Bootstrapping
Statistics & analysis
A resampling method that estimates the stability of a statistic by re-drawing thousands of samples from the data — used in PLS-SEM to test path significance without normality assumptions.
Bootstrapping estimates the sampling distribution of a statistic by repeatedly resampling the observed data with replacement — typically 5,000 or 10,000 times — and recalculating the statistic in each resample. The spread of those estimates becomes the standard error and the confidence interval.
Its appeal is that it assumes nothing about normality. That makes it the standard approach for statistics with unknown or awkward sampling distributions, most notably indirect effects in mediation analysis and all path significance testing in PLS-SEM, where no parametric formula exists.
Report the number of resamples, the interval type (percentile, or bias-corrected and accelerated) and the interval itself. An effect is supported when the confidence interval excludes zero. Because resampling is random, set a seed if you need your results to reproduce exactly.
Where it's used
- Significance testing of path coefficients in SmartPLS (5,000+ resamples)
- Confidence intervals for indirect (mediation) effects
Software used
Guides on this topic
Guides that use this concept
All guidesNot sure where to start?
Book a free 15-minute consult. We'll map your next three steps — no obligation.