1

Im carrying out a meta-analysis where I have had to carry out my own wilcoxon test on the data. I am now calculating the independant effect sizes for each study to manually input to my excel spreadsheet.

When I have carried out my Wilcoxon test this is my input:

rear<-read.csv(file.choose())
names(rear)
shapiro.test(rear$PI)
wilcox.test(rear$PI~rear$ï..RearingCond)

and my output just provides W = 82, p-value = 0.26

how do i now calculate my effect size and variances for this?

mdewey
  • 16,541
  • 22
  • 30
  • 57
Aims
  • 11
  • 1

1 Answers1

0

The simplest solution since you have the raw data is to carry out a Student's $t$ test and use the mean difference and its standard error from that as your effect size.

See also the warnings in the answer given in this Q&A How to convert from Mann-Whitney U to Z (or other effect size)?

mdewey
  • 16,541
  • 22
  • 30
  • 57