Hi there,

I am struggling with my thesis. I need to test in SPSS if there is a significant interaction effect.

I have four groups, and they all fill in a survey. The only difference of the groups is the casus about what the survey is going. In every casus are different sets of evidence (it's a case about rape). I ask the participants of each group if they want to convict the person in the casus or not. I found significance effects between the four groups, but now I have to know if men/women differ on how they response. In other words if there is an interaction effect of sekse on the answers: to convict or not to convict. I know this is more easy with a interval dependent variable (than you can test with multivariable ANOVA I guess?). But my dependent variable is dichotomous (convict/acquit). I try a lot but can't find how I can test this. On internet I read some things of binary logistic regression but there isn't much information and I am not sure if this is the correct test.

Sorry my English is worse. If you have more questions please ask. I hope someone can help me out!

Julia