[color=#000000] Hi Angela,[/color]
[color=#000000]Regarding your first question, yes, in [i]Setup.Conditions [/i]you would simply define your three conditions self/other/case (optionally you may also define the baseline condition), and each condition will encompass data from multiple runs/sessions. Then, when you define your gPPI first-level analysis, simply select the three conditions (self/other/case; there is no need to manually define the regressors that go into gPPI as that will be handled by CONN automatically). If you want to double-check, in the [i]analyses (1st-level) [/i]tab you may switch the 'source timeseries' option to 'first-level analysis design matrix' and that will display the design matrix that the gPPI analysis will use[/color]
And regarding your second question, in option (1) you should upload a [b]single[/b] .txt file, which will include the timing reference for all subjects, sessions and conditions (of course some of these definitions may apply to multiple subjects/sessions/conditions and some will be specific to individual subjects/sessions/conditions; see "help conn_importconditions" for details about the format of this file)
Best
Alfonso
[i]Originally posted by Angela Fang:[/i][quote]Hello,
I have a couple of questions about setting up a gPPI analysis in conn. I have 3 runs of an event-related fMRI task with 3 active conditions (self, other, case) + baseline, per subject. Is it correct that with this design, I would need to include the following regressors: self, other, case, baseline, time series of the seed, PPI_self, PPI_other, PPI_case, and PPI_baseline? When I'm setting up the conditions, would I just set up the self, other, case, and baseline conditions for each subject (with 3 sessions/runs each)? I want to make sure that treating each run as separate sessions for each subject won't create multiple PPI terms and will instead average the data across all 3 runs that contribute to a single PPI term representing all 3 runs. Can you confirm that this is correct?
Secondly, I'm also wondering how to upload unique timing files for each run in the Setup tab (rather than manually entering all of the onsets and durations for each run for each subject). In the Conditions tab, it looks like there are 3 options for uploading condition info from text files- (1) a Conn-legacy .txt or .csv file, (2) a single BIDS-compatible .tsv file, or (3) a BIDS-compatible .tsv per subject/session. I have a paradigm .txt file that is unique for each run, so would it work to upload the timing file for the entire run even though it includes info about all of the event types- in other words, can I upload a separate .txt file for each subject/run (using option 1), or would I have to create separate event-specific timing files?
Thanks for your help!
Angela[/quote]