That is usually not the way you want to handle dichotomous variables. For example, if you use Male=1 and Female=2, you are implying that Females have a factor of two greater thickness (or whatever) relative to males. Maybe it works if the factor only has two levels. But why not set it up as a discrete factor? QDEC was not really meant to handle a wide range of designs, so you might be better off going to the command line (mri_glmfit) stream
On 11/14/2016 07:03 AM, Dominic Dwyer wrote:
Hi Prof. Greve,
Thanks for the response and sorry for the confusion. I mean can we use dichotomous variables in the "Nuisance Factors" box of qdec? i.e., if we don't explicitly define them as categorical variables. I was thinking that if it's a GLM then it should be ok?
We want to do this because we have more than one dichotomous variable that we want to control for.
Many thanks, Dom
*Dom Dwyer, PhD* Psychiatry & Psychotherapy LMU Munich
A: Nußbaumstr. 7, Munich, 80336 E: dominic.dwyer@med.uni-muenchen.de mailto:dominic.dwyer@med.uni-muenchen.de T: +49 (0)89 4400 55880
I'm not sure what you mean. Dichotomous variables can be used. Our tutorial has an example of using gender. On 11/9/16 5:05 AM, Dominic Dwyer wrote:
Dear FS experts, Sorry for my confusion here, but I'm wondering why dichotomous nuisance variables (e.g., gender) can't be used in qdec? I was thinking that if it uses a generalized linear model then it should be ok? I read other posts and understand why discrete factors can't be used with more than 2 levels, but was thinking that dichotomous variables should be different.NB. we're trying to run a simple two group comparison, but we two dichotomous variables that we want to control for. Many thanks for your help, Dom -------------------------------------------------------------------- *Dom Dwyer, PhD* Psychiatry & Psychotherapy LMU Munich A: Nußbaumstr. 7, Munich, 80336 E: dominic.dw...@med.uni-muenchen.de <http://med.uni-muenchen.de> <mailto:dominic.dw...@med.uni-muenchen.de> T: +49 (0)89 4400 55880
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer