what type of analyses to use if one of the dependent variables is omitted when another dependent variable takes a certain value?

Question

I have one independent variable (IV) and two dependent variables (DV1 and DV2). DV1 can take value 0 or 1. DV2 can take a value between 1 and 7 only if DV1 is 1. If DV1 is 0, then DV2 is omitted.

As an example, suppose the IV is gender (i.e., male vs female). DV1 is 1 if the individual visited New York and 0 otherwise. DV2 represents how much this individual likes New York on 1-7 scale.

I am totally confused about how to do this analysis. I tried multivariate regression but not sure if that's the way to go. I'd appreciate any help.

Answer 1

If you only have 1 independent variable and 2 dependent variables that are effectively mutually exclusive, then I don't see any advantage to trying to do anything besides two separate analyses, one for each dependent variable. In your example, you could first estimate the probability of visiting New York as a function of gender, then model the opinion about New York as a function of gender.

With a more complicated set of predictors you might want to explore things like hurdle models that are designed for this type of 2-part process for outcomes.