You might try to sort students out by asking a question on the order of:  What is the most effective strategy you use for preparing for an exam?   This would allow you to distribute the students who prepare stacks of 3 x 5 cards and the high-lighters evenly among the other "types."     

Also, it seems that this list has previously suggested creating sorting question based on what counts as achievement in a particular class.   I think that in a history survey (yes, a far cry from med school, I know!), for example, a reasonable mix could be achieved by asking how a final grade should be weighted:  factual knowledge, evidence of improvement, participation, willingness to challenge a prevailing interpretation.   In other words,  getting at achievement from your own course's perspective might be more important, especially because TBL, at least in my experience, challenges students to think and behave in new ways.

good luck

Penne Restad
Department of History
Univ. of Texas
Austin, TX