Research Lecture2 Research Approaches
Research Lecture2 Research Approaches
Research Designs/Approaches
Type Purpose Time frame Degree of control
High
Examples
Experimental
Test for cause/ current effect relationships Test for cause/ Current or past effect relationships without full control
Quasiexperimental
Comparing two types of treatments for anxiety. Moderate Gender to high differences in visual/spatial abilities
Research Designs/Approaches
Type Purpose Time frame
Current (crosssectional) or past Past & current
Degree of control
Low to medium
Examples
Examine relationship between two variables Examine the effect of past event on current functioning.
Low to medium
Relationship between studying style and grade point average. Relationship between history of child abuse & depression.
Research Designs/Approaches
Type Purpose Time frame Degree of control
Low to moderate
Examples
Examine relat. Future betw. 2 var. predictive where 1 is measured later. Examine Future change in a var. over time in overlapping groups.
Low to moderate
Relat. betw. history of depression & development of cancer. How motherchild negativity changed over adolescence.
Research Designs/Approaches
Type Purpose Time frame
Current
Degree of control
None or low
Examples
Survey
Qualitative
Assess opinions or characteristics that exist at a given time. Discover potential relationships; descriptive.
Past or current
None or Low
Describes a particular situation or phenomenon. Hypothesis generating Can describe effect of implementing actions based on experimental research and help refine the implementation of these actions.
Correlational Design
Measure two variables
Study methods and grade-point average
Correlational Studies
cannot be experimentally manipulated (e.g., IQ and occupational status). Disadvantage is that it cannot determine causality. Third variable may account for the association. Directionality unclear
Correlation Scatterplot
Correlation Scatterplot
Correlational Designs
Variable of interest is not subject to direct manipulation but must be chosen after the fact. E.g., Define two groups of people according to a certain characteristic (e.g., history of trauma) and measure how they respond in terms of anxiety to a certain stimulus (e.g., watching violent film). Limitation self-selection bias, cohort effects may explain the effect.
2nd BP Screen
Personality Study
Personality Study
46.2 (9.2)
135.1/ 93.9 (9.2/5.1)
46.2 (8.2)
118.7/ 76.3 (11.5/5.5)
46.4 (8.3)
135.8/ 93.8 (8.2/3.4)
45.8 (8.0)
118.5/ 75.7 (10.3/4.8)
0.79 (8.5)
-3.0 (9.4)
-2.0 (9.4)
-2.6 (8.2)
Awareness of hypertension status confounds assessment of the association between personality characteristics and hypertension.
Due to hypertension labeling effect; or Due to self-selection bias
Advantage is that it is an efficient way to identify possible group differences because you can study them at one point in time. Disadvantage is that you cannot rule out cohort effects.
Longitudinal Design
Gathers data on a factor (e.,g. confidence) over time. Advantage is that you can see the time course of the development or change in the variables
Confidence increasing with age.
Cohort-Sequential Design
time.
Advantage very efficient and reduces some of the biases in the cross-sectional design since you can see the evolution of change over time. Disadvantage cannot rule out cohort bias or the problem of the unidentified third variable accounting for the change.
Naturalistic Observation
Aims to unobtrusively observe behaviour in the natural setting. Observing in the natural setting enables one to minimize or eliminate the problem of artificial behaviour in response to being studied (i.e., reactivity effects). One variation is being a participant observer (e.g., undercover agent).
Naturalistic Observation
Advantages
Observe the natural phenomena (not artificial)
Disadvantages
Observer bias Reactivity in subjects Ethics
Quantitative approach to integrate and describe results across a range of independent studies. Enables you to combine the probability (p) value for statistical tests over a number of studies. Enables you to determine the effect size of the independent variable (e.g., treatment group) across studies.
Survey Research
Collecting standarized information from people using an interview or self-report format. Typically survey knowledge or opinions. To standarized the information one uses a questionnaire with set questions. Ideally the questionnaire has been validated. Representativeness of the sample is very important.
Survey Methods
Interviews
Advantage - Comprehensive, ensure participant
understands the question, minimizes missing data, enables clarification of unclear responses
Disadvantage expensive, people more like to
refuse participation, can be risky for interviewer, interviewer may bias the responses.
Face-to-face interviews
Expensive and time-consuming
Telephone interviews
Need to use random-digit dialing to reach both
Mail
Return rate is usually low (20-30%).
Types of Questions
Open-ended
E.g., Can you tell me about your typical
Close-ended
E.g., How do you typically meet someone to date? Introduced by someone Social event In university class or place of work At a bar Through sports or other athletic events
Sampling
Population is everyone in your population of interest. Sample is some proportion of the population. Haphazard sampling convenience sample Random sampling
There is always some degree of sampling error.
Qualitative Methods
of the data Multimethod can use interviews, photographs, natural observation, archives, etc. It is typically conducted in persons natural environment.
Qualitative Methods
Transcripts
Experimental Designs
Examines differences between experimentally manipulated groups or variables (e.g., one group gets a certain drug and the other gets a placebo). At minimum, experimental (independent) variable has two levels (e.g., drug vs. placebo).
Advantage is that you can determine causality. Disadvantage is cost and many variables cannot
Method of Agreement
instances of the event Problem is you may inadvertently overlook a significant variable.
Method of Difference
Identify the different effects produced by two
situations that are alike in all ways but one. Fairly robust and strong method.
instances of the event Form hypothesis based on observations Test hypothesis using method of difference
than two situations that are alike in all ways but one. E.g., Compare two active drugs to a placebo
Experimental Design
Because it is so difficult with human behaviour to demonstrate causation unequivocally, some argue that a theory or prediction can only achieve the status of not yet disconfirmed (Popper, 1968). Our scientific efforts are directed at finding the causal factors rather than the cause per se.
identify where mistakes are likely. Take the steps necessary to minimize error.
Verbal statements
Responses to questionnaires Physiological responses.
Eliminate all factors that could influence the outcome other than the factor being manipulated. Control is needed to infer causation. All conditions are kept constant except one; the manipulated variable. The variable of interest is varied in order to test its effect.
Experimental Method
Advantages
Strength with which causal relationships can be
inferred. Ability to manipulate one or more variables. Proven to be a very useful and robust scientific method (i.e., withstood the test of time).
Experimental Method
Disadvantages
Tight controls often produce artificial
conditions that could limit the generalizability of the findings (i.e., internal vs. external validity trade-off). Time consuming. Expensive. Human behaviour is very complex and cannot be fully studied using experimental methods.
because participant has more familiarity with the measure and so answers more truthfully.
(e.g., society because more affluent generally) and this influences the participants responses.
Maturation
Natural developments in the participant account
Instrumentation
If pre and post tests are not equivalent in all
ways (e.g., difficulty, readability) then differences observed may be due to instrumentation differences rather than due to your experimental manipulation.
Subject problems
Selection bias (e.g., participation rate). Attrition (e.g., only motivated subjects stay in
the experiment).
Subject variables
Selection bias.
Attrition bias
Artificial conditions
E.g., In order to measure a subjects blood
pressure in response to a well-fined stressor you bring him/her into the laboratory but his/her response in the laboratory may not reflect how his/her blood pressure would really respond under stress in his natural environment.
Let me know
If there are any topics from todays lecture that need fuller explanations. Anything you particularly liked about the lecture (todays or others as we go along). Anything you particularly disliked about the lecture (todays or others as we go along).