help > the way to calculate the correlation
Showing 1-4 of 4 posts
May 8, 2017 03:05 AM | Jay Chen - National Cheng Kung University
the way to calculate the correlation
Dear Andrew:
Is the correlation calculated by NBS based on the Pearson correlation or Spearman rank-order correlation ? Thanks for help.
Best Regards
Jay Chen
Is the correlation calculated by NBS based on the Pearson correlation or Spearman rank-order correlation ? Thanks for help.
Best Regards
Jay Chen
May 8, 2017 05:05 AM | Andrew Zalesky
RE: the way to calculate the correlation
Hi Jay,
Pearson. Correlation is computed using the GLM, which is effectively Pearson.
Andrew
Originally posted by Jay Chen:
Pearson. Correlation is computed using the GLM, which is effectively Pearson.
Andrew
Originally posted by Jay Chen:
Dear Andrew:
Is the correlation calculated by NBS based on the Pearson correlation or Spearman rank-order correlation ? Thanks for help.
Best Regards
Jay Chen
Is the correlation calculated by NBS based on the Pearson correlation or Spearman rank-order correlation ? Thanks for help.
Best Regards
Jay Chen
May 9, 2017 04:05 AM | Jay Chen - National Cheng Kung University
RE: the way to calculate the correlation
Dear Andrew:
Thank you for your reply. The reason I asked this question is that my advisor thinks the behavior scores are discrete and the Spearman rank-order correlation should be used to perform the correlation analysis between the network strength and the behavior scores. I have read all the related threads in this archive. I believe my advisor is wrong but I don't know how to explain to her. Could you give me some hints? Thanks for help.
Best Regards
Jay Chen
Thank you for your reply. The reason I asked this question is that my advisor thinks the behavior scores are discrete and the Spearman rank-order correlation should be used to perform the correlation analysis between the network strength and the behavior scores. I have read all the related threads in this archive. I believe my advisor is wrong but I don't know how to explain to her. Could you give me some hints? Thanks for help.
Best Regards
Jay Chen
May 10, 2017 12:05 AM | Andrew Zalesky
RE: the way to calculate the correlation
Hi Jay, this seems to be a question
that might be more appropriate for a forum on statistics.
You could simply populate your design matrix with the rank of the behaviour scores (rather than the actual scores). This won't be exactly Spearman, but might alleviate some of the concerns.
Andrew
Originally posted by Jay Chen:
You could simply populate your design matrix with the rank of the behaviour scores (rather than the actual scores). This won't be exactly Spearman, but might alleviate some of the concerns.
Andrew
Originally posted by Jay Chen:
Dear Andrew:
Thank you for your reply. The reason I asked this question is that my advisor thinks the behavior scores are discrete and the Spearman rank-order correlation should be used to perform the correlation analysis between the network strength and the behavior scores. I have read all the related threads in this archive. I believe my advisor is wrong but I don't know how to explain to her. Could you give me some hints? Thanks for help.
Best Regards
Jay Chen
Thank you for your reply. The reason I asked this question is that my advisor thinks the behavior scores are discrete and the Spearman rank-order correlation should be used to perform the correlation analysis between the network strength and the behavior scores. I have read all the related threads in this archive. I believe my advisor is wrong but I don't know how to explain to her. Could you give me some hints? Thanks for help.
Best Regards
Jay Chen