| Course | DPH 870 Evidence-Based Practice and Advanced Research Methods |
|---|---|
| Module | Module 3 |
| Paper type | Research design paper |
| Length | About 1,101 words, 7 pages |
| Format | APA 7 student paper |
| School | Aspen University |
| Program | Doctor of Public Health |
| Updated | September 2026 |
Free sample paper for DPH 870 Module 3
The Right Tool for the Question: Matching Research Designs to a Community Health Worker Evaluation
Student Name
Doctor of Public Health Program, Aspen University
DPH 870: Evidence-Based Practice and Advanced Research Methods
Instructor Name
Month Day, Year
The Right Tool for the Question: Matching Research Designs to a Community Health Worker Evaluation
Every design has questions it suits best. Which one fits depends on what the question asks, whether the intervention can be assigned, what data exist and what decision the evidence will inform. This paper compares designs for a DPH capstone asking whether blood pressure improves when a health department deploys community health workers, and it justifies the design chosen.
What the Question Requires
The question asks about an effect: whether participation causes better control than would have occurred without it. Answering requires a comparison group similar to participants, measurement of outcomes after exposure and control of confounding. The program was already running when the capstone began, and the health department needs an answer within two years.
Designs Compared
The table compares candidate designs.
| Design | Causal strength | Feasibility here | Fit |
|---|---|---|---|
| Individually randomized trial | High | Low: program already open to all | Poor |
| Cluster randomized trial | High | Low: all neighborhoods already served | Poor |
| Stepped wedge | Moderate to high | Low: rollout complete | Poor |
| Interrupted time series | Moderate | Moderate: needs stable monthly data | Partial |
| Matched comparison with clinical data | Moderate | High: records available | Good |
| Prospective cohort | Moderate | Low: too slow | Partial |
| Qualitative interviews | Not causal | High | Good for how and why |
Randomized Designs
Randomized trials, whether of individuals or clusters, provide the strongest protection against confounding. A trial of community health worker support in primary care practices showed the value of this design (Kangovi et al., 2018). Here, however, the program was open to all eligible residents before the capstone began, so withholding it for a trial was neither feasible nor ethical.
Quasi-Experimental Designs
Interrupted time series compare trends before and after a program; they work well with long, stable series of population data. Matched comparison designs compare participants with similar nonparticipants. Both provide plausibility evidence when randomization is not possible, which Victora et al. (2004) argue is often the appropriate level of evidence for public health decisions.
Emulating a Target Trial
Observational analyses can be strengthened by designing them to emulate the randomized trial one would ideally run, specifying eligibility, time zero, treatment strategies, outcomes and follow-up. Hernán et al. (2008) showed that analyzing observational data this way reproduced trial results that conventional analyses had missed. The capstone defines time zero as enrollment for participants and an equivalent clinic visit for comparison adults.
The Chosen Design
The chosen approach pairs participants and nonparticipants by propensity score on clinical and demographic factors, framed as a target trial emulation. It fits the question, uses available data, can be completed in the time available and provides evidence at the plausibility level needed for the county's decision.
Threats to Validity
The main threat is confounding by unmeasured factors such as motivation or social support, which could make participants look better than comparison adults for reasons unrelated to the program. Other threats include measurement differences across clinics and missing outcome data. Sensitivity analyses, including an E-value for unmeasured confounding, address these threats.
Adding Qualitative Methods
Quantitative results show whether the program worked; qualitative interviews can show how and why. The capstone will include a small set of interviews with participants and community health workers to explore what helped and what got in the way, following complex intervention guidance that values understanding mechanisms and context (Craig et al., 2008). About 15 interviews are planned, enough to identify common themes without overburdening participants.
Mixed Methods Integration
Findings from both strands will be integrated in the discussion: for example, if participants with more visits had greater improvement, interviews may reveal whether trust, medication help or home monitoring explains the pattern.
Ethical Considerations
The design uses existing records, reducing burden on participants. Interviews require informed consent. The design avoids withholding a program from people who need it, an ethical advantage over a randomized trial in this setting.
What the Design Cannot Answer
The design cannot fully rule out confounding, cannot estimate long-term outcomes such as strokes and may not generalize to other counties. These limits will be stated clearly so decision makers understand the strength of the evidence.
Levels of Evidence for the Decision
The county's decision is whether to continue and expand the program. For that decision, evidence that outcomes improved among participants compared with similar adults, and that the improvement is plausibly due to the program, is sufficient. Probability-level evidence from a trial would be stronger but is not attainable and is not required to act responsibly.
Time Zero and Follow-Up
Defining when follow-up begins avoids bias. For participants, time zero is enrollment; for comparison adults, it is a clinic visit in the same month at which they met eligibility criteria. Outcomes are measured 12 months later in both groups. Misaligned start times can create immortal time bias, in which one group appears to do better simply because of how time is counted.
Data Availability as a Design Constraint
The design depends on existing records, which were not collected for research. Blood pressure readings vary in technique, and some social factors are not recorded. These constraints limit what the design can adjust for, which is why sensitivity analyses and qualitative data are included.
Stakeholder Views on Design
Health department leaders were asked whether a waitlist design, in which some eligible residents would start later, was acceptable. They declined because demand was high and delaying services felt unfair. Their view confirmed that an observational design was the right choice for this setting.
Documenting the Design Choice
The methods chapter will explain why randomized designs were rejected, why a matched comparison was chosen and how its weaknesses will be addressed. Committees expect this reasoning to be explicit, and future readers need it to judge the evidence.
Transferability
Findings from one county may not transfer directly to others. Describing the setting, program and population in detail allows readers elsewhere to judge whether results apply to them, an approach that complements statistical generalization.
Design and Analysis Together
Design choices constrain analysis. Because the capstone uses matched pairs, the analysis must account for matching, and because confounding is the main threat, sensitivity analyses are built in from the start. Choosing design and analysis together avoids discovering too late that the data cannot answer the question. The analysis plan is therefore drafted alongside the design section and reviewed by the committee at the same time.
Conclusion
Matching design to question means weighing causal strength against feasibility, ethics and timing. For the community health worker capstone, a matched comparison framed as a target trial emulation, supplemented by qualitative interviews, offers the best balance, providing plausibility evidence the county can use while acknowledging its limits.
References
Craig, P., Dieppe, P., Macintyre, S., Michie, S., Nazareth, I., & Petticrew, M. (2008). Developing and evaluating complex interventions: The new Medical Research Council guidance. BMJ, 337, Article a1655. https://doi.org/10.1136/bmj.a1655
Hernán, M. A., Alonso, A., Logan, R., Grodstein, F., Michels, K. B., Willett, W. C., Manson, J. E., & Robins, J. M. (2008). Observational studies analyzed like randomized experiments: An application to postmenopausal hormone therapy and coronary heart disease. Epidemiology, 19(6), 766-779. https://doi.org/10.1097/EDE.0b013e3181875e61
Kangovi, S., Mitra, N., Norton, L., Harte, R., Zhao, X., Carter, T., Grande, D., & Long, J. A. (2018). Effect of community health worker support on clinical outcomes of low-income patients across primary care facilities: A randomized clinical trial. JAMA Internal Medicine, 178(12), 1635-1643. https://doi.org/10.1001/jamainternmed.2018.4630
Victora, C. G., Habicht, J.-P., & Bryce, J. (2004). Evidence-based public health: Moving beyond randomized trials. American Journal of Public Health, 94(3), 400-405. https://doi.org/10.2105/AJPH.94.3.400
DPH 870 Module 3 instructions, in plain terms
Matching designs to research problems is named in Aspen's DPH 870 catalog entry, and the third module's instructions stay behind the classroom login, so this sample weighs designs for one question. Design papers generally ask you to state what the question requires, compare options and argue for one while admitting its weaknesses. List the question's requirements first. Compare several designs in one table on the same criteria. Explain why stronger designs were not feasible. Describe how the chosen design handles its main threat. Consider adding qualitative methods. State what the design cannot answer, and explain how you defined when follow-up starts for each group. Say who was consulted about design options. Name the level of evidence your decision needs.
Inside the DPH 870 Module 3 example
The paper first states what the question requires and presents a four-column table of seven designs. Randomized and quasi-experimental designs, target trial emulation, the chosen design and threats to validity follow, then qualitative methods, mixed methods integration, ethics and what the design cannot answer. Levels of evidence, time zero, data availability, stakeholder views, documenting the choice, transferability and design with analysis follow, and a margin comment notes that stating requirements first makes the choice follow from the question. Each rejected design is explained in terms of the program's circumstances, not only its general weaknesses. Ethics and stakeholder views are woven into the design choice.
Reading the DPH 870 Module 3 grading rubric
Instructors reward design papers that tie the design to the question, compare options evenly, face threats honestly and consider ethics. Sources include the Kangovi trial, work on plausibility evidence, an early target trial emulation and complex intervention guidance, cited in APA format. Every design in the table is rated on identical criteria. Time zero is defined to avoid immortal time bias. Stakeholder views on a waitlist design show real-world judgment graders notice. It also explains that plausibility evidence is enough for the county's decision, matching evidence to purpose rather than chasing an unreachable standard. Adding interviews shows attention to how and why the program works, not only whether it does. Defining time zero shows technical care.
Common DPH 870 Module 3 mistakes, and how to avoid them
A frequent mistake is choosing a familiar design without comparing alternatives, or claiming causal certainty from an observational study. Compare at least four designs. Explain feasibility and ethics honestly. Define time zero. Plan sensitivity analyses for confounding. Add qualitative methods if mechanisms matter. If target trial emulation is new, a tutor can walk you through specifying the trial you would ideally run. End with the design's main limitation and how readers should weigh your findings because of it, then ask your committee whether that trade-off is acceptable before you collect data. Keep your design comparison table in an appendix for the committee. State the evidence level your decision needs.
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 Aspen University 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.
More DPH 870 and Doctor of Public Health sample papers
- DPH 870 Module 1: The Steps of Scientific Research
- DPH 870 Module 2: Formulating the Research Problem
- DPH 870 Module 4: A Documented Search Strategy
- DPH 870 Module 5: Critical Appraisal of Key Studies
- DPH 870 Module 6: A Synthesis Matrix by Theme
- DPH 870 Module 7: Writing the Literature Review Argument
- DPH 870 Module 8: Chapter 2 of the Capstone
- DPH 840 Module 6: Health Economics Applied to a Decision
- DPH 890 Module 3: Discussion and Implications for Practice
- DPH 801 Module 3: Interpersonal Theories and Influence
- DPH 820 Module 8: Funding and Sustaining a Policy
DPH 870 Module 3 questions, answered
What does DPH 870 Module 3 usually ask for?
Aspen's DPH 870 covers matching designs to research problems, so a design comparison and justification is typical. Follow your classroom prompt.
When is a matched comparison design appropriate?
When randomization is not feasible and data allow finding nonparticipants similar to participants on important characteristics.
What is target trial emulation?
Designing an observational analysis to mimic the randomized trial one would ideally conduct.
Where can I find a free DPH 870 Module 3 sample paper?
The research design paper is available here, including a table rating seven designs on causal strength, feasibility and fit.
How do you choose a research design in DPH 870 Module 3?
State what the question requires, compare designs on causal strength, feasibility and ethics, and justify the choice with its main threats and how they are addressed.