<p>If you adjust the initial analysis (based on census) by post #264 (number of students coming from each region), you find the Ivy’s are more national than most other schools. MIT is the clear winner(101, 99, 108, 91) but Harvard has 117NE and Yale has 120 NE while Wash U has 218 from the MW, Duke has 202 from the South, and USC has 288 from the West…</p>
<p>*Fully agree that “lower” schools are going to be far, far more regional in scope. *</p>
<p>and we could do another few pages explaining why some regions don’t necessarily value the NE elites (for their own children) and why that makes sense. And why they don’t care if the NE values the NE schools for their own NE children. And how upper SES probably values pretty much the same schools nationwide.</p>
<p>In a thread where one point is The world of prestige is not just defined by the northeast view, I don’t see what value ranking schools by order of prestige makes. Maybe they aren’t “lower” schools. Maybe they just appeal to a different audience? Can we have a discussion about geographic diversity in colleges without also discussing prestige? sincere question. Maybe it isn’t possible. I have no idea.</p>
<p>Can you explain what you did so I can track? </p>
<p>My original numbers were already weighted to census. Are you now taking “only the universe of those attending top 20 schools” and treating THAT as the basis, versus census pop? </p>
<p>If that’s the case, I think what your numbers mean is … “Within ONLY the world of students who attend top 20 schools, MIT is most equally regionally representative of the geographical dispersion of THAT world.”
That happens to be a NE-skewed world, since NE-ers disproportionately make up that world, AND disproportionately more of those seats sit in the NE, but yes, MIT might indeed be more “nationally representative” of that world. Am I tracking correctly with what you did? </p>
<p>If that’s the case, then you baked it in that all NE schools by definition will better represent the “national” distribution of that pool of students, because you’re comparing each school to a NE-weighted universe. I don’t know anyway in which your definition could result in any other region school being more nationally representative. You skewed your comparison to NE students. </p>
<p>Note I didn’t weigh census regions equally, just by pop. I’ll do it what I think is “your way” and see what I find. It’s an interesting metric, just a different one. </p>
<p>
</p>
<p>someone said this upthread - the issue with this is that there are far more kids qualified and capable of doing the work in the most selective colleges than they have room for. As stated earlier to figure out exactly how many capable, qualified graduates are in every state and normalize that for state population and uni would be virtually impossible to do. One cannot just assume that per capita the NE has more “smart kids”…the only B/W is that those NE selective colleges take more local region kids - but we’re already said the same about the big public unis like Michigan and Virginia - there are more Michigan kids at UofM than any other region even though perhaps (and I don’t know) there are less kids in the midwest than other regions. You can’t presume therefore that Michigan kids are smarter because there are “more of them” which is what is being surmised by the earlier post comment. </p>
<p>That is exactly what I did, and you are right, these numbers are biased to the NE. The initial numbers also had a bias inherent in the data. You could take all of the NE schools out of your analysis, and students coming from the NE would still index over 100. This is pretty amazing considering all of the seats in the NE that these students are filling.</p>
<p>You admit the analysis is flawed, that your data is not what you would like, but still want to draw conclusions.as if your data is accurate. The conclusions drawn in this thread are not supported by the data shown.</p>
<p>How did my initial numbers have bias when I took the top schools and indexed to census? If I had cherry picked schools, yes, but I didn’t. </p>
<p>Because the census isn’t the right population.</p>
<p>Alittle more Thursday fun - Forbes smartest cities (and there are many of these generally overlapping in results) - the premise is often issued that smart people produce smart children ergo, perhaps the cities with the smartest populations produce the smartest children?</p>
<p><a href=“The Smartest Cities In America”>http://www.forbes.com/2008/02/07/solutions-education-smartcities-oped-cx_apa_0207smartcities.html</a></p>
<p>And here’s more fun about urban vs. rural vs. suburban etc.</p>
<p><a href=“Bloomberg - Are you a robot?”>Bloomberg - Are you a robot?;
<p>and to rip a statement from another of these survey stories:</p>
<p>
</p>
<p>
</p>
<p>That actually makes the Mass cutoff at 223 more impressive than Wisconsin’s 208. I would assume that the students taking the test would be more skewed toward the college-bound students. The fact that Mass had more than 3 time the students take the test than Wisconsin and still scored significantly higher is significant. I would doubt that those 17,000 in Wisconsin would be mostly less prepared, less academically focused students. </p>
<p>Here’s a little thought experiment: what if a student came on CC and posted the following: What I really value is a high-achieving student body that comes from all over the country–and the world–and isn’t full of people from a single part of the country. What schools should I consider?" Setting aside all the posts explaining to the student how this was a silly, pointless criterion, and the others explaining that what matters is finances, etc., what would you tell such a student? To my mind, the data set out in the beginning of this thread–mushy at the edges as they are–would cause me to give a different answer to this question than I would have before I read this thread.</p>
<p>I would have assumed that the schools with the best national reputations among selective schools would also have the most national student bodies. And it’s not completely untrue–some of those schools still have quite a few people from out of their home regions. But there are a number of schools that I would have assumed were more regional in terms of the student body than they actually are. Wash U is a prime example–I would have assumed they were still working to develop their out-of-region appeal, but it appears that they’ve pretty much achieved it.</p>
<p>
</p>
<p>At most schools, yes, but not all. There are a handful – which tend to be in the NE, btw – that do meet the need of internationals.</p>
<p>bp: I redid the analysis “your way” for both the uni and LAC lists, and I got the same numbers you do. </p>
<p>Using your criteria – representative WITHIN THE POOL OF STUDENTS ATTENDING THESE ELITE SCHOOLS – MIT is indeed the most “representative” university – with indexes of 101, 99, 109, 91. </p>
<p>On the LAC side, Wellesley has the most representative body WITHIN THE POOL OF STUDENTS ATTENDING ELITE LAC’s - with indexes of 100, 91, 108, 103. </p>
<p>So by your criteria, MIT and Wellesley “win the game” of being most representative of an elite student body, * even though 41% of MIT and 49% of Wellesley student bodies are from the NE. * </p>
<hr>
<p>There is some merit to your formula, but what this does is “water down” the home region index for ALL NE schools. It also raises the “home region skew” for ALL other schools in all other regions – since (say) a midwest school’s “natural” skew to its home region is now MAGNIFIED by the fact that relatively few of these students IN THIS SET OF SCHOOLS OVERALL are from the midwest (the same also holds for the southern and western schools .</p>
<p>In other words, because the seats are so heavily northeast in the first place, you are “rewarding” schools that naturally skew to the northeast by lowering their index (comparing them to a bigger number) and “penalizing” schools that naturally skew to any other region (comparing them to a lower number). That may be fine, but you have to at least acknowledge it. </p>
<p>*********|
Not convinced?</p>
<p>Here’s a live example. I’ll pick Harvard and Stanford as I think they are comparable bellwethers.</p>
<p>The actual % of students from each region are as follows;
(Northeast, Midwest, South, West)
Harvard: 48% … 13% … 17% … 22%
Stanford: 19% … 10% … 14% … 57% </p>
<p>Makes intuitive sense, right? Even without taking into account size of the regions, you’ve got roughly half of each coming from the home region - the next largest from the opposite coast – the smallest from the MW/South. That passes an intuitive sniff test. </p>
<p>Using MY methodology where I index to size of the regions (23% NE, 22% MW, 32% S, 23% W), the indexes I get are:
Harvard: 208, 58, 54, 96
Stanford: 81, 45, 44, 249</p>
<p>So … I conclude: H and S both skew to their home regions … then average/a bit below below avg to the opposite coast – and lowest to the midwest and south. H’s a bit more nationally representative than S, but neither are highly representative. </p>
<p>Using YOUR methodology where you index to size of the student body who attends these schools (41% NE, 16% MW, 19% S, 24% W), the indexes become:
Harvard: 117, 80, 91, 92.
Stanford: 45, 63, 74, 239.</p>
<p>So … YOU conclude: H does a far better job of getting a representative student body – because they’re better at representing a student body that is already heavily represented to the NE. In other words, you’ve penalized Stanford’s “natural skew” to the West because elite students as a whole don’t skew to the West. </p>
<p>Which do other people on this thread makes more logical sense? </p>
<p>I welcome additional analytical thoughts. </p>
<p>"I would have assumed that the schools with the best national reputations among selective schools would also have the most national student bodies. And it’s not completely untrue–some of those schools still have quite a few people from out of their home regions. But there are a number of schools that I would have assumed were more regional in terms of the student body than they actually are. "</p>
<p>Thank you, Hunt. This is a great hypothetical, and gets to what I’m trying to get at. </p>
<p>“If you adjust the initial analysis (based on census) by post #264 (number of students coming from each region), you find the Ivy’s are more national than most other schools”</p>
<p>I think you want to correct this to say … “you find that the Ivy student bodies are distributed in a way more similar to the pattern of distribution of all of the students who attend elite universities.” Not “more national.” The pattern of distribution of all of the students who attend elite universities is not equal nationally, which I demonstrated many posts ago. </p>
<p>
</p>
<p>I did not comment on NJ --did not follow that part of the conversation-- but I think that a student from Camden who attends Johns Hopkins should not contribute very much to a study that purports to establish an index of national reach. And neither should a student from Baltimore attending Penn. To be clear, it the thick brush of relying on four very broad regions that renders the entire effort suspect on the one hand.</p>
<p>On the other hand, there is a basic thesis that is, IMHO, suspect, and that is none other than the expected percentage of local or regional student at a school that has a national or worldwide appeal and becomes the yardstick of the effort. For instance, if one would track all packages of FedEx by hub, the conclusion might be that the company is a regional powerhouse because of its Memphis location. In the same vein, a school that distributes 30 percent of its student body as In-state and the remaining 70 percent throughout every state in the nation should NOT be viewed as purely local or even regional. You can have a strong local AND national presence at the same time. Stanford would be the poster child for the latter, with HYP in the same boat. </p>
<p>There are many reasons why the high ratio of local students exists at most schools. A school can have a very high percentage of applicants and enrollees for a narrower band and still be considered a national powerhouse. </p>
<p>Again, I discussed elements such as the distances from the schools. In the case of WUSTL/WashU I pointed out students from Nashville are in the academic backyard of St Louis. Just as it might be the case for students as far as Texas or Georgia, it might represent the most selective school in driving distance. To keep it simple, the narrow focus on the extremely large regions of the census is belying the simple facts relating to college admissions and attendance. To add to TPG point about El Paso, please consider that it takes less time to travel to Los Angeles than it’d take to travel to Austin or College Station. I am sure that there are countless similar cases in various parts of the country. </p>
<p>For the record, I believe that BClintonK reached a different conclusion regarding Harvard’s reach when he used the distribution of SAT scores sent by students to various schools. </p>
<p>HTH </p>
<p>So, the NE schools are admitting lots of students from the NE. Who are the non-NE schools admitting from the NE? When you look at your pool of schools not from the NE, the index is >100 for the NE for these schools. You are right, I am biasing the data for the NE. You are penalizing the schools in the NE when this is where a larger portion of students come from than represented by the census.</p>
<p>Why is Stanford at 81 for the NE while 45 and 44 for the other two out of regions? Why is WashU 130 for the NE while 58 and 71 for the other two out of regions? In both of these cases, most students had to travel farther from the NE to get to the school than the other two regions. Just looking at schools not in the NE, one could conclude that more students come from the NE than supported by the census. This can be seen in the data, and should not be ignored when drawing conclusions.</p>
<p>On PSAT scoring - The cutoffs keep going up in states where the number of test takers increases. The cut off is not based on the number of test takers but allocated seats in each state for NMSF. </p>
<p>So if a State has 100,000 seniors and only 10,000 take the test and the state has 500 NSMFs allocated, the cut off for that state tends to be much lower than another state with 100,000 seniors, 45,000 test takers and also has 500 NMSFs.</p>
<p>A larger population taking the test results in a higher cutoff mainly because there are more people who have reasoning abilities in a larger population group than a smaller one. </p>
<p>If you could study the movements of a large pool of “elite students,” if would be pretty interesting. But you can’t do this by just looking at top 20 research universities and LACs, because there are other strong draws for some of these elite students, such as honors colleges at flagships, and merit opportunities. The data about PSAT cutoffs suggests that the NE may have a higher incidence of elite students, but I suspect the connection may be more to SES than to geography.</p>
<p>Or simply that more people take the test in a particular state since the cut-offs are tied to volume. It really doesn’t say anything about the the college headed population as a whole unless I’m missing something. </p>