Dichotomous outcome variable

[Pandelis Perakakis]

Hello group!

My EMA data (29 days, 5 measurements per day) includes several psychosocial processes and I would like to test how they relate to a dichotomous outcome variable that codes whether a certain psychological threshold has been reached or not. This variable is measured with the same density (5 times a day) but there are few occasions above the threshold in each subject (mostly zeros and very few ones).

Does it make sense to use gimme with the dichotomous threshold variable as an "outcome"? If yes, are there other details I need to consider?

All best,

Pandelis

[Katie Gates]

Hi Pandelis, 

Currently GIMME can't handle dichotomous variables as an outcome. A different estimation approach would be needed, and initial tests show that the recovered models aren't reliable using this test.

hope this helps, 

Katie