Hello Community,
I am running glm and multinomial logistic regression but the results not compatible, I highlighed the only in the docx file,
while havejob has a positive effect (odd > 1) on accessing loan in multinominal loan it seems it has a negative effect(
Well I can't tell from what you sent what your DVs were in the 2 regressions. However, the two methods have different requirements for the DVs. GLM: Continuous Multinomial logistic: Categorical with more than 2 unordered categories. Thus you would not expect the results to be compatible since the models would be different. HTH D. Booth
Yes but my coeffs are the same for the models and y = formal.loan or informal.loan or both.loan or no.loan for multinominal
and y= having a loan or not for glm / so if all the relative ratios are under 1 in multinomial how the same ratio is above 1 in glm / considering the difference is huge
Sorry that is not how it works. See the attached for an introduction to the literature. Best, D. Booth
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression