential. Some result in flaws embedded
in the models and others occur in the modeling process. Our list includes
blind spots, myths, “logic,” and scale. We also name an overarching concern:
culture. It can influence all the preceding items and more. Next, we identify
some quality review techniques for program logic models. Through this
sequence, we name the problems often associated with models and attempt to
offer some remedies.

Context Challenges

Common Pitfalls: Blind Spots and Myths

Generating high-quality models through modeling is not easy, but both the
products and the process do offer significant benefits. Some discomfort
among participants during modeling can be an indicator that the process is
useful. This discomfort probably reflects the twists, turns, flips, and angles of
new or different thinking. It is important to be aware of blind spots (which we
all have). Individuals have blind spots and small groups can have collective
ones they protect and promote through interaction among participants. Blind
spots are simply unintentional omissions in our thinking or commissions of
error that happen because of habit (e.g., snap judgments, confidence,
experience enables them). Eventually, these blind spots show up in our
models because models represent how we think.
Prevailing myths are another pitfall to guard against. Common examples
of myths include “access equals use,” “knowledge equals action,” “activities
equals outcomes,” and “send equals communication.” These “myths” end up
as embedded flaws in models because they are not precisely named and
subsequently discarded. Almost every organization, department, or unit
perpetuates some myths. We think it’s how people defend and rationalize bad
habits. These habits are often a way of coping with the organization, a boss, a
funder, or another entity. Avoiding the pitfalls of blind spots and myths
through disciplined and intentional discovery can contribute to model quality
and eventually to program effectiveness.
Logic, Scale, and Specificity
Logic models often display a logical sequence or chain of events, but this is
not the same as feasibility. The content and order of activities as well as other
elements in a model can make sense, but scale is often an obstacle. Scale is
about the relative size of a given effort. In particular, we mean the
relationship between the results sought and the quality, volume, frequency,
and other characteristics of the intervention or “doing” that is described.
For example, a logic model could name “a just, sustainable community”
as its impact. Strategies to secure that intention might include some nonprofit
capacity development and leadership development. It is logical to think these
strategies might eventually contribute to a just, sustainable community. But
these strategies alone are insufficient to achieve the desired impact. There are
other essential strategies necessary for this recipe to yield “justice and
sustainability.” This theory of change model is not plausible.
Subsequently, as activities are selected (in association with the capacity
and leadership development strategies), they are far too dilute to have
influence. A program, if deployed as described, would be unable to make
progress against such an ambitious impact. This is particularly true given the
3 to 5 years allotted to most grant-funded programs to secure results
(outcomes and impact). Engaging a few leaders and nonprofit organizations
is inadequate to the huge ambition of a just, sustainable community given the
enormous number of variables that can intervene. The program logic is not
feasible.
If the scale of effort does not match short-term outcomes, then its relative
paucity can overwhelm chances of success. Given limitations, it is best to
revise the model and indicate something that is appropriate. The first error in
plausibility is compounded by a second error of feasibility. This example
illustrates why logic, alone, is not an adequate quality test for models. It is
important to note that sometimes theory of change models are plausible
although the program logic model is not feasible. We believe success is more
likely when plausibility and feasibility are evident in both models.
Given limited human and financial resources along with a time frame, it’s
important to specify your program or change initiative carefully. Making
these choices is difficult but very important. We applaud audacious and
substantial change agendas. However, success is more likely secured if your
work is discrete and seeks near-term outcomes that, if accomplished, will
contribute to results. For example, consider ending childhood obesity. This is
an important and big result that reflects a complex systems challenge. Rather
than taking on all of this substantial responsibility, perhaps the program effort
that reflects your niche is focused on a single aspect such as more physically
active middle school children. A reduction in childhood obesity will be more
likely if you describe a selected population of children, their gender, their
school district, and details of the activities that will constitute your project.
These choices focus your results and intended outcomes. They also help
inform the best prescription to deliver the dose needed to secure success. We
know from experience that specificity is linked to success. Specificity does
not preclude scale or ambition; it can complement those issues. We suggest
that efforts with limited resources and/or relatively short time horizons be
focused on outcomes that are earlier in the sequence on the path to the vision
of the future you want to create. These more proximal outcomes can help
define a contribution to the desired results. The strength of logic models lies
in articulating the contribution to desired impact.
Probability statistics underscore our point: With fewer variables, the
likelihood of the optimal combination increases. As the variables increase in
number, the likelihood of success declines. We advocate for “smaller bites.”
It simply means picking which results are likely and moving backward from
those results while employing a feasibility lens to determine the activities
needed to develop plans, programs, or related efforts.

Politics, Persuasion, and Perception

Sometimes errors in scale are authentic oversights. Other times, the politics
of context are the origin. It is quite common for agendas that include
marketing or positioning to supersede what’s feasible. Models can be very
persuasive. Something that is simple and clearly described in a graphic
format is compelling. Models often help to persuade others because they
clarify meaning and provide a logical organizer consistent with how people
think, from actions to consequence. Depending on intended use, beware of
models that are vague, general, and lacking in focus. Grandiose hyperbole in
a model can indicate it is a marketing tool. Because logic models are socially
constructed, it is important to beware that they may represent a shared view
—but not necessarily factual truth or reality. The model can display “the”
truth or perception of those who created it—or a combination of both. It is an
error to assume every model is an evidenced-based prescription. The rigor of
models in this matter varies considerably and reflects the context of their
purpose and construction.
We indicated early in the book that models are co-created to ensure
shared meaning. Exchange with and among participants during the model
creation and modeling helps to generate common language. This process also
ensures a display that is both intentional and commonly understood. A model
that conveys consistent messages among those who created it may not have
the same meaning to others because they missed the dialogue, exchange, and
iterative processes. An important challenge in this group work is avoiding
simple compromise. Compromise may satisfy the intragroup dynamics and
politics, but it is not the same as strategy. Models need to reflect strategic
thinking and choices. Keeping the focus on strategic choices to secure the
intended end results is crucial to model quality and ultimately to program
effectiveness.
Besides feasibility, we think program logic models should aim at utility.
For this reason alone, models will look different depending on the co-creators
and the intended model use. This situational aspect applies to the level of
detail in a given model, too. Program logic models can be complete but vary
considerably in their content. What one group determines as a useful level of
detail may be inadequate for another. Others may choose to model certain
aspects of their program with different level of detail. Chapters 6 and 7
display some of this variance in format and use in case profiles and examples.
The social construction of models means they will inevitably reflect
assumptions, expectations, use, and other context features.

A Learning Culture and External Review

How models are created influences their format and content. We’ve already
mentioned the composition of the review group. In addition, the culture
(norms, values, beliefs) of those gathered to create and/or improve the model
is very important. The best culture for improving models is one focused on
learning from experience. This implies lots of variation and versions of
models that will help secure the one (eventually) used for plans and
execution. Mistakes are critical to learning, and it’s important that “mistakes”
are welcomed. People often need the subjective processes of critical review to
identify errors and eventually secure remedies. Nobody draws a “perfect”
model on his or her initial attempt. We believe a perfect model is not
attainable regardless of the number of attempts. Any model simply represents
a snapshot of the best thinking at a given point in time.
Processes to test logic models can retain and engage the original
stakeholders who first drafted it. However, models improve most
dramatically when introduced to colleagues or external sources that are not
directly invested in the models’ elements or their implementation. In this
way, “disinvested” parties can offer objective critiques, which often uncover
blind spots and identify weaknesses, flaws, leaps of faith, ambiguities,
“cockamamie,” and fiction.
When inviting an external review, it’s important to be aware that people
may see and read models differently than those who created them. An
external review can help provide great insight to both improved model
content (e.g., resources, activities, outputs) and relative display (by virtue of
placement and arrows that indicate relationship among content). Moving
from general to specific, ask those doing an external review a series of
questions. For example, initial questions might discover more about how the
parts fit or do not fit together, what are other ways the same results could be
achieved, and if the resources are adequate. The review might also inquire
about the assumptions embedded in the model and ask about the underlying
theory of change it has been built on.

Quality Techniques

Modeling
Most ideas, projects, or programs can be characterized in their “life” to
include four simple stages: design, implementation, evaluation and
adaptation. We suggest that modeling is most useful when done in the
creation stage and during evaluation, but models can be used at any stage for
different purposes. Getting things right at the start can be very important to
ultimate results and is a key influence to subsequent stages. Modeling can be
thought of as a review process that occurs prior to implementation or
execution. It is done to improve thinking and the models that reflect thinking.
Time and effort spent in this work can have enormous return on investment
through the influence on the program itself. The steps in modeling are draw
and test. This construct is displayed in Figure 4.1.
As a program, project, or idea is created, we suggest it gets drawn as a
model. The “draw” step is satisfied when all elements of a program model
(see Chapter 3) are present. Completion of this step means resources,
activities, outputs, outcomes, and impact are named. This provides an
opportunity to graphically display the thinking behind how the ideas framed
in the theory of change will be implemented as a program. Many efforts with
logic models quit at this point. However, through modeling, you can move
quickly to dialogue to process the content and the “tangles.” Tangles
represent areas of confusion or where some in your group think a choice is
wrong, confusing, or poorly specified. Modeling is the process that guides
model improvement.

Figure 4.1 Relationship of Modeling to Program Life Cycle

In this chapter, we begin to name how and what can test (or explore)
model quality. We believe this testing can help improve models. The
subsequent versions of models that result from literal and figurative tests are
products of modeling. This process can yield benefits to the specific idea or
project as well as the individuals engaged as a work group. It is important to
be aware that many external issues influence modeling. We describe some of
those issues, but our list is not exhaustive.
Testing Model Quality: SMART and FIT
In a conscious testing effort, one way to explore the quality of a model is to
apply SMART principles to it. SMART is a mnemonic used since the early
1980s to set objectives:

Specific: what to do is clear enough to act on and is connected to

outcomes.

Measurable: the content can be both quantified and qualified.

Action oriented: the content is selected to provoke change in

awareness, knowledge, skill, and/or behavior.

Realistic: the content is both plausible and feasible.

Timed: the content specifies a duration and illustrates the time-

dependent sequence of outcomes for progress toward results.
If just a few people apply SMART principles to the outcomes portion of a
program logic model, it will probably surface several flaws and, at the very
least, some good questions. It may also raise concerns that are unique
reflections of the training, experience, perspectives, and style among those
contributing to the review. We suggest that a SMART analysis be applied to
both the parts and the whole. In other words, it can be used to examine
selected individual strategy strands and activities, as well as the model as a
comprehensive effort.
A second way to explore or test the quality of a model is to consider FIT.
FIT is an acronym that was first coined by the American College of Sports
Medicine as FITT (frequency, intensity, time, and type). Although originally
developed to prescribe exercise, we have modified it for use in our practice to
quantify program delivery. It stands for

Frequency of occurrence,

Intensity or strength of the given effort,

Targeted at a specified market or audience.

 FIT principles offer an easy way to consider the process elements of the
program, project, or idea you display in your model. They translate to
important questions about the adequacy of dose relative to intended results.
The frequency reference implies this question: Does the project (idea or
program) repeat, occur with appropriate volume, or happen often enough so
that results are likely? The intensity reference implies this question: Does the
project (idea or program) have enough depth or concentration that results are
likely? The target reference implies this question: Does the project (idea or
program) aim at an appropriate and specific audience or market segment?
In effect, FIT can remedy “treatment trickle,” one thing known about
resolution of problems or generating a change in awareness, knowledge, skill,
or behavior: The treatment (or intervention) needs to be both appropriate (the
right choice) and adequate (the right dose). The wrong choice won’t help get
the results you want. The right choice has more chance of securing results,
but there needs to be enough of it (volume) focused on appropriate targets to
ensure results. “Treatment trickle” means that a problem or challenge is
underdosed or an effort that is too broad or diluted to have impact. Feasibility
is enhanced when we tackle a desired change with the right strategies as well
as appropriate depth of intervention. Treatment trickle is common because
people prefer to give everything or everyone “some” rather than make hard
choices about who or what gets enough! It is often a reflection of limited
resources, vague goals, or organizations with an activities focus (see Figure
4.2).

A “Mark Up”
In Figure 4.3, we revisit the logic model introduced in Chapter 1 for the
Community Leadership Academy program. We suggest a technique that’s
often used in the legislative process as working drafts of language for a
regulation or authorization are generated. It is called a “mark up.” We adapt
the legislative mark up to raise important questions about model quality by
applying SMART and FIT principles. Other elements, including context and
technique questions, can also be used. This discovery is aimed at changing
the model in constructive ways that reflect evidence, strategic choices, and
better thinking. Using a disciplined approach to modeling captures an
important opportunity for models to mature in quality.
 Figure 4.2 Modeling as Quality Review

Frequently, those who participate in developing a model judge how well

it addresses the SMART and FIT criteria differently from how it is assessed
by an external audience. This is because external reviewers may lack the
context and shared meaning that emerge during development. Not all the
meaning in a model is explicitly captured, displayed, or explained. The
degree to which a model is SMART or FIT is highly contextual and
situational. Ratings will depend on who the model is developed for, the
model’s purpose, and the level of detail the model is intended to
communicate. If it is important that your model be self-explanatory under all
conditions and/or viewed as credible to a wide variety of external audiences,
then consider supplemental narrative. Remember, the model is a summary; it
is usually not intended to tell the whole story.
In the Figure 4.3, questions generated from the application of SMART,
FIT, and other process principles are placed on the model for discussion and
resolution. The provoking dialogue that will likely occur will in turn result in
changes to the model content based on new understandings, evidence, and
other contributions. Your review of this model may raise different questions
about its feasibility.
Figure 4.4 uses the same mark up process to identify questions and
concerns about content in the health model. As a constructive effort to
critique and refine, quality testing is an important but often overlooked step.
In both of these mark ups, FIT questions aim primarily at the quality of
process or intervention dose. SMART questions are focused on the feasibility
of outcomes. It is easy to see that as the model increases in complexity, the
challenges to success increase, too. In a comprehensive effort, more pieces or
parts must align and work together to secure the intended impact. The real
time and context between activities and outcomes means many external
issues also have the potential to influence the outcomes.

Figure 4.3 Community Leadership Academy Program Logic Model

Mark Up

Figure 4.4 Health-Improvement Program Logic Model Mark Up

 Typically, groups vary on how much emphasis they place on each of the
characteristics of quality models described here. As with model interpretation
and review, context and composition of the group frequently determine how
the difficult distinction is made.
This also points to crucial issues in execution. When a model
(representing an idea, project, or program) is evidence based, strategic, and
well designed, implementation can still fail its potential. Although logic
models are often used to inform evaluation, modeling at this stage is rarely
done (see Chapter 5). Although outside the scope of this text, great execution
is absolutely essential to results.
Quality Questions
The models you (and your colleagues) build reflect your collective thinking.
Nobody would argue with this assumption: Best thinking is a critical input to
any work and its likely success. As you consider the design or plans for your
project (program or idea), the following questions reflect quality
considerations. They include application of the SMART and FIT principles.
You might construct other questions based on your own experience or
particular subject matter expertise. Think about ways to conduct discovery
around blind spots, myths, and other culture concerns. Regardless, it is
important that deliberate efforts are invested in the maturation of a model.
Modeling provides an opportunity to test, improve, and revise models. This
process helps develop the thinking and the program that the model represents.

Quality Questions for Program Logic Models

1. Are the results specified with shared meaning among all

stakeholders?
2. What research, practice, and theory are grounding for our choices in
activities? Is there a relationship between the program logic model
and a theory of change?
3. Did we challenge the recipe of activities and outcomes to ensure
feasibility given the challenges we seek to resolve?
4. What activities have worked under similar conditions? What current
conditions might influence selection of activities? Is there evidence
that these activities are best?
5. Is the model FIT?
6. Are target audiences clearly defined?
7. Does the interaction among activities contribute to intended
outcomes?
8. Do the outcomes feasibly contribute to intended results?
9. Is the model SMART?
10. Do we have adequate resources and time to accomplish intended
results?
A Quality Model
Figure 4.5 displays characteristics for model quality. It assembles key
narrative from this chapter into a graphic. The figure describes two important
standards for model quality: plausibility and feasibility. The quality
characteristics for theory of change models are noted (as in Chapter 2), where
the focus is on the relationship between strategies and results. The quality
characteristics for a program logic model are captured by FIT and SMART
principles. We suggest the application of this in a mark up is one way to work
with colleagues in modeling. We think, taken together, theory of change
models that are plausible and program logic models that are feasible can
contribute to effectiveness.

“Better” Decisions
Earlier in the text, we asked three questions about effectiveness:

Are you doing the right work?

Can you make better decisions?
Are you getting superior results?

The second, about making “better decisions,” is highly central to

modeling. In effect, we apply this query to strategy and activity selection
relative to results. How you choose to focus time, energy, talents, and
resources in relation to your specified success is related directly to
effectiveness. There are many ways to secure a named intention. Discarding
strategies/activities that are peripheral, modest contributors, or less than
optimal focuses limited resources. Further, specificity about strategies and
activities contributes to the results you and your colleagues intend to secure.
Ambiguity should be avoided. Eventually, “right work” is also about the
strategies and activities in the program logic model. Giving conscious
attention to the criteria employed in selecting strategies for the theory of
change model and then the specific activities in the program logic model can
make a big difference in securing results. “Right work” gets clarified and
better decisions considered through the process of modeling. Models and
their versions can develop discipline as well as standard practices that
contribute to new understandings about what will generate progress on
results.

Figure 4.5 Factors and Features of Quality Models

IN SUMMARY

Although frequently overlooked, modeling is a helpful technique to improve

your chances of success, because great design and plans come from great
models. We suggest iterative versions of models are co-created through a
disciplined process that tests and retests quality. Models can be greatly
influenced by unintentional omissions (blind spots), myths, politics,
persuasions, and perceptions. Logic, scale, and specificity all have a bearing
on models, too. A simple way to explore model quality is the application of
SMART and FIT principles. We suggest a mark up as a good way to
critically review your program logic model.

LEARNING RESOURCES
Reflection

1. Given how subjective program logic models are, what are the
implications for the outside “reader” of a model? What does a model
that will be read and perhaps used by those other than those who
constructed it have to communicate?
2. What role might politics, persuasion, or perception play in how a model
might be created, tested, and improved? How do these issues influence
model quality and use?
3. What prevailing myths might influence choices in your workplace or
family? How do blind spots influence choices?
4. How might the improvement process for a simple, single-site project
model be different from that for a more complex multisite, multilevel
initiative? What concerns should the model development team be sure to
address, and what aspects of the model will be most important to
communicate?
5. Can a complex, comprehensive program be effectively modeled with a
single diagram? Why or why not? How would you approach a task like
this?

Exercises

1. Imagine what a strand focused on marketing might look like to promote

an innovative new hearing aid. Make an assignment for individuals to
illustrate this strand. Give them all the same amount of specified
financial resources over 12 months and tell them the desired outcome is
maximum sales volume. Then, ask them to name their target market and
key activities in the marketing strategy. Compare and contrast versions
each participant creates. What assumptions, knowledge, or evidence
accounts for the differences in how the marketing strands are
constructed? What criteria are useful for reconciling these different
illustrations?
2. Consider the questions raised in Figure 4.3 about the Community
Leadership Academy program logic model. How would you answer
them? What would the model look like once those questions had your
 answers? Draw it and discuss the implications of your response to the
mark up.

Consider the questions raised in Figure 4.4 about the Health

Improvement Program. How would you answer them? What would the
model look like once those questions had your answers? Draw it and
discuss the implications of your response to the mark up.
3. Draw a logic model for learning a foreign language. Ask a colleague to
conduct a mark up using SMART and FIT on the models. What might
change? How? Why?

References and Supplemental Readings

Texts
The Joint Committee on Standards for Educational Evaluation. (1994). The
program evaluation standards: How to assess evaluations of educational
programs (2nd ed.). Thousand Oaks, CA: Sage.
Van Hecke, M. L. (2007). Blind spots: Why smart people do dumb things.
Amherst, NY: Prometheus.

Journal Articles
Alter, C., & Egan, M. (1997). Logic modeling: A tool for teaching critical
thinking in social work practice. Journal of Social Work Education,
33(1), 85–102.
Doran, G. T. (1981). There’s a S.M.A.R.T. way to write management’s goals
and objectives. Management Review, 70(11), 35–36.
Dwyer, J. (1996). Applying program logic model in program planning and
evaluation. Public Health and Epidemiology Report Ontario, 7(2), 38–46.
Israel, G. D. (2010). Using logic models for program development (AEC
360). Gainesville, FL: University of South Florida, IFAS Extension.
Retrieved December 7, 2011, from http://edis.ifas.ufl.edu/wc041
Julian, D. (1997). The utilization of the logic model as a system level
planning and evaluation device. Evaluation and Program Planning,
 20(3), 251–257.
Renger, R. (2006). Consequences to federal programs when the logic-
modeling process is not followed with fidelity. American Journal of
Evaluation, 12(27), 452–463.
Rush, B., & Ogbourne, A. (1991). Program logic models: Expanding their
role and structure for program planning and evaluation. Canadian
Journal of Program Evaluation, 6, 95–106.

Internet Resources
Burke, M. (n.d.). Tips for developing logic models. Retrieved December 7,
2011, from http://www.rti.org/pubs/apha07_burke_poster.pdf

In addition to practicing the review steps on your own models, there are
many other examples of logic models to work from available on the Internet.
For several different approaches, see the following:
Duigan, P. (n.d.) Outcomes model listing. Retrieved December 8, 2011, from
http://www.out comesmodels.org/models.html
Tucson: University of Arizona Cooperative Extension. (2009). Logic models.
Retrieved December 7, 2011, from
http://extension.arizona.edu/evaluation/content/logic-model-examples
Capable communities: Examples. (2011). East Lansing: Michigan State
University. Retrieved December 7, 2011, from
http://outreach.msu.edu/capablecommunities/examples.asp
SEDL. (2009). Research utilization support and help: Logic model examples.
Retrieved December 8, 2011, from
http://www.researchutilization.org/logicmodel/examples.html
 PART II

Applications
 5

Logic Models for Evaluation

T
his chapter focuses on using logic models as the architecture for
deeper engagement of stakeholders in discussion about evaluation
design. Logic models inform the development of several elements of
evaluation design. Logic models are a powerful device even if they
have not been used for program planning. This chapter covers selected
concepts useful to an evaluation consumer.
 LEARNER OBJECTIVES
Describe the contributions logic models can make to evaluation design
Use a logic model to focus evaluation on high-value information needs
Use a logic model to provoke dialogue on both process and results indicators
Identify how logic models can be used to increase effectiveness

Getting More Out of Evaluation

Connecting Management With Measurement

So far, logic models in Chapters 2 through 4 have been shown as significant
tools to assist in improving the thinking behind the design and/or planning of
your efforts. At this point, remember that the model is just an illustration. A
model at the program planning stage functions much like the map an explorer
would use to guide a journey—to chart new direction. It highlights the
portions of program operations as well as the pathways and milestones you
believe are most critical to achieving desired results. The assumptions that
were used to construct the model during program planning need to be put to
the test. As your program is implemented, there are ample opportunities to
observe and collect information about what works and what does not.
Evaluation, whether informal and anecdotal or formal and highly technical,
supplies that vital information.
Evaluation is not just performed by evaluators. All day long, people
determine (and use) information about processes and connect it to results.
They do this to assess progress and for the purpose of improvement. We all
have at least once in our lives looked in the mirror, gasped, and made some
adjustment, either right then or at some future point. In this situation, the
result could be how we or others perceive our image. The process would be
any of those actions we take to achieve the image we desire (haircut,
wardrobe, grooming, etc.). Any change in our actual appearance will be
reflected in the mirror. Evaluation in isolation will not improve effectiveness
—it is just an activity.
 Funders, program staff, and participants are typically referred to in
evaluation circles as stakeholders. We use this term in earlier chapters when
explaining their role in logic model development during program planning.
During program design, funders and program staff serve as the program
architects and content experts, with participants as the experts in community
context and benefit. However, when the time comes to use logic models for
evaluation, the stakeholder role changes somewhat. During evaluation,
stakeholders also play a key role as information users although they are not
often experts at evaluation. This is why we refer to them as evaluation
consumers in this chapter. Evaluation must be understood, appreciated, and
used by evaluation consumers if it is to have any influence on improvement.
It is only when evaluation consumers take a more active role in connecting
what they do with what they get that they can benefit from their evaluation
investment.
The logic model serves as the focal point for discussion about evaluation
because it displays when, where, and how to look for the information most
needed to manage the program and determine its effectiveness. Although this
chapter focuses on using logic models to inform evaluation design, they can
be used in a number of other ways to benefit programs as well. Because logic
models show the key elements of your program, they can also be employed
for focusing your internal communication or external dissemination
messages, audiences, and products. In addition, logic models point out the
categories of data sources, expertise, and documents essential to include
when developing knowledge management systems to support program
operations.
When the model of your program is used to guide evaluation, it then
becomes more of a navigational aid (informs direction) and a dashboard
(informs data collection and gauges progress). The crucial step in improving
effectiveness, however, is to ensure the evaluation provides appropriate and
timely feedback to management. It follows then that management is
sufficiently evaluation literate and takes the time to reflect on the meaning
and significance of evaluation data. For evaluation to be truly useful, its
findings must be relevant, understood, and applied by program staff. Logic
modeling enables evaluators, program staff, and funders to develop shared
understanding about what the evaluation will cover, its purposes, and how the
information collected will be used. The way to get more out of evaluation is
to plan for its use right from the beginning.
 Evaluation typically includes the collection, analysis, interpretation, and
reporting of data needed to inform the decisions of those who fund, provide,
or administer programs. Anytime assessment occurs, there is opportunity to
apply the information gathered to improve the effectiveness of programs,
policies, personnel, products, and organizations. We believe that one key to
effectiveness is to couple program design and implementation with
evaluation. Logic models help accomplish this because they provide a set of
clear, measurable, and realistic program processes and outcomes. If these are
overly optimistic or are not measurable, the program may neither be able to
operate as planned nor demonstrate that it has been effective.
A decade ago, logic models were used almost exclusively to plan and
execute evaluations. Over time, evaluators and program staff have learned
that there are benefits to building the evaluative thinking that the logic model
process brings right from the start. A logic model for evaluation is like the
“bookend” or mirror image of one for planning and implementation. They
both display the same content. By using a logic model to inform evaluation
design, program staff, funders, and evaluators can have a clearer picture of
what their evaluation will measure and produce.
This chapter does not prepare you to conduct an evaluation. Instead, it
points out the key areas where a logic model can support program staff,
funders, and evaluators during evaluation design. They do this by focusing
decisions on what the evaluation will include and by planning for the use of
evaluation information to support learning and program improvement. This
chapter shows how the architecture of logic models can be used as the
platform to guide discussions about why, where, when, and for whom
evaluation activities should occur to generate maximum utility. Through
evaluative thinking and enhanced awareness about evaluation, models can
support a more active role for program staff and funders in shaping
evaluation to better meet their learning and accountability needs.

Evaluation for Effectiveness

In simple terms, evaluation often involves a critical review of the strategies
and activities (the “doing”) selected and their implementation and results (the
“getting”). There are two basic types of evaluation: summative and formative
evaluation. The following metaphor (paraphrased from Michael Scriven) is
widely used to illustrate the difference between these two types of evaluation.
When the cook tastes the soup, that is formative evaluation, but when the
guests taste the soup, that is summative evaluation. Also, summative
evaluation is sometimes used to determine causation (that the program
actually caused the effect) in preparation for replication or scaling up an
intervention.
A summative evaluation is typically done for the purpose of
accountability and determines whether intended results were achieved. It
generally responds to the question: What difference did we make? So this
type tends to emphasize the “getting” or outcomes and impact side of the
model. Summative evaluation tends to be retrospective in that it usually
occurs at the conclusion of a program. For example, the summative
evaluation of an educational initiative might focus solely on student
achievement. Or a summative policy evaluation might look only for changes
in specific legislation.
A formative evaluation is most often used for the purposes of learning
about what works and/or program improvement. It generally responds to the
question: How can we be more effective? Formative evaluation focuses
prospectively on improvement by looking at both the “doing” and the
“getting” sides of the model. It looks at the relationship or “logical”
connection between processes and outcomes throughout the program
duration. Building on the educational initiative example above, a formative
evaluation might also examine and address the factors that influence student
achievement, such as quality of instruction, curriculum, and/or assessment.
Similarly, a formative policy evaluation might examine the quality of the
policymaker education efforts and look for changes in attitudes or knowledge
as essential first steps toward legislative action. Evaluation findings might
show program leadership and staff that their message is not persuasive
enough, that their contact is too brief, or that too few policymakers are
reached and might suggest that they change their approach accordingly.
These two types are complementary, but the lines of distinction between
the two are often highly contextual and blurred. The information both
approaches can produce is needed to improve results. Regrettably, most
evaluations are conducted for compliance with requirements set by funding
sources, including governments and grantmakers. Many times, such
evaluations are not used by program managers, funders, or other
stakeholders. Typically, these evaluations are focused on verification that
dollars were spent in approved categories and that target audiences received
services in the numbers anticipated; very little focus is on the difference
made. So these evaluations often capture and codify information already
known by the program staff. Sometimes they are highly technical and
academic and explore aspects that have little relevance to the day-to-day
operation of a specific program.
Improved effectiveness requires that evaluation consumers have the
information they need. Achieving that requires some degree of engagement
by evaluation consumers in the evaluation design process. Logic models
facilitate communication between evaluation experts and the consumers they
serve. They communicate the essence of a program as envisioned to inform
evaluation. Conversely, they also communicate the essence of the evaluation
to inform program.

Evaluation Design Basics

Where Consumers Add Value

Typically, evaluators rely on an evaluation plan to communicate the details of
what they propose to do. Evaluation consumers frequently use the evaluation
plan as the starting point for evaluation contracting. An evaluation plan
should describe what information the evaluation will secure, what purpose it
will serve, and for whom. In addition, it often includes methods for data
collection, a schedule for tasks (or a timeline), personnel qualifications,
reporting, and cost. There is often quite a bit of negotiation from the original
request for proposals, initial proposal, and final evaluation plan. It is during
this back and forth between the evaluation experts and consumers that
evaluation-literate consumers can best take action to focus their evaluation
investment on effectiveness. Readers interested in detailed information on
evaluation are referred to the Supplemental Readings list at the end of this
chapter.
Figure 5.1 shows the basic elements of an evaluation design. Reading
from left to right, an evaluation design first and foremost needs to specify the
purpose of the evaluation. Purpose includes the intended use and who will
use it. As you recall, formative evaluation serves program improvement and
summative serves the purpose of accountability. However, both types of
evaluation have the same design elements. After that, the design should
describe the information needed to address the questions inferred by the
expressed purpose. Next, the indicators that further specify the quality and
quantity of information are developed. From the indicators, the range of
methods and types of appropriate analyses can be determined. Indicators also
inform the schedule and type of reporting because they show when data
might be available and how best to communicate results.
This chapter is limited to information needs, questions, and indicators
because this is where evaluation consumers are most able to add value. These
three elements define the core content the evaluation needs to address if the
evaluation is to be used as intended. Evaluation consumers are the program
experts and can inform evaluation design in ways that keep the purpose the
evaluation must serve in the forefront. With this focus, those charged with
conducting the evaluation (whether external consultants or internal staff) are
better able to select appropriate methods to capture pertinent information and
work with consumers to agree on reporting style and timing. Those
conducting the evaluation are counted on to provide methodological and
communication expertise.
Evaluation has some of the same limitations that program design,
planning, and implementation face: time, talent, and money. These limits
mean that there are important choices to make relative to information needs
and processes to secure information. Questions like the following become the
foundation for evaluation design:

What are the priority areas for inquiry?

What are the key questions about those areas that we need to answer?
What is “fair” to expect in outcomes given resources and effort in the
program?
What will be used as indicators of sufficient effort (process) and
progress (outcomes)?

The evaluation design process pursues answers to these (and other)

questions. So how does the use of logic models support evaluation consumers
in advocating for their interests during design?

Where Logic Models Add Value

Because a logic model (whether a theory of change or a program model)
illustrates the relationship between planned or actual work and results, it
supports evaluation through commonly understood visual architecture. This
architecture assists in the formulation of key questions and indicators about
the program and its effects over time. It does this in ways that encourage
evaluation use that targets improved effectiveness. Subsequently, appropriate
data collection methods and other aspects of the full evaluation plan can be
identified. All too often, groups are eager to rush into measurement without
first determining what should be measured and why.

Figure 5.1 Evaluation Design Elements

Because only limited resources are usually available for the evaluation, it
is important to identify who the evaluation users are and determine what they
need to know. Generally, there is lots of discussion about what they want to
know or could know. Evaluations are rarely allocated resources that provide
for a thorough examination of all program elements and their relationships as
expressed in a model. Logic models and modeling (which display versions or
aspects in greater detail) can help explore options and point to the most
strategic choices for evaluation investment. Sometimes the evolution of an
evaluation design is a long dance.
At the outset, clear determinations of users and their uses are important
considerations. Knowing your audiences and their information needs will
support good choices and focus your evaluation so that it has optimal utility.
In practice, the functional objective is to specify what information is essential
and secure an evaluation that discovers and delivers in response to that need.
The logic model and modeling process provide the architecture against which
evaluation experts and consumers can decide. The power of evaluation is
harnessed when the findings and analysis generated are applied to the work
examined. With logic models as the framework for design decisions,
evaluation can provide critical feedback loops about the progress of a
strategy, program, initiative, or organization toward its desired results.
Evaluation consumer participation in the logic model development
process (whether during program planning, evaluation, or both) helps to
ensure that the evaluation services they procure address their needs. The tools
and processes of logic modeling provide the opportunity to build common
language and understanding with their evaluation partners about what will be
included in the evaluation and how the information will be used.
Stakeholders, in the role of evaluation consumers, need to know enough
about the evaluation design process to have input on the questions to be
addressed and the evidence that will be used to determine success. Given that
the logic model is the graphic representation of the program’s key processes
and outcomes, consumers can then easily identify and advocate for those
aspects of the model most important from their perspective to manage and
measure.
While the reasons and expectations for evaluation can vary, we are
predisposed to utility. This requires a clear determination of who needs to
know what about the program and to what end? Without logic models to
portray a shared understanding of the evaluation, it may serve some or none
of your audiences. For evaluation to make its full contribution to performance
management and effectiveness, it is important to design the evaluation as a
resource that can support the learning of those for whom its use is intended.

A Design Example

Two Kinds of Learning

We use theory of change and program logic models to plan programs and
their evaluations because together they better stimulate conversations and,
subsequently, learning about those aspects of a program most essential to
success. One important concept frequently overlooked is that in order to use
evaluation to improve effectiveness, you need to engage in learning.
Organizational learning and development experts often speak of two
types of learning that are needed to improve program effectiveness. The first
type of learning examines the current program and uses evaluation
information to improve the program as it was originally envisioned. This type
of learning, technically referred to as “single loop,” does not question the
assumptions and evidence that underlie the program design. The original
thinking is a given and is not examined. This type of learning is focused on
simple changes in the quality and quantity of activities directed toward
producing increasingly better versions of what you originally intended. In this
case, evaluation is used to test and improve the “logic” behind
implementation. Effectiveness is improved incrementally.
The second type of learning examines whether the original program
design is sound. This type of “double loop” learning can transform an effort
into something dramatically different from what was initially intended. Here,
when activities do not seem to produce intended results, the strategies
themselves or the likely outcomes may be questioned and altered. Evaluation
that serves this type of learning is used to test the theory of change
(connection between strategies and results) and improve the logic behind the
design and/or model that drives implementation. For readers interested in
more information about single and double loop learning, see the
Supplemental Readings list provided at the end of this chapter.
In this section, we use the general Community Leadership Academy
(CLA) program logic model example introduced earlier to show how models
can be used to develop and communicate evaluation design that serve both
types of learning.

Key Evaluation Questions

In Chapter 1, using the example of the CLA program, we introduced model
types: theory of change, program, and evaluation. Before we review those
models, it is important to understand that key evaluation questions develop
from simple questions. These questions are “What did we do?” and “What
did we get?” They are asked at the level of the theory of change for the
program as shown in Figure 5.2. Evaluation helps us examine the links in the
logic underneath the planned work and its connection to desired results. By
this link, we mean “testing” whether the theory of change as originally
modeled actually described what happened. Does the original thinking that
asserted “If we do X, Y, and Z, we will get the results we want” hold true?
Here, the model serves double loop learning by helping focus attention on the
bigger picture.
 Before any effort is invested in improving the program as it is currently
designed, the question of whether it was the “right” thing to do should be
determined. By this, we mean two things. First, relative to goal attainment,
we would ask, “Is it aligned with broader organizational goals?” Second,
relative to the activities selected, “Were they among the best choices we
could have made to achieve our desired ends?” This is particularly important
in those instances in which logic modeling was not used during program
design.
The CLA theory of change model (Figure 5.2) helps us explore potential
areas for focused discovery and learning. The basic “do” and “get” questions
help to shape the information the evaluation can deliver. Recall, this program
had two strategies: curriculum and experiences.
In the CLA example, key evaluation questions were co-constructed with
the client, who sought external evaluation services. This client, a foundation,
had sponsored the program for almost 20 years. It and other stakeholders
wanted to know what difference this effort had made and ways the program
could be improved. The program design is simple. It assumes participants (as
an input) and provides greater detail about the two strategies named in the
theory of change: leadership curricula and experiences. An obvious first
question (see Figure 5.2) is whether leadership curriculum and experiences
are the right strategies in a leadership program aimed at community
development. It is important to focus evaluation on testing this basic
assumption (or “link”) about the “right work” because it is the foundation for
the program design. Figure 5.2 shows this first question on the theory of
change model because it is “testing” at the strategy level.
It is also important to note that the CLA, as well as other projects,
initiatives, and organizations, operates in a larger system. For the CLA, it is a
specific community. Organizations operate in other layers of reality besides
geography. This might be a sector or a country. These contexts certainly have
issues, both barriers and facilitators that can exert tremendous influence on
impact. In the CLA example, it may be the abundance or lack of participants.
Or perhaps it is a culture reluctant to invite program graduates into
community work. Organizations are subject to many other kinds of influences
in their external environment, like labor practices, taxes, or consumer whims.
It is worth mentioning here because sometimes evaluations aim at discovery
relative to barriers and facilitators, and these are not always specifically
stated in a theory of change model. When it is possible to include the
examination of these influences in evaluation, they can support improvement
and/or help explain results.

Figure 5.2 Community Leadership Academy (CLA) Theory of Change

Evaluation Question

In Figure 5.3, the program logic model is used to determine the other key
questions central to evaluation design. In this display, we indicate those key
questions that test the implementation logic. This information can be used to
determine areas for improvement and to increase the likelihood or magnitude
of effect. The key questions are placed near links of logic (areas of the
model) that specify where deeper discovery about implementation might
yield relevant information. It is important to note that the questions about
outcome and impact need to be addressed for both types of learning. Both
theory of change and program logic models show the same information, just
in different detail as well as for different purposes. Ultimately, the evaluation
design for the CLA addressed these five key questions:
1. Is the Academy doing the right things?
Question 1 is about the “recipe” for the program. It seeks information
about program content (strategies as well as the resources, activities, and
outputs). It attends to discovery about these, their interaction, and
contribution to results. This exact query is placed on the theory of change
model (see Figure 5.2). The question is hidden in the program logic model,
where the program view has considerably more detail.
2. Is the Academy doing things right?
Question 2 is about the implementation quality or execution of the
selected program content.
3. What difference has the Academy made among participants?
Question 3 focuses on how individuals may have changed because of
their Academy experience.
4. What difference has the Academy made across the community?
Question 4 examines the changes that could be attributed to the
community because of the program.
5. What are the ways that community needs can and should be
addressed by the CLA?
Question 5 seeks other information that can help inform a better or an
improved program. This might be by improving strategy and/or
implementation.
These questions are very typical but highly general program evaluation
questions. In some form, they may even have universal application because
they represent common areas of interest about any program, project, or
initiative. These questions can also be the basis for more precise inquiry or
subquestions in each area. Subsequently, data are collected to respond to
questions.
Theory of change and program models for this effort share the same
intended impact: “community development.” Before evaluation and during
planning, it could be useful to ensure shared understanding of what
“community development” means and what it would look like if the program
were successful. Does “community development” mean full employment, a
vibrant arts culture, effective schools, all of these, or something else?
Similarly, on the CLA theory of change model, note that the outcome of
“more and better leaders” precedes this desired impact. Assuming that “more
and better” means an increased number of designated leaders with skills, then
we could infer skill changes among Academy graduates. Arriving at shared
understanding of what the terms used in the models actually mean helps
determine how they can be measured. Questions like these help evaluators
and evaluation consumers address the “black box” issues facing many
programs. Logic models are ideal tools to use to dissect policies and
programs into their constituent parts. This way, the overall explanation of
what is expected to occur (and, to some extent, why) can be more coherent.
The next place where evaluation consumers can provide insight into
evaluation design is in the development of indicators. Program logic models,
in particular, can be used to develop and display quite specific definitions of
the evidence that evaluation experts and consumers agree is needed to
“indicate” progress from strategy to results during implementation. To inform
effectiveness, indicators of strategy and results are needed.

Indicators
We all are familiar with the indicator lights on the dashboard of our cars.
These lights call our attention to specific automotive functions or
performance issues, and typically they inform corrective steps. A logic
model, when used to improve strategy and results, is similar to the dashboard
in this example. An evaluation will typically focus primarily on
monitoring/measuring the output and outcome elements of a logic model;
thus, the output and outcome elements serve as the indicators of program
performance. We need indicators to help us understand whether we are
making progress. However, as most change does not occur instantly, it is
important to have gauges that show progression over time. Indicator
development is the step between the development of a logic model and the
specification of the metrics (data points) and methods that the evaluation will
use.
Indicators are the evidence that will verify progress (or lack of) for a
given output or outcome. They can be real measures of the concept or
surrogates, which are also referred to as proxy indicators. Proxy indicators
are indirect and represent the concept. The number of woman-owned
businesses is a real indicator of gender equity in a community. Proportion of
women in the Chamber of Commerce is a proxy indicator for the same
concept. Proxy indicators are used when a direct measure is unavailable.
Both kinds of indicators, those for outputs and those for outcomes,
provide confirming or disconfirming information about progress toward
impact. In this text, process indicator refers to those indicators selected to
gauge progress against the outputs. The process indicators are the evidence
you will collect to show what you “did.” We use the term outcome indicator
to distinguish those indicators of progress toward results (may include
outcomes and impact). The outcome indicators are the evidence that you will
collect to show what you “got.”
For example, in a model about mine safety, you would need indicators of
your efforts to achieve mine safety (“do,” the process) and indicators that
safety has been achieved (“get,” the outcome). You might use a live (or dead)
canary as an indicator of air quality (one of the many outputs needed to
achieve mine safety). Here, the canary in a cage would be a process indicator.
Alternatively, if we are focusing on mine safety as an outcome, accident
reduction could be among the many outcome indicators selected. Similarly, if
great hitters are important in winning baseball games, then batting averages
are an output. Here, things like batting averages and type of hits would be
process indicators. Games won would be an outcome indicator.
There is quite a bit of variability in the level of detail and complexity of
the concepts reflected in output and outcome statements. In practice, the
specification of output and outcome statements is often blurred with indicator
development. In the text that follows, we explain the concepts of process and
outcome indicators using the CLA example. We take the relatively broad
output and outcome statements shown on the CLA program logic model
(Figure 5.3) and split it into process (Figure 5.4) and outcome (Figure 5.5)
portions. In these two figures, we illustrate the first stage in developing
process and outcome indicators needed to inform evaluation design.
To move the logic model from illustrating program design to serving as
the framework for evaluation, the outputs need further specification to create
the indicators of whether the activities occurred as intended. For a program to
achieve its intended results, it is important to have information about both the
quantity and quality of the activities as well as the availability of resources to
support the work. This is important because the concept of “dose” has a
direct influence on effectiveness and your ability to improve your programs,
if you think of your program as a treatment or intervention, much like a
vaccination might be. How much of your program is actually delivered, who
and how many participate, over what time, and how “good” each activity is
all play a role in whether a program makes progress toward its intended
outcomes and impact.

Figure 5.3 Community Leadership Academy Program Logic Model

With Key Implementation Questions

This information, if collected, can be used to monitor the program to

ensure appropriate dose and to explain why progress toward desired results is
or is not made. Outputs are concrete aspects of your program that you can
adjust, as needed, to amplify your progress. The FIT acronym introduced
earlier applies here as well. Establishing process indicators of the frequency
(how often), intensity (quality and duration), and targets (with whom) for
your activities can go a long way to giving you the leverage you need to
boost program performance. In the case of the policy example we used
earlier, it is easy to see that a brochure left in the state house lunchroom is
less likely to produce results than is a series of lunchtime conversations, one
on one with the key house leaders. The brochure is a relatively weak dose of
policy-maker education (i.e., single shot, indirect delivery, no follow-up)
compared to a more comprehensive relationship-building strategy.
When we apply these concepts to the activities in the CLA example, it is
important to specify further the expected outputs. The following formative
questions, applied to Figure 5.3, may prompt your thinking about this model.
They can be tailored and used with groups to explore the process indicators
for models you develop:

What outputs would you expect or need to see from the curriculum?
What outputs would you expect or need to see from experiences?
What outputs must occur to support subsequent outcomes?

Figure 5.4 shows the range of process indicators the CLA evaluation
identified as measures of the output or “dose” of the CLA curriculum and
experience. Notice that they specify the quality of curriculum and
experiences in addition to listing the typical participant counts and
satisfaction. Logic models used for evaluation typically display much more
detailed information than those used during program design. Based on your
thoughts about the questions above, what might be missing from this set of
process indicators? What questions about implementation dose or fidelity
might the CLA program not be able to address?
Notice that some of the process indicators are more specific than others.
If we were going to continue to develop a full set of metrics for this model,
the more complex indicators such as “instructional delivery quality” would
need to be parsed into smaller, more measureable pieces. Indicators like
“number and type of curriculum units” although more specific, would need
instructions on how exactly this would be measured. Typically, for
measurement purposes, you want your indicators to reflect a single concept
and not be multidimensional. However, this is beyond the scope of this text.
Recall that outcomes reflect the majority of the “getting” side of the logic
model. Outcomes are also time sensitive. They occur in a typically fairly
ordered sequence. This sequence or outcome chain illustrates the likely steps
between “do” and “get.” How tight or loose the order is will depend on the
type of program being modeled. Sometimes the model might or might not
show the specific connections from a given activity to each particular
outcome. Some programs lend themselves to the description of distinct
pathways from activities to outcomes, while others are more holistic and
show all activities leading to all outcomes. The degree to which
interdependencies are strictly defined and clear entry points are
predetermined can vary considerably. Most models represent a cluster of
outcomes that occur at a number of levels (individuals, organizations,
systems) from a combination of efforts. In any case, short-, intermediate-, and
long-term outcomes inform evaluation design because they indicate when and
where to look for evidence. This is particularly true when the program is very
complex. What is likely to happen first, and then what? Sometimes the
outcomes are sufficiently specified in the program logic model to guide
measurement, and other times the model needs to be adapted to serve
evaluation design.

Figure 5.4 Community Leadership Academy Process Indicators

 Developing evidence that would support your claims that an outcome has
been achieved is among perhaps the most important steps in teasing out the
specifics of your evaluation design. In individuals, changes in awareness,
knowledge, skills, and behavior are a common, expected sequence of
outcomes. In organizations or systems, the outcome sequence might include
changes in context (the environment or external influences on the program),
conditions (policies that govern practice, communication, and networks that
spread it), and/or culture (relationships, norms, and values). In some
organizations, however, like those in public health, system and organizational
changes are necessary precursors to change in behavior at the population
level. It is important to understand and display these distinctions when
developing models.
Turning our attention back to the CLA example, Figure 5.5 shows the set
of initial outcome indicators identified for the CLA evaluation. Notice that
the evaluation design does not state the community development impact with
any specificity. The evaluation focused on short- and intermediate-term
outcomes, in that these outcomes were more closely connected to the
program itself as a training initiative as well as the limited budget and time
frame available to conduct the evaluation. The CLA evaluation did not intend
to test the theory of change beyond the contribution of its curriculum and
experiences to more and better leaders. This level of specificity was sufficient
to garner agreement from evaluation consumers on the broad categories of
outcomes the evaluation would address. In addition, it is important to note
how these outcomes meet the SMART criteria for outcomes described in
earlier chapters. They are specific in that they name what will and can be
measured. They are action oriented and realistic by pointing to attributes that
demonstrate accomplishments that could occur. They are timed in that they
show the order in which they are likely to occur.
Notice in Figure 5.5 that the outcome indicators are quite broad and
clearly multidimensional. This is acceptable at the point in the evaluation
process at which decisions about what to evaluate are being made. To move
deeper into measurement, as was the case with the output indicators, more
specificity would be needed. For example, the “new leadership attitudes,
knowledge, skills, and behaviors” indicator is huge—four concepts, all of
which have many dimensions of interest. Taking “leadership attitudes,” you
might build on the knowledge that the CLA was based on the work of Robert
Greenleaf. You would be developing more detailed indicators that captured
the key points of his Servant Leadership approach such as the desire to
develop others and the importance of community.
In the overview CLA example (Figure 5.3), the model does not include
data collection methods, although logic models tailored for evaluation design
sometimes do. The questions stimulate deeper conversation about evaluation
design. In particular, the questions lead to discussion of possible data
collection, analysis, and reporting issues—this is where most evaluation
consumers can add the most value. Coverage of technical evaluation issues
beyond the development of basic design questions and indicators is outside
the scope of this text.
Data collection methods also influence costs, rigor, and limitations. So be
aware that it is important to take evaluation budget and capacity into account
when posing the evaluation questions. In moving from the questions to
specifying the methods, it becomes obvious, for most programs, that multiple
sources and approaches will be needed to secure information. A short list of
the most common data collection methods includes document review,
surveys, interviews, focus groups, and observation. Specifying methods in the
model helps reinforce the integrated information needs your data collection
tools will need to serve. This gives evaluation consumers an opportunity to
consider the “burden” they are willing to have their program bear to support
the planned evaluation (cost and their time). It also provides insight to the
evaluation team on how best to approach data collection in the context of the
program and its participants.

Figure 5.5 Community Leadership Academy Outcome Indicators

The construction of tools (e.g., survey questions, interview protocols) and

data analysis can require special expertise. It is important to realize that the
placement of questions on the model implies content they will address. In that
way, there can be agreement on the outline of what the evaluation will cover
and how it will accomplish its purpose. The specific details of measurement
and analysis that follow evaluation design are beyond the expertise of most
consumers. Thus, the use of logic models to inform these later steps is
beyond the scope of this chapter. Readers interested in more detail on
evaluation practice (e.g., theory, planning, implementation, and reporting) are
referred to the Supplemental Readings list provided at the end of this chapter.

Indicators and Alignment

The CLA model shows what sequence of changes in awareness, knowledge,
skills, and behaviors might need to occur to secure the intended impact at
some point in the future. It draws on a frequently utilized feature of logic
models. Program design and planning with models use a left-to-right, if–then
logic; this allows the developers to prospectively examine the pathways
whereby their efforts will achieve success. Evaluation, however, can be
diagnostic and more retrospective. This is particularly true when evaluation is
conducted for the purpose of program improvement. Here, analytic and
reflective thinking processes are used to examine and reconstruct the chain of
evidence from right to left. It tests the preceding content relative to embedded
and named expectations.
There is an additional value to using logic models in evaluation. If you
can demonstrate successful achievement of short-term outcomes, you can
then use the “logic” described in the model to reasonably assert that your
program could make a contribution to outcomes and impact that take longer
or that are too fuzzy to evaluate on a tight budget. What was chosen for
display and included in the CLA evaluation telegraphs volumes about some
of the assumptions underlying the program design and its evaluation. How
might the emphases chosen limit the use of evaluation information to
improve the program?
Picking appropriate indicators is important. Selecting indicators has
meaning relative to the output or outcome each reflects. For example, weight
loss could be a process or outcome indicator of improved health—if obesity
is a health challenge. In other circumstances, weight loss may actually be an
indicator of health concerns. If student achievement is the outcome sought by
a school district, then increased enrollment may not be a good indicator
(process or outcome). It could suggest something about the district’s financial
health, but it may not be the best indicator of student achievement.
 The CLA example demonstrates alignment of indicators in that it includes
a relatively robust set of process (output) indicators, fairly comprehensive
short-term indicators, and a few indicators of outcomes farther out but on the
path to community development. Notice that Figure 5.4 (process indicators)
and Figure 5.5 (outcome indicators) look different from the logic model
shown in Figure 5.3. These two figures illustrate the areas that the evaluation
would focus on, not the universe of all possible indicators. This is a display
of the information that stakeholders agreed would suffice as evidence of their
programs’ performance. Typically, as mentioned earlier in this chapter, in
evaluation, another level of detail would show the specific measures that
would be used to unpack each indicator. Many times, a table or matrix is used
to better manage display because complexity and level of detail can quickly
escalate at this stage.
Sometimes indicators are selected and used to communicate progress, but
they are not directly or completely, as in the case of the CLA example,
connected with intended impact. These “Trojan horses” can be useful for
positioning or marketing, but they can confuse authentic evaluation. It is
important to make a critical review of the alignment (direct connection, or
link) between a selected process or outcome indicator and the path you assert
leads to eventual impact.

Results Require Choices

Performance Standards
If expectations (or standards) for performance have been cited, then outputs
are an easy place to look for both fidelity (how close to plan) and level (dose)
of performance. Sometimes expectations are more detailed and qualified.
These are called performance standards. Securing better health may require a
particular quantity and quality of exercise. The number of hours and type of
exercise can be recorded for any given participant. In mature fields, like
education and health, we have considerable knowledge about what works
under what conditions. Sometimes our knowledge is precise enough that
performance standards have been established. As work is planned and
evaluated, standards can be helpful in the pursuit of desired results. The CLA
example did not set performance standards initially, but once the evaluation
design was complete and data were collected, the group would have the
information needed to set expectations for the next round of evaluation.
In the CLA example, new or improved skills among participants are
indicators of progress toward outcomes. They are one choice on which to
focus inquiry. This deliberate choice about focus can occur because the
program is displayed graphically. It is easier to see and choose among areas
that have explanatory potential when they are named and displayed in a
model (instead of narrative). Evaluation could determine whether or not
individuals gained new skills.
At any point of time during the program implementation, inquiry could
yield many possibilities. Perhaps, in the case of the CLA evaluation, one
discovers no new skills were learned or the skills learned weren’t relevant to
community development. Maybe skill development for individuals happened
but the individuals were never engaged in any community projects. Each of
these findings would have implications for program improvement.
Alternatively, evaluation could look at curriculum content or even at the list
of inputs: participants, faculty, marketing, or other areas. To manage cost and
effort in evaluation, choices must be made about where to focus the inquiry.

Quality Evaluation Designs

We believe a quality evaluation design should respond clearly to evaluation
consumers and their information needs. In other words, the questions named
have utility. As you consider evaluation through the architecture of a logic
model, the following questions reflect quality considerations. With
colleagues, you might develop additional questions relevant to your
workplace, project processes, or content. It is important in evaluation to recall
that any model is only one way to represent work. It is a view. During
evaluation design, it is good to be aware of the many options vying for
attention. The same hazards (blind spots, myths, and cultural concerns) that
influence circumstances during program design and planning may also
present themselves in evaluation. Think about ways to conduct discovery that
will identify and resolve these hazards. Multiple models or modeling in
service to evaluation provide opportunities to make choices and focus
inquiry.
 QUALITY QUESTIONS FOR EVALUATION
DESIGN

1. Are users of the evaluation specified and ranked relative to priority?

2. Have the information needs of users been specified? What assures
use?
3. Will the evaluation improve the work, determine its results, or both?
4. Will the evaluation’s key questions meet information needs?
5. Have process indicators been specified and selected for the
evaluation?
6. Are there any performance standards established for programs of this
type that can assist evaluation?
7. Have outcome indicators been specified and selected for the
evaluation?
8. Are the selected outcomes reasonable to expect relative to time and
other resources invested in the effort to date?
9. Does the model inform data collection methods?
10. Are stakeholders engaged adequately in the evaluation process
(participating when and if appropriate during design, implementation,
interpretation) to encourage their use of the findings?

A Quality Framework
Figure 5.6 shows a framework for program and evaluation quality. It
assembles the key points from the book’s first five chapters. Previously, we
described two important standards for model quality: plausibility (theory of
change and “could it work”) and feasibility (program logic and “will it work
under your specific conditions”). The quality characteristics for theory of
change models are noted (as in Chapter 2) where the focus is on the
relationship between strategies and results.

Figure 5.6 Factors and Features of Quality for Program and

 Evaluation Design

The quality characteristics for program logic models focus on the strength
of the relationship between activities and outcomes. They employ FIT
(frequency, intensity, and targets) and SMART (specific, measurable, action
oriented, realistic, and timed) principles (see Chapter 4). We suggest that
logic models are extremely valuable for evaluation design. This means the
process of modeling surfaces the most important information needs of
identified users. Logic models can support and assure that information
gathered is used in the pursuit of performance management and greater
effectiveness. We think a program, project, or organization is more likely to
achieve impact if relative theory of change models are plausible, program
logic models are feasible, and the evaluation models that test the underlying
assumptions of each are designed for practical use. Similarly, the ideas
presented in this chapter could easily be applied in a research design setting
—particularly in problem identification and in posing the research questions
or hypotheses. Evaluation and research both are inquiry and/or problem
solving in much the same way.

IN SUMMARY

In the first half of this book, we posited three questions about effectiveness:

Are you doing the right work?

 Can you make better decisions?
Are you getting superior results?

All of these questions, including the third one, require some evaluation
literacy. This chapter describes the evaluative thinking and processes logic
models can support when effectiveness is given deliberate attention during
evaluation. We hope readers will use logic models to contribute to the design
of evaluations that will answer these vital questions. They are significantly
different from “Are we busy?” These questions focus attention on
effectiveness rather than on efficiency or the accomplishment of a laundry list
of activities.
Both formative (improve) and summative (prove) evaluations are useful
for many reasons. Both of these approaches can help build understanding
about what works under what conditions. Because evaluation is a key
function in managing for results, this chapter explains how logic models can
assist evaluation design directed toward that end. Models help with decisions
about the most relevant information and its use. Identifying and choosing
among information needs and users focuses evaluation resources where they
are most needed to influence effectiveness. These steps are crucial in creating
a useful evaluation. Program evaluation and planning are “bookends” that
reflect the same thinking and thus share a common theory of change and very
similar program logic model views. Specifically, outputs and outcomes can
be very helpful gauges for monitoring and improving the status of your work.

LEARNING RESOURCES

Reflection

1. What are the strengths and limitations for evaluation when the logic
modeling process has already occurred during program development?
What about when it occurs after the program is under way?
2. What are the various ways that a theory of change and/or logic model
can be used to inform the development of an evaluation design?
3. How might the information needs of funders, grantees, evaluators, and
 participants be different?
4. What relationships exist among evaluation, logic models, performance
management, and effectiveness?

Exercises
1. Based on the program, project, or idea you mapped out in Chapter 4,
design the key questions and indicators for its evaluation.
2. Using the health improvement example in Figure 3.4, display your
version of key evaluation questions. Cite some process and outcome
indicators. Compare your approach to that of your colleagues.
3. If the evaluation for the CLA (see Figures 5.4 and 5.5) focuses on two
strategies and the impact, what items are completely overlooked and
could yield some important information?

References and Supplemental Readings

Texts
Adair, J. (2010). Decision making and problem solving strategies: Learn key
problem solving strategies, sharpen your creative thinking skills, make
effective decisions. Philadelphia: Kogan Page.
Argyris, C. (1993). Knowledge for action: A guide to overcoming barriers to
organizational change. San Francisco: Jossey-Bass.
Fitzpatrick, J. L., Sanders, J. R., & Worthen, B. R., (2010). Program
evaluation: Alternative approaches and practical guidelines (4th ed.).
Saddle River, NJ: Pearson.
Frechtling, J. (2007). Logic modeling methods in program evaluation. San
Francisco: Jossey-Bass.
Grantmakers for Effective Organizations. (2007). Learning for results.
Washington, DC: Author.
Greenleaf, R. K. (1977). Servant leadership: A journey into the nature of
legitimate power and greatness. New York: Paulist Press.
Hatry, H. (2007). Performance measurement: Getting results. Washington,
DC: The Urban Institute.
Kapp, S. A., & Anderson, G. R. (2010) Agency-based program evaluation:
Lessons from practice. Thousand Oaks, CA. Sage.
McDavid, J., & Hawthorn, L. (2006). Program evaluation and performance
measurement: An introduction to practice. Thousand Oaks, CA: Sage.
Patton, M. C. (2008). Utilization-focused evaluation (4th ed.). Thousand
Oaks, CA: Sage.
Ridge, J. B. (2010). Evaluation techniques for difficult to measure programs:
For education, nonprofit, grant funded, business and human service
programs. Bloomington, IN: Xlibris Corp.
Spitzer, D. R. (2007). Transforming performance measurement: Rethinking
the way we measure and drive organizational success. New York:
AMACOM.
Stufflebeam, D. L. L., & Shinkfield, A. J. (2007). Evaluation theory, models,
and applications. San Francisco: Jossey-Bass.
Wholey, J. S., Hatry, H. P., & Newcomer, K. E. (2010). Handbook of
practical program evaluation (3rd ed.). Thousand Oaks, CA: Sage.
Yarbrough, D. B., Shulha, L. M., Hopson, R. K., & Caruthers, F. A. (2011).
The program evaluation standards: A guide for evaluators and evaluation
users (3rd ed.). Thousand Oaks, CA: Sage.

Journal Articles
Adler, M. A. (2002). The utility of modeling in evaluation planning: The case
of the coordination of domestic violence services in Maryland.
Evaluation and Program Planning, 25(3), 203–213.
Bellini, S., Henry, D., & Pratt, C. (2011). From intuition to data: Using logic
models to measure professional development outcomes for educators
working with students on the autism spectrum. Teacher Education and
Special Education, 34(1), 37–51.
Carman, J. G. (2007). Evaluation practice among community-based
organizations: Research into the reality. American Journal of Evaluation,
28(1), 60–75.
Ebrahim, A. (2005). Accountability myopia: Losing sight of organizational
learning. Nonprofit and Voluntary Sector Quarterly, 31, 56–87.
Forss, K., Cracknell, B., & Samset, K. (1994). Can evaluation help an
organization to learn? Evaluation Review, 18(5), 574–591.
Hayes, H., Parchman, M. L., & Howard, R. (2011). A logic model framework
for evaluation and planning in a primary care practice-based research
network. Journal of the American Board of Family Medicine, 24(5), 576–
582.
Herranz, J. (2010). The logic model as a tool for developing a network
performance measurement system. Public Performance and Management
Review, 33(1), 56–80.
O’Keefe, C. M., & Head, R. J. (2011). Application of logic models in a large
scientific research program. Evaluation and Program Planning, 34(3),
174–184.
Zantal-Wiener, K., & Horwood, T. J. (2010). Logic modeling as a tool to
prepare to evaluate disaster and emergency preparedness, response, and
recovery in schools. New Directions for Evaluation, 126, 51–64.

Internet Resources
Administration for Children and Families. (n.d.) Evaluation toolkit and logic
model builder. Retrieved December 8, 2011, from
http://www.childwelfare.gov/preventing/evaluating/toolkit.cfm
Bureau of Justice Assistance. (n.d.). Evaluation and performance
measurement resources. Washington, DC: Author. Retrieved December
8, 2011, from https://www.bja.gov/evaluation/evaluation-resources.htm
Centers for Disease Control and Prevention. (2011). CDC’s evaluation
efforts. Retrieved December 8, 2011, from
http://www.cdc.gov/eval/index.htm
National Center for Mental Health Promotion and Youth Violence
Prevention. (2010). Safe Schools Healthy Students evaluation tool kit.
Retrieved December 8, 2011, from http://sshs.promoteprevent.org/sshs-
evaluation-toolkit/logic-model
The Pell Institute. (2011). The evaluation toolkit. Retrieved December 8,
2011, from http://tool kit.pellinstitute.org/
United Nations Population Fund. (2008). The programme managers
planning, monitoring and evaluation toolkit. Retrieved December 8,
2011, from http://www.unfpa.org/monitoring/toolkit.htm
 6

Display and Meaning

T
his chapter describes selected examples of logic model display and
the implications of choices relative to meaning and use. In brief
examples, we present models used in private and public sector
organizations. The variation in format and content is intentional.
These models, presented with some context, are provided to enrich readers’
experience and experimentation with features of display.
 LEARNER OBJECTIVES
Identify variations in model format and style
Recognize that models reflect culture and intended use
Explore what will and will not work in your organization
Explain why logic models are highly interpretive

Because logic models are socially constructed, perception, politics, and

persuasion are all substantial influences on them. As a graphic display of the
general approach to change or as a more detailed description of work, logic
models reflect intentional choices of their authors. In reality, models can be
compromised by the skills and experiences of their creators, along with the
context and purposes they serve. Our own models reflect these influences.
For example, sometimes clients do not ask for or want modeling (“improved
versions”). They simply request documentation of their current thinking.
Although Chapters 2 through 5 in this text have suggested quality features
and selected principles for creating models, the examples here vary in
adherence. As the use of logic models grows, it is possible that standards for
them will emerge and be commonly used. Please note that the models in this
chapter include model examples that have successfully served specific
purposes. We identify variation of some key graphic features preceding each
to alert the reader to options you might consider when creating models.

Variation and Learning

It is our experience that variation nearly always offers a rich field for
learning. In this chapter, the examples provide a chance to see actual models
and how they look with a range of content and functions. Most of the entries
are contributions from colleagues who frequently use logic models. One is
from our own work with clients. All are influenced by the conditions in
which they were created and the people who generated them. And, like most
models, all of them could be improved. These examples include both theory
of change and program logic models. In several illustrations, the models are
templates or umbrellas for subsequent design/planning and evaluation work.
They all serve as “organizers” for a shared understanding and a platform or
framework to advance other functions, such as communications, evaluation,
planning, and research. Prior to the entries, we identify and describe
variations of key features in the models presented.

Graphic Display
As logic models are tools that show and support critical thinking, the
selection of elements used in their display helps illustrate the subject content
in a dynamic way. Models avoid some of the interpretation that dense text
requires, but they simply are not immune to interpretation. Because logic
models convey relationships among elements, it is important to be conscious
of the use of boxes, lines, curved lines, circles, single- and double-headed
arrows, and other shapes in terms of their meaning. Further, their creation
occurs in context and has meaning for their creators, and this can vary as they
are read by others.
Models in the cases range from pictorial images with copy (Example 4) to
circular displays (Example 1) and the most common flowchart style that
employs text plus symbols and shapes that are read from left to right.
Elements of the models differ, too. Some include inputs, barriers, and
facilitators; others do not. Some use arrows, some just lines. Others use
neither of these. There is a substantial difference in comprehensiveness.
Some are general change recipes, while others offer detail adequate to operate
a program. In some cases, the models require the case narrative to understand
their content. In others, the models are quickly and completely understood
without external copy to support them. Examples 3 and 6 show both a theory
of change and a program logic model.
While the use of graphics to convey meaning can quickly become very
sophisticated, most people have had some experience with a model or
diagram that contains words and arrows. And all cultures have symbols that
convey meaning. Many people, North Americans, for example, understand
that a lightbulb means an idea, crossed swords means conflict, and linked
hands means harmony. However, these symbols are cultural and may have no
meaning or different meaning in another context.

Complexity and Meaning

The models shown here and used by practitioners worldwide differ
considerably in their complexity. Logic models are often used instead of or as
a complement to text because they can simplify and untangle the
relationships among elements. They also can be dense or very simple. Their
relative complexity is generally a reflection of the number and type of
relationships they are communicating.
The New York Healthy Weight (Example 5) Program articulates vision,
mission and assumptions. It specifies roles and names external partners. It
considers both environmental factors and personal determinants. In contrast,
the model that displays a multiyear, multistrategy preschool change effort
(Example 3) has different detail and a significant number of arrows to
communicate influences among features. The preschool model is used for
multiple functions: planning, managing, evaluation, and communication.
Examples 1 and 5 also demonstrate that models can vary in their relative
prescription, from general to very explicit, and may include target audiences.
Models are often used as either umbrellas for clusters of subprojects (see
more on this in Chapter 7) or as templates to guide alignment across large
organizations. How tight or loose they are as prescriptions will depend on
their intended use and the culture of the users.
Moreover, “reading” or interpreting the models in these cases requires
entry in different ways. Many are processed left to right, but others are top to
bottom or the inverse. Several suggest repeated activity via circular shapes or
symbols. And the most pictorial examples (2 and 4) offer several entry points
to the content displayed. There are substantial differences among the models
relative to the volume and placement of text.

Content, Uses, and Creation

The subject content for each of the models in our examples is distinct. They
include citizen scientists, a local food initiative, preschool policy change,
collaborative inquiry, weight management, and the nonprofit sector. Models
appear to work regardless of the subject matter content. Functionally, they
most often serve design/planning, evaluation, and communication. But we
have also seen models that guide research management, organization
development, learning, knowledge management, and training.
The model content is informed in various ways. All have multiple
stakeholders, and some organizations use theory of change and logic models
as standard practice. The development of most of the models was externally
supported. Often in real-world situations, the thinking behind a model and its
development is not visible to external audiences. Several cases specifically
reference theory, evidence, and literature, and they all rely on a continuum of
practice experience. While one model required engaging an artist, this is an
exception and intentional choice.
Note that Chapter 8 includes the Paint Product Stewardship Initiative
model. It offers an application with important features you could consider,
including lines, shapes, multiple font types, texture, balance, contrast, unity,
and color. It is a new generation of modeling because it provides extensive
supplemental resources that inform and extend what information the model
displays at first glance. Through hyperlinks and other software features, the
model can be accessed and used in multiple ways. The sequence of steps and
related processes with stakeholders to generate a model varies considerably.
The social construction and technical execution are very important,
ultimately, to the utility and quality of the model produced.

Model Benefits
In all of these cases, the models secure at least one important process
objective: a shared understanding of the work among stakeholders. They all
organize and display relationships among multiple features such as strategies,
activities, and results. And they all provide a common vocabulary and
framework for those involved in model creation. Some of the models support
operations and others are simply input to the creation of other models or a
framework that provides “tent stakes.” Regardless of scale (a project,
initiative, organization, or other), models can be an important anchor for
implementation, evaluation, dissemination, or other next steps because they
quickly convey the parameters and content of a bounded effort. Describing
the “it” is vital to prospective work. It serves as construct explication.
Some models describe an organization’s direct and indirect influence, and
several of the cases suggest the important implication of time as their models
parse outcomes in a sequence or the accompanying narrative references this
feature. Direct influence means that the organization can take actions that will
likely affect cited outcomes. Indirect influence is a reference to work that is
dependent on other organizations, individuals, or target markets to act in a
particular way before outcomes may occur. Time is a particularly important
feature to identify in a model and to look for when reading one. Time is not
often labeled in years but rather in generic qualifiers like “short” and
“intermediate.” These phrases can have very different meanings among
readers. Occasionally, definitive parameters for time are omitted
intentionally.

Alternative Approaches
Causal loop diagrams and logical frameworks (also known as logframes) are
two other approaches to modeling the connections between “do” and “get.”
Causal loop diagrams are used to display complex systems behaviors. They
highlight the influential forces acting on cause-and-effect relationships. They
also show patterns of how and why things change rather than a static
snapshot. They have much less text than traditional logic models and are
more schematic in appearance. They use interlocking circles, arrows, and
other symbols to display cycles. These types of models are most often used
by practitioners active in systems thinking and organizational learning.
Logical frameworks grew out of the Management by Objectives
movement in the 1970s. They are typically a four-by-three matrix. The rows
describe objectives/goals, purposes, outputs, and activities. The columns
address achievement indicators, verification means, and important
risks/assumptions. The construction process emphasizes testing the vertical
and horizontal logic. These frameworks are widely used internationally by
development agencies, nongovernmental agencies, and philanthropies.
In addition to using different elements, logical frameworks differ from
logic models in several important ways. Logic models are generative in that
they typically emphasize the desired outcomes or impact. In contrast, logical
frameworks begin with an analysis of the problem(s) and thus are a more
reactive approach. In logic models, the assumptions are propositions upon
which the strategies and clusters of activities are based. Alternatively, the
assumptions in logical frameworks are those conditions that must exist for the
program to be implemented. References for these alternative approaches are
provided at the end of this chapter.

Selected Examples
The following examples include both theory of change and program logic
models in different formats with different content and uses. We hope that
your exposure to these materials helps you to explore important choices as
you create models that are most useful to your work and stakeholders. These
interesting examples are shared to display relative diversity. Each and every
logic model is distinct—although there are some common features among
them. Most of the models and associated descriptions were contributed by
colleagues in academia, the government, and the private and nonprofit
sectors. This range provides multiple perspectives and contexts.
At the beginning of each, we suggest one way to read the model and offer
comment on selected features. In most examples, we share the model with
associated narrative (boxed copy) contributed by colleagues who were
involved in its creation and use. Last, we ask some thought-provoking
questions about the display, meaning, and use. Each example also includes
some additional resources.
All the models in the following cases are versions of an initial effort to
capture and communicate. When people read (or interpret) a model, they
should ask, “What is this telling me?” As you explore the examples, it is
valuable to consider how the context may have influenced the model. It may
also be useful to think how you and your colleagues would create models for
the purposes named. What revisions would you make and why? Small
changes, just moving a line or element to a different area in the display, can
be very significant. We encourage use of the Resources section at this
chapter’s end because it can help in using these examples for additional
learning.

Example 1: Eco Hub

This circular model (Figure 6.1), referred to by the creators as a logic map,
has two primary spheres of tasks that feed the cycle and are labeled as
“ongoing.” They are needs awareness and evaluation. These represent upper
and lower halves and are designated by color. They are centered at the core of
the graphic. The model uses an “earth tone” palette of blue, green, brown,
and related colors to signal a reference to its subject matter, ecology and earth
science. The font labeling tasks icons is clear, strong, and easily read.
Associated icons appear with each task. They are symbolic (e.g., magnifying
glass, clasped hands) and literal (a scientist from clip art). The arrow placed
adjacent to “inputs” signals that readers can begin review or access the model
at that point. The top half of the circular model is a blue pathway and
suggests a coherent group of tasks that are part of the citizen awareness
activities. They include partner building, training, data stream, and quality
assurance. Each task is qualified as interactive. The bottom half of the circle
is labeled “evaluation” and has three sections. It includes the outputs on a
gray pathway: visualization and report cards. The next section is “outcomes,”
on a green pathway underneath the copy. Finally, third is “strategic impact,”
which lies on a brown pathway. The three sections in this lower half are
denoted by new color. Between each of the eight tasks is a heavy arrow that
pushes the reader forward on the circular path. The dependent sequence is
clear and the color use is integral to understanding how the elements of the
model are related.

Figure 6.1 Eco Hub Program Logic Model

Source: Integration and Application Network, University of Maryland Center for Environmental
Science, 2010.

The logic map describes a project designed to develop a citizen scientist

network and increase environmental literacy. Read clockwise, the reader
begins at Inputs at the top left. The inputs, activities, outputs, outcomes,
and strategic impacts are linked together in a recurrent circular set of tasks.
The inputs and activities use an ongoing needs assessment to inform these
tasks. The outputs, outcomes, and strategic impacts utilize ongoing
evaluation and enhancement to optimize effectiveness. There are eight
activities in the logic map: interactive partner building, interactive training
modules, citizen scientist data stream, interactive quality assurance,
interactive data visualization, eco-health report cards, environmental
 literacy, and evaluation and dissemination. A series of icons links the steps
(inputs, activities, outcomes, and strategic impacts) together.
The model is used on the Eco Hub website to illustrate how the project
can develop a citizen scientists network and increase environmental
literacy. It relies on experience from many other examples of projects that
have successfully taken data, synthesized the information, had strategic
impacts, and effected change. It hypothesizes that this can also be done
with citizen scientists and monitoring data. This model was created by the
Integration and Application Network, a small group within the University
of Maryland Center for Environmental Science.

Questions

1. Are the icons in this model helpful or confusing?

2. What unique messages does the circular format convey?
3. How would you draw this model with boxes and arrows?
4. How does color help show relationships (see the following link)? What
limitations are presented when shown in black and white?
5. What’s missing in this model?
6. Would this model work in your organization? Why?

References
For more information about Eco Hub and to view the logic model in full
color, see http://www.eco-hub.org/about/.
Eco Hub is a project associated with the Integration and Application Network
(IAN) at the University of Maryland Center for Environmental Science.
See http://ian.umces.edu/(retrieved January 10, 2012).

Example 2: Wayne Food Initiative

The Wayne Food Initiative (WFI) elected a distinct and evocative format to
display their important community-based work to build a local, sustainable
food system (see Figure 6.2). The deep brown trunk and roots are drawn in
several pieces that convey defined elements of their work. Central to the
trunk of the tree are group values. These cite what the Initiative considers
central: youths, farmers, health, education, food system, community, equity,
and justice. Group values (what and whom) contribute to their strategies. The
WFI strategies first “connect & strengthen,” then “assess & strategize,” and,
third, “promote & grow local.” Above the strategies are primary tree
branches that illustrate program areas with specified tactics.
The four branches, read left to right, name farmer support, food
sovereignty, public campaign, and youth programming. In bright green leaves
shooting from the branches, there is a range of activities associated with the
specified program branch. For example, on the youth programming branch,
the reader is introduced to food assessment, by-and for-youth workshops,
garden, and public art projects as well as community reads.
On the left, below a branch, the model names “outputs” via a listing near
a bit of brilliant red “fruit” from the tree. This fruit symbol appears adjacent
to the green leaf tactics across the crown of the treetop, too.
On the far right, a small seedling is associated with an “outcomes” listing
that includes increased farmer sustainability, capacity, community food
sovereignty, public food system literacy, youth leadership, and youth food
system businesses.
Below the tree trunk, at ground level, the model identifies “acquired
resources.” This is a listing of financial capital from a variety of sources.
Below it, the feature identifies “partner resources.” These include a charitable
entity, facilities, knowledge, experience, volunteers, past success, community
leadership, diversity, commitments to fair food and farmers, imagination, and
vision.
The WFI model was completed by a community collaboration focusing
on local food systems as a process for community organizing. They involved
about 30 people led by a facilitator who completed it after several months of
small group meetings.
The model is effectively read bottom to top in a cycle fashion, from roots
up the tree to the fruit, down to fallen fruit seeding new trees. The tree
elements are symbolic to allow for ease of expression. Roots are resources,
reminding us that our work comes from strengths, not merely needs. The
trunk is the collective values that form the foundation as well as core
strategies. The large branches hold objectives and leaves, activities. The fruit
is literally the fruit of WFI labor, which always has the potential to seed new
work or new directions.
The model is grounded in community knowledge of asset-based
opportunities. This tree is both an external communication tool to explain
what work the collaboration is about and an internal image reflecting the
organization of the objectives and activities.

Figure 6.2 Wayne Food Initiative Program Logic Model

Source: Wayne Food Initiative, 2007.

Questions

1. Does the tree format help or hurt the intended messages for the Wayne
Food Initiative? How would this novel approach to display be received
where you work?
2. Is it helpful to distinguish strategies from program and tactics as this
model does?
3. What advantages and disadvantages do you see in a model in this
format?
4. Could you easily use this model to inform evaluation design? What
helps and what hinders?

References
For more information about Wayne Food Initiative, see
www.waynefoods.org and/or
http://www.cefs.ncsu.edu/whatwedo/foodsystems/waynefoodinitiative.html
Both retrieved December 12, 2011. Their fascinating logic tree in full
color can be found at
http://www.waynefoods.files.wordpress.com/2008/12/wfi-plan-tree-logic-
model.jpg (retrieved December 12, 2011).

Example 3: Promoting Preschool Change

This entry includes three models: a basic theory of change, an expanded
theory of change, and a program logic model. The case is about using
advocacy to achieve policy change. The basic theory of change model for this
case relies on applied political science research. In his well-known research
on agenda setting, political scientist John Kingdon (1995) describes what it
takes to get an issue on the public policy agenda. The public policy agenda is
the list of issues or problems considered by policymakers. Getting on the
agenda or positioning an item there as a priority involves several processes:
problems, proposals, and politics. Kingdon’s theory of change model for
policy is illustrated in Figure 6.3.
 In this basic model, the stream of problems, proposals, and politics occurs
concurrently for individual policy issues. Problems are the process of
persuading policymakers to pay attention. It is “making the case” and
defining the reasons a particular issue demands action. Proposals are the
process of remedy—their generation, debate, revision, and adoption. And
politics are the many factors that influence the political context. They include
climate, culture, and advocacy for and against proposals.
In Kingdon’s research, he found the success of agenda setting was
dependent on the three processes. While these three elements operate
independently, the actors in each process can overlap or interact. When at
least two of the processes intersect at a critical time, then a policy window
occurs. These windows represent an opportunity to advance policy on a
particular topic or issue.
The expanded theory of change (Figure 6.4) reads from left to right. It
shows the connection between the three elements of Kingdon’s theory
(shaded) with the strategies and outcomes specific to the David and Lucile
Packard Foundation’s Preschool for California’s Children grantmaking
program. Packard’s Research strategy addresses the policy problem stream.
Evidence from research will be used to think about and frame the problems
that quality preschool can address. The Leadership and Engagement strategy
focuses on the identification and development of policy solutions (proposals)
and the building of key leader support for these solutions. The Flagships and
Local Strategy serve further to increase visibility and demand. These in turn
influence the public and political support for the policy solutions. The three
elements of problem, proposals, and politics together open policy windows
from which quality preschool policies can emerge. A strong reliance on
communication as a support strategy is implied in this model.

Figure 6.3 Kingdon’s Policy Stream Convergence

Source: Coffman, 2007.

Figure 6.4 Application of Kingdon’s Theory to Packard Preschool

Program

Source: Coffman, 2007.

The policy stream convergence theory represented by this expanded

model is key to reading the program logic model we describe next.
Although ambitious and specific outcomes are named in the Preschool
Logic Model (Figure 6.5) at the far right, this model is typically read from
left to right. We begin with “inputs” and the column titled Strategies and
Activities. The second column details three primary areas of work, which
include Leadership and Engagement, Research, and Flagships and Local
Strategies.
Figure 6.5 Packard Preschool Logic Model
Source: Packard Foundation, 2007.

Leadership and Engagement will cultivate a strong and diverse group of

advocates for preschool, both statewide and locally. Research will support
projects that reinforce the evidence for preschool value. Flagships and Local
Strategies will provide examples of success through support for preschool
expansion, quality improvements, and constituency building. Outcomes are
parsed by time in three stages: short, intermediate, and long term. This model
includes a specific timeline for implementation and outcome achievement. It
includes process and outcome indicators shown as bulleted lists within each
of the shaded rectangles.
 This model ranks strategies and associated outcomes by emphasis. The
legend in Figure 6.5 shows that the highest strategic emphasis is shaded the
darkest (Intermediate-Term Outcomes, Dissemination). The original model
also used colored arrows to differentiate and depict relationships connecting
strategies/activities with their specific intended outcomes over time.
Assumptions about relationships are implied by the arrows among elements.
This quantity of arrows depicts a highly interactive and integrated body of
work. Arrowheads show that the majority of the relationships are one way,
indicating progress toward outcomes moving toward the right. Four named
outcomes indicate intentions for comprehensive and specific change in the
right-most column. Double-headed arrows here (Long-Term Outcomes)
illustrate the reciprocal and likely amplifying relationship between the
connected cluster of terminal outcomes and the overarching result of quality
preschool.

The Preschool for California’s Children logic model is a graphic

representation of the David and Lucile Packard Foundation’s preschool
grantmaking program. Within the Packard Foundation, Preschool for
California’s Children is referred to as a subprogram, and it is within the
broader Children, Families and Communities Program. It depicts our core
grantmaking strategies and illustrates how we expect funded activities to
produce outputs and outcomes that build toward the ultimate goal of
voluntary, high-quality preschool for all of California’s 3- and 4-year-olds.
The original logic model was created by program staff but was later
modified in collaboration with the external evaluation design process. The
logic model flows from a theory of change advanced by political scientist
John Kingdon, who posited that policy change occurs when three catalytic
elements (problem, policy, politics) come together at the same time.
Advocates must seize upon the window of opportunities when these three
streams converge.
This program logic model translates the theory of change into a
detailed plan of action, connecting activities to outputs to short-,
intermediate-, and long-term outcomes over time. It contains three core
strategies: Leadership and Engagement, Research, and Flagships and Local
Strategies. It is expected that the combination of outcomes from all three
strategies will generate increasing progress toward outcomes. It is
important to note that the three core strategies are all interconnected. For
example, in this logic model, if one follows the flagship strategy of Quality
Preschool Expansion and Demonstration horizontally across the page, one
observes a number of significant local outcomes related to access and
quality. However, more pivotal to achieving the big win, flagships will
also contribute “vertically” to leadership and engagement goals of
recruiting champions, generating media coverage, growing public support,
and providing models for policy development.
The Preschool Logic Model is not only a useful graphic for
understanding the effort, but it is also an important planning tool for
evaluation. Evaluators use logic models to provide a blueprint for
evaluation, interpret evaluation results, and facilitate contingency
planning. Based on the preschool logic model, the Harvard Family
Research Project evaluation team identified evaluation questions,
outcomes to be examined, and indicators to measure those outcomes. The
logic model also provides insight into timing, for setting prospective
benchmarks, scheduling data collection, and subsequently gauging
whether sufficient progress has been made. Where progress isn’t
happening, the anticipated connections between the boxes in the logic
model may not form, inducing us to make midcourse correc tions or
rethink parts of our strategy and modify the logic model. The program
logic model is malleable; activities and their associated indicators may
change in response to contingencies, but the fundamental theory remains
constant. This model is a revised version.
At the Packard Foundation, both theory of change and logic models are
essential components of good programmatic strategy development and
management necessary to help us achieve greater impact in our work.
Theory of change and logic models are developed, reviewed, and/or
revised for many reasons.
For us, a theory of change serves as a tool with three purposes:
planning, communication, and monitoring. First, developing a theory of
change encourages strategic thinking, as one clarifies the connections
between desired outcomes and the strategies and activities designed to
achieve them. It illuminates underlying assumptions, explores the larger
system in which the change will take place, and brings to bear relevant
evidence or experience in support of the assumptions.
Second, by making specific the desired outcomes, strategic pathways,
 and underlying assumptions, a theory of change promotes transparency
and provides a platform for engagement with stakeholders, particularly
grantees. Grantee partners should be able to understand the theory of
change and connect their work to it. A theory of change helps to clarify
roles and expectations.
Finally, the theory of change serves as a basis for the monitoring and
evaluating approach. Since the theory of change helps to describe the
assumptions that their work is testing, it points to what needs to be
assessed in a monitoring plan.
We believe that theory of change and logic models should be
developed by engaging multiple perspectives and drawing on existing
knowledge and experience. Grantees, stakeholders, experts, and
Foundation staff can and should be engaged. In defining the landscape,
outside experts can be very helpful in developing background papers or
other inputs that can inform the development of the framework. Given
their central role in implementing the theory of change, grantees must be
engaged at some stage. The subprogram staff lead is the primary architect
of the theory of change and logic model.

Questions

1. What features do you like and dislike in this model? Why?

2. What are other ways you can imagine the content of this model
displayed?
3. Would a model like this one get use in your workplace?
4. Is the comprehensiveness of this model helpful or overwhelming? Why?
5. What key evaluation questions are inferred by the model?
6. Could you prepare a request for proposal from this model that would
assist grantmaking? Why or why not?

References
For more about the David and Lucile Packard Foundation, see
http://www.packard.org.
Coffman, J. (2007). Evaluations to watch: Evaluation based on theories of the
 policy process. Evaluation Exchange, 13(1), 6–7. Retrieved December
12, 2011, from http://www.hfrp.org/evaluation/the-evaluation-exchange/
issue-archive/advocacy-and-policy-
change/evalua tion-based-on-theories-of-the-policy-process
Kingdon, J. W. (1995). Agendas, alternatives and public policies (2nd ed.).
New York: Longman.

Example 4: Collaborative Learning,

Inquiry, and Practice
Reading this distinct illustration-type logic model (Figure 6.6) could start at
several different points of entry along a path or a roadlike arrow from left to
right. The relative size and strength of the word impact draws attention to the
far right side of the display. Intellectually, the impacts listed for
participants/members encourage the reader to search out information about
the “CLIPs” to understand how this term is defined. Information that
typically on a logic model might be labeled as inputs, activities, and
resources is placed as precursors on the path that weaves across the diagram.
Moving to the far left corner, it feels natural to explore the “key features,”
“relationship emphasis,” and “the three inquiry steps” that provide
information about the program, much like the activities column in a more
traditional model. Notice the spiral encircling the person; this likely infers the
iterative nature of the inquiry process. Dropping down the model to the lower
left corner, the “supporting structures” are defined, and these are similar to
the type of information in an input column. The relationship “guiding
principles” follow, and these describe attributes of CLIPs that could be
interpreted as shorter-term outcomes for the CLIP process that contribute to
member impact.
This model also uses a vine to organize features in relationship to each
other. Male and female figures are strategically placed with positive,
energetic postures. Signs and pages with text are repeating features that are
used instead of boxes and arrows to draw the reader’s attention from left to
right. Several sizes and styles of font are also used in what could be
interpreted as a hierarchy of importance. This program draws on a broad
literature base, grounded in evaluative and appreciative inquiry, to validate
the strength and direction of relationships among inputs, activities, outcomes,
and impact as displayed.

Communities of Learning, Inquiry, and Practice (CLIPs) are self-selected

informal, dynamic groups of community college faculty and/or staff (and
sometimes others) who conduct an inquiry about a topic they identify and
see as important. The inquiry steps are (1) design the inquiry, (2) collect
data, and (3) make meaning and shape practice. Through participation in
the CLIP, members build their capacity to collaboratively conduct such
inquiries and enhance their professional practice. A member within each
CLIP serves as its Facilitator. An overall CLIP Guide positions the CLIPs
at the college and supports the CLIP work by handling coordination,
organization, and financial matters related to CLIP work.
CLIPs provide opportunities for community college faculty and staff to
collaboratively study issues of importance to them about student learning
and success. CLIPs add vitality to institutionally determined assessment,
program review, and planning processes. The CLIP process is being
developed through an evaluative research study at Bakersfield College
conducted by InSites (a nonprofit organization) under a National Science
Foundation grant (Grant Number REC-0335581). The purpose of the grant
is to develop the CLIP process as an evaluation capacity building process.
To construct the theory of change/logic model for the CLIP process, I
drew on the data I gathered as part of the research in my role as the first
CLIP Guide. I also used data gathered by our external formative
evaluators. Using these data, I worked with a visual artist to create the
model.
The diagram is used to help people who are involved in CLIPs keep
the big picture in mind concerning their work. It also is used to inform
others of the CLIP process (through conferences, meetings, and online)
and to recruit additional CLIPs. It has been made into a large poster so it
can be displayed on walls in meetings. We have given people individual
copies to reference to keep in mind the whole process and where they are
in that process. The diagram depicts individual CLIPs within a college.
This version is the second version of the model. Initially, Guiding
Principles were not included because they had not yet been determined.
 One challenge in developing this type of diagram is finding a visual
artist to work with you to create it. Another challenge is keeping it
updated, since it is expensive to have an artist redo it. It raises the question
of how long you can use a diagram when it doesn’t quite represent the
project anymore. We found that it was still very useful and we could
describe the changes that were emerging as the research continued. The
diagram was a fine tool for working with the people who were directly
involved in the project and having them see their experiences captured
visually. It helped them put all the pieces together and share their
experience with others. In this case, I think “a picture is worth a thousand
words” was very true. The shortest description we had when we first
developed the diagram was four pages. This visual with the human figures
gave the four pages life, personalization, and meaning. Its colorfulness
also attracted people’s attention.

Questions

1. Would this style of model have appeal in your organization? Why or

why not?
2. What skills and resources would be important in creating models in this
format?
3. What significance do the vine, signs, pages, and figures have for you?
Would others share your interpretation?
4. In what other ways could you display the content of this model?
5. How does this model compare and contrast to others in this chapter?

References
Online modules are available for people to learn how to implement CLIPs on
their campus. The modules are available free through the InSites website at
www.insites.org. The modules can be downloaded and adapted to other
contexts.
Brown, J., & Isaacs, D. (2005). The world café: Shaping our futures through
conversations that matter. San Francisco: Berrett-Koehler.
Hughes, P. M. (2004). Gracious space: A practical guide for working better
together. Seattle, WA: Center for Ethical Leadership.
Maki, P. (2004). Assessing for learning: Building a sustainable commitment
across the institution. Sterling, VA: Stylus.
Mohr, B., & Watkins, J. (2002). The essentials of appreciative inquiry: A
roadmap for creating positive futures. Waltham, MA: Pegasus
Communications.
Parsons, B. (2002). Evaluative inquiry: Using evaluation to promote student
success. Thousand Oaks, CA: Corwin.

Figure 6.6 Communities of Learning, Inquiry, and Practice Logic

Model
Source: Parsons, 2006.

Example 5: New York Healthy Weight

Model
This model (Figure 6.7) is included because it has a vertical orientation that is
read from top to bottom and is an interesting contrast to those with a left-to-
right orientation or a circular form. It also includes some specific references
to both structures and targets that make it distinct and useful. Color, shapes,
and a few arrows are used to organize the content and convey messages. The
long-term impacts (at the bottom of the model) reflect a very simple
prescription that is evidence based for weight management: “increase energy
expenditure and decrease caloric intake.”
Starting at the top left of the model, notice the sections of content are
organized by Principles, Inputs, Outputs, and Outcomes. Each section
includes one or more rows that begins with a subhead associated with that
row category and appears on a black field. To the right of Principles, the
vision for the work is cited: “All New Yorkers will achieve and maintain a
healthy weight.” Moving to the right, the mission is named in narrative and
some assumptions are also cited. This row offers an inclusive, high-level
description of the planned work.
Immediately to the right of the section labeled Inputs is a column that
identifies a structure charged with implementation of the planned work, the
Overweight and Obesity Prevention Planning Partnership. Moving further to
the right in this section, primary inputs of the New York State Department of
Health (NYSDOH) are identified. They “rest” on a box labeled CDC (Centers
for Disease Control and Prevention), which also specifies inputs. Centered in
this row is a gauge or dial feature that shows layers of structure that imply a
hierarchy. At the center is a core planning team topped by work groups with
specific focus on a steering committee and, at the highest level, a leadership
group embedded in the state department of health. Feeding this central core,
on the right, are a list of external partners and some standalone events that
will be produced. This row combines a complex structure of operation with
inputs (or resources). Just below it, in the same row, the authors identify
target sectors/settings for the planned work. The arrow between all the other
content in the Inputs row indicates the inputs are aimed at these targets.
The Outputs section has a single row that includes a list of 10 strategies.
It includes feasible approaches such as “Employ social marketing techniques;
Educate providers, leaders and other decision-makers; Develop, maintain and
utilize surveillance.” The bottom section includes three rows: proximal
determinants, behavioral change, and long-term impacts. The first, proximal
determinants, is divided into two gross categories of environmental factors
and personal determinants. Both categories named here, as well as behavioral
changes, could be useful in evaluation design and planning. The behaviors
this model indicates the Partnership work aims to influence are physical
activity, television viewing, and nutrition.

Figure 6.7 New York Healthy Weight Program Logic Model

Source: NYSDOH, 2010.

Questions

1. Why would it be important to include as much content about the

 structures of operation as this model does?
2. How and when would you use this model?
3. Are the vision and mission features helpful to include? What about the
assumptions? Why?
4. Would this model be useful to action planning? Why or why not?
5. How would you draw this model with a left-to-right or other orientation?
6. Is the level of detail in this model adequate for evaluation design and
planning? Why? Why not?
7. Is the top-to-bottom orientation a good approach? Why?

References
For more information about the Overweight and Obesity Prevention Planning
Partnership in New York, see New York State Department of Health
(2010). New York state strategic plan for overweight and obesity
prevention. Albany, NY: Author. Retrieved December 12, 2011, from
http://www.health.ny.gov/prevention/
obesity/strategic_plan/docs/strate gic_plan.pdf

Example 6: Evaluation System

Development
This example includes both a theory of change and a program logic model.
The Independent Sector (IS) theory of change model is read from left to right
(Figure 6.8). IS work begins with their efforts, as an organization, to
strengthen, lead, and mobilize their members, who influence the sector and
ultimately generate two primary outcomes.
This model employs arrows and text only. It is simple and displays an
explanation of how change is expected to happen. IS influence is
characterized in two ways: direct and indirect. The most direct influence is
that which IS has on its members; this is shaded the darkest for emphasis.
The indirect influences are those mediated by IS members on the sector as
well as those mediated by the sector on communities and society. Lighter and
lighter shading is used to show the changes in influence. This distinction is
one way to communicate visually which of the outcomes and impacts are
closest to the work of the organization. Thin double-headed arrows are used
to indicate an interactive and reciprocal feedback relationship, with IS
bridging members to sector to society.
When using models to design evaluations and evaluation systems with a
relatively short time horizon, we suggest the use of direct and indirect
influences. This device helps to keep the “do” and “get” relationship most
tightly coupled. This way, the later models can focus on those outcomes with
the strongest and closest connection to those aspects with the highest strategic
priority. The concept of enabling and disabling environmental issues is
captured in the arrows labeled Facilitators and Barriers.

Figure 6.8 Independent Sector Theory of Change Model

Source: Independent Sector, 2007.

The IS program logic model (Figure 6.9) is read from left to right, and
content is grouped in three areas relative to outcomes: emerging issues
(Stronger Communities), operations (Stronger IS), and signature work
(Stronger Non-profits). The priority areas are included because they have
special significance to staff. They reflect internal action plans and
accountabilities. Outputs result from strategies in the priority areas (here, a
large number of activities are subsumed). They contribute to IS outcomes.
The strength of communities (society), nonprofit organizations (members and
the sector), and IS (the organization) are all linked to the outcomes named in
the theory of change. This model displays the work of the entire membership
organization; thus, there is a wide variety of targets for outcomes (e.g.,
members, staff, policymakers, sector influentials). This model is used for
monitoring (process side, outputs) and evaluation (outcomes). The intent is to
be explicit and to show reflection processes for staff that connect data from
each side to inform the work of the whole.

Figure 6.9 Independent Sector Program Logic Model

Source: Independent Sector, 2007.

In this model, the ellipses are critical features that convey groups, flow,
and relative (internal) value. They show the strands of work from strategies
through to desired outcomes and impact. They intersect to show interaction
and integration among program elements. The Priority Areas column is a
custom element that is important for organizing information and meaning for
those creating and using the model. The model describes “progress toward
outcomes and impact” on the far right. It does not, intentionally, define time
in months or years. Progress toward outcomes is used to indicate that a
sequence of outcomes from awareness through to action is implied. The
broad outcome statements are unpacked in detailed indicator and data
collection tables not highlighted here. Arrows are not used. This is because of
the highly interwoven nature of the organizations work across departments.
All strategies contribute to all outcomes. The outputs and outcomes shown in
this model draw on a variety of communication, policy advocacy, and
individual behavior change theories. Resources, at the far left, are
synonymous with inputs and are essential to the organization’s work.

Independent Sector (IS) is a nonprofit, nonpartisan coalition of

approximately 600 charities, foundations, and corporate philanthropy
programs that collectively represent tens of thousands of charitable groups
in every state across the nation. Its mission is to advance the common good
by leading, strengthening, and mobilizing the nonprofit community.
As part of its commitment to continuous improvement, Independent
Sector secured grant funds to design and implement an evaluation system
prototype. The models discussed in this example are part of early efforts to
develop an evaluation system that will meet selected formative and
summative information needs of the organization.
To create a shared understanding about what could be included in an
evaluation system, IS senior staff and an IS board member, with support
from Phillips Wyatt Knowlton, Inc., created both a theory of change model
and a program logic model. Both types of models were generated over
several meetings with the staff, with modifications made as a result of
subsequent discussions that clarified various aspects of the organization’s
work. Also important to the development of the models were the
organization’s strategic business plan and annual departmental work plans,
as were the experiences and perspectives of staff.
Independent Sector’s theory of change model displays its distinctive
role in strengthening, leading, and mobilizing the nonprofit community.
This role is designed to support a more effective, more accountable sector,
which in turn can support a just, inclusive civil society, a healthy
democracy, and vibrant communities. This theory of change model
recognizes the direct influence of Independent Sector on its members and
the nonprofit community as a whole and, by working through these
organizations, the indirect influence IS has on the sector and society.
Defining the ways in which IS can have influence helps to determine
where to focus evaluation efforts so that the organization is assessing the
areas with greater potential for direct influence. The model also recognizes
the dynamic external environment in which IS operates, including the
issues that facilitate the organization’s efforts and the barriers that it faces.
IS’s program logic model includes the organization’s “priority areas”
that reflect the focus of staff efforts. Much of the work of IS takes place
across departments through integrated strategies and approaches. While
logic models generally offer a sequence of short-, intermediate-, and long-
term goals, the breadth and depth of IS work occurs on a far greater scale
than can be captured with the limited resources available for this project.
Rather than building the evaluation system around long-term goals, the
logic model and the evaluation process show outcomes as indicators of
progress rather than on a set timeframe. Using this model, senior staff were
able to focus the collection of evaluation data on selected outcomes. The
priority areas that were selected for the initial focus of the pilot evaluation
system were its work on public policy on behalf of the nonprofit
community; providing leadership on ethics, accountability, and
effectiveness; and building a strong membership base.
Prior to this project, Independent Sector had not formally used logic
models, although several staff members were aware of them in other
contexts. Staff found the graphic display to be useful in articulating the
way the organization works, its goals, and intended outcomes. The models
were also instrumental in underscoring the integrated nature of strategies,
activities, and tasks among IS departments. Several strand models,
displaying greater detail about a specific area (e.g., policy), were created to
support inquiry and dialogue around monitoring.
Creating and using the models has contributed to building evaluation
capacity with the staff as versions were built and combinations of elements
were assembled and recast. The logic model provided a shared
understanding of the organization’s intended outcomes and how it works
toward those outcomes. This, in turn, enabled external facilitators to
explore and design an evaluation system that could ensure that the
information gathered would be of use to IS management.

Questions
 1. Is the level of detail in the theory of change model adequate to explain
how change is expected to happen? Why or why not?
2. How would you draw a model representing the IS theory of change?
3. Are the relationships between the theory of change and program logic
model evident? Why or why not?
4. What are the advantages and disadvantages of not specifying time in the
program logic model outcomes?
5. Could you build action or project management plans from this program
model? Why or why not?
6. Is there enough information to generate evaluation questions from the
theory of change or program model?
7. Are the ellipses adequate in organizing the content, left to right, or is
more detail about relationships between activities and outcomes
necessary? Why or why not?

References
Creation of this model was led by Phillips Wyatt Knowlton, Inc. For more
information about Independent Sector, see
http://www.independentsector.org.

IN SUMMARY

Logic models describe and reflect thinking about programs. They are a
display of information and the relationships among elements that depends
largely on graphic presentation. In practice, logic models address a vast range
of content areas and formats. Some are simple and others are complex, even
dense. They are influenced by who creates them, their relative experience and
skills, culture, and intended use. Sometimes models are used as templates to
align and organize related work. The choices of elements used in a model are
significant in their interpretation. Often, models are read left to right. Circular
displays, top-to-bottom, and other orientations are increasingly common. This
chapter offers examples of real-use models with considerable variation.
 LEARNING RESOURCES

Reflection

1. Is there consistent use of symbols and shapes in the case models? How
do you ensure models are “read” or interpreted with the same meaning
by everyone?
2. Does your field or workplace have technical or cultural standards for
communicating that might influence your models?
3. What do the examples suggest about how models can be used to transfer
and diffuse ideas? What challenges would an organization face using
logic models as a communications tool? What benefits seem evident?
4. What do the cases suggest about the use of logic models in the context
of measurement? How can models support measurement and
evaluation?
5. Which applications are most like and most different from your current
use of models? How? Why?
6. What level of detail is most useful in a given model? Why?
7. How does color or lack of it affect the models? What about font type,
arrows, shapes, columns, rows, texture, icons, and other features?

Exercises
1. Select a case and conduct a mark up (see Chapter 4). What changes
would you make? Why? Compare the model you create with versions
created by colleagues. Discuss your differences. Which model do you
think is the best and why?
2. Divide the cases in this chapter among your colleagues and contribute
your analysis to the matrix below:
 Once this matrix is completed, discuss the variation among models.
Which feature choices might work best under what conditions?
3. Select a theory of change model from the cases and apply the
suggestions we offer in Chapter 2. How would the model change?
4. Select a program logic model from the cases and apply the modeling
suggestions we offer in Chapter 4. How would the model change?
5. With your colleagues, list the stakeholders in any case you choose.
Then, independently, cite with whom and what action steps you would
use to generate a program logic model. Compare and contrast your list
of stakeholders and sequence of steps with others. What rationales are
used to explain differences?
6. Choose any of the models shown here and draw a new version of it with
different format and features.

References and Supplemental Readings

Texts
Buzan, T. (2002). How to mind map: The thinking tool that will change your
life. New York: Thorsons/HarperCollins.
Craig, M. (2003). Thinking visually: Business applications of 14 core
diagrams. New York: Continuum.
Duarte, N. (2008). Slide:ology: The art and science of creating great
presentations. Sebastopol, CA: O’Reilly Media.
Horn, R. (1999). Visual language: Global communication for the 21st
century. Bainbridge Island, WA: MacroVU, Inc.
Lohr, L. L. (2003). Creating graphics for learning and performance: Lessons
in visual literacy. Upper Saddle River, NJ: Pearson Education.
Margulies, N. (2005). Visual thinking: Tools for mapping your ideas.
Williston, VT: Crown House.
Nast, J. (2006). Idea mapping: How to access your hidden brain power. New
York: John Wiley.
Neumeier, M. (2006). Zag. The number one strategy of high-performance
brands. Berkeley, CA: Peachpit Press, New Riders. Arnheim.
Pink, D. H. (2006). A whole new mind: Why right-brainers will rule the
future. New York: Riverhead Trade/Penguin.
Racine, N. J. (2002). Visual communication: Understanding maps, charts,
diagrams and schematics. New York: Learning Express.
Reynolds, G. (2012). Presentation Zen: Simple ideas on presentation design
& delivery (2nd ed.). Berkeley, CA: Peachpit Press, New Riders.
Arnheim.
Roam, D. (2008). The back of the napkin: Solving problems and selling ideas
with pictures. New York: Portfolio/Penguin.
Roam, D. (2011). Blah, blah, blah: What to do when words don’t work. New
York: Portfolio/Penguin.

Internet Resources
In addition to the other modeling resources cited in Chapters 1 through 5, see
the following:
Logical Frameworks (Logframes)
ACP-EU Technical Centre. (n.d.). Smart tool kit for evaluating information
projects, products and services. Waginingen, Netherlands: CTA.
Retrieved December 10, 2011, from
http://www.smarttoolkit.net/node/376
Asian Development Bank. (1998). Using the logical framework for sector
analysis and project design: A user’s guide. Mandaluyong City,
Philippines: Asian Author. Retrieved December 11, 2011, from
http://www.adb.org/Documents/
Guidelines/Logical_Framework/default.asp
Food and Agriculture Organization of the United Nations. (1999). Manual on
logframes within the CGIAR system. Retrieved December 10, 2011, from
http://www.fao.org/Wairdocs/TAC/
X5747E/x5747e00.htm#Contents
International Labour Office. (2006). ILO technical cooperation manual:
Development cooperation. Version 1. Geneva: Author. Retrieved
November 28, 2011, from http://www.ifad.org/evaluation/
guide/annexb/b.htm#b_1

Causal Loop Diagrams

Maani, K. E., & Cavan, R. Y. (2002). Systems thinking and modeling:
Understanding change and complexity. Auckland, NZ: Prentice
Hall/Pearson Education.
Pegasus Communications. (2004). Causal loop diagrams. Retrieved
December 10, 2011, from http://www.pegasuscom.com/cld.html
Williams, R., & Imam, I. (2007). Systems concepts in evaluation: An expert
anthology. Point Reyes, CA: EdgePress.

Mind Mapping & Visualization

Consultative Group on International Agricultural Research. (2011). ICT-KM
project: Knowledge sharing toolkit (includes a mindmapping wiki).
Retrieved December 12, 2011, from http://www.kstoolkit.org/home
Delightability. (2011). Big idea tool kit: 7 visual thinking tools for innovative
teams to do their best work. Retrieved December 12, 2011, from
http://www.thebigideatoolkit.com/?p=1075 Prezi. (2011). Mind mapping
& presentation software. Retrieved December 12, 2011, from
http://prezi.com/

Note
1. The David and Lucile Packard Foundation is a tax-exempt charitable
organization qualified under section 501(c)(3) and classified as a private
foundation under section 509(a) of the Internal Revenue Code. Packard
Foundation funds may be used to support some, but not all, of the activities
of grantees and others described in this logic model. No Packard Foundation
funds are used to support or oppose any candidate for election to public
office. No Packard Foundation funds are “earmarked” or designated to be
used for lobbying or “attempts to influence legislation” (as defined in section
4945(d)(1) of the Internal Revenue Code).
 7

Exploring Archetypes

T
his chapter suggests readers consider the potent value archetypes can
give to their own models. We understand archetypes as a tested,
general template for an intervention, program, or strategy. They are
generic versions that can advance your own models. Often, with
modification, they can inform your planning, evaluation, communication, or
other needs. Archetypes can also provoke new thinking and provide a quality
check that improves ideas.
 LEARNER OBJECTIVES
Describe the rationale for evidence-based models
Define a logic model archetype
Specify contributions an archetype can make to modeling
Name the limitations of archetypes

Why squander the knowledge we have about what works? Many

important services, products, and programs have been built on the good
efforts of others. For example, while automakers may change body styles
year to year, they repeat great headlight designs that are cost efficient,
aesthetic, and effective. Software programmers do this, too. Once a particular
code path is created that works well, it is often repeated as part of a
subsequent routine. These examples demonstrate good use of prior
knowledge with a highly positive effort-to-value (efficiency) and -impact
(effectiveness) ratio. Theory of change and program logic models can garner
some of these benefits by using archetypes. They offer a substantive contrast
to trial and error. Because many archetypes are evidence based and have been
tested, they can help jumpstart your modeling.

The Blank Page Challenge

As you start to think about how your planned work and intended results
might look on paper, a blank page sometimes feels like a steep challenge. In
many cases, there is no need to start with a blank page. Archetypes are a great
remedy for “model block” or “display paralysis.” In addition to getting some
shapes and words on paper, they can also contribute significantly to model
quality because they are more likely to secure intended results.
We define archetypes as commonly used templates that offer simple
evidence-based guides for action. An archetype often looks and feels just like
any theory of change or a program logic model. The qualification is that all or
some of the elements, relationships, and outcomes specified in the archetype
are tested and proven. They are grounded in research and/or evaluation that
specifically confirm the validity of the connections as drawn. We touched on
the concept behind archetypes in earlier chapters when we emphasized the
importance of grounding your models in evidence. In this chapter, we take
the next step by sharing some examples.
It is very likely that models relevant to your needs are already drawn and
available on the Internet, in books, or in journal articles. You might pick
simple archetypes from several bodies of research and combine them to
create an initial version of a theory of change or program logic model. Or you
might find a model and use it in its entirety with only some minor changes.
Archetypes can reflect any number of broad strategies-to-results pathways
and illustrate the detailed connections between activities and outcomes. They
can display common program strategies like collaboration, communication,
advocacy, professional development, sustainability, and a whole range of
other topics. They can be general or discipline specific (e.g., health,
education, public administration, environment).
For example, some of the models in Chapter 6, although the narrative
focuses on display, are built on archetypes. The Packard Preschool models
rely on evidence for their program efforts. Rather than invent a policy change
initiative totally from scratch, the Packard Foundation used Kingdon’s policy
stream theory and the body of research behind it as a theory of change to
guide their program design and its evaluation. Similarly, the New York
Healthy Weight model relies on evidence that informs both nutrition and
exercise as proven weight-management interventions. Although in this
chapter we encourage readers to build on the evidence-based models already
available, you may find the need to delve deeper into the literature to identify
the relevant evidence base for models you generate.
Archetypes can provide a framework for content that is revised to suit
your distinct context or used “as is” because they are already diagrams that
illustrate strategies likely to work. The features of a logic model archetype
vary just as theory of change and program models do (see Chapter 1).
Evaluation logic models can also be archetypes. These models typically
provide specific measurement guidance. They specify outcomes as well as
indicators and, often, data collection tools that offer optimal points for
evaluative inquiry, whether formative or summative.

Archetypes and Learning

In Chapter 6, we offered diversity in display as a rich field for learning. This
chapter suggests archetypes have considerable value, too. An archetype might
be thought of as a generic formula or as a recipe. For instance, a tried-and-
true recipe for banana bread passed on through generations always delivers
great-tasting banana bread. Carefully following the inputs and directions of a
particular recipe that is well proven or improved through frequent use can
provide sure results. Archetypes are like recipes in that they rely on a
measurable, proven set of ingredients. Execution of the recipe means results
are predictable. Someone just learning to cook might choose to follow a
recipe exactly but later might experiment more with the ingredients or the
sequence of steps.
As a jumpstart for your work, archetypes fill a blank page and identify the
elements and conditions that might be modified. Building on the banana
bread example, as most bakers know, recipes for banana bread can vary
slightly depending on both conditions and preferences. Any given recipe for
banana bread may require a different temperature depending on the pan size
or material or oven type. And, if you like nuts or chocolate, those additions
might be positive innovations, too. Sometimes dietary restrictions or allergies
require substitutions. While it is possible to have minor distinctions reflecting
skills or preferences, a recipe identifies inputs and activities to secure the
intended result. Likewise, it is possible to use experience and evidence to
improve a recipe. These variations in the recipe are still results focused; they
just respond to different palates. In real-world execution, a program or
change-effort recipe can vary because of many internal or external conditions
(e.g., budget constraints, culture, policy, skills, and staff).

Recipes for Change

Earlier, we described knowledge and assumptions as important quality
features for logic models. Similarly, archetypes rely on theory first and are
then substantiated by research and evaluation. They use evidence as the basis
for their recipe. Like reliable recipes, archetypes can support replication and,
in that sense, they can be used prescriptively. Alternatively, they can also
support innovation by providing a grounded but initial platform to generate
new ideas. And archetypes offer greater chances of success under a variety of
conditions because of this reliability. Building on prior evidence-based
examples with carefully selected adaptations or revisions is a smart way to
tackle modeling.
One area where archetypes could be particularly helpful is in planning for
and evaluating the vast number of programs and efforts that target the actions
of individuals and organizations. For example, there are many theories of
how and why individuals change their behavior that program staff and
evaluators often use (e.g., Stretcher and Rosenstock’s health belief model,
Itzak’s theory of planned behavior, or Prochaska and DiClemente’s stages
and processes of self-change; see Supplemental Readings at the end of the
chapter). Although these theories are not described in detail here, they may be
helpful to your work and do inform some of the examples we have selected
for this chapter and those in Chapter 6. They are just a few pieces of relevant
research that can be useful in identifying, creating, or using archetypes aimed
at planning for or evaluating behavior change.
We often use the early work of Kay Rockwell and Claude Bennett as a
starting point to build models that include individual behavior change as
outcomes. Their approach, now widely known as TOP (targeting outcomes of
programs), focuses on describing the sequence of outcomes in planning,
implementing, and evaluating programs. Notably, TOP models generally
show how individual change occurs. Although this work began more than 30
years ago, it endures as highly relevant to the newer practice of logic
modeling in that it articulated specific steps in change that help us focus on
those aspects of a program most closely related to effectiveness—whether the
change effort is among individuals, groups, or organizations. We see threads
of Rockwell and Bennett’s work in many of the models used as examples
here and those we see in our practice. It was using their simple recipe to
inform our models that gave rise to our thinking that similar archetypes could
be found or created for other content. Theoretical concepts, determined and
tested through research, can inform the content in your models.

Value of Archetypes
Archetypes, just like recipes, are important because if the same model is
repeatedly implemented, it can be used as a platform to inform learning about
how to improve implementation and results. This means we can work toward
precision so that when replication of results is sought, it is a real possibility.
Further, we can also “stand on the shoulders” of the good work done before
us and have it inform where we might improve a process or result. In effect,
this serves the development of knowledge. It advances our understanding of
what works under what conditions. Several mature fields, specifically health
and education, have archetypes that practitioners rely on because of their
proven, well-established content. It is more likely, in some situations, to get
the results sought by using an archetype already developed and tested in
contrast to starting from scratch.
In general, the archetype examples selected for this chapter serve either
individual/group change or communities and systems change. Archetypes can
contribute to both program planning and evaluation as they generate new
learning about intentional variations in their content or execution. When
program efforts require shared elements or evaluation needs to aggregate
impact, archetypes can provide an umbrella or framework for design. Often,
evaluation archetypes are linked to valid and reliable measures.

More Critical Thinking

Archetypes can provide substantial benefits because they are founded on
evidence and prior research. They can launch your own models by
capitalizing on prior testing and experience. Even so, archetypes are no
substitute for critical thinking that generates appropriate revisions. Critical
thinking is a key contributor to logic models because it employs intellectual
criteria like clarity, credibility, relevance, and significance. It helps models
(and consequently what is modeled) because it requires analysis and
evaluation, and it restructures thought patterns. We believe that it is important
to avoid the risk of actions based on flawed or false premises. Critical
thinking helps form judgments that reconcile evidence with good sense and
innovation. Archetypes are valuable to be aware of and use, but it is still
important to employ some of the quality steps we have suggested previously.
We provide extensive description of processes that engage critical thinking in
Chapter 4. Critical thinking can contribute substantially to strategy
development. In addition to using archetypes, it’s a good idea to employ
research literature and field practice in your models. It’s important to produce
a model that’s plausible, better if it’s feasible, and best if it’s strategic.

Selected Archetype Examples

The following archetype examples are drawn from a range of disciplines:
social science, health care, communication, management, and government.
Some aim at individual and group change, while others tackle communities
and systems. We think they offer an interesting view of how experts and
stakeholders describe their recipes for change. In each example, we introduce
the model, and then we provide some context about its use. References for
each example are also included. The health promotion and organizational
effectiveness examples include several models we consider archetypes.
We hope that individually and together, these archetypes are useful for
solving your blank-page paralysis. More important, they represent the rich
possibilities of existing models and the valuable contributions archetypes can
make to your efforts. As in prior chapters, a resources section after the
examples can be used to support additional learning. For this chapter, it
includes a short compendium of related text, journal, and Internet materials
that is organized by topic and is placed after the reflection and exercises.

Example 1: Federal Block Grants

Juvenile Justice and Delinquency Prevention

Reading from left to right, the Juvenile Accountability Block Grant (JABG)
Logic Model (Figure 7.1) begins on the far left with problem and
subproblem(s) columns. This model illustrates an initiative targeting the
improvement of system and program accountability as well as performance
improvement as solutions to the delinquency problem.
The top right side of the model shows how the federal program wants its
grantees to describe and measure types of efforts at the system and program
levels as well as how to quantify results (mandatory rectangles within large
downward-pointing arrows). At the system level, the JABG model shows the
relationship between types of system improvements (staff, infrastructure, and
training) and grant funds invested with number of programs, length of
service, and the use of graduated sanctions and best practices. These are
noted as being of particular importance in influencing rates of participating
youth who reoffend. At the direct service program level, the model shows the
relationship between the implementation of youth programs, the number of
youth served using graduated sanctions, completion of program requirements,
and the rate at which participating youth repeat offending.

Figure 7.1 Juvenile Accountability Block Grant Logic Model

Source: U.S. Office of Juvenile Justice and Delinquency Prevention, 1997.

It also specifies an objectives pathway (ovals, bottom right). Although

this is not frequently seen in models, it serves the purpose here to show how
the mandated output and outcomes measures provide evidence of improved
programs and systems in terms of specific and measurable accomplishments
such as capacities, accountability, efficiency, quality, and effectiveness.
The Office of Juvenile Justice and Delinquency Prevention (OJJDP)
competitively distributes federal funding to community and regional
nonprofit organizations. In OJJDP grant solicitations, logic models are
required as a key element of funding proposals. They are consistent with both
the Government Performance and Results Act of 1993 (GPRA) and the
Program Assessment Rating Tool (PART). While the models help explain
program design, they are specifically intended to help applicants select
relevant performance measures. The OJJDP encourages grant applicants to
use its designated generic logic model as a template to create a logic model
for applicants’ customized programs. OJJDP staff encourage the use of the
template to ensure the applicants’ goals correspond to those identified at
OJJDP, determine the program and purpose areas for appropriate assignment,
and select indicators that will show required performance data. In all, this
office has more than two dozen programs with logic models and associated
indicators for grantees to use.
The model in Figure 7.1 is tied to specific performance measures named
in the block grants program. They include but are not limited to graduated
sanctions, training, juvenile records system, information sharing,
accountability, risk and needs assessment, school safety, restorative justice,
probation, and others. The Juvenile Accountability Block Grant (JABC)
model and its clear performance measures are aligned with the specifications
laid out in the OJJDP generic logic model template. In this way, the OJJDP
offers grantees some important tools that support effectiveness. While not
every program model proposal will be the same, they must include elements
of a change prescription that secures specified outcomes all aimed at reducing
delinquency and improving juvenile justice.
If content specific to juvenile justice were removed, this model could be a
generic recipe for a grant proposal on any subject or change agenda. The
elements of the model (e.g., outcomes, goals, objectives, activities, and
outputs) are common features of any intervention program. It is especially
helpful that the model encourages identification of both the problem and the
subproblems to be addressed through appropriate activities. If you were
planning, managing, or evaluating programs in this content area, the
supplementary materials and more detailed logic models (specific to
particular programs) provided on the OJJDP website would help focus and
stimulate your thinking.
References
For additional information and more resources about the OJJDP work with
logic models, see
Logic models. (n.d.). Retrieved December 12, 2011, from
http://www.ojjdp.gov/grantees/pm/logic_models.html
The OJJDP generic logic model. (n.d.). Retrieved December 12, 2011, from
http://www.ojjdp.gov/grantees/pm/generic_logic_model.pdf
Performance measures. (n.d.). Retrieved December 12, 2011, from
http://www.ojjdp.gov/grantees/pm/grantees.html
Logic models and performance measures for title V formula grants. (n.d.).
Retrieved December 12, 2011, from https://www.nttac.org/index.cfm?
event=titlev

Example 2: Education Readiness and

Success

Pathways Mapping Initiative (PMI)

This model (Figure 7.2, called a “map” by its authors) describes the most
critical actions and goals that are proved to contribute to student success in
the third grade. This model is read both left to right and top to bottom. Goals,
specified in a numbered sequence from top to bottom, are cited for individual
children in the context of a community. Each of the six goals contributes to
third-grade success (far right). Several evidence-based categories of actions
for each goal are also specified. Indicators of third-grade success are listed.
This model shows the comprehensive systems view from which programs or
groups of programs might approach their work.
Assembling the best information from a wide array of resources across
diverse systems and disciplines is central to the approach used in the
Pathways Mapping Initiative (PMI). The PMI is an initiative of the Project on
Effective Interventions at Harvard University, funded by the Annie E. Casey
and W. K. Kellogg Foundations. The PMI tackles several concerns and what
works in America’s families and neighborhoods. The PMI has assembled a
broad and deep knowledge pool on selected issues. Through a process they
call “mental mapping,” the PMI “systematically applies reasonable
judgments and plausible interpretations” to evidence culled from experience,
theory, and evaluation findings. It is a knowledge management project aimed
at effectiveness.
The Ready for School and Succeeding at Third Grade map covers content
about school readiness. Each component is discussed and detail is provided
on the actions with examples, indicators of progress, elements of effective
implementation, rationale, and research evidence. This and other PMI models
are very comprehensive because of the quantity and quality of materials
behind them. It relies on knowledge to inform elements, assemble the general
recipe, and direct users to viable options that reflect their context to complete
a customized model. At the PMI website, click-through content for each area
of the map is accessible so that users can create a map that reflects their own
community conditions and circumstances.
Few other archetypes have this support in the breadth and quality of
content for each feature. This, as well as the rationale, indicators, and
attributes, is also distinct and substantially enhances the archetypes’ value.
Consistent with logic models, the Ready for School map has tremendous
utility for many tasks. It can help with assessment or planning for a school-
improvement effort, inform proposals, and identify examples of effective
strategies.
An allied effort that uses elements of the PMI archetype can be seen in
the Texas Early Learning Council theory of change model (see Figure 7.3). In
2009, the Texas governor created a 19-member Early Learning Advisory
Council to improve school readiness in Texas through targeted strategies that
focus in four priority areas:

Figure 7.2 Ready for School and Succeeding at Third Grade Theory of
Change
Source: U.S. Pathways Mapping Initiative, 2007.

Parental outreach and communications

Early childhood workforce and professional development
Collaborations and standards
Data systems and quality rating and improvement systems

These priorities correspond to the PMI goal areas 3 and 5, Supported and
Supportive Families and Continuity in Early Childhood Experiences. To
improve key aspects of early care and education in Texas, the council will
spend nearly $11.5 million (over 3 years) in American Recovery and
Reinvestment Act (ARRA) funds. Like similar structures and efforts in other
states, the Texas Early Learning Council is responsive to the requirements of
the Improving Head Start for School Readiness Act of 2007.
The Texas Council plans exceed expectations of the 2007 Act.
Subcommittees were formed to address the priorities areas that are driven by
stated needs. On specific tasks and goals, subcommittees will partner with
key stakeholder groups, national experts, and consultants to ensure high-
quality and relevant products are created.
Through council, staff, and contractor efforts, the Texas Early Learning
Council will make key strategic improvements to the Texas early care and
education multisector system. The council will post more than 20 requests for
proposals (RFPs) to accomplish a significant portion of the goals identified in
the model. Note the reliance on information exchange, reporting, and needs
assessment, which can assist relevance in the council’s actions.

References
In addition to school readiness, PMI offers pathway maps and other materials
for successful young adulthood, family economic success, and the prevention
of child abuse and neglect.
For additional information and more resources about the PMI, see:
http://www.cssp.org/publications/documents/pathways-to-outcomes
(Retrieved December 12, 2011) as well as:
Schorr, L. B., & Marchand, V. (2007). Pathway to children ready for school
and succeeding at third grade. Washington, DC: Pathways Mapping
Initiative, Project on Effective Intervention. Retrieved April 24, 2012,
from
http://www.familyresourcecenters.net/assets/library/109_3rdgradepathway81507.pdf
Texas Early Learning Council. (2011). Infant and toddler early learning
guidelines. Retrieved December 12, 2001, from
http://earlylearningtexas.org/umbraco/default.aspx

Example 3: Communications
Human Behavior Change
Frequently, program, project, and initiatives aimed at human behavior change
rely heavily on communications or a special discipline known as social
marketing. In effect, very few efforts can avoid having a communications and
marketing strategy if there’s an expectation that people will adapt in a
particular way or adopt a new practice. This model is a generic archetype we
created that builds on Prochaska’s transtheoretical model (TTM). TTM has
four stages: precontemplation, contemplation, preparation, and action (see
Figure 7.4).

Figure 7.3 Texas Early Learning Council Theory of Change

Source: Texas Early Learning Council (2011). Infant and Toddler Early Learning Guidelines.

This model archetype also builds on social marketing literature. Social

marketing is an application of communication principles and practices for a
social good or benefit. For example, health promotion campaigns began in
Australia nearly 30 years ago aimed at reducing tobacco use and increasing
the use of sun-block to reduce skin cancer. Social marketing is widely used in
the United States to reduce drunk driving and affect teen pregnancy, too.
Note that the model considers two targets for the planned work:
“prospects” and “participants.” This dichotomy separates those that may
engage with messages and materials and that do engage. For any change to
occur (adoption/adaptation), it is essential that a prospect become a
participant. Participants engage in action by sharing, applying, and then
displaying change in skills and behavior. The model shows the process of
how a prospect becomes a participant and how participants, in turn, may elect
to recruit new prospects. Reading from the left side and starting with
prospects, the model articulates a precondition of the subsequent steps. It is a
perceived need or utility. This vital step can stop any forward progression. A
savvy communicator will be sure that any message and medium delivered to
a prospect can resonate with the targets’ relative perceptions. Moving to the
right, the acquisition stage includes three steps: request/participate,
read/hear/interact, and understand. The underlying theory is that prospects
need more than information to take action. Sending a message does not equal
understanding or action on it. But if there’s interaction and understanding,
then it’s possible to affect an attitude and knowledge. New or different
awareness and capacity precede action.
In effect, this model explicates readiness for change and receipt of
information that may or may not, eventually, affect adoption/adaptation. It
can be used as a “script” when in the design and planning stages or to
evaluate a program with substantial communication processes.

References
Itzak, A. (1991). The theory of planned behavior. Organizational Behavior
and Human Decision Processes, 50, 179–211.
Mackenzie-Mohr, D. (2011). Fostering sustainable behavior: An
introduction to community-based social marketing. Gabriola Island,
British Columbia, Canada: New Society Publishers.
Maio, G., & Haddock, G. (2010). The psychology of attitudes and attitude
change. Thousand Oaks, CA: Sage.
Perloff, R. M. (2010). The dynamics of persuasion: Communication and
attitudes in the twenty-first century. New York: Routledge.
Prochaska, J. O., Norcross, J. C., & DiClemente, C. C. (1994). Changing for
good: The revolutionary program that explains the six stages of change
and teaches you how to free yourself from bad habits. New York: W.
Morrow.
Weinreich, N. K. (2011). Hands-on social marketing: A step-by-step guide to
 designing change for good. Thousand Oaks, CA: Sage.

Figure 7.4 Communication, Adoption, and Adaptation Theory of

Change

Source: Phillips Wyatt Knowlton, Inc., 2010.

Example 4: School Improvement

Family and Parent Engagement
Aimed at student success, for more than 15 years, the National Network of
Partnership Schools (NNPS) has been an important project of the Center on
School, Family and Community Partnerships at Johns Hopkins University.
Led by Dr. Joyce Epstein, the Center guides district leaders to develop
programs of family involvement and community connections. It is a leading
resource for knowledge and the foundation of many other efforts in family
involvement and parent engagement. Epstein’s work pioneered a continuum
that offers parents a range of participation from homework to systems
change. Her general theory of change relies on school and district leadership
that influences what parents do on behalf of children and the associated
impact of those actions on student success. Figure 7.5 provides a display of
this thinking and the associated research.
The NNPS website provides extensive resources about the programmatic
research that continues to build this field of study. Research with early, basic
studies, began in 1981 and grew over time as learning about the nature of
“nested leadership” occurred. This multilevel leadership model for school,
family, and community partnerships is composed of school, district, state,
and federal levels. To date, schools and districts have shown the most active
interest in family and community involvement. Epstein’s framework for
family and community participation specifies six types of activities:
parenting, communicating, volunteering, learning at home, decision making,
and collaborating with the community. At the school level, she recommends
an action team for partnerships composed of teachers, parents, administrators,
and other stakeholders who work together so that all parents are welcome and
involved in their children’s education.
The Harvard Family Research Project (HFRP), as well as many other
organizations, has built program models that articulate family/parent
engagement in schools nationwide. All of these university efforts are based
on research, similar to that illustrated in the Epstein model, which indicates
children are more likely to be successful in school if parents and families are
involved. HFRP models often cite four goals for the work of family
engagement that resonate with those illustrated in the Epstein archetype.
They include better student preparation for postsecondary success, family
members who are wise consumers and active partners with children,
transformed schools, and districts with intentional family engagement.
 Research conducted by HFRP substantiates the value of the family and
parent engagement archetype. The effects of more frequent and higher-
quality interactions between schools and parents at all levels include greater
trust and respect, increased social capital for children, and a school
community more supportive of success for every child. In addition, the
academic advantage for children whose parents are involved in their
schooling averages across a number of studies about .5 standard deviation for
overall educational outcomes, grades, and academic achievement.

Figure 7.5 School, Family, and Community Partnerships Theory of

Change

Source: Johns Hopkins University, 2011.

References
Epstein, J. L. (2011). School, family, and community partnerships: Preparing
educators and improving schools (2nd ed.). Boulder CO: Westview Press.
Epstein, J. L., et al. (2009). School, family and community partnerships: Your
handbook for action (3rd ed.). Thousand Oaks, CA: Corwin.
National Network of Partnership Schools. (2010). The Center on School,
Family and Community Partnerships. Baltimore, MD: Johns Hopkins
University. Retrieved December 11, 2011, from
http://www.partnershipschools.org
For additional information and more resources about the HFRP Family
Involvement Project (FIP) and its research, see
Coffman, J. (1999). Learning from logic models: An example of a
family/school partnership program (Reaching Results, January 1999).
Retrieved December 11, 2011, from http://www.hfrp.org/publications-
resources/browse-our-publications/
learning-from-logic-models-an-example-
of-a-family-school-partnership-program
Harvard Family Research Project. (2011). Bibliography of family involvement
research published in 2010. Retrieved January 6, 2012, from
http://www.hfrp.org/publications-resources/
publications-series/family-involvement-
bibliographies/bibliography-of-family-
involve ment-research-published-in-2010
Harvard Family Research Project. (2011). Family involvement. Retrieved
December 11, 2011, from http://www.hfrp.org/family-involvement
Redding, S., Langdon, J., Meyer, J., & Sheley, P. (2005). The effects of
comprehensive parent engagement on student learning outcomes.
Cambridge, MA: Harvard Family Research Project. Retrieved January 6,
2012, from http://www.hfrp.org/publications-resources/
browse-our-publications/the-effects-of-comprehensive-
parent-engagement-on-student-
learning-outcomes

Example 5: Public Health Research

Injury Control Research Center

For more than 20 years, the Centers for Disease Control and Prevention has
funded injury control research centers to build the research base in the field
of injury. Grantees include primarily academic research centers but also some
hospitals with a focus on research and training of students, public health
professionals, and other researchers. The model shown here relies on the
planned work and outcomes of the Injury Control Research Center (ICRC) at
the Centers for Disease Control and Prevention. Both program and evaluation
logic models were developed by Center staff (Yee and Kress), then vetted
among workgroups, evaluators, and other stakeholders.
An early model documented the primary inputs, activities, outputs, and
outcomes as described in the funding opportunity announcements (FOAs)
over a 7-year period beginning in 2000. The second model (Figure 7.6)
illustrates the elements included in the evaluation of the ICRC portfolio
(participating centers). This set of models, like that shown for Juvenile
Justice in Example 1, is of interest because together they show how funder
intent (what work/activities will be funded) and accountability (performance
expectations for those centers receiving funds) for a group of projects can be
addressed flexibly when working toward synergy and innovation. Although
these models are specific to the injury field, in principle, they could be
adapted to describe how a request for proposals for a large body of work
could be used to frame the criteria for grantee selection. They also highlight
one way to develop the evaluation and metrics needed to capture evidence of
return on investment right from the start or at some later point. Portfolio-level
evaluation is frequently used by major U.S. foundations and the U.S.
government to identify common measurement strategies across a group of
projects operating in different contexts but targeting similar outcomes and
impact. When a common measurement framework is used, it is possible to
aggregate or “roll up” results from individual projects to reflect on the
progress made by the whole.
Both logic models use columns to depict different domains. Different-
colored boxes are used around a group of activities, outputs, and outcomes.
Each item in the logic models can be considered its own domain. As one
moves out from individual items, the surrounding box indicates the level of
relatedness between items. Activities labeled as “Core Activities” are within
their own domain, and the Core Activity domain is within the Activity
domain. The overarching domains are then expected to influence
subdomains.

Figure 7.6 Injury Control Research Center Portfolio Evaluation Logic

 Model

Source: Centers for Disease Control and Prevention, Injury Control Research Center, 2009.

CDC; NCIPC: Office of the ADS. Findings from the injury control research centers portfolio
evaluation. Atlanta, GA: U.S. Department of Health and Human Services, 2009.

Arrows depict influence and interaction. The heavy, black arrows indicate
interactions that are known to exist and have measures. For example, it is
relatively easy to model the interaction between Research and Core Activities
inside the Activities domain. Smaller arrows indicate interactions that are
known to exist, but the authors are less certain of the pathways and measures.
Lighter-colored arrows between domains indicate interactions in which there
is still active learning about movement from one domain to the next.
Evaluation work was conducted over a 2-year period from 2007 to 2009.
As a review of a portfolio, the evaluation necessarily illustrates the actual
activities of the grantees. It expands on the first funding opportunity model
and articulates greater detail in the outputs and specifies short-term, long-
term, and ultimate goals. The three squiggly lines between Longer-Term
Outcomes and Ultimate Goals represent the black box of translation.
Since program benchmarks had not been identified, the evaluation model
focused on possible outputs and outcomes over the last 20 years of the
program. This approach allowed the evaluation team to “back into”
identifying contributions of the program to injury research and practice. One
recommendation the portfolio evaluation generated was for CDC to work
with the ICRCs to develop specific indicators for the program. Note that
assumptions that influenced and guided construction of the models and the
evaluation process are specified below each model. This contrasts with the
traditional method of showing context variables in a logic model. The
coauthors indicate this specification was more consistent with the mixed-
method approach used in the portfolio evaluation.
These models were built after review of progress reports from injury
control grantees as well as literature that indicates how research activities
move a field from research to practice. Then, key stakeholders checked the
embedded logic and assumptions.

References
For additional information, contact Sue Lin Yee and Howard Kress at the
Centers for Disease Control and Prevention–National Center for Injury
Prevention and Control.
Centers for Disease Control and Prevention. (2010). Findings from the Injury
Control Research Center portfolio evaluation. Atlanta, GA: U.S.
Department of Health and Human Services.
Dahlberg, L. L., & Krug, E. G. (2002). Violence—a global public health
problem. In E. Krug, L. L. Dahlberg, J. A. Mercy, A. B. Zwi, & R.
Lozano (Eds.), World report on violence and health (pp. 1–56). Geneva,
Switzerland: World Health Organization.
 IN SUMMARY

As tested, general templates for action, archetypes have great potential for
informing your work. They can test the quality of your original efforts and
generate new thinking. Archetypes are evidence based, so they can reliably
jumpstart your modeling. Archetypes can be thought of as recipes. They can
contribute to planning, managing, and evaluation. They are improved upon
by your own knowledge and experience because of your unique context and
conditions. The breadth of content in an archetype varies. They look different
and are often not referred to specifically as theories of change or logic
models. What is important is that they contain the information distilled from
an evidence base needed to illustrate the basic concepts in theories of change
or program logic models. Some represent a single strategy, while others cover
complex projects. This chapter provided examples of archetypal theory of
change and program logic models.

LEARNING RESOURCES

Reflection

1. In the absence of an archetype, what elements of knowledge can help

you jumpstart creation of a logic model?
2. Does your field or discipline rely on any archetypes? If so, are some
better than others? Why?
3. Name any challenges you might experience with using an archetype.
Why could or would an archetype fail?

Exercises

1. Go to the Juvenile Justice website and prepare a program logic model.

Compare and contrast with others who do the same. Was there fidelity to
 the template? Why? Why not? Select a model and discuss how it might
be evaluated.
2. Find or create a description for a health promotion program. Compare it
to the evidence-based New York Healthy Weight model (Figure 6.7).
What is the same, and what is different?
3. Locate a case study on family and parent engagement. Apply Epstein’s
model to it. Does the case you located reflect any of the elements on the
archetypical model? If not, why not?
4. Locate logic models for collaboration, sustainability, or marketing.
Would they qualify as archetypes? Is the research or evidence base
readily apparent? How could the one(s) you located be used in your
work?

References and Supplemental Readings

For this chapter, additional texts, journals, and Internet resources (when
available) are organized relative to the subject matter content of the examples
cited. This list is illustrative but not comprehensive. Many more resources
exist than are cited here.

Evidence-Based Models
Baruch, G., Fonagy, P., & Robins, D. (2007). Reaching the hard to reach:
Evidence-based funding priorities for intervention and research. New
York: John Wiley.
Center for the Study and Prevention of Violence. (2004). Blueprints for
violence prevention. Boulder, CO. Retrieved December 11, 2011, from
http://colorado.edu/cspv/blueprints/index.html
Coalition for Evidence-based Policy. (2003). Identifying and implementing
educational practices supported by rigorous evidence: A user friendly
guide. Washington, DC: Author. Retrieved December 11, 2011, from
http://www2.ed.gov/rschstat/research/pubs/rigorousevid/index.html
Department of Health and Human Services. (2009). Identifying and selecting
evidence-based interventions: Guidance document for the Strategic
Prevention Framework State Incentive Grant Program. Washington, DC:
Substance Abuse and Mental Health Services Administration. Retrieved
 April 24, 2012, from http://store.samhsa.gov/shin/content//SMA09-
4205/SMA09-4205.pdf
Duignan, P. (2004). Principles of outcome hierarchies: Contribution towards
a general analytical framework for outcomes systems (outcomes theory).
A working paper. Retrieved December 11, 2011, from
http://www.parkerduignan.com/se/documents/122f.html
Norcross, J. C., Beutler, L. E., & Levant, R. F. (2006). Evidence-based
practices in mental health: Debate and dialogue on the fundamental
questions. Washington, DC: American Psychological Association.
Promising Practices Network on Children, Families and Communities.
(2011). Programs that work. Retrieved December 11, 2011, from
http://www.promisingpractices.net/programs.asp
Proven Models. (2011). Proven models. Amsterdam, the Netherlands.
Retrieved December 11, 2011, from http://www.provenmodels.com
Rockwell, K., & Bennett, C. (2004). Targeting outcomes of programs: A
hierarchy for targeting outcomes and evaluating their achievement.
Retrieved December 11, 2011, from
http://digitalcommons.unl.edu/aglecfacpub/48/
Schorr, L. B., & Farrow, F. (2011). Expanding the evidence universe: Doing
better by knowing more. New York: Center for the Study of Social
Policy. Retrieved January 6, 2012, from http://lisbethschorr.org/doc/
ExpandingtheEvidenceUniverse
RichmanSymposiumPaper.pdf
 8

Action Profiles

T
his chapter demonstrates the amazing utility and vast application of
logic models. It includes model examples with tremendous variation
in subject content and display. Generally, these models have enough
detail to support design, planning, and management as well as
evaluation. In several instances, they supported multiple functions. These
“practice profiles” include models about civic engagement, corporate giving,
international development, public health, sustainability, human services, and
environmental leadership. This chapter displays the versatile functionality of
logic models.
 LEARNER OBJECTIVES
Describe the benefits and limitations of logic models in practice
Identify the rationale for model use in multiple contexts
Recognize and use concepts introduced in Chapters 1–7
Show how models display problems and support strategy, evaluation, and learning

Strategy, Evaluation, and Learning

Each of the following seven profiles is an exciting example of how people in
diverse roles used models to support the design, development, and
communication of important work. As often as models assist with evaluation,
they help significantly with strategy and tactics. They can be used to improve
and prove services and to provide a vision of intentional, hopeful change.
While never perfect or comprehensive, models offer an alternative to long
narrative and are easily revised to suit a change in context as well as adaptive
management. Models can be vital tools in learning—for individuals, teams,
and organizations.
We hope these examples inspire and encourage your work across a range
of issues. The profiles include: civic engagement, corporate giving, labor
practices in Eastern Europe, asthma management, sustainability,
homelessness, and an effort to minimize the destructive influence of paint
disposal.

Profile 1: Building Civic Engagement

What exciting invitation for a “civic life” could entice a hip population in a
progressive West Coast city?
More than 20 years ago, Seattle Works began as The Benefit Gang, a
motivated group of twenty-somethings who formed an organization dedicated
to involving their generation in the Seattle community. The citizen-led group
believed that people in their twenties desired community engagement but
needed alternatives to the service clubs and expensive charity balls attended
by their parents. Their leaders understood the importance of giving back and
sought means of community participation that matched their lifestyle.
Now broadly known as an influential resource, Seattle Works supports
energetic volunteer teams that have generated inventive programs recognized
for their impact on the community.

Through volunteer and leadership development opportunities, Seattle Works

connects young adults with a range of service options. In turn, these
volunteers become more civically engaged and take action in their
communities.
Volunteerism and civic engagement research points to participant-driven
decision making, relationship building, and a variety of opportunities for
participation as viable means to increase participation in community service,
philanthropic giving, and community action. Seattle Works’s targeted
population reflects a socio-demographic profile that seeks technology-
enhanced support and social networking. Seattle Works increases access to
information and opportunities, both of which are important factors in
developing engagement.
The model reads from left to right and uses bright green labels to cite
primary elements. These include: Opportunities for Engagement,
Connections, Learning, Inspiration, Action, Growth, and Goals. The model
suggests if opportunities are identified, then connections, learning,
inspiration, action, and growth will occur that lead to vital community goals.
On the far right, Seattle Works goals are positive changes in volunteerism,
philanthropy, leadership, and civic participation. Under the Opportunities for
Engagement column, there are five primary sections that could be
synonymous with strategies.
Those five include communicating, volunteering, giving, developing
leadership, and influencing. Each of these includes specific events, products,
or activities. The steps toward the ultimate goals are not linear but do build
from the initial connections and learning, and the frequent and recurring
involvement is what leads to growth and development. The graphic choice of
a “waving flag” (instead of a flat rectangle) was meant to symbolize the fluid
nature of people’s involvement and the motion of forward progress (see
Figure 8.1). The “messiness” in the middle of the model is denoted by white
space and curving arrows. This is intentional and is meant to communicate a
relative and intimate interpretation of an individual’s change process.
This evidence-based theory of change model was developed in a small
group that included Seattle Works staff. Through expert facilitation, a draft
model was created. Participants indicate the modeling process clarified the
unique work and mission of the organization. Subsequently, it was shared
with the Seattle Works board for discussion and development. It is used
primarily for communications with board members and other stakeholders.
In 2007, Seattle Works members were surveyed to determine the
influence of this organization. Findings from that data collection indicate
since becoming affiliated with Seattle Works, respondents volunteer with
greater frequency than the national average for comparable age groups;
demonstrate more civic engagement in political activities, particularly voting
rates; and the majority made a financial contribution to a charitable
organization at rates considerably higher than the norm for a comparable
population. In addition, respondents are more aware of service opportunities,
feel they contribute and are connected to their community, and indicate a
good fit for their time, skills, and passion.
Seattle Works was honored with a community service award from the
Municipal League of King County in 1996, the History Makers in
Community Service Award from the Museum of History and Industry in
1997, the YMCA of Greater Seattle’s AK Guy Award in 2002, and Seattle
Parks & Recreation’s Denny Award for Outstanding Volunteer Service in
2005 and was named an Innovation Hub by the HandsOn Network in 2011.
Will is a key assumption for Seattle Works programs. Their efforts rely
on an expectation that people are eager to step up as active community
participants early in their careers and adult lives. They believe a vibrant
Seattle depends on the civic engagement of young adults who, over time, will
continue to demonstrate their mettle as volunteers, voters, leaders, and
philanthropic investors bringing positive impact to their communities. After
more than two decades of operation, Seattle Works remains a highly regarded
organization that launches and supports vital civic capital in the Puget Sound.

References
See the Seattle Works website at www.seattleworks.org.
Creation of this model was led by Dawn Smart at Clegg & Associates.
Contact her via e-mail at [email protected].
 Profile 2: Better Corporate Giving
Childhood hunger in America is a significant challenge. It is likely to
increase as our population grows, climates change, and food prices rise.
In households across every state in our nation, every day, children face
inconsistent access to nutritious and adequate food. They don’t know if or
from where they will get their next meal. Hunger has broad implications for
human development: increased susceptibility to illness, cognitive and
behavior limitations, and associated impairment of academic achievement.
ConAgra Foods, via its charitable giving through the ConAgra Foods
Foundation, has chosen this cause and used logic models inside and outside
to align its important work. The focus is ending childhood hunger. ConAgra
Foods Foundation intentionally chose ending childhood hunger as its primary
cause in 2006. The giving program distributes funding nationwide, through a
dozen community intervention programs, and through far-reaching brand
promotions. In 2011, 2.5 million meals were distributed as a result of a 30-
minute news special combined with a company-led consumer campaign that
paired products purchased with donations (see
www.childhungerendshere.com). Over the past 20 years, ConAgra Foods has
led the charge against child hunger in America with donations of more than
$50 million and 275 million pounds of food. ConAgra’s community
involvement platform, Nourish Today, Flourish Tomorrow®, focuses on
ending hunger, teaching kids and families about nutrition, and improving
access to food.

Figure 8.1 Seattle Works Theory of Change Model

Source: Seattle Works, 2006.

Business and Social Interests

Aligning business and social interests isn’t a new idea. It’s a vital feature
of corporate social responsibility (CSR) and informs corporate giving.
ConAgra’s active social profile in association with food, specifically
childhood hunger, is logical. What is new is how ConAgra and other
corporate givers can tackle common work and how a select cause can become
a focal point for synergies across an enterprise. ConAgra assembled its
primary Washington, D.C., grantees with national reach to sharpen its
aggregated aim at ending childhood hunger with shared resources.
They used logic models and modeling to advance their plans and
evaluation. ConAgra considered their primary grantees as anchors for a
portfolio. Foundation staff wanted to document the current grants in relation
to each other, establish indicators that could inform progress monitoring, and
aggregate data to gauge outcomes. These vital summaries will allow them a
practical, fact-based format to review best bets for additional foundation
funding with existing or new grantees.
ConAgra Food’s operating principles are simplicity, collaboration,
imagination, and accountability. ConAgra employees are also expected to
display “leadership attributes,” specifically authenticity, vulnerability, and
courage. These principles and attributes were key to their foundation’s
approach with four important nonprofit partners: Feeding America, Share
Our Strength, Congressional Hunger Center, and the Food Research and
Action Center.

Feeding America, formerly America’s Second Harvest, is a nationwide

network of more than 200 local food banks supplying more than 60,000
community-based agencies. This network helps feed 37 million
Americans each year.
Share Our Strength mobilizes individuals and industries to fight hunger
and supports nutrition education.
Congressional Hunger Center focuses on domestic and international
antihunger leadership development.
Food Research Action Center influences public policy and coordinates
public–private partnerships to eradicate hunger and under-nutrition.

While all these organizations have active and long roles in antihunger
work, their staff had never convened to see or understand the roles each
played among key strategies supported through ConAgra funding.
Our firm used highly participatory processes to ensure that multiple
perspectives were expressed and reflected in any products. A thorough
review of internal and external ConAgra documents along with several phone
conferences were essential to inform a preliminary draft of both a theory of
change (TOC) and a program logic model. The TOC, shown in Figure 8.2,
remained largely unchanged over the project. It simply documented the
knowledge-based strategies that would most likely influence childhood
hunger.

Reducing Childhood Hunger

The ConAgra Foods Foundation theory of change, read from left to right,
identifies five strategies for current funding: public awareness and education,
public policy and thought leadership, public program utilization, direct
service (feeding efforts), and food donations. Emerging funding strategies
include cause marketing as well as innovation, replication, and social
enterprise solutions. By integrating these strategies where appropriate, the
expectations are that the U.S. food distribution systems will both improve
nutrition for children and increase food access. These outcomes will
ultimately contribute to the planned result of ending childhood hunger. This
logic model represents a framework for how planned work can be organized
by and with ConAgra Foods Foundation staff, corporate functions, grantees,
and other stakeholders. The key code identifies grantees by their contribution
to relevant strategy.
Next, grantees were approached about a meeting to articulate their
organizations’ work and contribute to a collective view that would inform the
ConAgra theory of change and program logic models. Prior to this meeting,
grantee representatives were asked to consider only their work (relative to
ConAgra strategies) with internal colleagues. This was designed to ensure
that inside discussions defined a shared understanding of responsibilities
associated with ConAgra support. Armed with this information, they could
then confidently articulate their representative portion vis-á-vis peers from
other grantee organizations.
Setting the stage for better understanding about models and vibrant
participation, we provided stakeholders with a brief and practical introduction
to logic models. Using adult learning techniques, we asked organization
representatives working in small groups to plan an ideal event by specifying
what they’d do and get. Then we deconstructed the activities and primary
strategies relative to intended outcome. This easy, kinesthetic activity offered
a simple way to practice transferring what they’d learned in the orientation to
action steps in co-creating a model. It anchored the essential elements of a
logic model. The latter allowed a review of common elements in relation to
planned results and introduced a quality continuum from plausible to
strategic.

Figure 8.2 ConAgra Foods Foundation Theory of Change Model

Source: ConAgra Foods Foundation, 2011.

In addition to some advance reading, this experiential learning helped

prepare participants for a critical review of the preliminary ConAgra logic
models. From the outset, the dual challenge was concurrent attention to both
program and measurement. To ensure utility and validity, it was critical that
both these purposes were considered in the development work. Initial
organizing questions included How and where did grantees “see” their
organization in the strategies ConAgra had funded to date? and what would
be appropriate indicators of progress against childhood hunger in the
ConAgra portfolio? The primary strategies and relative activities (program)
were tackled first. Through a facilitated process, an exhaustive list of grantee
activities was cited and grouped in strategies. It was important to name
strategies that held shared meaning. Then, relative to the activities and
strategies, their associated outputs and outcomes were identified.

Measuring and Managing

A version of the socially constructed model generated by the DC
antihunger grantees and their funder is shown in Figure 8.3. Note, again, the
intended result on the far right of this graphic is “improved children’s food
security.” While the version displayed here does not explicate the
assumptions for resources/inputs (far left), they can be generally identified as
well-managed grantee partners and financial capital, as well as supporting
functions like public relations, communications, marketing, product
promotions, and others. Because the model needed to support the creation of
a monitoring and measurement system, it was practical to identify reasonable
outputs and related short-term outcomes.
In this model (see Figure 8.3), the impacts (far right) are likely to occur if
the long-term outcomes do. In this way, a dependent chain of “if–then” steps
is projected from the cited strategies (at the far left). The model is not a
substitute for action planning that would detail by which grantee, when, with
whom, and how (tactically) each strategy plays out over time. But it does
provide a high-level road map to specify what information will be gathered
and what indicators will suggest progress. At a point in time, this anti-hunger
portfolio-level model reflects the aggregate investment and associated metrics
for selected ConAgra grantees.
Modeling, the process of creating multiple versions of a display, is
generative. Additions and changes to models are crucial as they adapt to
capture knowledge. While this case “backed into” a model as documentation
from existing plans and work, more often, modeling happens at the design
phase. Modeling adds tremendous value in an initial (and ongoing) convening
of multiple stakeholders to launch, manage, and evaluate projects, programs,
or change work.
The ConAgra models are dynamic and will change. Any model, like a
photograph, is simply a snapshot in time. To remain relevant, it must be
revised. As evaluation provides feedback about what’s working, what’s not,
and rate of return on investment, changes in the grantee portfolio can occur.
This “steers the ship” based on performance data and changing contextual
factors.
Alignment, Accountability, Action
These models and associated modeling supported plans and
communication with external partners. It was critical to citing grantee
accountabilities. It offered an important way to specify relative and shared
outputs from activities that would influence outcomes. This evaluation
capability is a challenge—often considered a “resource drain and distraction
for nonprofit organizations.”
Ultimately, the indicators from the four grantees were used to inform an
electronic reporting template. For ConAgra, it allows annual aggregation and
can be used to describe value for its corporate giving. Like other corporate
funders, ConAgra uses this information in internal communications and
planning as well as with external stakeholders. With a clear picture,
foundation staff can better manage emphasis within strategies that could have
yields as the context for antihunger efforts changes.

Figure 8.3 ConAgra Foods Foundation Portfolio Logic Model

Source: ConAgra Foods Foundation, 2011.

Some of the important exploration, documentation, and accountability in

this corporate giving example provides vital context for related internal CSR
efforts. At ConAgra, CSR translates to “Good for You, Good For
Community, Good for the Planet.” These planks address an enormous range
of issues, including food safety and quality, health and nutrition,
biotechnology, animal welfare, workplace, suppliers, community investment,
water stewardship, sustainable packaging, and climate change. Corporate
philanthropy is included in the community plank.
We created proprietary models of ConAgra’s cause and corporate giving
in relation to other corporate functions. Those models are early inputs for
strategic and structural decisions that will drive the childhood hunger cause
across the enterprise. Importantly, they provide an initial view of how
functional areas can cooperatively contribute to ConAgra’s CSR profile.
Explicating the potential synergies for functions like government relations,
sales, and supply chain relative to a specific CSR cause supports “shared
value” for multiple stakeholders.
As sectors converge, organizations of many types partner, and new
alliances form, talented professionals need tools and processes that improve
chances of success. Logic models and modeling offer great value to design,
strategic planning, monitoring, and evaluation. They can contribute
enormously to alignment and integration because they offer a picture that
displays these powerful principles. When people and organizations can
clearly see their role, it is more likely they can fully contribute. This case also
offers a gentle reminder that accountability is central to social change. It cites
not only the intentions of a large corporation but also its grantee partners,
who publicly called out their own work.
Ultimately, consumer awareness and action in communities all across the
country are necessary to progress against this devastating problem. Along the
way, savvy corporate funders and their colleagues will get further faster on
complex social issues with potent tools and processes.

References
This content is adapted from a feature article, “Corporate Giving Gets
Smarter,” in The Foundation Review, Spring 2012.
Kotler, P., Hessekiel, D., & Lee, N. (2012). Good works: Marketing and
corporate initiatives that build a better world … and the bottom line. New
York: Wiley.
Creation of this model was led by Phillips Wyatt Knowlton, Inc.

Profile 3: Kyrgyzstan Decent Work

Country Programme
From 2006 through 2009, the International Labour Organization (ILO)
supported a Decent Work Country Programme (DWCP) in the Kyrgyz
Republic with 40 community-based projects. In 2010, the ILO hired a team
of consultants to conduct an independent evaluation of its support to the
program.
Based on an extensive review of documents, evaluators drafted a logic
model (Figure 8.4) to conduct program design analysis. This model helped to
show the entire program concept, or theory, at a glance and to visualize some
gaps in the program logic. For example, the model showed a midterm
outcome that was not related to a program priority (see “other areas of work”
at roughly one o’clock on the orbital model).
Kyrgyzstan’s DWCP had three main priorities:

Priority A: Employment creation, skills, and employability for

women and men covers issues such as creating more job opportunities
for both young women and men, including improving their work
abilities, in order to increase their employability.
Priority B: Improving the national Occupational Safety and Health
(OSH) system focuses on upgrading policies, programs, and practices
pertaining to the Kyrgyz national Occupational Safety and Health
system.
Priority C: Reducing the decent work deficit in the informal
economy focuses on extending decent work reality and standards to
other sectors of work, in particular the informal economy.

Because the program theory included several chains of intended outcomes

contributing to three overlapping priority areas, a graphical representation
was used given that a narrative description may have been inadequate or
easily misinterpreted. The authors used a priority-centered “orbital” model
with mid-term outcomes on the low orbit and the short-term outcomes on the
high orbit. On this first, more complex model, the evaluators purposefully
used shades of gray and made the center (priorities) dark, mid-term outcomes
lighter, and short-term outcomes lightest. The importance of the three
priorities suggested they would be best in dark and placed at the center to
immediately attract the reader’s attention. The arrows show the theory behind
the DWCP as illustrated by outcome chains leading to each priority. To read
the model, begin from the outside and move toward the center.
 As the evaluation process unfolded, it became clear to the evaluators that
the core of the DWCP could be presented as a combination of subregional
project activities implemented in Kyrgyzstan that contributed to DWCP
outcomes. To visualize this finding, evaluators developed a simpler orbital
model that showed several projects that made major contributions to the
implementation of the DWCP (Figure 8.5). The project shaded gray in this
second model was the only “national” project (i.e., a project implemented
exclusively in Kyrgyzstan).
This model helped demonstrate the actual nature of the DWCP
(combination of activities implemented under independent subregional
projects) as opposed to the theory described in program documents and
illustrated in Figure 8.4. The two models were presented in the same way and
could be easily compared. The evaluators indicate this simpler version
“represented evidence in a way that helped the evaluation team to introduce
some findings in a clear and convincing manner.” This profile demonstrates
that models can be used to illustrate both theory and the actual program as
implemented. The relative contrast can be informative for operations staff.

Figure 8.4 Kyrgyzstan DWCP Theory of Change With Outcomes

Source: Independent Evaluation of the ILO’s Decent Work Country Programme for Kyrgyzstan: 2006–
2009.

Figure 8.5 Kyrgyzstan DWCP Theory of Change

Source: Independent Evaluation of the ILO’s Decent Work Country Programme for Kyrgyzstan: 2006–
2009. International Labour Organization, 2010.

This evaluation focused on the ILO’s strategic positioning in the country

and its approach to setting an ILO agenda as well as the composition,
implementation, and evolution of ILO national strategies as they relate to the
Decent Work Agenda. The evaluation team concluded that:

The DWCP in Kyrgyzstan was focused on the priorities jointly

developed by the ILO and its constituents. Those priorities were relevant
to the challenges Kyrgyzstan faced and were in line with key ILO
strategic documents.
Although it focused on the three clearly defined priorities relevant to the
country context, the program was not based on a clearly defined logic
model and was not logically coherent.
 The evaluation team used primarily qualitative methods to collect data on
the DWCP results. During data collection, the evaluation team reviewed 33
documents, interviewed 56 stakeholders, and conducted direct observation of
several ILO projects. Evaluators recommended that the DWCP develop a
coherent program logic based on priorities. The new DWCP program logics
should be coherent and focused on the priorities identified by the tripartite
constituents rather than on the existing regional projects with secure funding.

References
For additional information, contact Alexey Kuzmin at
[email protected] and Craig Russon at [email protected].
The evaluation report for this work is found at Independent Evaluation of the
ILO’s Decent Work Country Program: Kyrgyzstan: 2006–2009.
Retrieved December 22, 2011, from
http://www.ilo.org/public/english/bureau/program/dwcp/download/eval-
kyrgyzstan.pdf

Profile 4: Alabama Tackles Asthma

Twenty-three million adults and children suffer from asthma in the United
States, incurring an estimated $13 to $20 billion in direct medical costs and
indirect costs due to lost productivity annually. In Alabama, 1 in 10 residents
is diagnosed with asthma. It is a substantial chronic health issue with
significantly different prevalence for gender and race. Women experience
asthma at a rate twice that for men and have a nearly double rate of death.
African Americans have an incidence rate of 3 percentage points more but
double the rate of death for the Caucasian or white population.
Led by the Alabama Department of Public Health and co-created with the
Alabama Asthma Coalition, public health officials and a broad assembly of
stakeholders built a 5-year strategic plan to affect asthma. This plan was the
basis for an overarching and generic model that guided multiple committees
to author committee models with specific activities and outcomes.
The model generates its own evidence base for community- and statewide
programs (see Figure 8.6). A primary activity of the model is to act as a
surveillance system. It will collect data, establish baselines in areas where no
data exist, and map disparate populations and areas of poor air quality or high
emissions. This initially informs comprehensive annual burden reports. In
this way, the surveillance system enables other aspects of the program to rely
on data. Short-term outcomes include acquisition of baseline data as well as a
continuing stream of information that will be used to inform activities,
programs, and policymaking efforts.
Three gross areas are identified in the model, reading left to right: inputs,
outputs, and outcomes. The meta-model is intended to be generic; the color-
keyed letters refer the reader to specific committee pages for more detailed
descriptions of the activities and outcomes. The activities listed on this page
are intended to be implemented in the coming years, with selected activities
to continue throughout the cooperative agreement. These activities are jointly
performed by separate committees, from different viewpoints. For instance,
both the Community/School (C) Committee and the Environmental (E)
Committee will be working on implementing Tools for Schools in schools.
However, the C Committee will be working on the educational programs and
public awareness, while the E Committee will work from the indoor and
outdoor air quality aspect. Once the program has been implemented by those
committees, the project will be picked up by the Advocacy and Policy (A)
Committee to bolster statewide adoption and public support for asthma-
friendly policies and ordinances. The model also includes a specific list of
both assumptions and external factors.
Each committee has its own “nested” logic model that coordinates with
the meta-model but gives more details regarding each planned activity and its
intended outcomes, as well as tailored inputs, assumptions, and external
factors. Future logic models will include separate models for each
intervention planned during a particular time frame and thus allowing for
ease in adapting activities as needs change or evaluation shows the efficacy
of the planned interventions.
After providing an initial overview of logic models, Public Health staff
sent each committee labeled sheets (a template) for them to prepare relevant
and focused content for their committee models. In a social process, the
facilitator provided a chance for review of each model and its fit with a larger
view. This provided an important opportunity for engagement.
The logic model shown here has been used as a guide to where
interventions and activities are leading—the ultimate outcome—as well as
showing which committees have parts to play during the process. It has been
adapted and changed as there is discovery about what works and what
doesn’t. For example, funding has been limited in some of the work
associated with schools, but the “No Idling Campaign” exceeded
expectations. School bus drivers attended an in-service and signed a pledge
not to idle buses more than 5 minutes, resulting in gas cost savings and less
emissions in the environment.
In this case, modeling was an asset for several reasons: It helped connect
stakeholders, engaged vital expert contributions, provided a common
communication platform, and managed expectations for more realistic
timelines, resources, and program design. The nested models contributing to
a whole also ensured important alignment among work teams.
Note that the model includes short-, medium-, and long-term outcomes,
which are planned to reduce healthcare utilization due to asthma, reduce
disparities among those who suffer asthma, and improve quality of life for
patients and their families.
This exciting public health work to reduce asthma in Alabama was part of
the CDC National Asthma Control Program. Staff expect models will be
revised to serve future planning, monitoring, and evaluation needs as their
work continues through 2014.

Figure 8.6 Alabama Asthma Program Logic Model

Source: Alabama Department of Public Health and the Alabama Asthma Coalition, 2010.

References
Contact Debra Hodges (at [email protected]), Alabama
Department of Public Health. See also:
Williamson, D. E., Miller, T. M., & McVay, J. (2009). Alabama asthma
burden report. Montgomery, AL: Alabama Department of Public Health.
Retrieved December 22, 2011, from
http://adph.org/steps/assets/ALAsthmaBurden.pdf
 Profile 5: Resilient Communities
A “world of resilient communities and re-localized economies that thrive
within ecological bounds” is an exciting vision. This is the work of the Post
Carbon Institute (PCI). Created in 2003, PCI is leading the transition to a
more resilient, equitable, and sustainable world.
Alarming changes reflecting fundamental crises face our planet. Experts
in economics, ecology, political systems, social justice, public health, and the
environment can each cite complex challenges in their respective content
areas. As these challenges converge and interact, they affect every living
thing. Identifying those intersections for both vulnerabilities and
opportunities is vital to building a more resilient society. The PCI suggests
the following assumptions are essential in future planning:

None of our global problems can be tackled in isolation.

We must focus on responses not just solutions.
We must prepare for business unusual.

The PCI theory of change model (see Figure 8.7) is read

counterclockwise, beginning with mission and continuing through audiences,
strategies, focusing events, desired shifts, and impact.
The model uses variations in contrast to sequentially lead the reader
through the information. The stylized area around focusing events, crises, and
windows of opportunity was included to emphasize this area: an integral part
of strategy formulation/implementation that can be easily overlooked in logic
models/theories of change. The increased contrast around the area of impact
was chosen to add emphasis, implying the role the rest of the model serves in
contributing to impact. Authors elected to design the model in grayscale,
since it is so common to see beautifully color-rendered models lose
significance when they are printed—which is often in black and white.
Innovation Network staff created the model based on content gathered
from interviews with PCI staff, fellows, board members, volunteers, funders,
and peers. A literature review of assessment areas for similar thought
leadership organizations was conducted and also informed the work. The
model is based on interview theme analysis, field approaches to evaluation of
like entities, and a thorough review of PCI documents. Iterative feedback on
versions of the model contributed to its development.
Models provided an interactive and important approach to discovering a
representative consensus by stakeholders. The theory of change model was
presented in tandem with recommendations for monitoring and evaluation
approaches. It has also helped support refinement of strategy and related
work plans. While the model relies on evidence collected during interviews,
literature, and document review, it also helped generate a hypothesis about
the connection between organization strategies and desired outcomes. In this
regard, the model was largely successful.
The modeling process was an asset to the PCI. It created a way to capture
the many and varied perspectives of vital stakeholders about the
organization’s planned work. It enabled staff to find agreement on key
audiences, strategies, and desired shifts. It also helped staff to sharpen their
understanding and language regarding key points of the organization’s work
—generating a clarity and focus that was vital to their organization
development.

References
Additional detail regarding this model can be secured via contact with
Johanna Morariu at Innovation Network, [email protected].
For more on the Post Carbon Institute, see http://www.postcarbon.org/about/

Figure 8.7 Post Carbon Institute Theory of Change

Source: Innovation Network & PCI, 2010.

Profile 6: Sheltering Families

Michigan’s challenging economy has created structural unemployment and
increases in poverty. These conditions affect people in serious, life-altering
ways. Fortunately, there’s an important resource for homeless families in East
Lansing called Haven House. It provides emergency housing and support
services for one- and two-parent families with children. The shelter helps
families who are homeless prepare for permanent housing by developing and
promoting self-sufficiency, stability, and financial responsibility.
Through an applied experience, students in a Michigan State University
(MSU) evaluation course became acquainted with the services of Haven
House. The first model draft was created by working with the MSU professor
and describing what happens to clients when they come into shelter. This
initial work was given to the students, who then met with staff to ask
clarification questions and direct program questions. Several meetings and
associated modeling were required to edit and revise the display to accurately
represent the Haven House program.
 The model (see Figure 8.8) describes Haven House programs, including
case management. It portrays the evolution of a client in the shelter, barriers,
and possible outcomes and includes external and internal forces that affect
outcomes. Read from left to right, the Haven House model is based on the
typical elements of a program logic model with some important adaptations.
The program components appear in a single column, but this version of the
model explicates both the crisis response and the intervention to obtain
housing and its implementation. The organization’s response and specificity
in the service delivery raises important feedback issues about a key
constituency: Haven House clients. The original model uses color to code
elements of the plan to column headers. Note that in this display, immediate
and long-term outcomes are cited. Important context is also articulated.
As part of a broader evaluation plan, students focused on assessment of
client satisfaction as well as changes in knowledge and skill. For this reason,
they posted, on the model, constructs about client feeling and learning. These
constructs guided the creation of features that indicate intended feelings (e.g.,
hopeful, valued) and new skills like budgeting. To determine the influence of
Haven House, one aspect of the evaluation included a client survey. These
data were collected and analyzed for sharing with staff. They provided
important feedback and insight from the client’s perspective. One resident
shared, “I loved the experience. It was something I needed at 18 so that I can
grow, meet different people, and learn new ways to do different things.”
Another said, “Overall it’s a good program and if the rules are followed then
success will come.” The model is a valued communication tool in external
relations. It also helps staff to visualize their role and the barriers residents
face. Participants indicate the modeling process and associated evaluation
work were clarifying. Angie Mayeaux, Haven House director, says, “Much of
the services we provide are difficult to articulate. The modeling pushed staff
to really look at what they do and how they do it. Our staff also took some
pride in seeing their work captured in the logic model.”

References
For more information, see www.havenhouseel.org.

Figure 8.8 Haven House Program Logic Model

Source: Haven House, 2007.

Profile 7: Environmental Leadership

Paint can have significant unintended environmental impacts—contaminating
groundwater, harming fish and other aquatic life. Because it is combustible
and contains solvents, it is also considered a hazardous waste. While most
paint sold is now latex instead of oil based, managing leftover paint is a big
and costly challenge for Americans. Nationwide, households generate some
75 million gallons of leftover paint. This is about 10% of the amount of paint
purchased annually. At more than $8 per gallon, the estimated cost to manage
it is substantial. In Oregon, paint is the single largest contributor to household
hazardous waste programs.
In 2002, product stewardship for postconsumer paint began when paint
manufacturers, local state and federal agencies, and retailers, along with
consumer and environmental agencies, formed the Paint Product Stewardship
Initiative (PPSI). Facilitated negotiations by the Product Stewardship Institute
(PSI) helped to create an industry-managed postconsumer paint management
system. After many years, in 2009, Oregon became the first state in the
United States to enact a law that identifies product stewardship as the
preferred method to reduce environmental impacts and costs associated with
leftover paint.
Oregon’s statewide system for the collection of postconsumer latex- and
oil-based paint is based on six goals originally created by the PPSI:

Goal 1: The pilot project is a collaborative and cooperative process.

Goal 2: Establish a paint stewardship organization (PSO), which
operates under the direction of the paint industry.
Goal 3: Consumers (including painting contractors) generate no or less
waste paint and containers.
Goal 4: The statewide postconsumer paint management system should
be designed to ensure that it is environmentally beneficial, economical,
and convenient. With these considerations, the system should strive to
use methods highest on the following waste management hierarchy:
reuse, recycling (into paint or other products), energy recovery
(generally applicable to oil-based paint), and proper disposal.
Goal 5: Identify cost-effective alternatives for using postconsumer paint
products and explore means to expand the market for products
containing post-consumer paint.
Goal 6: Measure and evaluate the performance of the pilot project, and
ensure the results and learning that the evaluation generates are
transferable and relevant to the rollout of a national postconsumer paint
management system.

The PPSI formed an evaluation committee to ensure overall

accountability and implementation of the last goal.
The Oregon program is composed of a diversity of interconnected
systems, actors, and processes. The major components of the Oregon
program are the paint stewardship organization (PaintCare), the oversight by
the Oregon Department of Environmental Quality (OR DEQ), the paint
market, and the leftover paint management system (see Figure 8.9 for original
drawing).
There are multiple points of access to this model. A reader might start
from the left at the PPSI and freely flow through the graphic, gaining an
understanding of the flow of paint from manufacturer to recycling, exiting the
graphic on the right side where the learning from the PPSI pilot program is
transferred to other states in the United States. On the other hand, a user may
simply begin clicking on the numbered paint “splatter” evaluation questions
because it’s more fun. Through strategic placement of basic design elements
and principles of graphic design, the model allows the user to take and be
taken on a visual journey, free and flowing or abrupt and acute, both
representing practical, though abstract and ambiguous, program space and the
fluidity of paint.
The model (see Figure 8.10) is intended to be accessible and used at
many levels—from the surface, a user can glance at the primary actors,
processes, and systems involved, while closer inspection and a look below
the surface gives access to the program’s underlying research and data that
are embedded in the model and influence the shapes of the systems,
placement and engagement of actors and processes, and the connections
between program components. Hotlinks take the viewer/user to documents
for associated purposes. The model uses multiple colors to key attention to
functional areas. The refined model versions were created in Adobe
Illustrator.
The evaluation team developed a “fuzzy” logic model with the intent of
expanding the accessibility and use of the evaluation (and program) and the
evaluation process to a greater diversity of stakeholders over a longer time.
Authors of the model refer to it as “fuzzy” because it embraces fluid and
approximate reasoning and varied context and assumptions with the aim of
improving the capacity of models of program theory to navigate nonlinearity,
feedback loops, adaptive agents, and other key concepts of complexity
integral to the life cycle of environmental programs and policies. Subsequent
integration of Web 2.0, graphic design and arts, and data visualization with
traditional logic models gives the evaluator the capacity to embed an
unlimited type and quantity of content into a web-based model of the
program (see Figure 8.10).
The PPSI required considerable research, conceptualization, design, and
planning completed prior to program implementation. It relies on evidence
and generates some hypotheses.
The model was created before the program was implemented. The model
represents the theory of the program as agreed upon by the evaluation
committee. After the model was created and when the PPSI had a better view
of how everything related, the model influenced evaluation questions and
performance measures as well as the effort and the views of the evaluation
committee, the PPSI, and pilot program as a whole.
Model creators project constant adaptation of the model. It has been in
constant flux since inception. The graphic has generally remained constant
after consensus of the evaluation committee, but it is constantly evolving as
the committee completes its work and that information is embedded in the
model.
As a web-based tool, there is the option to integrate social media to
encourage constant feedback and discussion. Currently, there is a dedicated
Facebook page that users can access from the website to leave comments and
feedback, ask questions, or start discussions.
At the time of the evaluation committee’s reporting (1 year after program
implementation), much of the data collected establishes baselines of cost,
volume, consumer behavior, and so forth. Some of these findings influenced
the model. For instance, larger fees resulted in widening the green arrows,
finding out where exactly disposed paint goes requires adding an arrow
where there was none, and identifying the companies contracted for
transportation and recycling requires new pop-up boxes and additional text to
describe them.

Figure 8.9 Paint Product Stewardship Initiative Concept

Source: Paint Product Stewardship Initiative, 2011.

Figure 8.10 Paint Product Stewardship Initiative Logic Model

Source: Paint Product Stewardship Initiative, 2011.

References
Matt Keene, Policy Office, U.S. EPA, and Chris Metzner, a graphic artist,
were deeply involved with the development of the PPSI models. They can
be reached via email at [email protected] and
[email protected], respectively.
Visit the live website: Retrieved December 22, 2011, from
http://www.PaintStewardship Program.com
Download required scripts for pop-up boxes: Retrieved December 22, 2011,
from http://flowplayer.org/tools/demos/overlay/index.html
Oregon Department of Environmental Quality, Paint Product Stewardship.
Retrieved December 22, 2011, from
http://www.deq.state.or.us/lq/sw/prodstewardship/paint.htm

IN SUMMARY

Logic models are a potent tool for many reasons and multiple functions. They
are robust communication platforms that can anchor a shared construction
that eventually serves strategy development, monitoring, evaluation, and
learning. These field profiles offer a big range of subject matter content and
use. Each was created in a process that reflected particular circumstances.
They vary considerably in display and frame problems, both implicit and
explicit. The preceding chapters suggest ways to both test and improve their
quality.

LEARNING RESOURCES

Reflection
 1. What features of logic models are most common in the field profiles
shown in this chapter? Why?
2. Which model is most like the one you might create? Why does it
resonate with your communication style or purpose?
3. Which model is most difficult to interpret? Can you name the reasons?
Are there changes you would make to simplify or clarify it?
4. Which model represents work that’s most likely to garner the intended
results?
5. Can you articulate assumptions for each model? How would you cite the
problem(s) each solves?
6. Consider contextual barriers and facilitators for each model. Try to name
some for each.

Exercises

1. Revisit Chapter 4 and consider quality principles for each model. How
does this influence your perception of the model’s potential to describe
work and associated results? Are there changes you would make?
2. Explain the purpose of a given model and its content. Then ask two
small groups to draw a model. Compare it to the figure shown. What
differences are there? Why? Any improvements?
3. Prepare an evaluation design for the ConAgra Foods Foundation (Profile
2). How do the models help or hinder? What questions does the process
raise?
4. Try to locate an evidence base for each of the models. How does your
discovery inform corrections or edits to the models?

References and Supplemental Readings

Texts
Krogerus, M., Tschappler, R., & Piening, J. (2008). The decision book: Fifty
models for strategic thinking. London: Kein & Aber.
Nissen, M. E. (2006). Harnessing knowledge dynamics: Principled
organizational knowing and learning. Hershey, PA: IRM Press.
Osterwalder, A. (2010). Business model generation: A handbook for
visionaries, game changers and challengers. Hoboken, NJ: Wiley.
Pugh, K. (2011). Sharing hidden know-how: How managers solve thorny
problems with the knowledge jam. San Francisco: Jossey-Bass.
Rumelt, R. (2011). Good strategy bad strategy: The difference and why it
matters. New York: Random House.
Sibbett, D. (2010). Visual meetings: How graphics, sticky notes and idea
mapping can transform group productivity. Hoboken, NJ: Wiley.

Journal Articles
Astbury, B., & Leeuw, F. L. (2010). Unpacking black boxes: Mechanisms
and theory building in evaluation. American Journal of Evaluation, 31(3),
363–381.
Brandon, P. R., Smith, N. L., & Hwalek, M. (2011). Aspects of successful
evaluation practice at an established private evaluation firm. American
Journal of Evaluation, 32(2), 295–307.
Garcia-Iriarte, E., Suarez-Balcazar, Y., Taylor-Ritzler, T., & Luna, M.
(2011). A catalyst-for-change approach to evaluation capacity building.
American Journal of Evaluation, 32(1), 168–182.
Kundin, D. M. (2010). A conceptual framework for how evaluators make
everyday practice decisions. American Journal of Evaluation, 31(3), 347–
362.
Piggott-Irvine, E. (2010). Confronting evaluation blindness: Evidence of
impact of action science-based feedback. American Journal of
Evaluation, 31(3), 314–35.
Saari, E., & Kallio, K. (2011). Developmental impact evaluation for
facilitating learning in innovation networks. American Journal of
Evaluation, 32(2), 227–245.
Skolits, G. J., Morrow, J. A., & Burr, E. M. (2009). Reconceptualizing
evaluator roles. American Journal of Evaluation, 30(3), 275–295.

Internet and Other Resources

Center for Civic Partnerships. (2007). Organizational learning: Tips, tools
 and resources. Retrieved December 21, 2011, from
http://www.civicpartnerships.org/toolsRes.htm
Foundation Center. (2011). Tools and resources for assessing social impact.
Retrieved December 21, 2011, from
http://trasi.foundationcenter.org/about.php
Human Resource Development Council. (n.d.). Organizational learning
strategies. Retrieved December 22, 2011, from
http://www.humtech.com/opm/grtl/ols/ols2.cfm
Problem-solving Techniques. (2011). Problem-solving techniques portal.
Retrieved December 22, 2011, from http://www.problem-solving-
techniques.com/Problem-Solving-Tools.html
 Name Index

Alabama asthma program profile, 150–153

American College of Sports Medicine, 54
American Recovery and Reinvestment Act, 126
America’s Second Harvest (Feeding America), 141
Annie E. Casey Foundation, 124

Bennett, Claude, 6, 119–120

Berkowitz, Gale, xvi

Center on School, Family and Community Partnerships, 130

Centers for Disease Control and Prevention, 132–134
Chen, Huey, 5
CLIPs (Communities of Learning, Inquiry, and Practice), 103–105
Coffman, Julia, xvi
Communities of Learning, Inquiry, and Practice (CLIPs), 103–105
Community Leadership Academy (CLA), 6–13, 19, 29, 56, 70–81
Nonlinear Theory of Change Model, 29
Outcome Indicators, 79
Process Indicators, 77
Program Evaluation Model, 10
Program Logic Model, 8
Theory of Change Model, 7
ConAgra Foods Foundation, 139–146
Congressional Hunger Center, 142

Decent Work Country Programme in Kyrgyzstan, 146–150

DiClemente, C. C., 119

Eco Hub example, 92–94

Epstein, Joyce, xvi, 130
Feeding America (America’s Second Harvest), 141
Food Research Action Center, 142
Fullan, Michael, 5

Government Performance and Results Act, 1993 (GPRA), 123

Greenleaf, Robert, 78

Harvard Family Research Project, 130

Haven House, 155–156
Hodges, Debra, xvii

Improving Head Start for School Readiness Act, 126

Injury Control Research Center, 132–134
International Labour Organization, 146–150
Itzak, A., 119

Jones, Adrian, xvi

Juvenile Accountability Block Grant (JABG) Logic Model, 124

Keene, Matt, xvii

Kingdon, John, 96
Kress, Howard, xvi
Kyrgyzstan Decent Work Country Programme, 146–150
Kuzmin, Alexey, xvii

Logic Model Development Guide, The, 6

Management by Objectives, 91
Mayeaux, Angela, 155, xvii
Measuring Program Outcomes, 6
Metzner, Chris, xvii
Michigan, 155–156
Morariu, Johanna, xvii

National Asthma Control Program, 151

National Network of Partnership Schools, 130
New York Healthy Weight Model, 27, 89, 106–108, 118

Office of Juvenile Justice and Delinquency Prevention, 123

Oregon, 157–160

Packard Preschool, 89, 96–102, 118

Paint Product Stewardship Initiative, 90, 157–160
Palchinsky’s Principles, xii
Parsons, Beverly A., xvi
Pathways Mapping Initiative, 124–126
Ready for School and Succeeding at Third Grade Theory of Change, 125
Post Carbon Institute profile, 153–154
Prochaska, J., 119, 128
Product Stewardship Initiative model, 157
Program Assessment Rating Tool, 123
Program Evaluation Standards, 49
Project on Effective Interventions, 124

Reed, Kori, xvii

Reich, Kathleen, xvi
Rockey, Sherry, xvi
Rockwell, Kay, 119–120
Rosenstock, I., 119
Russon, Craig, xvii

Salisbury, Lois, xvi

School, Ready and Succeeding at Third Grade map, 124, 125
Schorr, Lisbeth, xvi
Scriven, Michael, 66
Seattle Works, 138–140
Servant Leadership, 78
Share Our Strength, 141
Smart, Dawn, xvii
Smith, Tara, xvii
Stretcher, V., 119
Texas Early Learning Council, 124, 126, 127
Thraves, Tes, xvi

United Way of America, 6

U.S. Agency for International Development, 6

Wayne Food Initiative, 94–96

Weiss, Carol, 5
W. K. Kellogg Foundation, 6, 124

Yee, Sue Lin, xvi

 Subject Index

Accountability in corporate giving, 144, 146

Action profiles, 137–162
about strategy, evaluation, and learning, 137–138
Alabama Asthma Coalition, 150–153
ConAgra Foods, corporate giving, 139–146
Haven House, 155-156
Kyrgyzstan Decent Work Country Programme, 146–150
Michigan, sheltering families, 155–156
Oregon, environmental leadership with paint, 157–160
Post Carbon Institute, resilient communities, 153–154
Seattle Works, civic engagement, 138–140
Action steps
in creating program logic model, 43
in creating theory of change logic model, 21–22
Activities
defined for program logic models, 36
models begin with results, 12
from strategy to, 40–41
Alabama asthma program profile, 148–153
Alignment and indicators, 80–81
Alternative approaches to modeling, 90–91
American College of Sports Medicine, 54
American Recovery and Reinvestment Act, 126
America’s Second Harvest (Feeding America), 141
Annie E. Casey Foundation, 124
Archetypes, 117–136
blank page challenge, 117–118
definition of, 117–118
examples of, 121–134. See also Examples of archetypes
learning and, 118–121
Assumptions in logic models
for program, 35, 37
 for theory of change, 21, 22, 24
Asthma program in Alabama, 150–153

Behavior change example, 126, 128–129

Benchmarking in theory of change logic models, 26
Blank page challenge, 117–118
Blind spots, 49, 82
Building logic models. See Creation of program logic models; Creation of
theory of change logic models
Business interests in corporate giving, 141–142

Causal loop diagrams, 90–91

Center on School, Family and Community Partnerships, 130
Centers for Disease Control and Prevention, 132–134
National Asthma Control Program, 151
Change recipes, 119–120
Childhood hunger action profile, 139–146
Civic engagement profile, Seattle Works, 138–140
CLA. See Community Leadership Academy
CLIPs (Communities of Learning, Inquiry, and Practice), 103–105
Collaborative Learning, Inquiry, and Practice example, 102–105
Communications example, human behavior change, 126, 128–129
Communities of Learning, Inquiry, and Practice (CLIPs), 103–105
Community Leadership Academy (CLA), 6–13, 19
in evaluation design, 70–74
in evaluation design, alignment of indicators, 80–81
nonlinear theory of change model, 29
outcome indicators, 75, 78, 79
process indicators, 75–77
program logic model mark up, 56
Complexity and meaning, 89
Compromise, 51
Computer software applications for model creation, 45
ConAgra Foods Foundation, corporate giving profile, 139–146
Congressional Hunger Center, 142
Consumers. See Evaluation consumers
Context
 challenges in improving program logic models, 49–52
features of, in theory of change logic models, 27
Corporate giving profile, ConAgra Foods, 139–146
Corporate social responsibility (CSR), 141, 146
Creation of program logic models, 38–45
action steps for program logic model, 43
activities, 40–41
group process, guiding, 44–45
health improvement example, 38–40
from strategy to activities, 40–41
Creation of theory of change logic models, 16–22
action steps in creating, 21–22
assumptions and knowledge in, 21
the big picture, 18
evidence-based, 17–18
plausibility, 17–18
preferences and styles, 17
realistic models, 20
results, 18–20
strategies, 18–20
Critical thinking, 120–121
Culture for improving models, 52, 82

Data collection, 78–79

Decent Work Country Programme in Kyrgyzstan, 146–150
Decisions, making better, 59–60
Desired results, 21
See also Results
Direct influence, 90, 108
Display and meaning, 87–116
about selected examples, 91
alternative approaches, 90–91
complexity and meaning, 89
examples. See Examples of display and meaning
graphic display, 88–89
model benefits, 90
subject matter content, 89–90
 variation and learning, 88–91
Doing the right work, 12, 29–30, 59–60, 72
Dose influence on effectiveness, 37, 76
Double loop learning, 70, 71
Draw step in modeling, 52–53

Eco Hub example, 92–94

Education readiness and success example, 124–126
See also School entries
Effectiveness
in benefits of logic models, 3–4
dose influence on, 37, 76
evaluation design and, 66–67, 70
and logic models, 12–13
and modeling program logic models, 48–49
Epstein, Joyce, 130
Evaluation consumers, 65
adding value, 67–68, 69, 78
Evaluation design, 64–86
as archetypes, 118
contributions of logic models to, 64–66
defined, 66
design basics, 67–70
for effectiveness, 66–67
formative evaluation, 67
indicators, 74–80
indicators and alignment, 80–81
learning, two kinds of, 70
performance standards, 81
and program logic models, 9–11
quality questions, 81–82
questions, 9, 68, 70–73, 81–82
summative evaluation, 66–67
Evaluation system development, 108–113
Evidence-based
in archetypes, 119
in theory of change logic models, 17–18
Examples
community leadership. See Community Leadership Academy (CLA)
concepts and terms explained, 6–13
health improvement program example, 19–20, 26–27, 38–40, 58
of program logic models, 6–9, 92, 95, 96, 107, 108, 110
of theory of change logic models, 6, 96, 108, 109
Examples of archetypes, 121–134
about selected examples, 121
communications, human behavior change, 126, 128–129
education readiness and success, Pathways Mapping Initiative, 124–126
federal block grants, juvenile justice and delinquency prevention,
121–124
public health research, Injury Control Research Center, 132–134
school improvement, family and parent engagement, 130–131
Examples of display and meaning, 91–113
about selected examples, 91
Collaborative Learning, Inquiry, and Practice, 102–105
Eco Hub, 92–94
Independent Sector evaluation system development, 108–113
New York Healthy Weight Model, 27, 89, 106–108, 118
Packard Preschool logic model, 89, 96–102, 118
Wayne Food Initiative, 94–96
Exercise strategy, 41–42
External review, 52, 55

Family and parent engagement example, 130–131

Feasibility
quality model features, 59
standard for model quality in evaluation design, 82–83
testing to improve model quality, 48–49
Federal block grants example, juvenile justice and delinquency prevention,
121–124
Feeding America (America’s Second Harvest), 141
FIT principles
in evaluation design, 76, 83
testing model quality, 54, 55, 57
Food Research Action Center, 142
Formative evaluation, 67

Government Performance and Results Act, 1993 (GPRA), 123

GPRA (Government Performance and Results Act, 1993), 123
Grant proposals, archetypes as, 123
Graphic display, 88–89
Group process in logic models
for program, 44–45
for theory of change, 26–27

Harvard Family Research Project, 130

Haven House action profile, 155–156
Health
health improvement program example, 19–20, 26–27, 38–40, 58
New York Healthy Weight Model, 27, 89, 106–108
program logic model mark up, 58
Housing for families in Michigan, action profile, 155–156
Human behavior change example, 126, 128–129
Hypotheses, 17

Idea maps, 16
If-then sequence, 7
Impact
defined for program logic model, 7
models begin with results, 10
in program logic models, 37, 43
Improvement of program logic models, 48–62
context challenges, 49–52
decisions to do the right work, 59–60
modeling and effectiveness, 48–49
quality model features, 59
quality questions, 58
quality techniques in modeling, 52–53
testing model quality with SMART and FIT, 53–54
using a mark up, 54–58
Improvement of theory of change logic models
 benchmarking, 26
doing the right work, 29–30
group process, 26–27
knowledge and assumptions, 24
multiple perspectives, 23–24
nonlinear theory of change logic models, 28–29
promising practices, 26
questions in reviewing, 27, 30
toggling, 25
Improving Head Start for School Readiness Act, 126
Independent Sector evaluation system development example, 108–113
Indicator development, 74
Indicators, 68, 73–80
and alignment, 80–81
outcome indicators, 74, 78, 79
process indicators, 74–77
proxy indicators, 74
Indirect influence, 90, 108
Influence, 90, 108
Information needs, 68, 69
Initiatives, 40
Injury Control Research Center example, 132–134
Inputs, defined for program logic model, 6
Intermediate-term outcomes, 36
International Labour Organization, 146–150

Juvenile Accountability Block Grant (JABG) Logic Model, justice and

delinquency prevention example, 121–124

Knowledge, in theory of change logic models, 21, 24

Kyrgyzstan Decent Work Country Programme profile, 146–150

Large-scale programs, 40
Learning
and archetypes, 118–121
Communities of Learning, Inquiry, and Practice (CLIPs), 102–105
 in evaluation design, single loop and double loop, 70, 71
and variation, 88–91
Legislation
American Recovery and Reinvestment Act, 126
Government Performance and Results Act, 1993 (GPRA), 123
Improving Head Start for School Readiness Act, 126
Legislative mark up, 55
Logical frameworks (logframes), 90–91
Logic in models, 50
Logic map, 92
Logic Model Development Guide, The, 6
Logic models
about, xii–xiii, 2–15
adding value in evaluation design, 68–70, 80
beginning with results, 10–12
benefits of, 3–4, 90
compared to logical frameworks, 91
computer software for creation of, 45
definition of, 4
effectiveness and, 3–4, 12–13, 48–49
for evaluation. See Evaluation design
examples. See Examples
historical background of, 5–6
improvement. See Improvement of program logic models; Improvement
of theory of change logic models
limitations of, 10
program logic models. See Program logic models
relationship of program and theory of change models, 5, 34–37
synonyms for, 4
theory of change. See Theory of change logic models
uses of, 4, 5
Long-term outcomes, 36

Management
connecting with measurement in evaluation design, 64–66
in corporate giving, 144
Management by Objectives, 91
Marketing, social, 126, 128
Marketing tool, models as, 51
Mark up, in program logic modeling, 54–58
Mayeaux, Angie, 155
Meaning. See Display and meaning
Measurement
connecting management and, 64–66
in corporate giving, 144
Measuring Program Outcomes, 6
Mental maps/mapping, 16, 124
Michigan, sheltering families profile, 155–156
Mistakes, 52
Modeling, 48–62
basic concepts, 3
context challenges in, 49–52
draw and test steps, 52–53
and effectiveness, 48–49
limitations of, 10
quality techniques, 52–60
Multiyear change efforts, 40
Myths, 49, 82

National Network of Partnership Schools, 130

Nested leadership, 130
New York Healthy Weight Model example, 27, 89, 106–108, 118
Nonlinear logic models
program, 37
theory of change, 28–29

Objectives pathway, 123

Office of Juvenile Justice and Delinquency Prevention, 123
Oregon, environmental leadership profile, 157–160
Outcome indicators, 74, 78, 79
Outcomes
defined for program logic model, 7, 36
models begin with results, 10, 12
in program logic models, 43
 time spans for, 36, 76–78
Outputs, defined for program logic model, 7, 36

Packard Preschool logic model example, 89, 96–102, 118

Paint Product Stewardship Initiative model, 90, 157–160
Palchinsky’s Principles, xii
Pathways Mapping Initiative example, 124–126
Performance standards
in evaluation design, 81
indicators of program performance, 74
Persuasion, 51
Planned work of program logic model, 37
Plausibility
in modeling and effectiveness, 49
quality model features, 59
and realistic models, 20
standard for model quality in evaluation design, 82–83
in theory of change logic models, 17–18
PLM. See Program logic models
Policy stream convergence theory, 96–97, 118
Politics of context, 51
Post Carbon Institute profile, 153–154
Preferences, in theory of change logic models, 17
Preschool logic model example, 89, 96–102, 118
Process indicators, 74–77
Product Stewardship Initiative model, 157
Profiles. See Action profiles
Program Assessment Rating Tool, 123
Program Evaluation Standards, 49
Program logic models
assumptions, 35, 37
creating, 38–45. See also Creation of program logic models
definition of, 5
dose, 37, 76
effectiveness and, 12–13, 48–49
elements of, 35–37
evaluation design and, 8–10. See also Evaluation design
 examples of, 6–8, 92, 95, 96, 107, 108, 110
improving. See Improvement of program logic models
modeling. See Modeling
nonlinear, 37
relationship to theory of change models, 5, 34–37
strategies, 34
Project on Effective Interventions, 124
Promising practices, in theory of change logic models, 26
Proofs, 17
Proxy indicators, 74
Public health research example, 132–134
Public policy agenda example, 96

Quality evaluation designs, 81–83

Quality techniques, 52–60
better decisions, 59–60
characteristics of quality model, 59
in modeling, 52–53
questions for program logic models, 58
testing model quality with SMART and FIT, 53–54
using a mark up, 54–58
See also Improvement of program logic models; Improvement of theory
of change logic models
Questions
for evaluation design, 8, 68, 70–73, 81–82
for program logic models, 58
for theory of change logic models, 27, 30

Realistic models, 20
Recipes, archetypes as, 119–120
Resilient communities action profile, 153–154
Resources, defined for program logic model, 6, 36
Results
choices are required, 81–83
models begin with, 10–12
in program logic models, 34, 43
in theory of change logic models, 18–20, 21–22, 26
 See also Outcomes
Right work, doing, 12, 29–30, 59–60, 72

Scale of effort, 50
School-improvement efforts, 26
School improvement example, family and parent engagement, 130–131
School, Ready and Succeeding at Third Grade map, 124, 125
Seattle Works, civic engagement profile, 138–140
Servant Leadership, 78
Share Our Strength, 141
Short-term outcomes, 36
Single loop learning, 70
SMART principles
in evaluation design, 78, 83
testing model quality, 53–54, 55, 58
Social interests, in corporate giving, 141–142
Social marketing, 126, 128
Software applications for model creation, 45
Specificity, 50–51
Stakeholders, defined, 65
Strands, 19, 40
Strategy
models begin with results, 12
in program logic models, 34, 38, 40–41, 42, 43
in theory of change logic models, 18–20, 22, 26
Styles, in theory of change logic models, 17
Subject matter content, 89–90
Summative evaluation, 66–67
Sustainable food system example, 94

Targeting outcomes of programs (TOP), 119–120

Templates. See Archetypes
Testing model quality, 53–54
FIT principles, 54, 55, 58
SMART principles, 53–54, 55, 58
test step in modeling, 52–53
Texas Early Learning Council theory of change model, 124, 126, 127
Theory of change (TOC) logic models
building, 16–22. See also Creation of theory of change logic models
definition of, 5
effectiveness and, 12–13
in evaluation. See Evaluation design
examples of, 6, 96, 108, 109
improving, 23–30. See also Improvement of theory of change logic
models
nonlinearity of, 28–29
relationship to program logic models, 5, 34–37
Time, featured on models, 90
Time spans for outcomes, 36, 76–78
TOC. See Theory of change logic models
Toggling, 25
TOP (targeting outcomes of programs), 119–120
Transtheoretical model (TTM), 128
Treatment trickle, 54

Value added
by evaluation consumers, 67–68, 69, 78
by logic models, 68–70, 80
Value of archetypes, 120
Variation, 88–91

Wayne Food Initiative example, 94–96

W. K. Kellogg Foundation, 124