DreamShake wrote:For a mass mailing survey...which is widely recognized as one of the poorest statistical measures available.
The gradations are too large (needs to be 1-10 or at least 1-7) and the response rate too low for this survey to be useful. Really, fuck US News for trying to pass off more shoddy statistics as legitimate information.
I agree 1-10 should be instituted instead of 1-5. When they use 1-5, a lot of individuals will feel guilty grouping some schools at the same number, but won't make appropriate changes because they feel like they would be cheating a particular school.
For example, someone may give schools 1-20 a (5), 21-50 a (4), 50-100 a (3), 100-150 a (2), and 150+ a (1). But, GW is not as good as HYS or even Georgetown/Cornell - the recruiter would not want to give GW a 4 because clearly GW is better than IUB. Also Temple and Villinova are better than Hofstra, but would Temple and Villianova deserve the same score as Fordham? A 1-10 gradiation could solve all of these problems.