| Course | HLTH 8560E Advanced Analysis of Secondary Data |
|---|---|
| Week | Week 6 |
| Paper type | Sample derivation |
| Length | About 560 words, 5 pages |
| Format | APA 7 student paper |
| School | Walden University |
| Program | PhD in Public Health |
| Updated | October 2026 |
Free sample paper for HLTH 8560E Week 6
Sample Derivation: From 39,156 Pooled NHANES Participants to 2,663 Adults With Diagnosed Diabetes, a Glycohemoglobin Value and a Complete PHQ-9, With Every Rule Counted and Every Removed Group Described
Student Name
Walden University
HLTH 8560E: Advanced Analysis of Secondary Data
Instructor Name
Month Day, Year
Sample Derivation: From 39,156 Pooled NHANES Participants to 2,663 Adults With Diagnosed Diabetes, a Glycohemoglobin Value and a Complete PHQ-9, With Every Rule Counted and Every Removed Group Described
Purpose
This derivation records each rule that stands between the full public release and the records the analysis uses. For every rule it gives the count remaining and describes the people the rule removed, so a reader can judge whether exclusions changed who the results describe (von Elm et al., 2007).
The Chain of Rules
Steps 1 and 2 define the population and are not losses. Steps 3 through 5 are losses: these 483 adults belong to the population the question describes but lack a measurement, so the next section compares them with the 2,663 who remain.
Table 1
Records Remaining After Each Rule, NHANES 2011-2012 to 2017-2018
| Step | Rule applied | Remaining | Removed | |
|---|---|---|---|
| 0 | All participants in the four demographic files | 39,156 | | |
| 1 | Age 20 or older at screening | 22,617 | 16,539 | |
| 2 | Told by a clinician they have diabetes (DIQ010 = 1) | 3,146 | 19,471 | |
| 3 | Examined in the mobile examination center | 3,027 | 119 | |
| 4 | Glycohemoglobin value present | 2,913 | 114 | |
| 5 | All nine PHQ-9 items answered 0 to 3 | 2,663 | 250 | |
Counts are unweighted and come from the author's merge of the DEMO, DIQ, GHB and DPQ files for cycles G through J. Per-cycle analytic counts were 575, 627, 728 and 733.
Who Each Loss Removed
Step 3 removed 119 adults who completed the household interview but never came to the examination center. They were older than those retained, with a mean age of 67.2 years against 61.8, and 21.8% carried the top-coded age of 80 against 9.6% of the analytic group. Nearly a third had no income to poverty ratio. Their absence likely leaves the oldest and least connected adults with diabetes underrepresented, which the examination weight is designed to offset only in part (Chen et al., 2020).
Step 4 removed 114 adults who were examined but have no glycohemoglobin value. This group differed most in composition: 50.9% were non-Hispanic Black, compared with 25.9% of those retained, and 57.9% were women. The documentation does not explain the missing values, so I cannot say whether they reflect refused blood draws, failed samples or something else. This is the loss most likely to bias comparisons by race and ethnicity, and the results will say so.
Step 5 removed 250 adults with no complete PHQ-9. Of these, 223 have no depression screener record at all, and the rest refused or did not know at least one item. They had higher glycohemoglobin, a mean of 7.74% against 7.47%, more insulin use, 34.0% against 28.5%, and nearly twice the share at the age cap. If people who skip the screener are also more often depressed, the main comparison will understate the difference, a pattern Sterne et al. (2009) describe for data that are not missing at random.
What the Derivation Settles
The analytic group is 2,663 adults, of whom 369 score 10 or higher on the PHQ-9. All design variables are kept for the full file, so variance estimation will treat the analytic group as a subpopulation rather than deleting the other records. The losses at steps 4 and 5 will be reported in the limitations, and a sensitivity check will repeat the main estimate with partial PHQ-9 records scored.
References
Chen, T. C., Clark, J., Riddles, M. K., Mohadjer, L. K., & Fakhouri, T. H. I. (2020). National Health and Nutrition Examination Survey, 2015-2018: Sample design and estimation procedures (Vital and Health Statistics, Series 2, No. 184). National Center for Health Statistics.
Sterne, J. A. C., White, I. R., Carlin, J. B., Spratt, M., Royston, P., Kenward, M. G., Wood, A. M., & Carpenter, J. R. (2009). Multiple imputation for missing data in epidemiological and clinical research: Potential and pitfalls. BMJ, 338, Article b2393. https://doi.org/10.1136/bmj.b2393
von Elm, E., Altman, D. G., Egger, M., Pocock, S. J., Gøtzsche, P. C., & Vandenbroucke, J. P. (2007). The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. The Lancet, 370(9596), 1453-1457. https://doi.org/10.1016/S0140-6736(07)61602-X
What the HLTH 8560E Week 6 instructions ask for
Students document how they moved from the full dataset to the records their analysis uses. Faculty expect each inclusion or exclusion rule stated precisely, applied in a stated order and followed by the number of records remaining. A flow table or diagram is standard. The most important part, and the one most often skipped, is a description of the people each rule removed, compared with those retained on characteristics that matter for the question. Rules that define the population should be distinguished from rules that lose members of it through missing data. Faculty also expect the derivation to keep design variables for the full file so variance can be estimated correctly, and to say what the losses could mean for the results.
Inside the HLTH 8560E Week 6 example
A purpose paragraph cites the STROBE statement by von Elm and colleagues for reporting participant flow. Table 1 lists six steps, from all 39,156 participants to the 2,663 retained, with remaining and removed counts, and a note gives the analytic count for each cycle. A short paragraph separates the two defining rules from the three losses totaling 483 adults. Three paragraphs then describe each loss: the unexamined adults by age, top-coding and missing income, with Chen and colleagues cited on weighting; the adults missing glycohemoglobin by race and sex; and the adults missing PHQ-9 data by glycohemoglobin and insulin use, with Sterne and colleagues cited. A final section gives the 369 adults with symptoms and the planned sensitivity check.
HLTH 8560E Week 6 rubric: what earns full marks
The grade depends most on whether removed records are described, not just counted. A derivation that reports a final sample size with a list of exclusions shows procedure; one that compares the removed with the retained shows judgment about bias. Faculty check that counts add up from step to step and that rules are precise enough to reproduce. Credit rises when the derivation distinguishes defining rules from losses and names the likely direction of bias for each loss. Keeping the full design for variance estimation, instead of deleting records, is expected in complex survey work and earns credit when stated. Honest reporting of an unexplained loss, such as missing laboratory values with no stated cause, is rewarded over speculation. A closing link to the analysis plan is valued.
Common HLTH 8560E Week 6 mistakes, and how to avoid them
Derivations often present a tidy funnel of numbers with no word about who fell out at each step. A reader then cannot tell whether the final sample still represents the intended population. To write a stronger derivation, list your rules in the order you apply them and record the count after each. Mark which rules define your population and which lose people who belong in it. For every loss, compare the removed group with the retained group on age, sex, race and ethnicity and one or two variables tied to your outcome. Say which way each difference could push your results. Do not delete records from a complex survey file; flag them so variance estimation keeps every cluster. If you want a derivation like this for your own sample, one can be prepared from your rules and counts.
Get a HLTH 8560E Week 6 example written to your instructions
Provide the Week 6 prompt and rubric, your inclusion criteria and the file's starting size if you know it. The derivation, with a flow diagram, coded rules and a kept-versus-excluded comparison, returns inside 24-48h, and your first request carries no fee. Its counts are placeholders made up for the sample, never figures from any real release.
More HLTH 8560E papers and PhD in Public Health samples
- HLTH 8560E Week 1: Dataset Orientation Post
- HLTH 8560E Week 2: Dataset Selection Memo
- HLTH 8560E Week 3: Codebook Review
- HLTH 8560E Week 4: Survey Round Discussion Post
- HLTH 8560E Week 5: Analytic Variable Map
- HLTH 8560E Week 7: Survey Analysis Plan
- HLTH 8560E Week 8: Weighted Results Section
- HLTH 8560E Week 9: Recode Peer Replies
- HLTH 8560E Week 10: Data Constraints Memo
- HLTH 8560E Week 11: Journal Manuscript Draft
- HLTH 8551 Week 6: Research Question Draft
- HLTH 8475 Week 6: Reach Measure Peer Critique
- HLTH 8402 Week 6: Missing Voices Peer Critique
- HLTH 8392 Week 6: Shared Burden Phrase Reply
HLTH 8560E Week 6 questions, answered
Why flag records instead of deleting them?
Because estimates from a complex survey depend on the full design. Standard errors are computed from variation between clusters within strata, and removing records can leave some strata with a single cluster or misstate that variation entirely. Subpopulation or domain statements in survey software keep every record in the design while restricting each estimate to the group you care about.
Should missing values be excluded or imputed at this stage?
The derivation records them; the analysis plan decides. At this stage the job is to show how many adults lack each value and how they differ from the rest. If the plan later imputes, the derivation already holds the counts and comparisons that justify it. Excluding silently now would close off an option before its costs were known.
What belongs in the flow diagram versus the text?
The diagram carries the sequence and the counts; the text carries the exact rules and the reasons for their order. A diagram with long captions becomes hard to read, and text without a diagram hides the arithmetic. Most rubrics expect both, cross-referenced, with each box in the diagram matching a numbered rule in the table.
Where can I find a free HLTH 8560E Week 6 sample paper?
This page presents the complete Week 6 sample derivation, counting every rule from 39,156 pooled NHANES participants to 2,663 adults with diabetes and describing who each loss removed, with three references.
Which loss is most likely to bias the results?
The 114 adults with no glycohemoglobin value were about half non-Hispanic Black, double the share among those retained, so comparisons by race and ethnicity are the most exposed. The 250 missing a complete PHQ-9 had higher glycohemoglobin, which could shrink the main difference.