Something's Wrong With This Columbia Statistic

Colleges and Universities A-Z Columbia University
1

<p>This is a ranking of placement in top grad school placement.
The statistic is calculated using a Columbia class size of 1600+, which is clearly incorrect, and seems to be a CC + Barnard calculation, when it really should be CC + SEAS. CC + SEAS would make the total 1300, and increase the number of students attending top grad programs significantly, boosting Columbia probably slightly above Williams.
Am I completely off base on this?
Something is clearly wrong.</p>

<p><a href="http://www.wsjclassroomedition.com/pdfs/wsj_college_092503.pdf%5B/url%5D"&gt;http://www.wsjclassroomedition.com/pdfs/wsj_college_092503.pdf&lt;/a&gt;&lt;/p>

2

<p>It's probably true that they included Barnard then, though for all to be truly corrected you'd have to take into account the number that Barnard sends to top schools as well. Barnard actually fares pretty well with placement, so you'd be taking that number out too.</p>

3

<p>Right...but you would also have to add in SEAS students, which definitely is a greater number than Barnard.</p>

4

<p>Surely you'd have to look at how many actually applied in the first place for this to make any sense?</p>

5

<p>I think I saw the methodology for this. It was terrible. The 5 schools they chose weren't necessarily the top 5 schools. It doesn't factor in people who take several years off before they go to grad school. It doesn't factor in people who would have been able to go to one of these top programs but did not because they went into something (which hugely biases against Columbia because Columbia gives its undergrads so many more opportunities than other schools).</p>

<p>The comments are also asinine and seem to be written by a 14 year old. Take Harvard's comment -- Harvard has great grad school placement not because the Harvard degree is magical for a grad school adcom but because a Harvard alum is probably pretty smart and/or a great standardized test taker if they got into Harvard in the first place. And take Yale's comment -- Harvard law is 3 times the size of Yale law and much easier to get into so the numbers cited indicate anything but a lack of school loyalty.</p>

6

<p>
[quote]
you would also have to add in SEAS students, which definitely is a greater number than Barnard.

[/quote]
</p>

<p>Where do you get that the SEAS numbers would be "definitely" a greater number than Barnard? </p>

<p>I agree with Brand that to be an accurate representation of Columbia University grads who go on the grad schools, Barnard would need to be included along with SEAS, CC, etc. Though I would be curious to know Barnard's numbers alone. I do know they have a reputation for doing very well in this area.</p>

7

<p>In response to Columbia2002's post: sheesh, if that was indeed the methodology, I agree. It is just silly.</p>

8

<p>Methodology from <a href="http://www.wsjclassroomedition.com/college/feederschools.htm#rankings%5B/url%5D"&gt;http://www.wsjclassroomedition.com/college/feederschools.htm#rankings&lt;/a&gt;&lt;/p>

<p>


Traditionally, college rankings have focused on test scores and grade averages of kids coming in the door. But we wanted to find out what happens after they leave -- and try to get into prestigious grad schools.</p>

<p>We focused on 15 elite schools, five each from medicine, law and business, to serve as our benchmark for profiling where the students came from. Opinions vary, of course, but our list reflects a consensus of grad-school deans we interviewed, top recruiters and published grad-school rankings (including the Journal's own MBA rankings). So for medicine, our schools were Columbia; Harvard; Johns Hopkins; the University of California, San Francisco; and Yale, while our MBA programs were Chicago; Dartmouth's Tuck School; Harvard; MIT's Sloan School; and Penn's Wharton School. In law, we looked at Chicago; Columbia; Harvard; Michigan; and Yale.</p>

<p>Our team of reporters fanned out to these schools to find the alma maters for every student starting this fall, more than 5,100 in all. Nine of the schools gave us their own lists, but for the rest we relied mainly on "face book" directories schools give incoming students. Of course, when it comes to "feeding" grad schools, a college's rate is more important than the raw numbers. (Michigan, for example, sent about twice the number as Georgetown, but it's also more than three times the size.) So our feeder score factors in class size.

</p>

<p>Given this, I'll retract my comment that "It doesn't factor in people who take several years off before they go to grad school." Everything else I posted would be correct.</p>

9

<p>They don't include Barnard with Columbia; Barnard is on the list by itself, lower down (with a typically inane comment). </p>

<p>But c2002's right, the methodology is absolutely stupid. Huge selection bias, both in who goes to that school to begin with, so it's not a measure of value-added at all, and in terms of who applies (the score factors in class size, not the applicant pool size/admissions rate to these grad programs). Would you rather go to a school where everyone applies to the top business programs, and ten get in, or one where 5 people apply and they all get in?</p>

<p>And yeah, totally crap list of "top" schools, especially considering that some of the schools in the ranking are hurt because students there might have a geographic or home-school preference (why are NONE of Stanford's schools considered top five? Business? Heck, LAW?).</p>

<p>Basically, this list is a good measure of which schools have a lot of students that want to get into these 15 programs and eventually do. Not much else.</p>

<p>I'm also personally offended by calling professional school graduate school. ;) It's very misleading of the WSJ.</p>

10

<p>I don't think they intended on including Barnard with Columbia, but the figure they used to represent class size definitely included Barnard.</p>

11

<p>Given the methodology described by Columbia 2002, to determine the alma maters (either school submissions or school facebook entries), AND given the fact that Barnard grads have degrees issues by Columbia U., there is no reliable way they could have distinguished among Barnard/Columbia grads for the study, because it depends entirely on the subjective determination of the underlying report. That is, it is very possible that some of the reporting colleges listed their Barnard matriculants as "Barnard" and some listed them as "Columbia".</p>

12

<p>


Because these opportunities don't happen for people from other top 15 colleges right? Please. </p>

<p>And if you haven't realized, Columbia's professional schools are included in TWO of the three categories surveyed, and it is a given that professional schools traditionally take the largest fraction of the incoming class from the university's undergraduate school. Williams, Stanford (very questionable, but unrelated to this argument), Amherst, Princeton, Duke, and Swarthmore are all placed before Columbia, yet they have no prefessional schools listed on the survey. </p>

<p>Please stop this whining. Columbia lost out on this "ranking"-- get over it.</p>

13

<p>
[quote]
Because these opportunities don't happen for people from other top 15 colleges right? Please.</p>

<p>And if you haven't realized, Columbia's professional schools are included in TWO of the three categories surveyed, and it is a given that professional schools traditionally take the largest fraction of the incoming class from the university's undergraduate school. Williams, Stanford (very questionable, but unrelated to this argument), Amherst, Princeton, Duke, and Swarthmore are all placed before Columbia, yet they have no prefessional schools listed on the survey.

[/quote]
</p>

<p>You seem to have poor logical skills. If you cannot see that the methodology is flawed because it is premised on the poor assumption that the same percentage of students from each college are interested in going to law/biz/med school, then I cannot help you.</p>

14

<p>
[quote]
Please stop this whining. Columbia lost out on this "ranking"-- get over it.

[/quote]
</p>

<p>C2002's rankings: 1) Columbia. 2) Harvard. 3) Princeton. 4) Yale</p>

<p>HYP lost out in my stupid, made-up, arbitrary flawed ranking. Get over it.</p>

<p>Do you see the poor logic of your insinuations?</p>

15

<p>But your insinuation that Columbia graduates are NOT interested in professional schools is unfounded, way moreso than my implication that they are unlike students from any other top colleges bar caltech. CC and SEAS are both liberal arts colleges at their heart, and both require the completion of a core curriculum to graduate. If you look carefully, even MIT, which is not known as a feeder into law and med schools, places above Columbia in this survey.</p>

16

<p>
[quote]
But your insinuation that Columbia graduates are NOT interested in professional schools is unfounded,

[/quote]
</p>

<p>That's not my insinuation. My point is that they may not be interested in professional schools AT THE EXACT SAME LEVEL as that of other schools -- an assumption on which the WSJ survey depends.</p>

<p>
[quote]
CC and SEAS are both liberal arts colleges at their heart, and both require the completion of a core curriculum to graduate. If you look carefully, even MIT, which is not known as a feeder into law and med schools, places above Columbia in this survey.

[/quote]
</p>

<p>And this has what to do with the price of tea in China? It's fluff and a way for you to avoid the fact that the survey makes a huge assumption that you cannot support.</p>

17

<p>No point in continuing this debate. Apparently you've never heard of common sense.</p>

18

<p>
[quote]
Apparently you've never heard of common sense.

[/quote]
</p>

<p>Nice ad hominem for lack of any substantive response justifying the correctness of the flawed methodology that you are attempting to defend. Common sense is an ability to see a blatant flaw in a survey's methodology.</p>

19

<p>......................... ......................... ................... __ . <strong>....................... ......
......................... ......................... ................../’<code>: :</code>\,---<code>.„................ .....
......................... ......................... ................/|,-‘</code>¯¯<code>(o)_ \,----,,,...........
..................„~*¯¯</code>” \,.............. <em>„</em>.............( <code>(o),,/</code> ¯ : o : : :o <code>-, ......
................./ : · :(“, ~;’¯¯¯”\, (_,--</code><code>”~,......\ : : : : : : : : : : : : : : :</code>.....
.................| : · :,/<code>,-~;~··-„,/ (‘</code> <code>)/· ,\....,/</code> : : :<em>„„,: : : : : : :: : : : ...
................| : · / ,/<code>,--\ \&#8217;</code>c\,---„1 ‹’--(</em>,/ ./ : : :,/<code>· · ): : :: : : : : : : : |...
.................) : µ’</code> .\ (c) <code>¯ . . ·</code>), . ,-~<code>..... \ : : :| · · '”\, : : : :: : : : : ::/...
.............../</code> ,/~-, .<code>;;</code>„-„,</strong>,./, <code>/.......: : :\ · · · · *-.,: : : : : :,-&#8216; ....
...............-/¨;--;~’ `”*-=,=- ,) . ,/...........› : \, · · · · · · ¯&#8221;~---,--......
...„„.........\, · ‘, · . . . \&#8222;_,/ ,/;-;_ ............ /&lt;/code&gt;| ./ · · · · · · · : · : „-“..........
.( :·&lt;code&gt;\,-~*&lt;/code&gt;¯ · · ·&lt;code&gt;¯&lt;/code&gt; &lt;code&gt;~--*--~;/&lt;/code&gt;~&lt;em&gt;--, . 1 ./ : |, . . . .---~”`...............
. \·:··:”</em>~-,,„„_<em>„„,.-~” <code>¯</code>/ / · · · · · ‘\,) , / : :| · : : · ·/....................... ....
...<code>”’~-,„„„„„„,,~‘</code>.( · · · ,.</em><em>| | · · · ·, <code>&#8222;/ ,/</code>: : : / · · · ·| ......................... .
......................... ”””<code>· · ·</code>’~;-,„,,</em>)”<code>_-&#8216; : : : / · · · · · | ......................... .
......... (</code><em>”-,„,-”¯¯”<code>-;&#8222;· · · · · · , /'</code><code>,-~&#8221;</code>¯: : :/ · · · · : |........................ ....
.......... | : :(,;-===-„, <code>\,· · · ,-</code>| · ·/ : : : | :,/<code>· · · · · ·| ......................... ..
......... . \ : :\, · · · · \\ ·</code>. ·)· / · / : :\ : :<code>~,_ · · · · · :: | ......................... ..
............ \\,_</code>~.,_„,.-</em>\, <code>/,/&#8222;/</code> ,/ : : :<code>&#8217;;-.&#8222;_ : ¯-, · · · /,....................... .....
..............</code>\,,<code>&#8221;| : :</code>-,„<em>„))’<code>"</code> ·,/</em>„,~<em>’<code>· · ( · , ,</code>)·,-; <code>’\,..................... .
............. ,.¬-,--\ : : : : : : : : / · · · · · · · · · · · ·,,-“</code>.....’-„,-‘........ .............
..............| |: :</em>:\ : : : : : : :<em>/ „</em>__<em>„„„„,---~*<code>......... ......................... .......
............./ / : : :</code>~-„„„„„,.;;<code>,,.--</code>...... ......................... ......................... ....
............| | : : : :|¯ : ,/ ¯ ......................... ......................... ......................
.............\,\,</em>,„./--- ~`....................... ......................... ..... ........................
......................... ......................... ......................... ..... ......................... ....</p>

20

<p>Wow...^</p>

<p>
[quote]
My point is that they may not be interested in professional schools AT THE EXACT SAME LEVEL as that of other schools -- an assumption on which the WSJ survey depends.

[/quote]
</p>

<p>Why does the level need to be "exact?" I think anyone making that assumption would be looking at things wrong, but it is safe to assume that the percentage is similar at Columbia. Either way, because the exact number of students in each school applying to the "top" law/biz/med programs was not used, these rankings should be considered nothing more than an approximation when determining a school's success at placement. Going back to the original post, the fact that Columbia is one or five or however many spots behind Williams says nothing to me - in the end, I've learned that they both place exceptionally well, but it would be foolish to take this ranking (especially given its numerous flaws) for gospel.</p>