I apologize if my phrasing is un-educated, I don't have much of a stats background. I'm struggling to find info answering my question because I lack the terminology to accurately describe it.
Let's say that I was reviewing data on mp3 player sales, and wanted to see if certain combinations of categorical data resulted in returns.
Things like color, memory size, brand. I don't simply want to see the relationship between returns and the color red, but rather which combinations indicate a likely return.
For example, imagine that a certain manufacturer uses a different factory to produce each color mp3 player - and the factory that produces blue ones has a defect that shows up 3 months later and is causing returns. I'd be able to see a relationship between that color and brand with returns in that example. Does that make sense?
What test could I leverage for this? My jumping off point is a chi-square, except that I can't see the interplay between the different variables with that. I haven't been able to figure out where to go from there.