BRM Unit-3
BRM Unit-3
Non-parametric tests are statistical methods used to analyze data when the assumptions of
parametric tests are not met, particularly when the data is not normally distributed or when
the sample size is small. Unlike parametric tests, non-parametric tests do not rely on specific
distributional assumptions about the population from which the sample is drawn. These tests
are also known as distribution-free tests.
The Wilcoxon Rank Sum Test, also known as the Mann-Whitney U test, is a non-parametric
statistical test used to determine if there is a difference between two independent and
unpaired groups. It is used when the assumptions of a parametric test, such as the t-test,
cannot be met due to non-normally distributed data or when the data is measured on an
ordinal scale.
Purpose:
The Wilcoxon Rank Sum Test is employed to assess whether there is a statistically significant
difference between the distributions of two independent groups.
Assumptions:
1. Independence: The observations in one group are independent of the observations in
the other group.
2. Random Sampling: The data are collected through a random sampling process.
3. Ordinal Data: The data should be at least ordinal, meaning the values have a
meaningful order.
Hypotheses:
Procedure:
1. Ranking: Combine the data from both groups, rank them from smallest to largest,
and assign ranks.
2. Sum of Ranks: Calculate the sum of ranks for each group.
3. Test Statistic (U): The test statistic U is calculated based on the smaller sum of ranks.
It can be compared to critical values from a statistical table or used to calculate a p-
value.
4. Decision Rule: If the p-value is less than the chosen significance level (commonly
0.05), you reject the null hypothesis in favor of the alternative hypothesis.
Interpretation:
A significant result indicates that there is a difference between the two groups.
Example:
Suppose you want to compare the test scores of two different teaching methods. You collect
data from two independent groups, apply the Wilcoxon Rank Sum Test, and find a p-value
less than 0.05. In this case, you would reject the null hypothesis, suggesting that there is a
significant difference in test scores between the two teaching methods.
The Wilcoxon Rank Sum Test is a robust alternative to the t-test for non-normally distributed
data or when the assumptions of parametric tests cannot be met. It is widely used in various
fields, especially in situations where the data is measured on an ordinal scale or when sample
sizes are small.
Kruskal-Wallis test
The Kruskal-Wallis test is a non-parametric statistical test used to determine whether there
are any statistically significant differences between the medians of three or more independent
(unrelated) groups. It is an extension of the Mann-Whitney U test, which is used to compare
two independent groups. The Kruskal-Wallis test is employed when the assumptions of
parametric tests, such as one-way ANOVA, cannot be met.
Purpose:
The Kruskal-Wallis test is used to investigate whether there are differences in the distribution
of a continuous dependent variable across three or more independent groups.
Assumptions:
Hypotheses:
Procedure:
1. Ranking: Combine the data from all groups, rank them from smallest to largest, and
assign ranks.
2. Sum of Ranks: Calculate the sum of ranks for each group.
3. Test Statistic (H or χ²): The test statistic is based on the sums of ranks and follows a
chi-square distribution. This statistic is used to determine whether the observed
differences in ranks are likely due to chance.
4. Critical Value or P-value: Compare the calculated test statistic to critical values
from a chi-square distribution table or use it to calculate a p-value. If the p-value is
less than the chosen significance level (commonly 0.05), you reject the null
hypothesis.
Interpretation:
Example:
Suppose you are comparing the exam scores of students from three different teaching
methods. The Kruskal-Wallis test can help determine whether there is a significant difference
in exam scores among the three teaching methods.
Friedman Test
The Friedman test is a non-parametric statistical test used to detect differences in treatments
across multiple related groups. It is an extension of the Wilcoxon signed-rank test and is
suitable for situations where the data is measured on an ordinal scale and the assumptions of
parametric tests (such as repeated measures ANOVA) are not met.
Purpose:
The Friedman test is employed to determine whether there are any statistically significant
differences in the distribution of a continuous dependent variable measured under different
conditions or at different time points across related groups.
Assumptions:
1. Related Groups: The observations are collected from the same individuals or related
entities.
2. Random Sampling: The data should be collected through a random sampling
process.
3. Ordinal Data: The dependent variable should be at least ordinal, meaning the values
have a meaningful order.
4. Independence: Observations within each group are assumed to be independent.
Hypotheses:
Procedure:
1. Ranking: For each individual or entity, rank the measurements from each condition
or time point.
2. Calculate Ranks: Calculate the average rank for each condition or time point across
all individuals or entities.
3. Friedman Test Statistic (χ²): The test statistic is based on the differences between
the mean ranks and follows a chi-square distribution. This statistic is used to
determine whether the observed differences in ranks are likely due to chance.
4. Critical Value or P-value: Compare the calculated test statistic to critical values
from a chi-square distribution table or use it to calculate a p-value. If the p-value is
less than the chosen significance level (commonly 0.05), you reject the null
hypothesis.
Interpretation:
Consider a study measuring the performance of individuals on a task under three different
conditions. The Friedman test can help determine whether there is a significant difference in
performance across these conditions.
If the Friedman test indicates a significant difference, further post-hoc tests or pairwise
comparisons may be conducted to identify which conditions or time points differ from each
other.
Introduction to Research
The need for research is significant across various fields and disciplines, driven by the
following key factors:
1. Advancement of Knowledge:
o Research is essential for expanding our understanding of the world,
uncovering new information, and pushing the boundaries of existing
knowledge. It contributes to the accumulation of facts, theories, and concepts
that form the basis of academic and intellectual progress.
2. Problem Solving:
o Research provides a systematic approach to understanding and addressing
problems. It plays a crucial role in identifying solutions to challenges faced by
societies, industries, and individuals.
3. Innovation and Creativity:
o Research fosters innovation by encouraging exploration and experimentation.
It is a catalyst for new ideas, technologies, and approaches, driving progress in
various fields such as science, technology, medicine, and the arts.
4. Evidence-Based Decision Making:
o In policy-making, business strategy, and various professional domains,
research provides the evidence needed for informed decision-making. It
ensures that decisions are based on data, analysis, and a thorough
understanding of the subject matter.
5. Continuous Improvement:
o Businesses and organizations use research to evaluate and improve their
products, services, processes, and strategies. This iterative process of
assessment and refinement is crucial for staying competitive and adapting to
changing environments.
6. Academic and Educational Progress:
o Research is fundamental to the academic mission. It contributes to the
development of educational curricula, supports teaching and learning
processes, and provides a foundation for students to engage with and
contribute to their fields of study.
7. Social and Cultural Development:
o Research helps to document and understand societal changes, cultural
dynamics, and human behavior. It contributes to the development of social
sciences, humanities, and other disciplines that explore the complexities of
human societies.
8. Validation of Existing Knowledge:
o Research plays a role in validating or challenging existing theories and
concepts. It helps ensure the reliability and accuracy of information,
promoting a more robust foundation for understanding the world.
9. Health and Medicine:
o In healthcare, research is crucial for discovering new treatments, improving
diagnostic methods, and understanding the underlying causes of diseases. It
contributes to medical advancements that enhance patient care and overall
public health.
10. Technological Advancements:
o Research in science and engineering leads to technological breakthroughs,
driving advancements in fields such as electronics, materials science,
information technology, and more.
11. Global Challenges:
o Research is essential for addressing global challenges such as climate change,
poverty, and public health crises. It provides the knowledge and solutions
needed to tackle complex issues that affect societies on a global scale.
12. Quality Improvement:
o In various industries, research is conducted to enhance the quality of products
and services. It involves studying consumer preferences, market trends, and
technological advancements to deliver better outcomes.
Customer journey maps visually represent the entire end-to-end experience that users
have with a product or service. This technique helps designers identify touchpoints,
pain points, and opportunities to improve the overall journey.
3. Prototyping:
4. User Testing:
User testing involves gathering feedback from actual users to evaluate the usability
and overall experience of a product or service. This iterative process helps designers
refine and improve their designs based on real user insights.
5. Storytelling:
Storytelling is used to create narratives that guide users through an experience. This
technique helps in conveying the brand message, creating emotional connections, and
ensuring a cohesive and memorable user journey.
6. Service Blueprints:
Service blueprints outline the entire service delivery process, including customer
interactions and behind-the-scenes operations. This technique helps in visualizing and
optimizing service experiences.
7. Empathy Mapping:
Empathy mapping involves understanding and visualizing the thoughts, feelings, and
behaviors of users. This technique helps designers empathize with users, leading to
more user-centric and emotionally resonant designs.
8. A/B Testing:
A/B testing (or split testing) is a quantitative technique that involves comparing two
versions of a design to determine which one performs better. It is often used to
optimize user interfaces and experiences.
9. Gamification:
Gamification involves incorporating game-like elements, such as rewards, challenges,
and competition, into non-game contexts. This technique is used to enhance
engagement and create more enjoyable experiences.
Sensory design considers how different sensory elements (sight, sound, touch, smell,
taste) contribute to the overall experience. This approach is essential in designing
environments and products that engage multiple senses.
Responsive design ensures that digital interfaces adapt to different devices and screen
sizes. This technique is crucial for providing a consistent and user-friendly experience
across various platforms.
Cultural probes involve providing participants with tools (such as cameras, journals,
or activities) to gather insights into their daily lives and experiences. This qualitative
method helps designers understand cultural nuances and user contexts.
Plagiarism
Plagiarism is the act of using someone else's ideas, words, or work without proper attribution
or permission, presenting it as one's own. It is considered a serious ethical violation in
academia, professional settings, and creative industries. Plagiarism undermines the principles
of intellectual honesty, originality, and the fair acknowledgment of contributors to knowledge
or creative endeavors.
Forms of Plagiarism:
1. Direct Copying:
o Replicating someone else's work, whether it's a sentence, paragraph, or entire
document, without proper citation.
2. Paraphrasing Without Attribution:
o Rewriting someone else's ideas or text in different words without providing
proper acknowledgment.
3. Self-Plagiarism:
o Submitting one's own work that has been previously published or submitted
for another purpose without proper citation or permission.
4. Uncited Quotations:
o Using quotes from sources without placing them in quotation marks or
providing the necessary citation.
5. Inadequate Citation:
o Failing to credit sources properly, either by not citing them at all or by
providing incomplete or inaccurate citations.
6. Collaboration Violations:
o Presenting collaborative work as individual work without acknowledging the
contributions of others.
Consequences of Plagiarism:
1. Academic Penalties:
o In academic settings, plagiarism can result in severe consequences, including
failing a course, academic probation, or expulsion.
2. Legal Ramifications:
o Plagiarism can lead to legal consequences, especially in cases of copyright
infringement. Authors and creators have legal rights to their work.
3. Professional Repercussions:
o In professional settings, plagiarism can damage one's reputation, career, and
credibility. It is considered a breach of professional ethics.
4. Educational and Career Impact:
o Instances of plagiarism can have long-term effects on educational and career
opportunities. Employers and academic institutions value integrity and may
view plagiarism as a serious offense.
5. Loss of Trust:
o Plagiarism erodes trust between individuals, whether it's among peers,
colleagues, or mentors. It undermines the principles of honesty and integrity.
Graphs
Histogram
Components of a Histogram:
1. Bins or Intervals:
o The range of values in the dataset is divided into intervals or bins. These bins
represent the classes or ranges of values that the data is grouped into.
2. Frequency:
o The frequency of each bin corresponds to the number of data points falling
within that specific range. It represents how often values occur within each
interval.
3. Bars:
o The histogram is constructed using bars, where each bar represents a bin, and
the height of the bar corresponds to the frequency of data points in that bin.
The bars are typically adjacent, and there is no gap between them.
1. Data Collection:
o Gather the dataset for which you want to create a histogram. Ensure that the
data is continuous, and identify the range of values.
2. Choose the Number of Bins:
o Decide on the number of intervals or bins you want to use. The choice of the
number of bins can impact the appearance of the histogram, and there are
various methods for selecting an appropriate number.
3. Calculate Bin Width:
o Determine the width of each bin by dividing the range of values by the number
of bins. The bin width is essential for correctly representing the distribution.
4. Create Frequency Distribution:
o Count the number of data points falling into each bin and record the
frequencies.
5. Construct the Histogram:
o Draw the x-axis (horizontal axis) to represent the variable's values and the y-
axis (vertical axis) to represent the frequencies. Use the bins and frequencies
to construct the bars. The bars should be of equal width and adjacent.
6. Label Axes and Title:
o Label the x-axis and y-axis with the variable name and frequency,
respectively. Include a title that provides information about the dataset or the
purpose of the histogram.
Interpretation of Histograms:
Central Tendency: The center of the distribution can be identified by locating the
peak or highest point of the histogram.
Dispersion: The spread or variability of the data is indicated by the width of the
histogram.
Skewness: The direction and degree of skewness can be assessed by observing the
tail of the histogram.
Outliers: Histograms can help identify potential outliers or unusual observations in
the dataset.
Pie Chart
A pie chart is a circular statistical graphic that is divided into slices to illustrate numerical
proportions. Each slice represents a proportionate part of the whole, and the size of each slice
is proportional to the quantity it represents. Pie charts are commonly used to represent the
distribution of categories within a dataset, showing the relationship of each part to the whole.
1. Slices:
oThe circular chart is divided into slices, where each slice represents a category
or subgroup of the data. The size of each slice is proportional to the quantity or
percentage it represents.
2. Central Angle:
oEach slice is associated with a central angle, measured from the center of the
circle. The central angle is proportional to the percentage of the total dataset
that the category represents.
3. Whole Circle:
o The entire pie represents the whole dataset or 100%. The sum of all the slices
should equal the entire circle.
1. Data Collection:
o Gather the dataset for which you want to create a pie chart. The data should be
categorical, with each category contributing to the total.
2. Calculate Percentages:
o Calculate the percentage of each category in relation to the total. This is done
by dividing the quantity of each category by the total quantity and multiplying
by 100.
3. Determine Slices:
o Assign each category a slice in the pie chart. The size of each slice is
determined by its percentage of the total.
4. Draw the Pie Chart:
o Use the central angles to draw each slice, starting from the center of the circle.
The slices should add up to 360 degrees, representing the entire circle.
5. Label the Chart:
o Label each slice with the category name and its corresponding percentage to
provide clarity and context.
A response surface plot is a graphical representation that illustrates the relationship between
multiple independent variables and a response or output variable in a mathematical or
statistical model. This type of plot is particularly useful in the context of experimental design
and optimization, allowing researchers to visualize the behavior of a system as the input
variables change.
1. Experimental Design:
o Conduct a well-designed experiment where multiple factors are varied
systematically. Use statistical techniques like factorial design to efficiently
explore the factor space.
2. Data Collection:
o Collect data on the response variable for each combination of factor levels in
the experiment.
3. Fit a Model:
o Fit a mathematical model, often a polynomial regression model, to the
collected data. This model describes the relationship between the independent
variables and the response.
4. Generate the Response Surface Plot:
o Use the fitted model to generate the response surface plot. Plot the two
independent variables on the x and y axes, and the response variable on the z-
axis.
5. Analyze and Interpret:
o Analyze the response surface plot to understand how changes in the
independent variables impact the response variable. Look for peaks, valleys,
and trends that provide insights into the system's behavior.
Contour Plot
1. Experimental Design:
o Conduct an experiment with two independent variables and a response
variable, systematically varying the factors.
2. Data Collection:
o Collect data on the response variable for each combination of factor levels.
3. Contour Plot Creation:
o Use statistical or computational tools to generate a contour plot based on the
collected data. Software like MATLAB, Python with libraries such as
Matplotlib or Seaborn, and others are commonly used for creating contour
plots.
4. Interpretation:
o Analyze the contour plot to understand how changes in the independent
variables affect the response variable. Pay attention to the contour lines and
their spacing.
Key Considerations:
1. Effect Size:
o The effect size is the magnitude of the difference or relationship that the study
aims to detect. Larger effect sizes generally require smaller sample sizes to
achieve statistical significance.
2. Significance Level (Alpha):
o The significance level (often denoted as α) is the probability of rejecting a true
null hypothesis. Common values are 0.05 or 0.01, representing a 5% or 1%
chance of making a Type I error, respectively.
3. Power (1 - Beta):
o Power is the probability of correctly rejecting a false null hypothesis.
Researchers typically aim for a high power level, often 0.80 or 0.90, to
minimize the risk of Type II errors.
4. Variability (Standard Deviation):
o The variability or standard deviation of the data influences the precision of the
estimates. Higher variability usually requires a larger sample size.
5. Type I and Type II Errors:
o Type I error occurs when a true null hypothesis is incorrectly rejected. Type II
error occurs when a false null hypothesis is not rejected. The balance between
these errors is considered in sample size calculations.
6. Study Design:
o The type of study design (e.g., randomized controlled trial, observational
study) and the statistical tests being used impact sample size requirements.
7. Desired Precision:
o Researchers may specify the desired level of precision in estimating
parameters or testing hypotheses, influencing the required sample size.
8. Ethical and Practical Constraints:
o Ethical considerations and practical constraints, such as budget and time
limitations, play a role in determining a feasible sample size.
Power of a Study:
Introduction: Power is the probability that a study will correctly reject a false null
hypothesis, indicating the study's ability to detect a true effect or relationship if it exists. A
study with high power is more likely to produce significant results when a true effect is
present.
Key Components:
1. Effect Size:
o Larger effect sizes increase the power of a study. A substantial effect is more
easily detected than a small one.
2. Sample Size:
o As sample size increases, statistical power also increases. Larger samples
provide more precise estimates and increase the chances of detecting true
effects.
3. Significance Level (Alpha):
o A lower significance level (α) increases the stringency of the test, making it
more challenging to achieve statistical significance but also reducing the risk
of Type I errors.
4. Variability (Standard Deviation):
o Lower variability increases the power of a study. If the data points are tightly
clustered around the mean, effects are easier to detect.
5. Study Design:
o The study design and the chosen statistical test impact the power of a study.
Well-designed experiments with appropriate controls often have higher power.
6. Type I and Type II Errors:
o The balance between Type I and Type II errors influences the power
calculation. An increase in power often comes with an increase in the risk of
Type I error and vice versa.
7. Desired Power Level:
o Researchers typically aim for a power level of at least 0.80 (80%) or 0.90
(90%) to ensure a high probability of detecting true effects.
Balancing Act:
Trade-Offs:
o There is often a trade-off between sample size, effect size, and power. Larger
sample sizes increase power, but they may also be impractical or costly.
Sensitivity Analysis:
o Researchers may conduct sensitivity analyses to assess how changes in effect
size, sample size, or other factors impact the study's power.
Pilot Studies:
o Conducting pilot studies can help refine sample size estimates and assess the
feasibility of the study.
1. Title Page:
o Includes the title of the report, the author's name, the date, and other relevant
details.
2. Abstract/Executive Summary:
o Summarizes the main points of the report, providing a concise overview for
readers.
3. Table of Contents:
o Outlines the structure of the report, listing sections and subsections with page
numbers.
4. Introduction:
o Sets the stage for the report, introducing the topic, objectives, and the scope of
the study.
5. Methodology:
o Describes the methods and procedures used in collecting and analyzing data.
6. Findings/Results:
o Presents the main results or findings of the study, often using tables, charts, or
graphs.
7. Discussion:
o Interprets the findings, discusses their implications, and may compare results
to existing literature.
8. Conclusion:
o Summarizes the key points, draws conclusions, and may provide
recommendations for further action.
9. Recommendations:
o Offers suggestions or actionable recommendations based on the report's
findings.
10. References:
o Lists all sources cited within the report.
11. Appendices:
o Includes additional materials, such as raw data, supplementary information, or
detailed analyses.
Audience:
o Consider the target audience and tailor the language and content accordingly.
Clarity and Conciseness:
o Use clear and concise language to convey complex information.
Structure:
o Organize the report logically, with a clear flow from introduction to
conclusion.
Visual Elements:
o Use visual elements such as graphs and charts to enhance understanding.
Citations:
o Properly cite sources to give credit and provide a basis for further reading.
Editing and Proofreading:
o Edit and proofread the report to ensure accuracy and clarity.
Data Presentation:
1. Bar Graphs:
o Represent data using bars of varying lengths to compare quantities.
2. Line Graphs:
o Display trends or relationships over time using connected data points.
3. Pie Charts:
o Illustrate the proportion of different components within a whole.
4. Tables:
o Present data in rows and columns for detailed reference.
5. Scatter Plots:
o Show the relationship between two variables using individual data points.
6. Histograms:
o
Depict the distribution of a continuous variable using bars.
7. Heatmaps:
o Represent data values in a matrix using color intensity.
8. Box-and-Whisker Plots (Boxplots):
o Display the distribution of a dataset, including median, quartiles, and potential
outliers.
Audience Understanding:
o Tailor the visual representation to the audience's level of expertise and
familiarity with the subject.
Simplicity and Clarity:
o Keep visuals simple and ensure they convey information clearly.
Appropriate Use:
o Choose the right type of visual representation for the specific type of data
being presented.
Consistency:
o Maintain consistency in formatting and labeling across visuals for easy
comparison.
Interactivity (if applicable):
o Consider using interactive elements for digital presentations to engage the
audience.
Narrative Support:
o Use visuals to support the narrative of the report or presentation.
Protocol:
Cohort Studies:
1. Prospective Nature:
o Cohort studies are often prospective, meaning that data is collected from the
cohort over time.
2. Defined Exposures:
o Participants are grouped based on their exposure to certain factors, and
outcomes are assessed over the follow-up period.
3. Calculation of Relative Risk:
o Relative risk (RR) can be calculated to quantify the association between
exposure and outcome.
4. Strengths:
o Cohort studies are strong for establishing temporal relationships and assessing
cause-and-effect associations.
5. Weaknesses:
o They may be resource-intensive and require long follow-up periods.
Observational Studies:
Introduction: Observational studies are research designs where researchers observe and
analyze subjects in their natural environment without intervening. These studies are valuable
for exploring associations, identifying patterns, and generating hypotheses.
Types of Observational Studies:
1. Cross-Sectional Studies:
o Data is collected at a single point in time to assess the relationship between
variables.
2. Case-Control Studies:
o Individuals with a particular condition (cases) are compared with those
without the condition (controls) to identify factors associated with the
condition.
3. Cohort Studies:
o Follows a group of individuals over time to assess the relationship between
exposures and outcomes.
Observational studies are versatile and applicable in various fields. However, they do
not establish causation, and the presence of confounding variables can affect results.
Experimental Studies:
Experimental studies provide strong evidence of causation but may lack external
validity. Ethical considerations also play a crucial role in experimental study design.
Introduction: Clinical trials are rigorous experiments designed to evaluate the safety and
efficacy of medical interventions, such as drugs, devices, or therapies, in humans. The design
of clinical trials is crucial for obtaining reliable and valid results.
1. Phase I:
o Focuses on safety and dosage in a small group of healthy volunteers.
2. Phase II:
o Assesses efficacy and side effects in a larger group of individuals with the
target condition.
3. Phase III:
o Confirms efficacy, monitors side effects, and compares the intervention to
standard treatments in a larger, diverse population.
4. Phase IV:
o Post-marketing surveillance to monitor long-term effects and collect additional
information after the intervention is approved for public use.
1. Randomization:
o Randomly assigning participants to groups minimizes bias and ensures
comparability.
2. Blinding:
o Blinding (single or double) prevents biases in the assessment of outcomes.
3. Placebo Control:
o Placebos help control for the placebo effect and provide a baseline for
comparison.
4. Sample Size Calculation:
o Adequate sample sizes are crucial for detecting meaningful effects and
ensuring statistical power.
5. Informed Consent:
o Participants must provide informed consent, understanding the risks and
benefits of participation.
6. Ethical Approval:
o All clinical trials must be approved by an ethical review board to ensure
participant safety and adherence to ethical standards.
7. Endpoint Selection:
o Defining clear and relevant primary and secondary endpoints ensures accurate
assessment of the intervention's effectiveness.