RE: gPPI in block and event-related designs.

[color=#000000]bump -[/color]

I'm still unclear on whether or not I need to include the 'control' condition explicitly in the gPPI analysis or if it is appropriate to select only 'drug', and lump control/fixation together as the baseline.

Any help greatly appreciated.

Cheers,
Patrick
[i]Originally posted by Patrick McConnell:[/i][quote]Hello,  

I have two tasks for which I am looking to model functional connectivity, one being a slow block design and another being a fast event-related design.  I'm still struggling a little to understand the differences between gPPI and weighted-HRF (or hanning-weighted for that matter) and subsequent interpretation of results.  

In the block design, control and drug blocks are alternated with fixation blocks in between.  I am interested in connectivity between regions that is different between the control and drug blocks.  If I am modeling this with gPPI, do I only select the 'drug' condition?  And here, do the results indicate where connectivity is different during drug than during control/fixation?  

In the event design, there are multiple no-go variables and I am interested in looking at connectivity during correct no-go trials while controlling for connectivity in correct rare-go trials.  Since there are other conditions of non-interest (e.g., go correct), how would I model this with gPPI vs. weighted-HRF?  

Any help greatly appreciated.  

All the best,
Patrick[/quote]