| Course | NU704 Biostatistics, Epidemiology, and Social Determinants of Population Health |
|---|---|
| Unit | Unit 6 |
| Paper type | Secondary data appraisal |
| Length | About 880 words, 6 pages double-spaced |
| Format | APA 7 student paper |
| School | Purdue University Global |
| Program | Doctor of Nursing Practice (DNP) |
| Updated | October 2026 |
Free sample paper for NU704 Unit 6
What a County's Data Can and Cannot Say: A Unit 6 Secondary Data Appraisal for Diabetes Outcomes Around Dubuque
[Student Name]
Purdue University Global
NU704: Biostatistics, Epidemiology, and Social Determinants of Population Health
Unit 6 Secondary Data Appraisal
[Instructor Name]
[Date]
The county figures are illustrative examples of what each source provides; the data sources and studies cited are real.
The Question
The project team wants to describe the diabetes burden in the hospital's county for the project proposal: how common diabetes is, how many deaths are attributed to it and whether the picture differs from the state. These are population questions that hospital records cannot answer, since they cover only patients who come to the hospital.
First Source: CDC WONDER Mortality Data
CDC WONDER provides online access to national vital statistics, including deaths by underlying and contributing cause, from death certificates, with queries by county, year, age, sex and race (Centers for Disease Control and Prevention [CDC], n.d.-a).
Strengths: the data cover essentially all deaths, are coded consistently and allow age-adjusted rates that can be compared with the state.
Limitations: diabetes is often a contributing rather than underlying cause, so counts by underlying cause understate its role, and certificate completion varies. For a single county, counts can be small. To protect privacy, CDC WONDER hides any county count under ten, and it marks a rate built on fewer than twenty deaths as unreliable. A county the size of Dubuque's may have too few diabetes deaths in one year for a stable rate.
Use: the team will combine five years to obtain a stable age-adjusted rate, query both underlying and multiple-cause deaths, and report the state rate alongside for context.
A single year of a single county's deaths is a story, not a rate.
Second Source: BRFSS Survey Data
The BRFSS is an annual telephone survey of adults in every state that asks, among many topics, whether a doctor has ever said the respondent has diabetes (CDC, n.d.-b). State estimates are published each year, and county estimates are available in some states or through modeled small-area estimates.
Strengths: it is the main source of diabetes prevalence estimates at state level, includes risk factors such as physical activity and obesity and allows trends over time.
Limitations: diagnosis is self-reported, so undiagnosed diabetes is missed. Pierannunzi and colleagues' review judged its prevalence figures broadly in line with other questionnaire-based surveys, while noting larger gaps when compared with surveys that measure people physically (Pierannunzi et al., 2013). County estimates rest on small samples or on statistical models that borrow strength from other areas, so their precision is limited and their confidence intervals wide.
Use: the team will report the state prevalence with its confidence interval and the modeled county estimate with an explicit note that it is model-based.
Checks Before Any Figure Is Used
Before a number from either source enters the proposal, the team will run four checks. First, read the source's own documentation for the year in question, since definitions and survey methods change; the BRFSS, for example, changed its weighting method in 2011, so earlier and later estimates are not directly comparable. Second, confirm the geography: a county query should match the county the hospital serves, not the metropolitan area. Third, look at the denominator and the count together, because a rate that seems striking may rest on a handful of events. Fourth, record the exact query or table, the date of retrieval and any suppression or reliability flags, so that another person could reproduce the figure. These checks take minutes and prevent the most common errors in using public data, which are comparisons across incompatible years and small numbers presented as stable rates.
Fit With the Project's Question
Neither source describes the project's patients directly. Mortality data describe deaths, not readmissions; survey data describe prevalence among adults who answer the phone, not patients leaving a hospital on insulin. Both are useful for context, explaining why diabetes matters in this county, but neither can serve as a baseline or comparison for the project's outcomes, which will come from the hospital's own records.
Illustrative Use
To show how the sources would appear in the proposal, the team drafted a sentence with placeholder values: "Over five combined years, the county's age-adjusted death rate with diabetes as underlying cause was similar to Iowa's, and modeled estimates suggest about one adult in ten has diagnosed diabetes, with a wide interval." Actual figures will be retrieved directly from the sources at the time of writing and cited with the date of retrieval.
Rules for Responsible Use
Five rules follow from the appraisal. Combine years when counts are small. Report intervals with every estimate. Name whether a figure is observed or modeled. Do not compare a county estimate with a state figure as if both were equally precise. And never present population data as if they described the project's patients.
Conclusion
CDC WONDER and the BRFSS can describe the diabetes burden around the project's hospital, each with characteristic limits: small counts and cause-of-death coding for mortality data, self-report and modeled precision for survey data. Used with combined years, intervals and clear labels, they provide context for the project; used carelessly, they could make a county look different from its true situation.
References
Centers for Disease Control and Prevention. (n.d.-a). CDC WONDER. U.S. Department of Health and Human Services. https://wonder.cdc.gov
Centers for Disease Control and Prevention. (n.d.-b). Behavioral Risk Factor Surveillance System. U.S. Department of Health and Human Services. https://www.cdc.gov/brfss
Pierannunzi, C., Hu, S. S., & Balluz, L. (2013). A systematic review of publications assessing reliability and validity of the Behavioral Risk Factor Surveillance System (BRFSS), 2004-2011. BMC Medical Research Methodology, 13, Article 49. https://doi.org/10.1186/1471-2288-13-49
Reading the NU 704 Unit 6 instructions
The sixth unit commonly asks students to appraise one or more public data sources for a population health question related to their practice. Prompts expect the question stated, each source described with its collection methods, definitions and coverage, strengths and limitations identified and a plan for how the data will be used responsibly. Instructors reward appraisals that address small numbers, suppression rules, self-report bias and the difference between observed and modeled estimates, and that say plainly when a source cannot answer the question. Linking the appraisal to the student's project, by explaining what context the data provide and what they cannot replace, is often expected. Common sources include CDC WONDER, the National Center for Health Statistics and the BRFSS. Citations should include the data sources themselves and at least one study examining the reliability or validity of a source.
Inside the NU704 Unit 6 example
The appraisal states the project team's question and why hospital records cannot answer it. CDC WONDER is described, its strengths listed, then its limits: diabetes as contributing cause, certificate variation and small county counts with suppression rules, under a highlighted line about a single county's year. Its use is set as five combined years with state comparison. The BRFSS section covers self-reported diagnosis, Pierannunzi and colleagues' validity review and modeled county estimates, with a margin note on stating use as well as flaws. A section of four checks, from documentation to recording the query, guards against common errors. A fit section explains why neither source can be the project's baseline. An illustrative sentence shows how the data would appear in the proposal, and the appraisal ends on five rules, the subject of a second margin note.
Reading the NU704 Unit 6 grading rubric
Data appraisals are graded on a clearly stated question, accurate description of each source, identification of strengths and limitations, attention to small numbers and modeling, fit with the question and a plan for responsible use. Description credit requires collection method, definitions and geographic coverage. Limitations earn the most when they are specific, such as suppression thresholds or self-report bias, rather than generic. Graders reward recognition that a source may not answer the question at all, and a clear statement of what it can provide instead. A plan for responsible use, including combining years, reporting intervals and labeling modeled figures, shows practical skill. Pre-use checks and documentation of queries demonstrate rigor. Citations of the sources and of a validity study, with a clear structure, complete the rubric.
Common NU704 Unit 6 mistakes, and how to avoid them
Appraisals lose marks most often when they describe a source's features from its website without connecting them to the question. Another frequent error is quoting a single-year county rate as if it were stable, or comparing a modeled county estimate with an observed state figure as equals. Some students overlook that cause-of-death coding can understate a condition like diabetes. Others present population data as a baseline for a hospital project, which mixes different populations. Failing to record the query and retrieval date makes figures impossible to check. Begin by writing your question in one sentence and the population it concerns, then ask whether each source describes that population at all. A source that fails that test can still give context if labeled honestly. Your question and the sources you were assigned can become a custom appraisal.
Get an NU704 Unit 6 example written to your instructions
Send the question, the data source and the appraisal instructions with the rubric. A custom appraisal explains what the data can show, where they fall short and how to use them responsibly. It is free the first time and arrives within 24-48h. 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 NU704 papers and DNP samples
- NU704 Unit 1: Biostatistics in Practice Post
- NU704 Unit 2: Regression Interpretation Paper
- NU704 Unit 3: Social Determinants Analysis
- NU704 Unit 4: Statistical Test Selection Worksheet
- NU704 Unit 5: Causal Inference Paper
- NU704 Unit 7: Measures of Association Problem Set
- NU704 Unit 8: Significance Interpretation Paper
- NU704 Unit 9: Data Analysis Plan
- NU704 Unit 10: Statistical Literacy Reflection
- NU713 Unit 6: Disparities Analysis, Heart Failure
- NU703 Unit 6: Writing Seminar Reflection
- NU733 Unit 6: Policy Brief to the City Manager
- NU702 Unit 6: Interprofessional Team Charter
NU704 Unit 6 questions, answered
What public data sources does NU704 use?
Common ones include CDC WONDER, the National Center for Health Statistics, the Behavioral Risk Factor Surveillance System and state health data. Each has strengths and limits.
What should an appraisal of secondary data cover?
Who collected the data and how, definitions, completeness, small numbers, time lags and whether the data fit the question.
Can secondary data answer every question?
No. An appraisal should say plainly when the data cannot answer the question asked.
Where can I find a free NU704 Unit 6 sample paper?
This page holds the full NU704 Unit 6 secondary data appraisal. It appraises CDC WONDER and the BRFSS for describing diabetes around a composite Dubuque hospital and sets checks and rules for using public data responsibly.
Can CDC WONDER give a reliable diabetes death rate for one county in one year?
Often not. Counts under ten are suppressed and rates based on fewer than twenty deaths are flagged as unreliable, so the appraisal combines five years to produce a stable age-adjusted rate.