Is there a model/test for a sequence of binary dependent variables?

[Daanmh]

Hi all,

I am not sure if this is the right place to ask this question, but as I am planning to use RStudio to run my tests let's give it a try. I am currently working on my thesis which is about the possible effect of several touchpoints on purchases. However, my database consists of a variable for purchases in general and a variable for purchases at the focal company (so the company you 'wanted' them to purchase at). Therefore my model has two binary dependent variables: a binary dependent variable of purchase in general YES/NO and a second binary dependent variable of purchase at focal company YES/NO. So the second dependent variable can only be 'reached' if the first one was a yes: there was a purchase in general.

I know there are some models/tests where you can use assumptions in your dependent variable (like the Tobit 2 and the Hurdle model), but the second part always has to be continuous in these models.

Thus, my question is whether there is a model/test where I can use a sequence/assumption that has to hold (purchase in general) for my two dependent variables that are both binary.

Thanks in advance!

[technocrat]

Hi, and welcome!

Is it really a model question? Let df be your data frame with Y_0, Y_1, X_0 ... X_n. Why not

df2 <- df %>% select(-Y0)

which gives you a data frame in which the first condition is always satisfied and you can run a logistic regression on the target.