Broadley, the General Linear Model, especially multiple regression, factor analysis, and structural equation modeling. In terms of substantive interest, I'm fascinated my measures of personality such as the Big 5 and how those traits may not be completely orthogonal, thus calling the theory into question. I also have some experience with latent class and mixture models, although I'm not completely sold on those. If I study those, I will mainly be interested in how the number of groups is derived...so more methodological in that sense.
I'm not positive what I want to study yet, which I see as a good thing about the Quant field as a whole. As a professor put it, you probably don't even know that what you will study exists yet, since the undergrad Quant experience can only be so broad.
How about you?