Dear Alfonso and other CONN Users,
When analyzing the effect of my variable of interest on resting-state ROI-to-ROI connectivity using CONN results explorer, I obtain a beta value of 0.03 and an uncorrected p-value of 0.014. However, when I extract the raw connectivity values from Z in the resultsROI_Condition001.mat file and run multiple linear regression analysis in R (using the same covariates and excluding the same subjects), I get the same beta value (0.03), but a p-value of 0.003. Do you have any insight on why these p-values would differ between the two programs? Does CONN use another type of statistical test?
Thank you!
Miranda
When analyzing the effect of my variable of interest on resting-state ROI-to-ROI connectivity using CONN results explorer, I obtain a beta value of 0.03 and an uncorrected p-value of 0.014. However, when I extract the raw connectivity values from Z in the resultsROI_Condition001.mat file and run multiple linear regression analysis in R (using the same covariates and excluding the same subjects), I get the same beta value (0.03), but a p-value of 0.003. Do you have any insight on why these p-values would differ between the two programs? Does CONN use another type of statistical test?
Thank you!
Miranda