search

jacobaparker / PRET

Pupil Response Estimation Toolbox
GNU General Public License v3.0
16 stars 1 forks source link

Modelling experimental design with condition dependent events latency #1

Open AlexLepauvre opened 8 months ago

AlexLepauvre commented 8 months ago

Dear Jacob,

I am currently attempting to use your toolbox to model pupil response to a dual task experiment. In my experiment, we presented subjects with a visual followed by an auditory stimuli. The SOA between these two events was varied to investigate the effect of task relevance on the psychologically refractory period. Accordingly, I have 4 experimental conditions, in which the timing of the auditory event differs. My goal is to estimate the latency of the pupil response to the auditory event to investigate whether an increased delay in pupil response is observed at shorter compared to larger SOAs.

I was wondering what is the best approach to model such a design. It is my understanding that it is not possible to have separate eventtimes for each condition within the same pret_model object. Is the best approach therefore to model separately each condition?

Thanks in advance for your help,

Best regards,

Alex

jacobaparker commented 7 months ago

Hello Alex,

You are correct in that the pret_model object only allows you to specify a single set of event times, so from a practical perspective, that is the right thing to do. Given that you are interested in latency differences between conditions, it probably makes the most sense to fit a separate model to each condition individually anyway.

Best, Jacob

AlexLepauvre commented 7 months ago

Dear Jacob, thanks a lot for the feedback, this is very helpful :)

Best regards,

Alex