Mark Blumenthal flags some interesting crosstabs from the Washington Post's indispensable numbers guru, Jon Cohen (or as we like to say around here, Jon Cohn, Jr.). Cohen has broken down the white vote into college and non-college educated groups in 14 primary states. One thing that immediately jumps out at you is how well Obama did in Virginia among non-college voters, suggesting real progress for him. On the other hand, Obama did only okay among this group in Maryland, which voted on the same day. I think the difference might have something to do with the white-ethnic vote I mentioned earlier this week, though who knows.
The other interesting thing is that, with the exception of Missouri and Virginia, Obama doesn't really run up big margins among white college grads, even in states where you'd think he would (again, like Maryland.) Not entirely sure what to make of that--other than, you know, Hillary is basically tied with Obama in national polls, so she must be convincing some college-educated whites to vote for her. (Mark asks a similar question.)
P.S. Mark thinks one source of variation could be advertising. Sounds plausible to me. We'll see how that shakes out over the next few weeks, when Obama will presumably have a money advantage.