NURS 8310 Week 2: Strengths and Limitations of Secondary Data Sources — Guide and Example

Population Health Problem: Adult Obesity Prevalence in Rural U.S. Counties

Adult obesity is a measurable, data-rich population health problem with well-documented rural-urban disparity, making it a defensible Week 2 selection. National Health and Nutrition Examination Survey (NHANES) data from August 2021–August 2023 places age-adjusted adult obesity prevalence at 40.3%, with severe obesity at 9.4% and significantly higher rates in adults ages 40–59 than in younger or older cohorts (age-adjusted prevalence of U.S. adults age 20 and older with obesity was 40.3%, including 9.7% with severe obesity and another 31.7% who were overweight). The association of interest for this worked example: the relationship between rurality and obesity-related mortality among U.S. adults aged 25 and older.

Data Sets and Variables (40 pts)

Data Set 1 — Behavioral Risk Factor Surveillance System (BRFSS) / PLACES

BRFSS is a CDC-administered, state-based telephone survey of the noninstitutionalized adult population, conducted annually across all 50 states, D.C., and three U.S. territories (BRFSS supported 26 modules in 2024 and is reviewed annually by a comprehensive body of published scholarly studies confirming its validity and reliability). Its small-area-estimation derivative, PLACES, models BRFSS responses down to the county and census-tract level.

Variables needed for this association:

• Self-reported BMI category (calculated from height/weight items)
• County FIPS code (for rural-urban linkage via Rural-Urban Continuum Code or RUCA)
• Healthcare access items: routine checkup interval, primary care provider status, cost-related care delay
• Sociodemographic stratifiers: age, sex, race/ethnicity, education, household income

Data Set 2 — National Health and Nutrition Examination Survey (NHANES)

NHANES combines structured interviews with clinical examination and laboratory measurement, making it the only one of the three data sets that uses measured, not self-reported, height and weight (measuring height and weight to estimate obesity prevalence is important because self-reported estimates underestimate weight and overestimate height).

Variables needed for this association:

• Measured BMI (examination component)
• Examination sample weights (required for nationally representative estimation given the multistage probability design)
• Education level, age group, sex (NHANES 2021–2023 cycle reports obesity prevalence stratified by all three)
• Geographic region (NHANES does not support county-level estimates, a limitation addressed below)

Data Set 3 — CDC WONDER Mortality Files

CDC WONDER provides public, de-identified death-certificate data coded by ICD-10, enabling direct mortality-rate calculation rather than prevalence estimation alone.

Variables needed for this association:

• ICD-10 underlying/contributing cause codes (E66.0–E66.9 for obesity-related mortality)
• Age-adjusted mortality rate (AAMR) per 100,000, standardized to the 2000 U.S. population
• Urbanization classification (NCHS 6-level urban-rural classification scheme)
• Year, sex, race/ethnicity, census region

Validity Assessment (40 pts) — The Highest-Risk Criterion

BRFSS validity and prior use

BRFSS reliability and validity has been assessed through systematic peer review and cross-survey comparison against the National Health Interview Survey, NHANES, and other federal data collections (many BRFSS questions are derived from other national surveys, allowing prevalence estimates to be compared and validated against them). Confirmed prior peer-reviewed use includes a 2021 Preventive Medicine study analyzing 2016–2019 BRFSS data to assess gender disparities in healthcare access and cost-related medication non-adherence (Daher et al., 2021), and a 2024 Human Vaccination & Immunotherapeutics database analysis of BRFSS-derived adult vaccination coverage.

A known limitation directly relevant to this association: BRFSS is a dual-frame (landline/cellular) telephone survey, and response rates vary meaningfully by state — the 2022 cycle carried a 45.0% national response rate, ranging 22.8%–66.8% by state (the 2022 cycle of BRFSS had a response rate of 45.0%, with a state range of 22.8%–66.8%).

NHANES validity and prior use

NHANES is the federal reference standard for obesity prevalence precisely because it substitutes clinical measurement for self-report, eliminating the social-desirability bias inherent to telephone or interview-based surveys. Confirmed prior peer-reviewed use includes NCHS Data Brief No. 508 (Emmerich, Fryar, Stierman, & Ogden, 2024), the federal government’s own August 2021–August 2023 obesity prevalence report, and a 2025 ScienceDirect sociodemographic disparities analysis comparing 2017–2020 and 2021–2023 NHANES cycles by income, education, and insurance status.

A known limitation: the 2021–2023 cycle’s examination response rate fell to 25.6%, and the NHANES sample (roughly 5,000–8,860 adults annually) is sized for national, not state-level, estimation (the 2021-2023 cycle of NHANES had a response rate of 34.5% for the interview and 25.6% for the examination, with an analytic sample of 8,860 adults) — meaning NHANES alone cannot answer a county-level rurality question and must be paired with a geographically resolved source such as BRFSS/PLACES.

CDC WONDER validity and prior use

CDC WONDER mortality data has substantial, recent peer-reviewed precedent specific to obesity-related mortality. A 2026 Obesity Science & Practice analysis used CDC WONDER to evaluate obesity-related mortality trends across nearly six decades (1968–2025), finding age-adjusted mortality rates increased from 1.15 to 3.32 per 100,000 (Ahmed et al., 2026). An earlier 2024 Cureus analysis of the same database covering 2010–2020 similarly documented rising obesity-related mortality with clear age, sex, and racial stratification (Achara et al., 2024). Both studies disclose the same limitation that applies directly to this assignment: CDC WONDER depends on death-certificate coding accuracy, and contributing-cause reporting for obesity is inconsistent across states and coding practices, which can understate true mortality burden.

Access and Permission Challenges (35 pts)

A generic “data access takes time and money” statement scores in the Fair-to-Poor range. A defensible response names the specific mechanism:

• Restricted-variable suppression. County-level identifiers in public-use BRFSS files are recoded when a state has too few rural counties to prevent re-identification — the 2023 BRFSS comparability documentation confirms that single-county nonmetropolitan categories are recoded to an adjacent category specifically as a disclosure-avoidance measure (states with a single county in a nonmetropolitan category require recoding to an adjacent category as a disclosure-avoidance measure). This directly limits county-level rural analysis using the public file alone.
• Restricted Data Center applications. Where county-level NHANES or BRFSS linkage is required beyond what public files allow, researchers must apply through the NCHS Research Data Center, which involves a formal proposal, IRB documentation, projected analysis plan, and a multi-week to multi-month federal review cycle before access is granted.
• Small-cell suppression in CDC WONDER. CDC WONDER automatically suppresses mortality counts under 10 per geographic/demographic cell to protect privacy, which restricts rural-county-level mortality analysis where obesity-related deaths are infrequent in any single year.
• Sample-weight and survey-design competency. All three sources require complex-survey statistical methods (replicate weights, Taylor series linearization, or jackknife variance estimation) rather than simple descriptive statistics — a methodological barrier as real as a permission barrier, since improperly weighted analysis produces biased estimates even when access is granted.

Comparison Table: Three Data Sets at a Glance

Strengths and Limitations of Secondary Data Sources

John Smith

Walden University

NURS-8310: Epidemiology & Population Health

Dr. Susan White

06/30/2026

Strengths and Limitations of Secondary Data Sources Example

The secondary sources of data are important in the population health studies as they present timely, cost-effective, and detailed facts about the trends of population health, disease burden, and risk elements in a wide variety of populations. Such datasets are particularly useful to researchers who do not need the financial and logistical constraints that a primary data collection would necessitate to analyze large-scale health issues (Friis and Sellers, 2021).

The population health issue addressed in this paper is obesity among adults in the United States, which is a leading cause of chronic diseases including diabetes, cardiovascular disease, and some cancers. This paper aims to examine three sources of secondary data applicable to this issue and discuss their variables, validity, and research issues.

Population Health Problem: Adult Obesity in the United States

Obesity is one of the most challenging yet chronic population health issues in the United States that affects adults. National surveillance reports indicate that prevalence of obesity is still on the increase, causing morbidity to increase, quality of life to be decreased, and healthcare expenditure to increase (Mehta, 2023). The trend, determinants, and disparities of obesity are critical to the formulation of effective interventions and the public health policy. The secondary data sets that reflect the data on weight status, demographics, behavioral risk factors, and environmental effects are essential in the accurate description of the magnitude and the extent of this issue, especially in relation to the high-risk subpopulations (Friis & Sellers, 2021).

Data Source 1: CDC Behavioral Risk Factor Surveillance System (BRFSS)

Variables Needed to Examine the Association

BRFSS is a telephone-based survey of health consequences that is among the largest and constant surveys in the world and offers a broad range of data regarding health behavior of adults every year. The variables necessary to analyze obesity are height and weight measured self-reportedly to compute Body Mass Index (BMI), physical activity, diet, demographics (age, sex, race/ethnicity, education, income), chronic diseases, and socioeconomic indicators. These variables enable the researcher to learn about associations between obesity and behavioral, demographic, and social determinants.

Validity of the Data Set

BRFSS has been extensively validated and is commonly applied in peer-reviewed studies. According to the Centers of Disease Control and Prevention (CDC), it is considered a primary source of data to track the national and state level health indicators (CDC, n.d.). Although self-reported data will result in measurement bias (especially underreporting of weight) due to the large sample size and standardized methodology will increase reliability. BRFSS has been established to be valid as many studies have conducted research on obesity prevalence, risk factors, and disparities in health.

Data Source 2: National Health and Nutrition Examination Survey (NHANES)

Variables Needed to Examine the Association

NHANES offers high quality and nationally representative data that are integrated with interviews and physical examination. The important variables used to determine obesity are height and weight measured objectively, waist circumference, dietary intake, laboratory biomarkers, physical activity, sociodemographic variables, and clinical histories. The NHANES enables more precise calculations of obesity as compared to self-reported measures, used in other data collections, and makes it possible to conclude about connections between obesity and health outcomes more strongly.

Validity of the Data Set

According to the National Center of Health Statistics (NCHS, 2015), NHANES is one of its flagship data sets and certifies its long-term use in high-impact research and federal health reports. NHANES is regarded as a gold standard, as physical examination and laboratory tests are performed by trained specialists. This significantly enhances measurement validity. NHANES has assisted in the establishment of research based on obesity patterns, nutritional risk variables, and chronic illness occurrence, with an abundant validation by peers. Its greatest weakness is the smaller sample size than that of BRFSS, which can limit subgroup examination.

Data Source 3: World Health Organization Global Health Observatory (GHO)

Variables Needed to Examine the Association

WHO Global Health Observatory offers international comparative data, which could be used to analyze obesity in a global context. The variables of interest are country prevalence rates of obesity in adults, risk factors within the diet, physical inactivity, economic factors, and environmental health variables. These variables enable the researcher to contrast the obesity level in the U.S. with the other countries across the world as well as analyzing the influence of the structural, environmental, and socioeconomic aspects that are not equal across all counties.

Validity of the Data Set

The GHO is maintained by the World Health Organization (2021) based on the information provided by governments of countries and certified by the global reporting standards. GHO data have been highly mentioned in international health-related studies, comparative chronic disease studies, and global burdens of disease publications. Nevertheless, the quality of the data can differ depending on country of reporting infrastructure and updates might not occur as frequently as existent national-level U.S. data sets. Nevertheless, the GHO is an authoritative and credible source of information globally.

Challenges in Identifying or Securing Proper Data Sets

There are a number of problems that are usually faced by researchers in handling secondary data sources. First, data sets may not be available and complete over time, particularly where variables required to be associated are not available or measured in a different manner across years. Second, restrictions may present a serious barrier in terms of access. Although BRFSS and GHO data is publicly available, NHANES and other NCHS data sets could entail restricted-use agreements, Institutional Review Board (IRB) review, or data-use contracts, extending the research process.

Third, when choosing a data set, researchers have to assess the validity of data and the risks of bias. For instance, BRFSS uses self-reported measures, which can be used to underestimate the prevalence of obesity.

NHANES, on the other hand, is more precise with smaller samples thus complicating the comparison of subgroups. Fourth, technical proficiency would be necessary to work with large data sets, complicated weighting methodologies and statistical analysis needed to produce nationally representative estimates. Lastly, ethical and privacy aspects, particularly in data sets that involve sensitive health or demographics data, might demand formal permission or secure information surroundings. Consequently, when determining the most suitable data set, it may be necessary to compromise the quality of data, its access and methodological practicability.

Conclusion

Secondary data sources offer critical avenues in population health research by allowing access to vast, high quality and cost-effective data on current pressing public health issues like obesity in adults. BRFSS, NHANES and WHO Global Health Observatory have their own advantages in terms of scale, quality of measurement and global comparability, but also suffer the disadvantage of some limitations in terms of data accuracy, availability and complexity. The knowledge of these strengths and limitations can guide the researcher to identify the right data source that can be used to answer research questions. Finally, the use of high-quality secondary data contributes to making effective decisions and enhancing evidence-based practice in the field of public health.

References

Centers for Disease Control and Prevention. (n.d.). Data.CDC.gov: Home. https://data.cdc.gov/

Friis, R. H., & Sellers, T. A. (2021). Epidemiology for public health practice ( 6th ed.). Jones & Bartlett.

Mehta N. K. (2023). Obesity as a main threat to future improvements in population health: Policy opportunities and challenges. The Milbank quarterly, 101(S1), 460–477. https://doi.org/10.1111/1468-0009.12635

National Center for Health Statistics. (2015). Resources for researchers. https://www.cdc.gov/nchs/nchs_for_you/researchers.htm

World Health Organization. (2021). WHO Data collections. [Data sets]. https://www.who.int/data/collections

Common Mistakes That Cost Points

• Naming a data source generically (“the CDC”) instead of the specific data set (BRFSS, NHANES, CDC WONDER, PLACES) — this alone caps the data-set criterion at Fair.
• Treating “validity” as a definitional paragraph rather than producing evidence of prior peer-reviewed use, which is what the rubric explicitly asks for.
• Listing access challenges as generic research barriers (cost, time) instead of source-specific mechanisms (cell suppression, RDC application, geomasking).
• Selecting three data sets that all measure the same thing (e.g., three prevalence surveys) rather than pairing prevalence, clinical-measurement, and outcome data — which weakens the “association of interest” framing the rubric rewards.

{
“@context”: “https://schema.org”,
“@type”: “FAQPage”,
“mainEntity”: [
{
“@type”: “Question”,
“name”: “What population health problems work best for NURS 8310 Week 2?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Problems with at least three independently maintained federal or state data sources — obesity, diabetes, maternal mortality, opioid-related mortality, and cardiovascular disease all have BRFSS, NHANES, and CDC WONDER coverage.”
}
},
{
“@type”: “Question”,
“name”: “Does ICPSR count as a secondary data source for this assignment?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “Yes. ICPSR hosts archived secondary data sets accessible through Walden’s library access, and ICPSR-hosted studies typically include their own documented prior-publication history.”
}
},
{
“@type”: “Question”,
“name”: “Why does NHANES alone not work for a county-level analysis?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “NHANES is designed and weighted for national-level estimation only; its examination sample size is too small to support state or county estimates, so it must be paired with a geographically resolved source like BRFSS/PLACES.”
}
},
{
“@type”: “Question”,
“name”: “What’s the difference between BRFSS and PLACES?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “BRFSS is the underlying survey; PLACES is a CDC small-area-estimation model built from BRFSS responses that produces county- and census-tract-level estimates the raw BRFSS survey cannot support.”
}
},
{
“@type”: “Question”,
“name”: “How many peer-reviewed sources does the validity section require?”,
“acceptedAnswer”: {
“@type”: “Answer”,
“text”: “The rubric does not set a fixed number, but Excellent-tier responses document at least one confirmed prior peer-reviewed use per data set plus the data set’s own technical documentation.”
}
}
]
}

Resources

Achara, K. E., Iyayi, I. R., Erinne, O. C., Odutola, O. D., Ogbebor, U. P., Utulor, S. N., Abiodun, R. F., Perera, G. S., Okoh, P., & Okobi, O. E. (2024). Trends and patterns in obesity-related deaths in the US (2010–2020): A comprehensive analysis using CDC WONDER data. Cureus, 16(9), e68376. https://doi.org/10.7759/cureus.68376

Ahmed, M., Deshmukh, F., Bakr, M., Patel, S., Hossain, M., Cheema, A. H., & Ishtiaq, A. (2026). Trends and disparities in obesity-related mortality among U.S. adults: A CDC WONDER analysis (1968–2025). Obesity Science & Practice, 12, e70146. https://doi.org/10.1002/osp4.70146

Centers for Disease Control and Prevention. (2024). Behavioral Risk Factor Surveillance System: 2024 summary data quality report. U.S. Department of Health and Human Services.

Centers for Disease Control and Prevention. (2024). Comparability of data: BRFSS 2023. U.S. Department of Health and Human Services.

Daher, M., Al Rifai, M., Kherallah, R. Y., Rodriguez, F., Mahtta, D., Michos, E. D., Khan, S. U., Petersen, L. A., & Virani, S. S. (2021). Gender disparities in difficulty accessing healthcare and cost-related medication non-adherence: The CDC Behavioral Risk Factor Surveillance System (BRFSS) survey. Preventive Medicine, 153, Article 106779. https://doi.org/10.1016/j.ypmed.2021.106779

Emmerich, S. D., Fryar, C. D., Stierman, B., & Ogden, C. L. (2024). Obesity and severe obesity prevalence in adults: United States, August 2021–August 2023 (NCHS Data Brief No. 508). National Center for Health Statistics. https://doi.org/10.15620/cdc/159281