If there are preferences, they're nonobvious, and it's not clear how you would go about discovering them. (I haven't found anything more than students' speculation by Googling around.)

Here's one: http://www.nytimes.com/2010/07/19/opinion/19douthat.html

And some of the hits in http://scholar.google.com/scholar?q=college+admissions+odds+extracurricular look promising. ('odds' is there to help weight towards quantitative studies which would be using odds ratios as their effect size, since you're either admitted to a particular college or not, and a binary effect size like odds ratio is what most statistical approaches would use.)