Interpreting Quantitative and Qualitative Results in Public Health Studies

Quick Answer: Quantitative interpretation focuses on point estimates, confidence intervals, p-values, and effect sizes; qualitative interpretation focuses on themes, context, and trustworthiness criteria (credibility, transferability, dependability, confirmability). Mixed-methods studies triangulate both to converge on a defensible conclusion.

Interpreting Quantitative Results

A point estimate (a mean, proportion, RR, OR) is the best single guess for a population parameter, but it is always flanked by uncertainty. A 95% confidence interval (CI) is interpreted as: if the study were repeated many times, 95% of the resulting intervals would contain the true parameter. A common misreading is that there is a 95% probability the true value lies in this specific interval — once computed, the parameter is either inside or outside. The width of the CI reflects precision: wide intervals come from small samples, high variability, or both. A CI that crosses the null (RR = 1, OR = 1, difference = 0) means the result is not statistically significant at that alpha level.

A p-value is the probability of observing data as extreme as or more extreme than what was observed, assuming the null hypothesis is true. A p-value below the pre-specified alpha (commonly 0.05) leads to rejecting the null; it does not measure the probability that the null is true, nor does it measure the size or importance of an effect. A very small p-value from a huge sample can correspond to a trivial effect size. Always report and interpret the effect size (the magnitude of the difference or association) alongside the p-value and CI.

Interpreting Qualitative Results

Qualitative findings are not summarized by means and p-values; they are summarized by themes, patterns, narratives, and contextual explanations. Common analysis approaches include thematic analysis (identifying recurring themes across transcripts), grounded theory (building theory inductively from data), phenomenology (lived experience descriptions), and content analysis (systematic coding of text). Findings are typically supported by direct participant quotes that illustrate each theme.

Qualitative trustworthiness replaces the quantitative notions of validity and reliability with four criteria from Lincoln and Guba:

• Credibility (parallel to internal validity): do the findings represent the participants' realities? Established through prolonged engagement, triangulation of sources, member checking, and peer debriefing.
• Transferability (parallel to external validity): can findings transfer to similar contexts? Established through thick description of the setting, participants, and context so readers can judge applicability.
• Dependability (parallel to reliability): are the findings consistent and the process auditable? Established through an audit trail and external review.
• Confirmability (parallel to objectivity): do the findings reflect the data, not researcher bias? Established through reflexivity, audit trails, and confirmability audit.

Mixed-Methods Interpretation and Triangulation

Mixed-methods designs combine quantitative and qualitative components. Convergent mixed methods collect both forms simultaneously and compare findings: do the numbers and the narratives agree? Explanatory sequential designs collect quantitative first, then use qualitative follow-up to explain unexpected quantitative results. Exploratory sequential designs start with qualitative exploration, then build a quantitative instrument or test hypotheses. Triangulation is the process of comparing evidence from different sources or methods; results can be convergent (both methods agree), complementary (each method adds a distinct facet), or divergent (methods disagree and the discrepancy itself becomes the finding to investigate).

A common interpretation error is treating qualitative findings as hypothesis-generating only and quantitative findings as the real truth. Both carry inferential weight; a well-conducted qualitative study with thick description and member-checked themes can support strong conclusions about context, acceptability, and mechanisms that numeric data cannot capture. Conversely, a quantitative study with a biased sampling frame can produce precise but misleading estimates. Public health reports should explicitly state the type of evidence, its strengths and limitations, and how convergent or divergent findings were reconciled.

Effect modification and confounding also shape interpretation. Effect modification (interaction) means the association differs across strata of a third variable (e.g., a vaccine works well in adults but not elderly); it should be reported as stratum-specific estimates, not a single pooled estimate. Confounding means a third variable distorts the exposure-outcome association; it is addressed through study design (restriction, matching, randomization) or analysis (stratification, regression adjustment). A result that flips direction after adjusting for a confounder (Simpson's paradox) demands explicit discussion in any report.

Worked example of confounding: A cross-sectional study finds coffee drinkers have higher pancreatic cancer rates (RR = 2.0). After adjusting for smoking — a confounder because smoking causes pancreatic cancer and coffee drinkers are more likely to smoke — the adjusted RR drops to 1.1. The apparent association was confounded by smoking. Worked example of effect modification: A vaccine trial finds overall efficacy of 70%, but stratified analysis shows 90% efficacy in adults under 65 and only 20% efficacy in adults over 65. Reporting only the pooled 70% would mask the important finding that the vaccine's effect differs by age group. The correct approach is to report stratum-specific estimates and test for interaction, not to pool across strata. This distinction — confounding distorts and should be removed, effect modification is real and should be reported — is a frequent CPH exam theme.