When to use this
Use this when theory specifies which indicators measure each reflective construct.
Why it matters
CFA separates measurement specification from structural hypotheses and provides multiple sources of fit and validity evidence.
Step-by-step instructions
1. Review and approve constructs and indicators.
2. Open the measurement-model builder.
3. Define the reflective factor structure and identification.
4. Run CFA on the intended immutable dataset version.
5. Review convergence, loadings, residuals, global fit, reliability, and validity evidence.
6. Revise theory explicitly rather than modifying the model automatically.
Expected outputs
An immutable CFA result with fit indices, loadings, residual diagnostics, reliability, validity, and specification provenance.
How to interpret the results
CFI, TLI, RMSEA, and other indices are complementary evidence, not universal pass/fail rules. Consider estimator assumptions and substantive plausibility.
Common mistakes
Do not chase fit by adding unsupported correlations. Do not describe fit as proof of validity or use formative constructs in a reflective CFA.
Was this article helpful?
Your response helps the methodology team improve this documentation.