Help using Base SAS procedures

logistic regression with overdispersed categorical variables

Reply
N/A
Posts: 0

logistic regression with overdispersed categorical variables

I am running several models in logistic regression. One of my models has only two categorical variables, for the model fit I obtain the following:

Hosmer and Lemeshow Goodness-of-Fit Test

Chi-Square DF Pr > ChiSq

0.0000 1 1.0000

My variables are overdispersed (they are binary variables (yes/no) and have mostly outcomes of only one type). My question is, what does this model fit tell me about my model? is it an acceptable fit? or does this tell me there is something wrong with my data and my model? should I exclude the model from my analysis?
Thank you.

Jose
Ask a Question
Discussion stats
  • 0 replies
  • 103 views
  • 0 likes
  • 1 in conversation