Dear Sadjad Your DCM analysis will only work if there is an experimental effect for it to explain. It's fine to extract ROIs at the single subject level without a strict significance threshold. However, if you're not seeing anything for p < 0.05 uncorrected, your experimental effects might be quite weak or non-existent in those subjects. Are you only getting the flatlining in subjects without SPM results? Best Peter -----Original Message----- From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Sadjad Sadeghi Sent: 09 June 2016 11:13 To: [log in to unmask] Subject: [SPM] Flat line in DCM analysis Dear Peter & DCM expert, I am doing the DCM analysis for a group of 15 subjects and for 9 of them I got a flat line as the predictions for 3 regions that I've included in my analysis. I have used a p-value of 0.5 for extracting the time series for 3 different regions, as for smaller p-values e.g. 0.05 I had no activation for one of my regions and I was made to increase this value. Do you think should I exclude those subjects from my group analysis and is that right to use this p-value for extracting the time series? I am looking forward to your help. Best Regards, Sadjad