I migrated over here from the SCRalyze board and I'm sorry to ask a somewhat redundant question from what I asked on there. I'm currently attempting to implement a paradigm that uses a long CS presentation (potentially 12-14s long) and longish ITIs (14-16s). From reading the board and the manual it seems as if a DCM model is preferrable to situations of long stimulus presentations but was just curious as to if it is appropriate for long ITIs as well. It seems as if this may be the case but don't want to throw out the idea of a GLM if it is a better option.
Also I was able to plot my SCR from data collected earlier (16, 20, and 24s ITIs) and it seems as if there are many instances of the SCR not completely returning to baseline before the onset of the next trial (this is particularly problematic with the 16s ITIs but also happens in some of the 20 & 24s ITI trials as well) and, if this is the case, it seems almost inevitable that 14-16s ITIs will disallow for a full return to baseline as well. You stated previously that this isn't necessarily an issue with PsPM and it seems like people are using the toolbox in situations where the ITI is as short as 5s. Is this true here? Or, I suppose, is this true for both the GLM and DCM modeling options? I'm just curious because we're using an electric shock to evoke SCR on CS+ trials but we're shocking on ALL CS+ trials so there's no way to use non-reinforced CS+ trials for analyzes like they do in 50% reinforcement paradigms.
I truly appreciate your help and I apoligize for the semi-redundant question but this toolbox may have solved a big issue for me.
Thank you.
Lauren
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
PsPM takes care of overlapping responses (loosely speaking, by subtracting the tail of the previous response from the current one).
Regarding the stimulus presentation, the question is what are you looking for? If you expect evoked responses to the stimulus onset (or offset) then GLM is the right thing for you. If you expect a response (or several) sometime during the presentation, then DCM is the right thing. If you don't know what you are looking for, you could also try the SF-DCM approach which handles up to around 1 response every 2 seconds, and flexibly estimates their timing and amplitude.
Dominik
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Dominik:
I migrated over here from the SCRalyze board and I'm sorry to ask a somewhat redundant question from what I asked on there. I'm currently attempting to implement a paradigm that uses a long CS presentation (potentially 12-14s long) and longish ITIs (14-16s). From reading the board and the manual it seems as if a DCM model is preferrable to situations of long stimulus presentations but was just curious as to if it is appropriate for long ITIs as well. It seems as if this may be the case but don't want to throw out the idea of a GLM if it is a better option.
Also I was able to plot my SCR from data collected earlier (16, 20, and 24s ITIs) and it seems as if there are many instances of the SCR not completely returning to baseline before the onset of the next trial (this is particularly problematic with the 16s ITIs but also happens in some of the 20 & 24s ITI trials as well) and, if this is the case, it seems almost inevitable that 14-16s ITIs will disallow for a full return to baseline as well. You stated previously that this isn't necessarily an issue with PsPM and it seems like people are using the toolbox in situations where the ITI is as short as 5s. Is this true here? Or, I suppose, is this true for both the GLM and DCM modeling options? I'm just curious because we're using an electric shock to evoke SCR on CS+ trials but we're shocking on ALL CS+ trials so there's no way to use non-reinforced CS+ trials for analyzes like they do in 50% reinforcement paradigms.
I truly appreciate your help and I apoligize for the semi-redundant question but this toolbox may have solved a big issue for me.
Thank you.
Lauren
Hi Lauren
PsPM takes care of overlapping responses (loosely speaking, by subtracting the tail of the previous response from the current one).
Regarding the stimulus presentation, the question is what are you looking for? If you expect evoked responses to the stimulus onset (or offset) then GLM is the right thing for you. If you expect a response (or several) sometime during the presentation, then DCM is the right thing. If you don't know what you are looking for, you could also try the SF-DCM approach which handles up to around 1 response every 2 seconds, and flexibly estimates their timing and amplitude.
Dominik
Perfect! I'll stop bugging you now.
Thanks for your patience and your time. I truly appreciate your help.