I apologise in advance for the basic nature of this question but I'm very new to this and no one at my university has used short ISIs in event-related SCR research before. I'm beginning testing soon and just wanted to check that my planned procedure is appropriate, in particular to analyse the data using PsPM.
All timings are average ISIs with a jitter of +/- 30%.
Paradigm 1, each trial:
Emotive and uncertain situation presented (one sentence, different conditions) for 6 seconds, a rating gathered during this time.
Outcome (positive or negative) revealed after this 6 seconds, displayed for 5 seconds.
Fixation 3 seconds.
Paradigm 2 (coin flip gambling task), each trial:
Magnitude of the amount to win / lose on that trial presented for 4 seconds before input accepted to bet heads or tails.
Self paced additional jitter based on how long it takes to press button to bet.
Anticipation period, waiting for outcome of gamble, 4 seconds.
Outcome (win or loss) presented for 3 seconds.
In addition to a general question about whether these timings are appropriate, I'd appreciate advice on whether to use GLM or DCM analysis - it seems DCM may be better due to the longer nature of the stimuli. As this is my first time using SC measures I'd like to analyse a pilot before collecting all data.
I'm also concerned whether the lack of independence between the different parts of each trial presents an issue for the analysis. For example, when a greater amount is at stake in the gamble, SCR to all parts of that trial would be expected to increase.
Any advice is very much appreciated.
Many thanks,
Jo
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
You could make the second one accessible to GLM analysis by shortening the presentation times slightly. In general, GLM is more robust (and much faster to run).
If your hypothesis is that sympathetic arousal across the entire trial scales with the amount at stake, this is easy to test. There is no neural/biophysical reason to expect that, so if that's the case it would be most likely due to a psychological mechanism.
Dominik
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I just have another question about event markers. I'm presenting my stimuli in MATLAB which pushes event markers to the paralell port, to Biopac and I'm recording data using Biopac Student Lab which can output a .mat file to read into PsPM.
I've worked out how to do all of that but not sure about the best way to mark my data as I need each trial to have a unique marker rather than labelled by condition. For one of my tasks I have 120 trials and each trial has 3 events (onset of stimulus, onset of anticipation, reveal of outcome) so I can't label each uniquely as only have values up to 255.
Any advice on a way to label my trials which will enable me to analyse them separately in PsPM would be much appreciated.
Many thanks,
Jo
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
PsPM does not use your marker labels - you have to provide trial labels independently from the markers. That is, if you use markers PsPM will just give them a running index, and you then assign all trials in one conditioning by referring to this index.
This means you don't need to worry about trial labels at all - but you need to keep track of the experiment structure independently from the psychophysiological data.
Hope this helps
Best
Dominik
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Dear all,
I apologise in advance for the basic nature of this question but I'm very new to this and no one at my university has used short ISIs in event-related SCR research before. I'm beginning testing soon and just wanted to check that my planned procedure is appropriate, in particular to analyse the data using PsPM.
All timings are average ISIs with a jitter of +/- 30%.
Paradigm 1, each trial:
Emotive and uncertain situation presented (one sentence, different conditions) for 6 seconds, a rating gathered during this time.
Outcome (positive or negative) revealed after this 6 seconds, displayed for 5 seconds.
Fixation 3 seconds.
Paradigm 2 (coin flip gambling task), each trial:
Magnitude of the amount to win / lose on that trial presented for 4 seconds before input accepted to bet heads or tails.
Self paced additional jitter based on how long it takes to press button to bet.
Anticipation period, waiting for outcome of gamble, 4 seconds.
Outcome (win or loss) presented for 3 seconds.
In addition to a general question about whether these timings are appropriate, I'd appreciate advice on whether to use GLM or DCM analysis - it seems DCM may be better due to the longer nature of the stimuli. As this is my first time using SC measures I'd like to analyse a pilot before collecting all data.
I'm also concerned whether the lack of independence between the different parts of each trial presents an issue for the analysis. For example, when a greater amount is at stake in the gamble, SCR to all parts of that trial would be expected to increase.
Any advice is very much appreciated.
Many thanks,
Jo
Dear Jo
for both paradigms, DCM appears suitable.
You could make the second one accessible to GLM analysis by shortening the presentation times slightly. In general, GLM is more robust (and much faster to run).
If your hypothesis is that sympathetic arousal across the entire trial scales with the amount at stake, this is easy to test. There is no neural/biophysical reason to expect that, so if that's the case it would be most likely due to a psychological mechanism.
Dominik
Dear Dominik,
Thank you so much for your reply.
I just have another question about event markers. I'm presenting my stimuli in MATLAB which pushes event markers to the paralell port, to Biopac and I'm recording data using Biopac Student Lab which can output a .mat file to read into PsPM.
I've worked out how to do all of that but not sure about the best way to mark my data as I need each trial to have a unique marker rather than labelled by condition. For one of my tasks I have 120 trials and each trial has 3 events (onset of stimulus, onset of anticipation, reveal of outcome) so I can't label each uniquely as only have values up to 255.
Any advice on a way to label my trials which will enable me to analyse them separately in PsPM would be much appreciated.
Many thanks,
Jo
Dear Jo
PsPM does not use your marker labels - you have to provide trial labels independently from the markers. That is, if you use markers PsPM will just give them a running index, and you then assign all trials in one conditioning by referring to this index.
This means you don't need to worry about trial labels at all - but you need to keep track of the experiment structure independently from the psychophysiological data.
Hope this helps
Best
Dominik