I have two groups of data, each consisting of 100 correlation coefficients.

I am interested in essentially seeing if the correlations in group 1 are stronger than group 2. However, I am unsure of the proper statistic to use.

I know I cannot take the mean correlation coefficient value for each group and then run a t-test, however that is essentially what I am trying to achieve if this makes sense.

  • $\begingroup$ You have two data sets, and each data set has 100 correlation coefficients? How were these coefficients obtained? $\endgroup$ – user2974951 Jan 17 at 8:07
  • $\begingroup$ Essentially I have 200 total trials in my study (100 in each condition), each with 4 dependent measures. I took a time series of each of those trials for each dependent measure (say device A and B) and got a correlation coefficient for each trial. Now I want to assess whether the correlations are stronger in Condition 1 vs 2, which would be the first 100 trials vs the second to assess whether the effect of the condition impacted the strength of the correlations. $\endgroup$ – JDM1992 Jan 17 at 16:48
  • $\begingroup$ Why can't you use a t-test for this? Have you checked the assumptions? Are the samples dependent? Basically you would use all 100 correlation coefficients from both samples and run a t-test to check whether the mean of one sample is higher than the other. $\endgroup$ – user2974951 Jan 21 at 13:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.