| Course | CHL 640 Evaluating Community Health Initiatives (CHL/640) |
|---|---|
| Week | 4 |
| Paper type | Evaluation design paper |
| Length | about 1,151 words, 4 double-spaced pages plus title page and references |
| Format | APA 7 student paper |
| School | University of Phoenix |
| Program | MPH |
| Updated | September 2026 |
Free sample paper for CHL 640 Week 4
No Randomization, Still Credible: Choosing a Matched Comparison, Interrupted Time Series and Staggered Rollout to Evaluate Walk and Eat Well
[Student Name]
University of Phoenix
CHL/640: Evaluating Community Health Initiatives
Week 4 Assignment
[Instructor Name]
[Date]
The initiative, comparison groups, data series, rollout schedule and analysis choices are composites written for a model paper modeled on Yakima, Washington; research findings come from the sources cited.
In its review of the coalition's full proposal, the foundation's evaluation reviewer asked a direct question: if participants lose weight and stores sell more produce, how will the coalition know Walk and Eat Well caused it rather than a new clinic program, a price change or the weather? This paper answers by choosing an evaluation design.
The Range of Designs
Designs differ in what they can show. A post-test only design describes results but cannot show change. A pre-post design shows change but cannot rule out other causes. Adding a comparison group helps separate program effects from outside trends. Interrupted time series use many measurements before and after launch to detect changes in level and trend. Randomized trials, when feasible, provide the strongest evidence of cause.
Why Not Randomize
The Diabetes Prevention Program trial randomized adults with elevated glucose and found that intensive lifestyle intervention reduced diabetes incidence by 58% compared with placebo (Diabetes Prevention Program Research Group, 2002), the benchmark for the lifestyle component. But randomizing east side residents was set aside. Residents and the coalition opposed withholding services from neighbors who asked for them. Store and path changes affect everyone who shops or walks, so they cannot be randomized to individuals. And the budget could not support a trial.
Natural Experiment Guidance
Rigorous alternatives exist. New Medical Research Council guidance explains that natural experimental studies can provide convincing evidence of impact even when effects are small or take time to appear, but that researchers need a good understanding of the process determining exposure, careful choice and combination of methods, tests of assumptions and transparent reporting (Craig et al., 2012). The team designed the evaluation with those principles.
Understanding How People Get Exposed
The guidance stresses understanding who gets the intervention and why. Enrollment in lifestyle groups depends on referral, interest, schedule and language; walking depends on proximity to the path and work hours; store exposure depends on where families shop. The team mapped these routes so that comparison groups could be chosen from people with similar chances of exposure, and so that differences in who enrolls could be measured rather than assumed away.
Design One: Matched Comparison for Weight and Activity
For weight and activity, the team uses a pre-post design with a matched comparison group. The comparison group comes from the health center's prediabetes registry: east side patients who were referred but did not enroll, matched to participants on age, sex, language, baseline weight and clinic. Weight is available from clinic visits for both groups.
Strengths and Limits of Design One
The comparison helps separate program effects from trends in the clinic population, such as a new medication or clinic program. Because both groups are drawn from the same clinics and neighborhoods, outside changes should affect them similarly. Its main weakness is selection: people who enroll may be more motivated than those who do not. The team will adjust for measured differences and compare early and late enrollees to assess how much motivation might explain.
Design Two: Interrupted Time Series for Stores and Path Use
Store sales and path counts are measured monthly, allowing an interrupted time series. The team will gather store produce sales for twelve months before the store program and eighteen months after, and path counts for six months before and eighteen months after the walking groups launch.
Analyzing the Time Series
A tutorial on interrupted time series explains that the design is valuable for population-level interventions implemented at a clearly defined time, recommends specifying the expected impact model in advance and describes segmented regression analysis with attention to overdispersion, autocorrelation, seasonal trends and time-varying confounders (Lopez Bernal et al., 2017). Specifying in advance whether sales should jump, climb gradually or both keeps the analysis honest. The team expects a gradual rise in sales and an immediate jump in evening path use.
Handling Seasonality
Seasons matter: produce prices and supplies change through the harvest, and evening walking falls in winter. The time series will include a full year before launch where possible and adjust for month, and store results will be compared with sales at two similar stores outside the program.
Design Three: Staggered Rollout
Lifestyle groups start at different church sites over twelve months for practical reasons. The staggered start provides another comparison: residents at later sites, measured before their groups begin, serve as a comparison for residents at earlier sites. The rollout order was set by church readiness, not by any expectation of results.
Threats to Internal Validity
Key threats include selection, since motivated people enroll; history, such as a new clinic diabetes program or food price changes; maturation, since weight may change with age; regression to the mean, since people may enroll when their weight is unusually high; and attrition, since dropouts may differ from completers. Each design choice addresses at least one of these.
What the Design Cannot Show
Even combined, these designs cannot prove cause with the certainty of a randomized trial. Unmeasured differences in motivation may remain, and outside events could coincide with launch. The evaluation report will state these limits plainly and describe results as evidence consistent with, or not consistent with, program effects. Honesty about limits protects the coalition's credibility with funders and residents.
Threats to External Validity
Results may not apply to other communities with different food environments, cultures or partners. The team will describe the context in detail so that others can judge applicability.
Qualitative Evidence on Causation
Participants' own accounts can support or challenge causal claims. Exit interviews will ask participants what, if anything, changed their eating and activity, and whether other programs or events played a role. If participants credit a different clinic program or a family change, the team will weigh that alongside the quantitative results.
Sample Size
With about 180 participants a year and a matched comparison of similar size, the evaluation can detect a difference of about 3 percentage points in mean weight change between groups with reasonable confidence, smaller than the 5% expected among completers.
Measuring Exposure
Following the guidance on natural experiments, the team will record each participant's exposure: sessions attended, walks joined and purchases at partner stores, so that results can be examined by dose.
The Final Design
The final design combines a matched comparison group for weight and activity, interrupted time series with comparison stores for sales and path use and staggered rollout for lifestyle groups, with exposure measured for each participant and analyses specified in advance.
Conclusion
The foundation reviewer's question now has an answer. A randomized trial was set aside for ethical, practical and budget reasons, but natural experiment guidance supports combining rigorous nonrandomized designs. Matched comparisons, interrupted time series with attention to seasonality and a staggered rollout together make it far harder for other explanations to account for any change the evaluation finds.
References
Craig, P., Cooper, C., Gunnell, D., Haw, S., Lawson, K., Macintyre, S., Ogilvie, D., Petticrew, M., Reeves, B., Sutton, M., & Thompson, S. (2012). Using natural experiments to evaluate population health interventions: New Medical Research Council guidance. Journal of Epidemiology and Community Health, 66(12), 1182-1186. https://doi.org/10.1136/jech-2011-200375
Diabetes Prevention Program Research Group. (2002). Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. New England Journal of Medicine, 346(6), 393-403. https://doi.org/10.1056/NEJMoa012512
Lopez Bernal, J., Cummins, S., & Gasparrini, A. (2017). Interrupted time series regression for the evaluation of public health interventions: A tutorial. International Journal of Epidemiology, 46(1), 348-355. https://doi.org/10.1093/ije/dyw098
What the CHL 640 Week 4 instructions ask
The fourth CHL 640 assignment commonly asks students to choose an evaluation design. Prompts may ask students to compare designs such as post-test only, pre-post, pre-post with comparison group, interrupted time series, stepped or staggered rollout and randomized trials, match designs to evaluation questions, identify threats to internal and external validity, explain how the chosen design addresses them within ethical and practical limits and describe sampling and sample size. Some versions ask students to justify why a randomized trial was or was not used. Address that directly if so. Strong papers choose different designs for different questions, explain what each can and cannot show, name specific threats and describe concrete steps to reduce them.
How this CHL 640 Week 4 example is built
A foundation reviewer's question, how the coalition will know the program caused any change, opens the paper. Designs are compared from weakest to strongest. A randomized trial is considered and set aside for ethical and practical reasons, with the Diabetes Prevention Program trial as the benchmark. Natural experiment guidance frames the alternatives. For weight and activity, a pre-post design with a matched comparison group from clinic records is chosen. For store sales and path use, interrupted time series using monthly data before and after launch are chosen. A staggered rollout across church sites adds a comparison. Threats to validity, what the design cannot show, sample size and the final design close the paper.
CHL 640 Week 4 grading rubric: where the points go
The evaluation design week is typically graded on sound design choices matched to questions and honest attention to validity. Graders look for designs compared, the chosen design justified, reasons for using or not using randomization, specific threats to internal and external validity, strategies to reduce bias, comparison groups or time series explained, sample size considered and limits acknowledged. Methodological guidance on natural experiments and time series strengthens the paper. Combining designs to answer different questions earns credit. Naming threats specific to the program also earns marks. Orderly sections and exact APA entries earn the remaining credit. Papers that choose a pre-post design without discussing its weaknesses usually score lower.
CHL 640 Week 4 help: mistakes to avoid
Many CHL 640 Week 4 papers choose a simple before-and-after design and never mention what else could explain the change. Start from your questions: some need only counts, others need causal evidence. Consider a comparison group from people who were eligible but did not enroll, matched on key traits. For outcomes measured often, such as monthly sales or path counts, consider interrupted time series. If sites start at different times, use the rollout as a comparison. Name specific threats, such as seasonal patterns, selection or other programs, and explain how each design choice reduces them. Finally, estimate whether your sample is large enough to detect the change you expect.
Related CHL 640 sample papers
Other CHL 640 week samples
- CHL 640 Week 1: Evaluation Purposes and Stakeholders
- CHL 640 Week 2: Evaluation Questions
- CHL 640 Week 3: Indicators and Data Sources
- CHL 640 Week 5: Data Collection and Analysis
- CHL 640 Week 6: Using and Sharing Results
More MPH sample papers
- CHL 620 Week 4: Coalition Building
- CHL 630 Week 4: Program Budget
- MPH 510 Week 4: Professional Values and Standards
- MPH 520 Week 4: Life-Course Perspective
CHL 640 Week 4 questions, answered
What does CHL/640 Week 4 usually ask for?
The fourth evaluation paper commonly asks students to choose an evaluation design, compare alternatives, justify the choice, address threats to validity and consider sample size.
Where can I find a free CHL 640 Week 4 sample paper?
Open the evaluation design paper above free; a note beside each choice explains the reasoning. Share your evaluation questions; the first paper is on us.
What is an interrupted time series design?
A design that compares the level and trend of an outcome measured repeatedly before and after an intervention begins, often analyzed with segmented regression, to estimate the intervention's effect.
What is a natural experiment?
A study of an intervention whose timing or allocation was not controlled by researchers, such as a policy or program launch, analyzed with methods that account for how exposure occurred.
Why not use a randomized trial for a community program?
Randomizing residents may be unacceptable when a community wants services for everyone, impractical for place-based changes such as store improvements and too costly, so rigorous nonrandomized designs are used instead.
Write yours, or have the desk draft it
This paper is an original model document written by our desk, not a submitted student paper and not an official University of Phoenix document. Read it for the moves, then write your own to the instructions in your classroom. If you want one built to your exact prompt and rubric, the first custom sample is free and arrives in 24 to 48 hours.
Request this one custom, free · All CHL 640 week samples · All courses