Transformation of samples that don't meet normal distribution and equal variance?

[marwa38]

Hi great team

I am using your great tool as a webpage here https://fbmn-statsguide.gnps2.org/

It would be nice to add to the deep analysis approach of your tool, option to transform data in case it is not meeting parametric assumptions. Then add another step again to check for normality and equal variance after tranformation before deciding to run next step in your workflow (ANOVA&Tukey's as parametric or Kruskal_Walis&Dunn's tests as non-parametric). That could be especially helpful in the case as the dataset samples could be meeting only one of the asssumptions but not the other, where transformation will most probably make the samples meeting both of the normality and equal variance assumptions.

Not sure if there is a need to allow to do both parametric and nonparametric based on the samples as not all are meeting the assumptions of the parametric test). For example this dataset:

Thanks
Marwa