I'm using a Negative Binomial model to analyses the effect of a scale (NR) and another binary variable (Condition) on a count outcome. When I add both variables to the model as main effect, they are not significant. However, when I add the interaction (Condition:NR) to the model, they become significant. I hear this is a common phenomenon but I'm not sure what steps to take to interpret it.

Also, are there any further tests I can use to interpret the interaction? Because the Negative Binomial outputs don't say much overall.

Without interaction: enter image description here

With Interaction: enter image description here


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.