4 ms·
I've though before that a data driven approach would be an interesting attack on this problem. Take the applications, then 4 years later (or 10 or 20) compare
by jderick 18y ago
I've though before that a data driven approach would be an interesting attack on this problem. Take the applications, then 4 years later (or 10 or 20) compare it to the students' records and see what factors correlated best with success. There really is a treasure trove of data here that is just going to waste.
- tokenadult 18y agoThat's an interesting suggestion. I think some college admission offices do that. I've heard MIT describe one admission factor it looks for: students challenging themselves with HARD classes in high school, which correlates better with success in upper division courses in MIT than does having a perfect high school grade point average. http://talk.collegeconfidential.com/massachusetts-institute-technology/582157-report-mit-information-session-saturday-4-october-2008-minneapolis.html http://talk.collegeconfidential.com/massachusetts-institute-... Another thing MIT looks for is baby-sitting experience. (I'm not making this up.)
- snprbob86 18y agoI think public schools are obligated to publish some of this data by law, but we really should get some from private schools too. Should it become a widely available data set, you'd see a few wikis/apps/startups/etc overnight.