Dear Alfonso and Conn users,
First of all, I just want to say thank you so much for making the forum such a helpful community, and I learned a lot from communications on here. I'm trying to do some MVPA analysis on my resting state data, and have a couple of questions.
1) When reading the literature, I see some times people recommend to conduct MVPA analysis on unsmoothed data, given the potential effect of neighor voxels. When doing the MVPA analysis in conn, did it take the smoothed data or unsmoothed data? If it is using smoothed data, could you comment on why and whether you think smoothing will make a big difference for MVPA?
2) I have 154 participants, based on the recommended rule of thumb, the number of factors should be 1:5 to participant number. Does it mean I need to identify about 30 factors in my data? I also read your comment somewhere on the forum that usually 3-4 factors should be a good number. So could you advise on how many number of factors should I use in this case?
3) Dimensionality reduction. I know the default value is 64, and you said 64 is the minimal and optimal number. Could you explain to me what does this number do?
Thank you so much!
Sophie
First of all, I just want to say thank you so much for making the forum such a helpful community, and I learned a lot from communications on here. I'm trying to do some MVPA analysis on my resting state data, and have a couple of questions.
1) When reading the literature, I see some times people recommend to conduct MVPA analysis on unsmoothed data, given the potential effect of neighor voxels. When doing the MVPA analysis in conn, did it take the smoothed data or unsmoothed data? If it is using smoothed data, could you comment on why and whether you think smoothing will make a big difference for MVPA?
2) I have 154 participants, based on the recommended rule of thumb, the number of factors should be 1:5 to participant number. Does it mean I need to identify about 30 factors in my data? I also read your comment somewhere on the forum that usually 3-4 factors should be a good number. So could you advise on how many number of factors should I use in this case?
3) Dimensionality reduction. I know the default value is 64, and you said 64 is the minimal and optimal number. Could you explain to me what does this number do?
Thank you so much!
Sophie