Clear all

2021 Day 3 - choosing a GLM model

Neta Weitzman
Posts: 5
Topic starter
Active Member
Joined: 9 months ago

Hello and thank you again for this excellent course.

I'm trying to implement everything I’ve learned to my data and I’m not sure which GLM model should I choose. 

In my study participants are doing the n-back task for 30 minutes (includes n-back blocks and 'rest blocks') each block lasts 33 sec in which participants see ongoing stimuli (and then they rest for 33 sec). 

1. In the t-range should I choose -2 33 as parameters, or longer (35/40)? I guess HBO takes time to decrease, so maybe it will be better to choose longer block duration so I won't miss any hrf increases? 

2. In the idxBasis tried both the gaussian model and idxBasis=3 and results pattern was pretty similar, only that when idxBasis=1 y axis values were much lower than when I chose idxBasis=3. Does the values between different hrf models similar? If it is, do you think it means that the third model fits my results and I should use it rather than the gaussian?

3. If in the idxBasis i'm choosing 3, what should I put in paramsBasis (considering my study design)?

Thank you so much in advance,