Page 1 of 1

unable to use a factor in linear regression

Posted: Sun Jul 21, 2019 8:01 pm
by anthonyb
I'm receiving an error message: "contrasts can be applied only to factors with 2 or more levels"
However, I'm using a factor that has 2 levels, and I verified that there are several observations in each. Unfortunately, this is with proprietary data for consulting work, so I can't share the data file. I've tried to show the relevant information in the attached screen shot. Any ideas about what I'm doing wrong, or is this a bug?

Re: unable to use a factor in linear regression

Posted: Sun Jul 21, 2019 11:45 pm
by jonathon
hi,

yeah, it's hard to say if it's a bug or not, without see the data sorry. the problem persists if you drop out the covariates?

you could try running the same analysis with GAMLj and see if you get the same result.

kind regards

jonathon