Friday, May 2, 2008 - New servers are in! Click-in for more info!
VoyForums

VoyUser Login optional ] [ Contact Forum Admin ] [ Main index ] [ Post a new message ] [ Search | Check update time | Archives: 1234567[8]910 ]
Subject: Re: Columbia's Admit Rate This Year Is Either 10.0% Or 12.9%


Author:
In Defense Of Intellectual Honesty
[ Next Thread | Previous Thread | Next Message | Previous Message ]
Date Posted: 21:04:14 04/09/08 Wed
In reply to: AsiaSunset 's message, "Re: Columbia's Admit Rate This Year Is Either 10.0% Or 12.9%" on 14:53:12 04/09/08 Wed

Asia,

It appears to me that the two of us are debating this topic as vigorously as we have been because each of us desires the satisfaction of the other party admitting that he is wrong. Will you grant me the honor of admitting that you are wrong about whether it is the 31st percentile or the 15.5th percentile which determines an Ivy schools AI threshold?

I cannot help but observe how quickly you were willing to accept my explanation of how statisticians calculate standard deviations once it came from a third party, namely Phil. You were originally unwilling to accept my interpretation of the AI threshold because my version did not comport with what [youve] been told repeatedly which has been [thrown] around the board for years. You could have looked up a detailed definition of standard deviation on any number of websites, but you did not do so. Is it possible that you placed too much faith in your pre-existing view of the world?

Similarly, you are unwilling to accept my conclusion about whether 2350 Barnard students are included in the calculation of Columbias AI minimum because this is not consistent with the information [you] have about the AI hierarchy (although admittedly its 2 or 3 [years] old), so [youre] skeptical. I ask you with all due respect, is it possible that your information about the AI hierarchy is wrong?

You have already conceded that Columbias AI floor excluding Barnard likely ranks sixth in the Ivy League, tied with Penn. How much more inconsistent with your information about the AI hierarchy would it be if Columbia ranked seventh, as the Lions do when including Barnard?

You have already further conceded it is likely that the board scores of Barnard student-athletes are included in the calculation of Columbias aggregate AI, but you continue to hold out that the AI minimum is calculated WITHOUT the inclusion of the Barnard student body as a whole. Dont you see how inconsistent that logic is?

You would have the AI threshold calculated with one methodology but the determination of whether that AI threshold is satisfied calculated with a different methodology.

[ Next Thread | Previous Thread | Next Message | Previous Message ]

Replies:
[> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> Subject: Re: Columbia's Admit Rate This Year Is Either 10.0% Or 12.9%


Author:
AsiaSunset
[Edit]

Date Posted: 16:06:36 04/10/08 Thu

Phil is fairly established here as a long time poster and one who has established himself as both knowledgable and credible. Importantly he posted a link. I tried to follow the link. While it didn't work, it did cause me to goggle the subject matter. The math was far too complex for me, but the pictures weren't. So - I while didn't mean to dismiss you, the combination of the poster and the support he offered won the day.

The new information provided was interesting. I think the 25% data I posted is still the best public data we have from a directional standpoint, but I think the dynamics at the 16th percentile of each school is probably less predictable than at the 31st percentile.

As of 2/3 years ago the AI hierarchy was H,Y, P, Pe, D, Col,Br and Cornell. The 3 at the top were in a cluster separated from the next 3. Those next 3 were clustered as to render the differences almost non existant. I was told this by a very senior member of the Penn Athletic Dept. I'm not aware of the exact hierarchy based on the AI's of the current freshmen classes at the schools. It may have shifted a bit. As I said above, the differentials in some of the clusters were very small.

I still question your math. I don't think the impact of adding the Barnard students is proportional to the enrollment % of their student body when combined with Columbia College and Fu students. And I think the drag on the Columbia AI would be even more extreme now that I understand it is the 16th percentile that is relevant. At that percentile we are really talking for the most part about athletes, under represented minorities and other types of special admits and it seems logical that brand and financial aid and other preferences like location have a very heavy influence on the make up of the bottom part of our classes. I think Harvard is at the top of the food chain here and I doubt Barnard is competitive with any of the schools in the middle cluster and probably not even Cornell at the 16th percentile.
[> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> Subject: Columbia's Admit Rate This Year Is 10.0%


Author:
In Defense Of Intellectual Honesty
[Edit]

Date Posted: 23:42:12 04/11/08 Fri

Asia,

I see that your sense of pride prevents you from giving me the satisfaction of seeing you use the W word in regards to the 31st percentile. I labor under no such restrictions and I will take this opportunity to tell you and any other readers who have followed this thread to this point, I was wrong about whether statistics from Barnard students are included in calculating Columbias minimum AI.

When you and I first began going back and forth on this topic, I made a few phone calls. Today, I finally heard back from somebody who knows the answers to the questions we have been debating. To wit, here is the unalloyed final determination:

1. Barnard student-athletes have their board scores and class rankings included in the calculation of Columbias Academic Index.

2. Columbias mens and womens sports teams must achieve an AI above the same minimum AI.

3. Columbias minimum AI is calculated using only the board scores and class rankings of Columbia University (i.e., CC and Fu) students.

Thus, I was wrong about Point Number 3. In my own defense, my source agreed with me that it is wholly inconsistent to include the statistics from Barnard student-athletes in calculating the AI, while excluding other Barnard women from setting the minimum AI which must be exceeded. But thats the way its done and I was wrong. More precisely, I guessed wrong.

To the Columbia supporters who posted earlier in this thread, given this new information, Columbias relevant admit rate for the purposes of discussing selectivity in Ivy sports is therefore not 12.9%. It is 10.0%, the combined admit for CC and Fu, but excluding Barnard.

After learning this new information, I went back to Columbias website. Interestingly, of the 282 women athletes wearing light blue, 39 (or 13.8%) are from Barnard, while only 13 (or 4.6%) are from the engineering school. So Barnard is actually three times as important as Fu in terms of contributing women to Lion sports teams.

Parenthetically, there is one single student-athlete from the School of General Studies. Of course, there is not a critical mass of General Studies students to include the GS in calculating the minimum AI. (And I do not know if GS students even earn degrees.) Following the same line of thinking, perhaps when originally setting the minimum AI policy, it was assumed that Barnard women would not participate in Columbia sports to a significant extent. If the percentage of women athletes from Barnard were to increase from 13.8% to, say, 29.6% (their proportion of the overall CC-Fu-Barnard student body), perhaps the minimum AI methodology would be revisited. As it is, as Observer posted on April 6, Barnard student-athletes are being held to a higher standard of academic achievement than their non-athlete peers.
[> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> Subject: Re: Columbia's Admit Rate This Year Is 10.0%


Author:
AsiaSunset
[Edit]

Date Posted: 00:38:44 04/12/08 Sat

If it makes you happy, I was wrong. I think that was not only very obvious, but also clearly communicated when I responded to Phil's post.


I also totally agree that it is only logical that Barnard student/athletes count in the AI calculation and against the maximum recruit numbers utilized by Columbia under the post 2003 restrictions. I specifically stated such in this thread and I don't believe there were any syntax errors that could have confused you in that post(see post 10:48 on Sunday).

I joined the thread because I was contesting your original assertition that the Barnard student body as a whole was used in the calculation of that magic one standard deviation below the mean number, computed by Columbia University for purposes of banding in football and as the AI target number all recruited athletes from Fu, the College and Barnard must average out above. The answer you got from Columbia not only seems logical, it jives with what I've heard about the AI hierarchy in the league and also what Chris Lincoln posited at the time he published his book (2004 I think). I do believe statements A,B and C in your post are likely the correct answers, but I also believe Statement A, the critical one, is 180 degrees from where you started several days ago.
[> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> [> Subject: Re: Columbia's Admit Rate This Year Is 10.0%


Author:
AsiaSunset
[Edit]

Date Posted: 00:51:13 04/12/08 Sat

I mean "C"


[ Contact Forum Admin ]


Forum timezone: GMT+3
VF Version: 2.94, ConfDB:
Before posting please read our privacy policy.
VoyForums(tm) is a Free Service from Voyager Info-Systems.
Copyright © 1998-2008 Voyager Info-Systems. All Rights Reserved.