Thanks for the additional info, Edward. Assuming you have a large enough sample to avoid over-fitting, you can use the NOMREG command to fit a multinomial logistic regression model. In the GUI: Analyze > Regression > Multinomial Logistic. Here's an example from the UCLA website to get you started:

• https://stats.idre.ucla.edu/spss/dae/multinomial-logistic-regression/

You may be able to find other examples online that include an interaction term. HTH.

How many levels does your categorical outcome variable have? If it has 3 or more levels, are they ordered or not? Thanks for clarifying.

Hi Bruce Weaver thank you for the prompt response! The categorical variable has five levels and they are not ordinal

Thanks for the additional info, Edward. Assuming you have a large enough sample to avoid over-fitting, you can use the NOMREG command to fit a multinomial logistic regression model. In the GUI: Analyze > Regression > Multinomial Logistic. Here's an example from the UCLA website to get you started:

• https://stats.idre.ucla.edu/spss/dae/multinomial-logistic-regression/

You may be able to find other examples online that include an interaction term. HTH.

Thank you Bruce Weaver that is so helpful.

I will take a look and get back in touch if I have any other questions - cheers!

On SPSS you go to Analyze: then click on Regression and then on Binary Logistic