10 November 2019 5 1K Report

I have 1 independent and 3 mediator variables to predict 2 dependent variables. 4 variables do not meet univariate normality (skewness/S.E < -3.29), histograms are skewed for those. Multivariate normality is met...Do I need to transform the data for these nonnormal ones? Can I use bootstrap for mediation with univariate nonnormal data?

Thanks

More Pınar Acet's questions See All
Similar questions and discussions