Marco Torchiano comments

Results 8 comments of


Marco Torchiano

Improve reference for effect sizes Cliffs delta

The paper mentioning the levels for effect size levels is: Exploring methods for evaluating group differences on the NSSE and other surveys: Are the t-test and Cohen’s d indices the...

feature request: Cohen's d and Hedge's g for Welch's tests

Thanks for your suggestion, I’ll look into the papers and think about enhancements Marco. > On 9 Nov 2018, at 22:16, Indrajeet Patil wrote: > > Recently a couple of...

Cliff's delta for repeated measures?

The function `cliff.delta()` currently does not allow computing effect size for paired samples, so the short answer is NO. If you need just an estimate of the delta statistic, you...

The confidence interval does not contain the point estimation value.

Sorry for the delay, I looked into your report today. I cannot reproduce it. Apparently the factor should be `X` and not `Y`, still it has 3 different levels.

Order of labels change the outcome of VD.A

Thanks for the report. I will look into it, I fixed a similar bug in another function.

Bug in geom_xspline

Just realized that there is a simple workaround. The problem is very likely due to the use of `line` instead of `path` as the basic geom. The workaround consists in...

Changed the parent prototype class for GeomXspline

The failure is not related to the proposed change there is a general problem in setting up the VM

cohen.d() is giving different CIs depending if paired is used or the equivalent 1-sample t-test. Also different to JASP and SPSS

Thank you for your bug report. Unfortunately I have no time right now to investigate the cause of the discrepancy. I'll try to look into it as soon as possible.