> Examining contrasts separately sounds like a good option, but will that tell me
> anything about the interactions?
No, it only gives a specific t value for each contrast, which does say something about the interactions, but not as direct as an overall F.
Gang