Configurational Strategy Analysis (fsQCA)

Regression estimates the average, independent net effect of each variable while holding others constant, assuming effects are additive and symmetric. fsQCA, following Ragin, treats causation as set-theoretic and combinatorial: it asks which combinations of conditions are sufficient for an outcome, allows several different combinations to produce the same result (equifinality), and treats the explanation of an outcome as distinct from that of its negation (asymmetry). It identifies sufficiency relations and recipes rather than effect sizes, making it suited to theories about configurations rather than isolated drivers.

Calibration converts raw measures into fuzzy-set membership scores from 0 to 1 using three substantive anchors - full membership, full non-membership, and the 0.5 crossover of maximum ambiguity. Ragin emphasizes that these anchors must be grounded in theory and external knowledge of what the condition means, not derived mechanically from the sample. Because every subsequent set operation depends on these memberships, poor or arbitrary calibration invalidates the analysis; calibration is where qualitative knowledge enters and where much of the method's rigor - and its sensitivity - resides.

Fiss introduced this distinction by comparing the parsimonious and intermediate fsQCA solutions. Conditions that appear in both are core: they are part of the essential causal recipe and show a strong, stable relationship to the outcome. Conditions appearing only in the intermediate solution are peripheral: they contribute to the configuration but are more dispensable. The distinction lets researchers describe organizational types in terms of their indispensable versus supporting elements, giving configurational findings clearer theoretical meaning than an undifferentiated list of conditions would.