catinthehat

Okay. From what I've heard, a perfect quant score gets faculty's attention. It doesn't mean that you get in, but if an interview means you're in the running, I would expect a person with a perfect score to get one. Now maybe an interview means the opposite, i.e. they can't decide b/c you're on the bubble. I'm just saying that we are not seeing the kind of large-scale interviews that you would expect if they were a filtering tool. Something doesn't make sense.

Well, you would think that 800 quant would at least warrant an interview...no?

I'm more suspicious than ever re. the Yale interviews. Here's why: 1) We just saw today that trolls monitor what we write about and post results accordingly. There's talk about Yale interviews, boom they show up. 2) This troll was obviously a bad one, i.e. more sympathetic, by posting everything at once. A hard-core troll would subtly post messages without giving us any clue they were fake to get a real laugh out of it. 3) Too few people have received interviews for that to be representative of Yale's acceptance pool. 4) People who should be very strong candidates for Yale have not received interviews, e.g. RWBG (GRE) or adblanche (got into Stanford). That doesn't make sense.

Isn't that what we said last weekend?

Something tells me it's over for the weekend.

Don't feed the trolls. Report as spam and get moderators in here.

I love how it's "Professor Beim," who does not exist.

It's really cruel. I don't know what kind of person would see this as funny...

Willing to bet it's another troll...

Yup, confirmed as a troll.

Princeton is moving. Everything goes at 4 PM. Insane.

There's a theory admit just now at Stanford and Princeton.

Someone was accepted to Stanford just now? EDIT: Make that two. I think the all-at-once hypothesis has been officially rejected.

I completely agree. The predictive value of GRE scores is almost certainly non-linear. Variation within certain score ranges probably has no predictive value at all. ETS sought to smooth these distortions with the new scoring system but as with any change, this introduces uncertainty. The effect of this uncertainty on the timing of admissions decisions likely varies by school, but the uncertainty remains. (I'm not sure that this uncertainty is ameliorated by more test takers because that gives no information as to the reliability of the test in predicting success in graduate school. Only comparing the new scores with doctoral students' success over time can reduce this uncertainty.) That's why extrapolation from last year's results--a common tendency on GC--is likely to be less helpful this time around. GC is also a fascinating study in human response to uncertainty. Many emphasize the link between silence and rejections, but nobody has mentioned Harvard accepting a number of students in April 2010 who had been previously rejected: http://thegradcafe.com/survey/index.php?q=political+science+harvard&t=a&o=&p=2. We crave certainty, even when we know that silence might actually be beneficial.

Good point. By "less information" I meant a more costly signal, i.e. if a given score on the old GRE had x predictive value, it now has x-c predictive value, where c is the increased cost from having a new, untested format and scoring system. The new GRE thus provides relatively less information because x-c < x.