 |
 |
Hi, I was going thru the mail list since im trying to use randomise and was a bit confused about some input parameters. Reading the randomise manual it says there that you need to use a design.con and a design.mat file to establish contrasts, etc of the model. Also I needed, as said in the email below, a concatenation of the cope images which is the filtered_func_data.nii.gz file in your grouplevel.gfeat output from Feat. In this email it says I can use that if it is a one-sample t-test design which I dont have since I am modelling a one factor four levels model. I am not sure what you mean with that and was wondering if you could clarify this. Also if I am right can I feed the design.con and design.mat files to randomise from any given Feat output model? as in run first Feat and then randomise from the output directory of Feat with the same design.con and design.mat files from feats output? Thank you for your help. Andres On Apr 19, 2011, at 1:19 PM, Jodi Gilman wrote: > Hello, > > What is the best way to get FEAT to generate a corrected p map? Should I use randomise, and if so, what should the input file be? (this is an fMRI dataset). > > Thank you, > > Jodi