0

I found other users using psych package corr.test function and getting corrected and uncorrected p.values, here and here...They used Pearson and Spearman Rho correlation with "adjust" argument.

But I am not getting the adjustments when kendall correlation method is used... I think it is not a data issue... the code is the following:

JustScores.Cor<-psych::corr.test(JustScores,method="kendall",adjust="none")
ap<-data.frame(JustScores.Cor$p)

I get the same values if I omit "adjust="none"" (default = "holm"), or change the adjust to "BH", "bonferroni", "BH", "fdr"...

Whatever what I am getting, it is not the uncorrected value, as it is not the same as the value presented in ggscatter (ggpubr package)and stats' cor.test...

cor.test(JustScores[,VarName1],JustScores[,VarName2],exact=TRUE,method="kendall")$p.value

I tried also the corr.p, but it delivers the same...

What am I doing wrong? Any idea?

Any help is appreciated.

hamagust
  • 728
  • 2
  • 10
  • 28

1 Answers1

0

I can not reproduce your problem. The raw probabilities are below the diagonal and the adjusted are above the diagonal of the matrix labeled Probability values.

William Revelle
  • 1,200
  • 8
  • 15