forum.smartpls.com

Hi,

could someone explain to me why Percentile Bootstrapping (SmartPLS Viedeo) should be use for MGA purposes
instead of bias-corrected Bootstrapping, please.

I don't understand the reason since bias corrected Bootstrapping is the default in SmartPLS 3
and there is quite no literature about that.

Thank you very much in advance
Best Raffaele

Generally, it is not advisable to use the non-overlapping confidence intervals as a criterion for MGA. It is a too strict criterion. You may also have significant differences when confidence intervals overlap slightly. With normal theory confidence interval they may overlap by half their length to maintain a 5% confidence interval. However, there is no such clear criterion in a PLS context.