EFA vs CFA: choose from the research decision, not the software menu

Exploration and confirmation answer different questions and require different commitments before estimation.

Use EFA when structure is genuinely uncertain

EFA can explore dimensionality and loading patterns, but the choices of estimator, rotation, factor retention, and item treatment still need justification.

Use CFA for a specified measurement model

CFA is strongest when indicators, factor relations, cross-loadings, residual relations, and identification are reasoned about before fit statistics are inspected.

Avoid calling a data-driven revision confirmatory

If the same sample generates and evaluates extensive changes, label the work exploratory and seek new data or resampling evidence where feasible.

Failure modes to check

• Choosing a method because it is familiar rather than because it matches the estimand and data structure.
• Treating assumptions as a generic pass/fail checklist instead of evidence about a specific model.
• Changing the workflow after seeing results without recording the decision or its effect on uncertainty.
• Reporting software output without tracing the claim to final code, diagnostics, and limitations.

Continue the decision path

Explore the related statistical consulting service, or discuss the decision in your own study.

Related decision guides

• Choosing a regression family from the outcome and data structure
• What to decide about missing data before regression