asimov wrote:
The fact that schools move around so much in BW is a result of the fact that the scoring clusters the schools so close, that small changes in results can sway the rankings. In another word, instead of the linear appearance of 1, 2, 3 ranking, the schools are more grouped like 1, 1.01, 1.04, 1.045, 1.05, 2.02, 2.021... The usefulness of the ranking is to see the clustering within the rankings, since those schools are likely to have similar performance.
One thing I do like about BW is that it doesn’t tamper with the survey “noise”. Whereas I strongly believe USNews does. For example, for law school ranking, the same 14 schools are ALWAYS ranked in the top 14. It is hard for me to imagine that a #15 school (UT, UCLA, Vandy) has NEVER outperformed a #14 (Cornell, Georgetown) school in the history of USNews ranking. There have been many law school scandals, changes in regional economic prominence since the inception of USNews ranking. NEVER once is hard to imagine. Underdogs can get lucky once or twice, just look at all the Cinderella teams in March Madness.
For me, BW is like a startup. It’s trying to reflect the changing business school climate. USNews is like IBM, staying the conservative course -- “no one ever got fired for buying IBM.”
hahaha ... funny comparison...
I read this article today by BW.
https://www.businessweek.com/articles/2014-11-10/best-business-schools-2014-methodology-for-ranking-schoolsIt explains the methodology in detail. I am really impressed to be honest, since that's a lot of data to handle. According to it the ranking of top schools is in almost correlation with the recruiter ranking received by them, as in top schools the student ratings / scores were pretty much bunched up.
However, the grouping structure theory doesn't really hold water in this specific case, since the vertical movement in ranking is too much (refer to the instances I quoted in last post). I will take your word regarding USNews though, since I don't much follow it and know nothing.