Bootstrapping

Bootstrapping resamples your dataset (with replacement) a configurable number of times -- 1,000, 2,000, 5,000, or 10,000 -- re-estimating the model on each resample to build an empirical distribution for every coefficient.

Results include standard errors, t-values, p-values, and confidence intervals for path coefficients, indirect effects, and other estimates that a single-sample run cannot provide significance for on its own.

A fixed random seed can be set for reproducibility, so a bootstrap run can be repeated later with identical results.