US NEWS Rankings: What Would They Look Like Without Peer Assessment Score?

<p>^^ I agree, vonlost (post #79). The schools that now get to cherry-pick the best students and then boast about how well those students do once they leave the hallowed halls of alma mater would want nothing to do with a ranking system that tried to measure what value they actually added to their alums’ success. Concededly, it’s a tricky thing to measure. But the bigger issue is that many of them fear they’d be exposed as riding on their students’ coattails, rather than vice versa which is the message they’ve been successfully selling up until this point.</p>

<p>Say Harvard had 2400 - 2400 = 0 and Podunk had 2400 - 2100 = 300. Both produce top grads, but is Harvard better?</p>

<p>Except I don’t think measuring before-and-after test scores would tell you much. Good test-takers are good test-takers. Generally speaking, those who do best on the SAT are likely to well on the LSAT, MCAT, GMAT, GRE, etc., regardless of where they get their undergraduate education. That shouldn’t be a surprising or interesting result. And very little in the course of an undergraduate education is going to prepare one to do well (or better) on those tests. They’re all just variants on academic “aptitude” tests, aimed more at measuring the native abilities of the test-taker, rather than measuring what one has learned or how one’s cognitive, analytical, and critical skills have improved.</p>

<p>You’d need some more sophisticated measure of the value added of an undergraduate education, normalized so that a school would not get to claim credit for the native abilities of its students. I’m not sure what that would be.</p>

<p>In general critical thinking, cognitive ability and such is innate. No school you go to is going to change that. They can hone your ability but one can do so much with available raw material</p>

<p>

</p>

<p>Another way to look at it is that this calls into question, whether there is any fundamental difference between top end schools where students of high ability can choose to learn as much as they can and improve themselves as much as possible. There is no substantial evidence that one schools teaching is better than the other.</p>

<p>The is a lot of hearsay on the internet though</p>

<p>^^ Yes, it would have no more value than considering SATs for entrance. Deltas would be expected for both skilled and unskilled test takers, so it could be apples to apples. It would be just one additional measure of quality, nothing definitive.</p>

<p>I remember reading a study of SAT submitters vs. non-submitters at test-optional schools. In some measures, such as 4 year graduation rates, the non-submitters out-performed the submitters, and in almost all others they were within the study’s margin of error (e.g. a .04 difference in average GPA -something like 3.31 vs. 3.35). The one measure in which submitters clearly out-perfomed non-submitters was graduate school admissions in test-reliant fields such as medicine and law.</p>

<p>(Please don’t hold me to the exact numbers. I’m too lazy to look up the study this morning. :))</p>

<p>All Jesuit schools would be much higher, Emory would be lower. Michigan higher</p>

<p>“In general critical thinking, cognitive ability and such is innate”</p>

<p>IMO, critical thinking is not innate. It is a skill that can be learned and continuously improved upon.</p>

<p>Although cognitive ability was thought to be natural, recent studies indicate that one’s physical environment also influences its outcome. </p>

<p>The challenge is to measure the growth in intellectual abilities and field specific knowledge & skills over the course of students’ time in college. That is what the USNWR or other rankings do not provide.</p>

<p>

</p>

<p>My bad agreed. I meant cognitive ability, I actually have never believed that critical thinking was innate and pretty much understand it is strongly training dependent. However cognitive ability is still innate, I can bet physical environment serves as a variable but some individuals have a maximum.</p>

<p>

</p>

<p>I still believe there are no fundamental differences in most top 20-25 undergrads and that such a measure will prove futile and just be a measure of the students innate cognitive abilities and drive. Smart students will be motivated to learn more in their field. As long as there is a library, and some sort of guidance in place they are fine.</p>

<p>^^You’re making the same fundamental mistake other USNews supporters make, the assumption that the same 20-25 colleges would carry over from one ranking system to the other. That hypothesis has yet to be tested.</p>

<p>

</p>

<p>Well tbh 20-25 was arbitrary. I dont know where the cutoff mark is. </p>

<p>However, people keep making request to measure academic outcomes but this is impossible because academic outcomes will still strongly be linked to cognitive ability. Smart motivated students will still get the best jobs, get the best test scores, gain the most out of college.</p>

<p>Hypothetically, if you send the best scholars to some school with individual average SAT scores in the low 500s and you think you bet you are gonna breed einsteins you are going to be losing lots of money. The scholar will just get tired of wasting their time setting exams that students just don’t understand.</p>

<p>My thesis is that the stratifications created by USNEWS and most rankings are imaginary.</p>

<p>Biggest difference is in student’s actual academic ability and motivation to learn. </p>

<p>Your teacher can teach but he/she can’t make you learn.</p>

<p>^^I think the alternate theory goes something like this: That there are schools where students exit knowing more than when they walked in the door. What are these schools? How much more knowledgeable were they? How much could be attributable to the school itself? I think ifyou could actually design such a study, the results would be far more interesting than the idea that smart people stay smart over time.</p>

<p>I think the hardest thing about measuring outcomes isn’t necessarily the measurement per se but defining the set of outcomes. What kinds of outcomes are most desirable? Even if we could agree they should have something to do with knowledge, well then what kind of knowledge should we measure? </p>

<p>In my opinion, admissions selectivity actually is one of the best metrics we have. If we assume that the most capable students are rational and reasonably well-informed (collectively), then we can expect they will tend to choose the best available colleges. It’s possible for mis-information or some mass psychological effects to distort their decisions ([Tulip</a> mania - Wikipedia, the free encyclopedia](<a href=“http://en.wikipedia.org/wiki/Tulip_mania]Tulip”>Tulip mania - Wikipedia)). However, in a democratic society mis-information eventually gets corrected by better information. In a free market, bubbles eventually pop.</p>

<p>

</p>

<p>I disagree with this to an extent.</p>

<p>It’s true that students should be academically qualified in order to get in to colleges and understand the material. However, there are really two types of “qualified” that we see today: academically qualified, and competitively qualified. Academically qualified would mean meeting the minimum requirements set by the university - the threshold at which the university feels students are prepared enough to grasp the material, and succeed. This level of qualification is generally much lower than the competitive qualification, which is more dynamic, changing year by year depending on the number of applications the schools receive.</p>

<p>And competitive qualifications have simply gotten ridiculous over the years in order for colleges to remain exclusive, and highly ranked. I would argue that nobody really needs a 4.0 GPA with perfect SAT scores in order to study something like relatively easy like Philosophy or English… it’s overkill. There are less competitive schools that offer challenging programs in Physics, Math, and Engineering, and most of those less competitive students still end up succeeding and doing well for themselves.</p>

<p>

</p>

<p>But then shouldn’t we assume the schools with the largest N of top-stats students are the best, rather than the schools with the highest median admissions stats? After all, more “rational and reasonably well-informed” high-stats (which we take as a proxy for “capable”) students are “choosing” the school where the largest number of them end up, aren’t they?</p>

<p>As an illustration: in the fall of 2012, UC Berkeley enrolled an entering class of 4,162. Of that number, 92% or 3,792 freshmen submitted SAT scores; of those, 56% or 2,124 had SAT M scores in the 700-800 range. That same year, Princeton enrolled 1,357 freshmen, of whom 89% or 1,213 submitted SAT scores; 80.3% of Princeton’s SAT-submitters, or 974 freshmen, had SAT M scores in the 700-800 range. So there were more than twice as many freshmen with 700+ SAT M scores at UC Berkeley as at Princeton.</p>

<p>Now of course it’s a little more complicated than saying twice as many high-stats students “chose” UC Berkeley than Princeton. It might be the case that more 700+ SAT M students applied to (and in that sense “chose”) Princeton and were waitlisted or rejected . . . but we don’t get to see those figures, which would give us a more complete picture of which schools high-stats (“capable”) students are “choosing.”</p>

<p>But I submit that if we take your proposition seriously–that the most capable students end up choosing the best schools and that therefore if we follow where the high-stats applicants end up it will tell us which are the best schools–then we need a far different metric than the one US News uses, middle 50% SAT/ACT scores. That metric deeply discounts the very large numbers of high-stats students who end up choosing schools like UC Berkeley, UCLA, UVA, Michigan, and UNC Chapel Hill, diluting their stats by averaging them against the lower-stats applicants who also choose to attend those large public institutions. A metric that accounted for the number of high-stats (“capable”) students who ultimately choose a school, rather than their percentage in the student body, would come closer to telling us what you claim “admissions selectivity” tells us, i.e., where the best students are choosing to go.</p>

<p>

</p>

<p>It is indeed a lot more complicated than using the Math %age 700-800 statistics:</p>

<p>For instance, the 25th percentile at Princeton is 700V and 710M. That means that only 25 percent did not score at 700 or above. In the case of Cal, the figures are 590V and 630M. </p>

<p>The corresponding 75th percentile are 790V and 800M at P, and 720V and 770M at Cal. </p>

<p>In so many words, when it comes to verbal achievement, the 75th percentile at Cal barely reached P 25th percentile. Something that would have been apparent when comparing the 700-800 percentage for verbal, namely 36 percent at Cal versus 76 percent.</p>

<p>And as we can see from comparing multiple CDS at the most selective schools, the weight placed on the verbal score can be seen as being given a more “deciding” factor as the ultra high math scores have become more and more frequent at highly selective schools.</p>

<p>All in all, the real story of selectivity is told at the 25th percentile.</p>

<p>No, xiggi, you’re trying to shift the ground back to average and median test scores. What tk21769 said was that top students choose top schools, and therefore you can determine which are the top school by following where the top students go. My point is simply that there are more top students, in absolute numbers, at a school like UC Berkeley than at a school like Princeton. So let’s add CR into the equation. In 2012, 36% or 1365 of UC Berkeley’s enrolled freshmen had SAT CR scores of 700+. That same year 76% or 1213 of Princeton’s enrolled freshmen had SAT CR scores of 700+. So there were more 700+ SAT CR-scorers at UC Berkeley than at Princeton.</p>

<p>The point is, focusing on averages and medians–whether the 75th percentile, 25th percentile, middle 50%, mean, or whatever–tends to obscure the fact that very large numbers of high-stats students enroll at the top public flagships; more, in fact, than enroll at a school like Princeton. The only difference is that in addition to a very large number of very high-stats students, the public flagships also enroll a very large number of students with less lofty stats. That brings down their averages and medians. But we shouldn’t lose sight of the fact that a very large absolute number and a very large percentage of the nation’s high-stats students are choosing the leading public flagships. The averages and medians can be very misleading if we let them deceive us into thinking (as I suspect was the premise of tk’s original comment) that the top private schools have the largest numbers of top-stats students. No, they have the largest percentages of top-stats students, but for most of them the N is actually quite small.</p>

<p>

</p>

<p>You obviously don’t work in marketing. :)</p>

<p>Stating the obvious, the more students who apply, the more selective a school will become (since it can only admit so many). Look at all the posts on this site that say things like, “Princeton sent me a catalog! They must want me. I wasn’t going to apply, but I will now.” Or look in your own student’s inbox to see how colleges encourage them to take a chance–waived application fees, automatic entrance into scholarship competitions, quick notification of decision if they apply by a certain date.</p>

<p>

</p>

<p>Evidently, that was my intention when I wrote: “All in all, the real story of selectivity is told at the 25th percentile.” Or was it?</p>

<p>Fwiw, a statement such as “there are more top students, in absolute numbers, at a school like UC Berkeley than at a school like Princeton” is a testament to a faulty use of statistics. In this case, you define “top students” as the students who happen to score above a 700 SAT score, and place special focus on the Math component. This seems to ignore that the “top students” are evaluated on a great number of factors, and that a 700 Math score might very well be considered as trivial if not a bottomline. </p>

<p>This is not different than the previous arguments made on behalf of schools such as Michigan, and other elite public schools. The simplistic “absolute” number of 700 SAT scores simply belies the fact that the student body at Cal or Michican is not interchangeable with Princeton’s, or HYS for that matter. The students who DO get accepted at such schools DO attend the private and more prestigious schools when given the opportunity. This fact has been documented in the past as hardly NONE of the cross-admitted students enroll at Cal over Stanford. Hardly none as fewer than a … dozen! </p>

<p>In summary, you are using a simple datapoint and give it excessive weight in the application process. The fact that a student could score 700 plus 700 (or 1400) has, perhaps unfortunately) become almost meaningless in the “high game” … just as being one of the 30,000+ valedictorians. </p>

<p>And, if the gross SAT really had so much importance, would you not agree that a class composed of 2/3 of students who failed to surpass a 700 on the verbal component ought to be quite different from one where about 1 out of four did the same? </p>

<p>Again, the real story of the selectivity of a school is not determined by its top 25 percent, but by how homogeneous the full class is. And on this issue, to be complete, you might want to address how different a graduating class is statistically at Princeton and Cal … after accounting for the massive number of JUCO transfers. </p>

<p>Very different schools. Very different missions!</p>

<p>A typical Berkeley undergrad will take a Political Science class with say 100 people and this class will reflect the academic aptitude of the wide range of students that UCB admits and not only the high achievers. At Princeton, everyone is an academic superstar so the overall academic environment will be richer.</p>

<p>Bclintonk, here’s the ultimate reason why you’re wrong; if there are more high-stat UCB students than Princeton students, then why are Princeton undergrads better represented even at the absolute level at Harvard Law, Harvard Business, Harvard Med, Stanford Law, Stanford GSB, etc.?</p>

<p>U of M Ann Arbor, UC Berkeley, and UVA should have as many alums at Harvard Law as a school like Princeton and Stanford do but they don’t. In fact, its not even close.</p>

<p>Statistics are misleading and Princeton’s student body is much stronger than UCB’s-in a way that statistics simply can’t measure.</p>

<p><a href=“http://www.nationalmerit.org/annual_report.pdf[/url]”>http://www.nationalmerit.org/annual_report.pdf&lt;/a&gt;
Princeton enrolled 181 National Merit Scholars in the Fall of 2012 while UC Berkeley enrolled 91. The next best state schools were UT-Austin with 57 National Merit Scholars and UMich with 46 National Merit Scholars.</p>