 |
 |
Hi Elif, You've already done the analysis that tells you there is a significant relationship between levels of activation and the questionnaire scores, so you could stop there... But if you want to plot the values in a scatter plot (e.g., to rule out outliers), then you can use Featquery -- but note that the effect size here will be inflated due to a lack of independence (you should check your prior analysis for errors or describe what you did more thoroughly because it has to be significant). So, if you want to know the true effect size, then you need to examine correlation in independent data. See this article for a nice discussion of these issues: http://www.ncbi.nlm.nih.gov/pubmed/20571517 Cheers, David On Oct 10, 2011, at 10:30 AM, Elif Alkan wrote: > Dear all, > > I want to calculate some correlations with questionnaire-scores and activation. I have employed a GLM with covariates, so that I could see,where in the brain is a sig. activation in relation to this score. I found activation in Insula and ACC. > > I would like to know, how high this correlation between this questionnaire score and activation Insula is. > > I ran featquery and then calculated regression and correlation analyses with featquery output and questionnaire score and there is never a sig. result. > > Can you please tell me how to calculate correlations between a behavioral measure and the activity in a specific region (in whole brain analysis, no ROI)? > > There must be a common reliable to way to do it, that I unfortunately don't know. > > Thank you for your help! > > Cheers > > Elif