BookmarkSubscribeRSS Feed
Choose Language Hide Translation Bar

Not applicable


Oct 24, 2013

Unstable Intercepts

I am attempting to compare the number of comments received in different categories for a control and experimental rubric using the Nominal Logistic Fit test, and a lot of the terms for my parameter estimates are unstable. I've been doing some reading (I'm very new at this) and that seems to mean that the category is rare, which would not surprise me. I'm just wondering if that is true and if I can still use confidently use the results. If not, is there anything else I could try? Thanks!