Hi Alfonso & all,

I'm new to CONN and wonder if anyone can help –

I have 3 main conditions (encoding, recall, rest) each in blocks of 8 trials and want to compare connectivity in each condition across 3 groups, after regressing out the main task effects.

However, we used sparse sampling with an effective TR=5 (2 second acquisition followed by 3 second silence/stimulus presentation). I'm not sure if this is typical, but counter-intuitively (to me), the acquisition of each trial's stimulus is actually at the start of the next TR/trial.

1) What is the best way to set this up in CONN? After reading previous posts, I think I select sparse acquisition on Setup.Basic tab, enter the effective TR=5, copy all my conditions into covariates/confounds for denoising, and select 'no weighting' rather than 'HRF weighting' in first-level analysis?

2) Given the sparse sampling (which starts with a volume acquisition), what is the correct way to set the onset times? E.g. if stimuli are presented at 2 seconds, then 7s, then 12s, all with 3s duration (see attached if helpful), do I need to adjust these timings for CONN to understand my design? I ask because I know that FSL (for instance) thinks in terms of TRs and assumes an instantaneous acquisition at the midpoint of the effective TR, so manual shifting of onset times is needed for sparse designs there.

Any advice would be very much appreciated!

Thanks in advance,
Cathy