Quantitative-Dominant Case-Focused Mixed Methods

A traditional case study can be entirely qualitative or descriptive. This design deliberately combines quantitative data (as the dominant strand) with qualitative data (as a secondary strand) within a bounded case, and requires explicit integration of the two types of evidence to produce meta-inferences. The case structure provides the unit of analysis; the mixed methods logic governs data collection, analysis, and integration.

Yes. Multi-case versions of this design are common in comparative evaluation or organisational research. Each case provides a distinct context in which the quantitative strand yields site-specific results, and the qualitative strand explains case-level variation. Cross-case integration then allows the researcher to draw conclusions that span the cases while respecting their individual boundaries.

Both concurrent and sequential timings are legitimate. In a concurrent version, both strands are collected during the same field period and integrated at analysis; this is efficient but requires careful coordination. In a sequential version (typically QUAN first), qualitative data are collected after preliminary quantitative results are available, so the interviews or observations can target specific patterns or anomalies — this is especially useful when the quantitative findings are surprising or require contextual explanation.

The most widely recommended technique is a joint display — a table or figure that places quantitative outcomes and qualitative themes side by side for each case or sub-group, making convergence, divergence, and complementarity visible. The meta-inference section of the discussion should then explicitly address what the combined evidence, taken as a whole, reveals that neither strand could reveal alone.

Because the qualitative strand is secondary and purposive, the sample is usually much smaller than for a standalone qualitative study. Enough participants or data sources to illuminate the quantitative findings is the guiding principle — often 5–15 interviews within each case, selected to represent relevant variation (e.g., high versus low performers on the quantitative measure). Independent saturation is not the goal; explanatory adequacy is.