Hello,

I am completing a Five Model Logistic Regression. All variables are categorical.

In the first model I have my DV (auxmood) which is coded as 1. Yes 0. No. Then I have my IV (rsatt) which is coded as 1. yes, 2. sometimes, and 3. No. In this model the odds ratios are all significant at the 0.01 level.

For model two, I added in a potential mediator, education. In this model the odds ratio of my IV have all decreased and are still significant At the 0.01 level. However, when I look at and interaction between my IV and the mediator, none of them are significant.

I have never done Logistic Regression mediators before, so I’m trying to understand what this means. would this mean that education is not a mediator due to the interaction not being significant?

thanks,

kat