| Course | HD560 Health Education Evaluation and Research |
|---|---|
| Unit | Unit 4 |
| Paper type | Validity threat analysis |
| Length | About 1,100 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | MS in Health Education |
| Updated | October 2026 |
Free sample paper for HD560 Unit 4
What the Lottery Still Leaves Open: A Unit 4 Validity Threat Analysis for the Within Earshot Hearing Aid Evaluation
[Student Name]
Purdue University Global
HD560: Health Education Evaluation and Research
Unit 4 Validity Threat Analysis
[Instructor Name]
[Date]
Within Earshot, Tamsen County and the scenarios described are composite; the validity framework, the federal hearing aid rule and the trial cited are real.
The Framework
Shadish et al. (2002) sort threats into four types. Statistical conclusion validity asks whether a relationship between the program and the outcome is detected accurately. Internal validity asks whether that relationship is causal. Construct validity asks whether the program and outcomes are what the evaluation says they are. External validity asks to whom and where the result applies. Random assignment addresses the most familiar internal threats. It does little for the other three types and leaves some internal threats open.
Why Go Beyond Randomization
The temptation in a randomized evaluation is to treat the lottery as the end of the validity discussion. Randomization is a strong protection, but it works at one moment, the draw. Everything that happens to the two groups afterward, how they are treated, who leaves, who finds help elsewhere and how outcomes are measured, can still introduce bias. A threat analysis that stops at the lottery leaves the evaluation exposed exactly where real programs most often go wrong.
What the Lottery Controls
Because a lottery decides who is offered a place, the two groups should be similar at the start in hearing, loneliness, health, income and motivation. Events such as the senior center's reopening affect both groups. Regression toward the mean affects both groups equally, since both were recruited the same way. These were the main problems with last year's report, and the design removes them.
Statistical Conclusion Validity
Power. The design's smallest detectable difference on the loneliness scale is roughly 0.6 points. A smaller real effect could be missed. Direction: makes the program look ineffective when it may help a little. Response: report the interval, not just a p value.
Multiple outcomes. Testing many outcomes raises the chance of a false positive. Response: one primary outcome named in advance, with secondary outcomes labeled as such.
Reliability of a short scale. A three-item measure is less reliable than a longer one, which widens the interval. Response: accept this for brevity, and report the scale's reliability in this sample.
Internal Validity
Contamination. Since 2022, federal rules have allowed hearing aids to be sold over the counter to adults whose own sense of their hearing loss places it in the mild to moderate range (U.S. Food and Drug Administration, 2022). Wait-listed applicants may buy their own devices, some for a few hundred dollars. Direction: shrinks the difference. Response: ask both groups at six months whether they obtained hearing aids elsewhere, and report the share.
Differential attrition. Wait-listed applicants receive nothing for six months and may stop answering calls. If those who drift away are lonelier, the wait-list's average will look better than it is. Direction: shrinks the difference. Response: two contact attempts at different times, a small gift card for each assessment and a comparison of baseline scores between those who stay and those who leave.
Resentful demoralization. Wait-listed applicants who know they lost the lottery may report feeling worse. Direction: inflates the difference. Response: tell applicants clearly that everyone will be served within six months, and send a monthly letter with the expected start date.
Spillover within households. A few married couples applied together. If one spouse wins and the other does not, the second may benefit from the first hearing better. Response: randomize couples as a unit.
Randomization balances the groups on the day of the lottery, but it cannot keep wait-listed people from buying their own hearing aids, losing interest or living with someone who now hears better, and each of those can quietly change the comparison afterward.
Construct Validity
What the program is. Within Earshot is not only hearing aids. It includes a fitting, three adjustment visits at the senior center and a monthly drop-in hour for troubleshooting. Any effect on loneliness could come from those visits and the people met there rather than from better hearing. Hearing intervention trials have used attention-control groups for this reason; one large trial compared a hearing intervention with a health education program of similar contact (Lin et al., 2023). Response: record each participant's visits and drop-in attendance, and test whether change in hearing handicap accounts for change in loneliness.
What the outcomes are. Self-reported hearing handicap may improve partly because participants expect it to. Response: report the objective hours of use from the devices alongside it.
External Validity
Applicants are people motivated enough to apply, in one county with one audiology partner, mostly English and Spanish speakers. The result may not apply to older adults who would not seek hearing aids, or to programs without follow-up visits.
Threats Considered and Set Aside
Two threats were considered and judged minor. Testing effects, in which answering the loneliness scale at baseline changes later answers, should affect both groups equally and cannot explain a difference between them. Instrumentation, a change in how outcomes are measured, is controlled by using the same assessors, script and phone procedure at both time points for both groups. Each is listed here because a threat analysis should show what it weighed and dismissed, not only what it kept.
How the Threats Interact
The threats do not act alone. Contamination and attrition both shrink the difference, so together they could hide a real effect of moderate size. Resentful demoralization pushes the other way. If the final difference is small, the report will need to weigh these against each other using the contamination and attrition data, rather than presenting the estimate as if no forces acted on it. The bundled program raises a separate question, not whether there is an effect but what it should be called.
Monitoring During the Evaluation
Some threats can be watched while the evaluation runs. The assessors will record, each month, the share of wait-listed applicants they cannot reach, so that rising attrition shows up early. The program manager will log any wait-listed applicant who calls asking about buying hearing aids. If either figure climbs sharply, the evaluator will brief the board before the six-month assessment rather than after, and record the briefing in the evaluation file.
Summary
Three threats most need attention: contamination and differential attrition, which would shrink the difference, and the bundled nature of the program, which affects what any difference means. Each has a response in the plan.
References
Lin, F. R., Pike, J. R., Albert, M. S., Arnold, M., Burgard, S., Chisolm, T., Couper, D., Deal, J. A., Goman, A. M., Glynn, N. W., Gmelin, T., Gravens-Mueller, L., Hayden, K. M., Huang, A. R., Knopman, D., Mitchell, C. M., Mosley, T., Pankow, J. S., Reed, N. S., ... Coresh, J. (2023). Hearing intervention versus health education control to reduce cognitive decline in older adults with hearing loss in the USA (ACHIEVE): A multicentre, randomised controlled trial. The Lancet, 402(10404), 786-797. https://doi.org/10.1016/S0140-6736(23)01406-X
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-experimental designs for generalized causal inference. Houghton Mifflin.
U.S. Food and Drug Administration. (2022). Medical devices; ear, nose, and throat devices; establishing over-the-counter hearing aids. Federal Register, 87(158), 50698-50762. https://www.federalregister.gov
Reading the HD 560 Unit 4 instructions
The fourth unit generally asks students to identify threats to the validity of their evaluation design and explain how each will be addressed. Prompts commonly reference the four types: statistical conclusion, internal, construct and external validity. Begin by stating what your design controls, then move to what it does not. For each threat, explain the mechanism, the likely direction of bias and a specific response. Include threats that arise after assignment, such as attrition or contamination. Note threats you considered and judged minor. Describe how threats interact and how you will monitor them during the evaluation. Many sections expect six to eight pages with the four types as headings.
How the HD 560 Unit 4 example is put together
The analysis opens with the framework and explains why it does not stop at randomization. A section credits the lottery with what it controls. Statistical conclusion threats cover power, multiple outcomes and a short scale. Internal threats cover contamination through over-the-counter aids, differential attrition, resentful demoralization and spillover between spouses, each with a response. The highlighted sentence states what randomization does and does not do. Construct threats address the bundled program and expectation in self-reports, citing a trial with an attention control. External validity, threats set aside, interactions and monitoring follow before a summary of the three most serious threats.
Reading the HD560 Unit 4 grading rubric
Strong validity analyses earn credit for accurate use of the framework, threats specific to the design, reasoning about the direction of bias, practical responses and attention to what happens after random assignment. Graders often deduct when an analysis lists every threat from a textbook table or treats randomization as solving everything. Direction of bias is a frequent differentiator, since it shows the writer understands each mechanism. Responses score well when they are concrete and affordable. Monitoring threats during data collection shows mature planning. Clear headings and correct APA citations of the framework account for the rest. Some instructors also reward an analysis that explains how threats interact, for example when contamination and attrition both shrink the difference.
HD560 Unit 4 help from the desk
Students often stop at randomization and declare the design free of threats. Others name threats and never say whether each would inflate or shrink the result. A common gap is construct validity: what the program actually consists of may be broader than its name. Some analyses offer responses that the budget could never support. Writing down what each group experiences week by week after assignment usually reveals the threats that matter. If your design differs, describe who is assigned to what and what happens to each group afterward, and a writer can help you sort the threats by type and direction. Asking what a wait-listed person might do while waiting is a quick way to find contamination.
Get an HD560 Unit 4 example written to your instructions
Send the design your HD560 evaluation uses and the Unit 4 prompt and rubric, and mention anything unusual about how the program runs. The free first custom sample comes back within 24-48h, true to your instructions, deriving each threat from that design and stating which way it would bias the result. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.
More HD560 papers and MS in Health Education samples
- HD560 Unit 1: Discussion Board Post
- HD560 Unit 2: Evaluation Question Paper
- HD560 Unit 3: Design Selection Rationale
- HD560 Unit 5: Measurement Plan
- HD560 Unit 6: Seminar Reflection
- HD560 Unit 7: Analysis Approach
- HD560 Unit 8: Ethical Review Determination
- HD560 Unit 9: Reporting Strategy
- HD560 Unit 10: Evaluation Plan
- HD550 Unit 4: Policy Brief
- HD545 Unit 4: Readiness Assessment
- HD520 Unit 4: Instrument Appraisal
- HD535 Unit 4: Construct Mapping Table
HD560 Unit 4 questions, answered
What are the four validity types?
In Shadish, Cook and Campbell's framework, statistical conclusion validity concerns whether a relationship is detected accurately; internal validity, whether it is causal; construct validity, whether the program and outcomes are what they are labeled; and external validity, whether the result generalizes. Many prompts accept a simpler internal and external division, so check yours.
Does randomization eliminate all threats?
No. It balances groups at the start, which removes selection and most time-related threats, but it cannot prevent differential dropout, contamination between groups, or changes in how staff treat each arm. Those arise after allocation, so a threat analysis for a randomized design concentrates on what happens once the lottery is drawn.
How do I know which direction a threat biases the result?
Ask what the threat would do to each group's outcome. Contamination gives the comparison group some benefit, which shrinks the difference. Losing less motivated comparison members could make that group look better at follow-up. Reasoning through each mechanism, and saying where the direction is uncertain, is what graders look for.
Where can I find a free HD560 Unit 4 sample paper?
The full Unit 4 validity threat analysis appears on this page, covering what a hearing aid lottery still leaves open.
How can contamination affect a wait-list comparison?
If people on the wait-list obtain the service elsewhere, such as over-the-counter hearing aids, the groups become more alike and the measured difference shrinks.