search

JoramSoch / MACS

MACS – a new SPM toolbox for model assessment, comparison and selection
GNU General Public License v3.0
22 stars 6 forks source link

question about 2nd level covariate #6

Closed munoztd0 closed 4 years ago

munoztd0 commented 4 years ago

Hello,

First, this is more of a question than an issue. I was wondering if there was a way to compare a model using a first-level modulator (as in a rating for each trial) VS a model using a second level ( as in an overall rating index) covariate using MACS ?

Thanks in advance,

David

JoramSoch commented 4 years ago

@munoztd0: 1) Bayesian model selection has an identical data requirement. This means, it only makes sense to compare different models operating on the same data (because it's trivial that different models operating on different data perform differently). As first-level models and a second-level model are operating on different data (measured BOLD signals vs. first-level parameter estimates), you cannot compare them to each other. 2) What you can do (I actually added this functionality not very long ago) is to compare different second-level models, e.g. one having a regressor for overall rating index vs. one not having it. 3) Similarly, you can of course also always run a comparison between two first-level models, e.g. one having the parametric modulator coding trial-wise ratings vs. one not having it. Hope this helps!

munoztd0 commented 4 years ago

Hey, thanks for your clear and detailed response!