# Logistic Regression: interpreting an interaction term

#### rclukey

##### New Member
Dear Forum,

This topic has been covered a lot here and in other forums, but I have yet to find a clear answer to this particular problem.

I'm trying to understand how to interpret the odds ratio of an interaction between a categorical variable with 6 levels (6 similar products) and a continuous variable (a score for each product). The DV is stated intent to purchase.

the model is specified as: logit = Const + B1 + B1*B2

*I'm using SPSS.

I have a significant main effect for the categorical variable, and significant interaction on 3 of the categories.

The odds ratios for those significant interactions are:
cat 1: 4.8
cat 2: 6.8
cat 4: 2.3
(all significant interactions have positive odds ratios)

What I can't understand is how this data relates to the reference category...

Is it saying that, for the significant interaction, that the odds of reporting a high intent are 4.8 times that the reference category?

It seems funny, because when I plot the probabilities, the reference category has the highest probability (~61%), while cat 1 is at 20%, cat 2 is 5%, and cat 4: 45%

Can someone please help me to interpret the odds ratios for the significant interaction?

Thank you!
Ryan