I would like to calculate the size of the effects using RStudio for writing an article.
I have longitudinal data taken in 4 time points: baseline, day 20, day 50 and recovery (one month after) in a confined environment. Some variables are parametric but most of them are non-parametric after using a Shapiro test and verification by a Brown-Forsythe test (homogeneity of variances). Also, I have two profiles (LPa versus HPa) of participants characterized at baseline on the median of the RMSSD (heart rate variability). The objective was to see the impact of time on the two cardiac biosignal profiles with a more vulnerable LPa group and a more protective HPa group.
For this, I used the nparLD library from Nogushi et al. (2012): An R Software Package for the Nonparametric Analysis of Longitudinal Data in Factorial Experiments. I chose this library because I needed a nonparametric test for longitudinal data that took into account two different groups.
Now I would need to calculate the effect sizes. I know that for parametric data, Cohen's d or Hedges' g are used. For non-parametric data, the r calculation is used. However, the latter applies for the Wilcoxon test or the Kruskall Wallis.
Do you know how I can calculate effect sizes for longitudinal (4 time x 2 profiles) non-parametric data? Knowing that I used the little known nparLD library which does not provide for the calculation of effect sizes?
I have searched a lot for the answer without results so I am relying on you.
I thank you very much for your answer, I am not a great connoisseur of R I manage by applying the libraries.