2.3 Primary, Secondary, Qualitative, and Quantitative Data

Four data distinctions

CHES candidates must sort data along two axes: source and type. Primary data are collected directly for the current assessment, such as a local survey, focus group, key informant interview, windshield ("drive-through observation") survey, listening session, observation checklist, or environmental scan. Secondary data already exist, such as Behavioral Risk Factor Surveillance System (BRFSS) reports, U.S. Census data, hospital discharge data, school attendance records, clinic quality measures, or a previously published Community Health Assessment (CHA).

Quantitative data use numbers to answer how much, how many, how often, or how rates differ; they estimate prevalence, compare groups, identify trends, and set measurable baselines. Qualitative data use words, narratives, and observations to explain meaning, barriers, beliefs, context, and acceptability; they answer why and how.

Choosing the right source

If a team needs fast background on rates, secondary data is efficient. If it needs current local barriers not captured in existing datasets, primary data is necessary. If it needs both size and meaning, mixed methods is best.

Strengths and limitations

Secondary data is usually cheaper and faster, but it may be outdated, too broad, missing key subgroups, or collected for a different purpose. Primary data can be tailored but demands time, permissions, recruitment, and careful design. Quantitative data shows patterns but cannot explain lived experience; qualitative data reveals context but cannot, by itself, estimate prevalence.

The exam tests whether the conclusion fits the data. A focus group with 12 parents can surface barriers and themes, but it cannot establish a countywide rate. A county prevalence rate shows burden, but it cannot explain why one neighborhood avoids a clinic. Match the inference to the data source.

Data quality questions

Before using any source, check five attributes:

• Relevance - do the data answer the current question?
• Timeliness - are the data recent enough for the decision?
• Credibility - is the source trustworthy and the method sound?
• Representativeness - are the priority population and subgroups included?
• Cultural fit - are the tools understandable, respectful, and accessible?

For primary data, also weigh literacy level, language, privacy, mode of administration, sampling method, and response burden. A long online-only survey misses people without reliable internet. An English-only instrument distorts findings in a multilingual community. A focus group on a sensitive topic needs skilled facilitation and confidentiality safeguards.

Triangulation and exam approach

The strongest assessments rarely lean on a single convenient source. They triangulate, comparing multiple sources to see whether findings agree; this raises confidence and prevents one dataset from driving a biased conclusion. When an item asks for the best method, mentally underline the assessment question. If the team asks "how common," pick a quantitative source. If it asks "why," pick a qualitative method. If it asks whether local findings match state trends, pick comparable secondary data. If it asks about assets, look for scans, inventories, partner mapping, and stakeholder input.

Mixed methods and sequencing

Many real assessments combine methods in a deliberate sequence. A team might begin with secondary quantitative data to size the problem and find affected neighborhoods, then run qualitative focus groups to learn why a pattern exists, and finally field a primary quantitative survey informed by what the focus groups revealed. This is an explanatory sequential approach: numbers first, then meaning, then confirmation. The reverse, exploratory sequential, starts with qualitative work to discover what matters and then builds a survey to measure it at scale.

CHES items rarely name these designs, but they reward recognizing that quantitative and qualitative data answer different questions and are strongest together.

Common data-source traps on the exam

• Treating a small focus group or a handful of interviews as if it produced a population rate.
• Using outdated secondary data (for example, a five-year-old national table) to make a current, local decision.
• Choosing an English-only or online-only instrument for a multilingual or low-connectivity community, then trusting the skewed result.
• Confusing primary with secondary based on recency: a brand-new dataset collected for someone else's purpose is still secondary, while a survey you design today for this question is primary.
• Counting outputs (flyers printed, classes held) as evidence of need or impact.

Documentation and reuse

Assessment data should be documented so it can be revisited. A clear codebook for surveys, dated source citations for secondary data, and de-identified summaries of qualitative themes let a team defend its conclusions later and feed the next planning cycle. Good documentation also supports the ethics expectations in Area VIII, particularly confidentiality and honest representation of findings. When an exam item asks how to strengthen the credibility of an assessment, options that add triangulation, document methods, or confirm data definitions are usually stronger than options that simply collect more of the same convenient data.