I'm actually an interloper here (a pre-law advisor who already attended law school and took/passed the bar ... and is now a social scientist with a decent background in stats). LSAT scores are correlated (ish - the correlation is not incredibly strong) with first year grades and bar passage. NOT other law school grades. Hmmm. And that correlation is not broken down between students who study a lot for the LSAT and students who only study a little for the LSAT. So it's entirely possible to me that students who study a lot for the LSAT (and get high scores) are the students who study a lot their first year (and get good grades) and study a lot for the bar (and get good grades). We call that an intervening variable: hard work.