Regression

Regression is available as a standalone analysis independent of the SEM model builder -- useful when the research question is a direct predictor-outcome relationship rather than a full structural model.

Multiple linear regression reports unstandardized and standardized coefficients, R² and adjusted R², VIF for multicollinearity, and a full ANOVA table. Logistic regression reports coefficients, odds ratios, and McFadden's pseudo-R² for binary outcomes.

Both support bootstrapping for robust standard errors when the classical assumptions are in doubt.