Comprehensive Ivy League v. non-Ivy League Thread

College Search & Selection
461

<p>DunninLA-
Good point. There are some disadvantages to having a more "diluted" student body. If you stay away from the football games, you can avoid most of those disadvanteges. </p>

<p>But, there are also advantages to having large numbers of very bright students. The culture of a school tends to be dominated and led by the most talented, like the culture of our nation. Having greater numbers of talented individuals adds to intellectual diversity, cross-fertilization of ideas, potential for new ideas and original discourse, potential for growth. An extreme example of a college with high average SATs: imagine being at a college with two students, you and Einstein (or whoever). Even that could grow stale. At the same time, large numbers of talented students increase your ability to find intellectual and social soul-mates. You simply need a critical mass of such talented individuals for all these things to happen.</p>

<p>A larger number of talented alumni will create a stronger alumni network and maintain/enhance the reputation of your school with their contributions to society. In classrooms, the questions and discussions are elevated to the level of the brightest students.</p>

<p>In second and third tier schools, the bottom 25% are gone after the first year or two. In the elite schools, even the bottom 25% are bright enough to succeed but I think they are elevated to a higher level as the years go by. I recall a student being admonished by a Professor freshman year for asking a "stupid" question. (Most Professors are more patient than that.) If you are in the bottom 25% at an elite school, you will succeed and probably grow intellectually and socially more than any other quartile. I have heard more than one person say that Cornell takes away your self-esteem and then gradually gives it back to you. Students who were outstanding in high school find themselves working hard to keep up.</p>

<p>Furthermore, I think you CAN choose with whom you associate after freshman year. You get to know other students in your major and you can take courses, work on projects and problem sets together. Everybody chooses their roommate and/or housemates. You decide which organizations to join, which activities to do, who to eat with. Most of all, school is fundamentally a solitary enterprise. You are the one who has to study and learn and take exams and write papers. Education is not very social. Everybody spends a lot of time alone with their books.</p>

462

<p>
[quote]
I did check Berkeley’s statistics and it says the average SAT scores are 1450, not 1350 as you mentioned and 99% of the students were from the top 10% of their high school.

[/quote]
</p>

<p>For Berkeley to have a 1450 SAT average, the school would need to dismiss about 75% of its current enrollment --and drop its athletic department. The 1450 is the 75% percentile, and 1200 is the 25% percentile. The 1350 figure was probably generous for the average SAT at Berkeley.</p>

<p>Why would the "99% of the students were from the top 10% of their high school" be any surprise considering the admission policies of Berkeley that render such figure quite irrelevant on a national scale.</p>

463

<p>
[quote]
how reliable are those data?

[/quote]
</p>

<p>On the student-related expenditure per full-time equivalent data, I've seen the comment that a college with an attached medical school or other high-expense program can look better than a college without one. I'm still investigating the data definitions, and will probably ask about this on the Common Data Set email list. I can't imagine that the overall spending per undergraduate student differs greatly between Cal and UCLA, but I'd have to pore over the line items in the budget of each campus to be 100 percent sure what to think.</p>

464

<p>I think it's stupid to avoid the bottom 75% or whatever amount of the school. At the schools being discussed in this thread, the kids in the bottom 75% are damn smart kids. Acting like the top 25% are the only people one should associate with to have a fulfilling experience in college is ridiculous. Locking oneself away in an ivory tower of the top 25% seems like it's one of the best ways to miss out on the college experience.</p>

465

<p>tokenadult:</p>

<p>
[quote]
Can you show impact from that? Where are the figures that show that that makes much of a difference in comparing colleges?

[/quote]
</p>

<p>I'm not sure where it is now, but a study was done some time ago that showed superscoring can add up to 50 or so points to a college's average SATs. Perhaps Alexandre knows about this.</p>

<p>A few notes:</p>

<p>1) Berkeley doesn't superscore.
2) While the 25th percentile ranges seem a little low in comparison, it's often because a student may have high scores in one or two of the SAT sections, but a lower score in another. It seems that (top) publics tend to emphasize the SAT less than privates; whereas privates will rate scores "very important," Cal rates it just "important." This is probably because research has shown that GPA is much more indicative of one's success in college than SAT. Hell, I think research at Berkeley showed that SAT II scores were more indicative than SAT I scores. In addition, Cal seems more forgiving on a single weak SAT section; if one is low but the others are higher, it's okay. But for privates, it tends to be that all three sections need to be more comparable; if one is weaker than the others, it tends not to fall as low as 590 (the 25th percentile for Berkeley). This goes along with the superscoring policies.
3) The 50th percentile is not necessarily the average. According to Cal's ranges, the 50th would be a 1990 or so, but the actual average is around a 2050 (from the UC site). This would be higher if superscored. This seems to support #2.
4) One of the consequences of not superscoring is that the difference between the 75th and 25th percentiles tends to be greater than 100 points. Usually, from the superscored SAT data (Stanford, Harvard, any top private), the difference is 100, but if you look at Berkeley's, it's much greater. From the newest SAT data:</p>

<p>CR: 590-710 (120 pt difference)
M: 620-750 (130 pt difference)
W: 590-710 (120 pt difference)</p>

<p>Assume the superscored data will yield a 100-pt difference and that if superscored, the 75th and 25th percentiles would be higher. One way I've seen this done is: first, minus 100 from the 75th percentile; second, find the difference between that and the 25th percentile; add that to find the new 75th percentile; then minus 100 from that to find the new 25th percentile. If done like this, it'd be:</p>

<p>CR: 630-730
M: 680-780
W: 630-730</p>

<p>That would put the 50th percentile at 2090; if you add about 50 points to the actual average (unsuperscored), it comes out to be about the same, so the 50th percentile and the average are more comparable. The jump seems legitimate given the very broad ranges from before.</p>

<p>5) Last I heard, Berkeley enrolled the highest number of 1500+ scorers (CR+M), more than even Harvard. But of course, there is a greater number of sub-1500 scorers on the SAT.</p>

<p>The difference between Berkeley and, say, Cal State Humboldt is that while both accept 1500+ scorers (the former tends to, but often doesn't), Berkeley manages to attract them to attend. That's why the top X students at Berkeley are equal to the X undergraduate body at a top private. Berkeley also has a "tail" of students--an extension of the elite 6000 or so--that aren't as statistically amazing, but still are excellent students.</p>

<p>
[quote]
I've seen the comment that a college with an attached medical school or other high-expense program can look better than a college without one. I'm still investigating the data definitions, and will probably ask about this on the Common Data Set email list. I can't imagine that the overall spending per undergraduate student differs greatly between Cal and UCLA, but I'd have to pore over the line items in the budget of each campus to be 100 percent sure what to think.

[/quote]
</p>

<p>Actually, I recently posted a discussion about this; in looking at the budgets of the two, I'm certain that UCLA receives more federal money because it has a med school. Same for UCSD.</p>

<p><a href="http://talk.collegeconfidential.com/university-california-los-angeles/440682-ucla-federal-revenue.html%5B/url%5D"&gt;http://talk.collegeconfidential.com/university-california-los-angeles/440682-ucla-federal-revenue.html&lt;/a&gt;&lt;/p>

466

<p>
[quote]
One of the consequences of not superscoring is that the difference between the 75th and 25th percentiles tends to be greater than 100 points.

[/quote]
</p>

<p>I doubt very much that that is a direct result of superscoring or not. More likely that reflects the difference between a college that draws all of its enrolled students from the top of the national score distribution and a college that admits some students who score nearer to the middle range of scores.</p>

467

<p>
[quote]
Actually, I recently posted a discussion about this; in looking at the budgets of the two, I'm certain that UCLA receives more federal money because it has a med school. Same for UCSD.

[/quote]
</p>

<p>Makes sense...I wonder how it compares when you have Cal + UCSF?</p>

468

<p>tokenadult:</p>

<p>
[quote]
I doubt very much that that is a direct result of superscoring or not.

[/quote]
</p>

<p>To me, it makes perfect sense. If a student took the SAT twice, and scored lower in one section but higher in the other two, but overall the score improved, the second score would be taken, despite the low single-section. That's why superscoring can make a difference. I believe my superscored SAT was 60 points higher than the highest single-sitting. (And I'm just one example.)</p>

<p>UCBChemEGrad:</p>

<p>
[quote]
Makes sense...I wonder how it compares when you have Cal + UCSF?

[/quote]
</p>

<p>I was actually think about this. Both UCSD and UCLA receive substantially more federal money because of their med schools. Berkeley's budget is half of UCLA's budget. But if you add Berkeley's and UCSF's budgets, the figure is much more comparable.</p>

<p>University</a> of California Financial Reports</p>

469

<p>It still remains to be shown just how many individuals have higher "superscored" SAT scores than single-sitting SAT scores. One anecdote is exactly one anecdote. I have asked College Board for comprehensive figures, but so far they have not provided those.</p>

470

<p>
[quote]
Berkeley's budget is half of UCLA's budget. But if you add Berkeley's and UCSF's budgets, the figure is much more comparable.

[/quote]
</p>

<p>I still need to look at the College Results methodologies more, but it appears from what you have just said that when colleges are compared by spending per student, that often means spending divided by ALL students, rather than spending per UNDERGRADUATE student, which is probably more relevant for choosing an undergraduate college to attend.</p>

471

<p>
[quote]
I have asked College Board for comprehensive figures, but so far they have not provided those.

[/quote]
</p>

<p>I think it'd be better to ask the universities themselves, unless College Board collects data for specific schools. The reason is that, I'd say, those who are accepted at and attend top schools tend to be those who improve their SAT more markedly. CB has already said that the average increase is only 20 or 30 points, but I'd bet anything that at schools like Berkeley, the students' average increase is a bit higher.</p>

472

<p>Only College Board has figures for EVERY test-taker, and that's what I'm trying to check: how many more, say, 2340 scorers there are on a "superscored" basis than on a single-sitting basis. Some number lower than the total number of test-takers :) achieves each level of scoring, by either basis, and College Board has published each year for at least a few years now how many distinct individuals in each high school graduating class reach each level of scoring, e.g., </p>

<p><a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/SATPercentileRanksCompositeCR_M_W.pdf%5B/url%5D"&gt;http://www.collegeboard.com/prod_downloads/highered/ra/sat/SATPercentileRanksCompositeCR_M_W.pdf&lt;/a> </p>

<p><a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/composite_CR_M_W_percentile_ranks.pdf%5B/url%5D"&gt;http://www.collegeboard.com/prod_downloads/highered/ra/sat/composite_CR_M_W_percentile_ranks.pdf&lt;/a> </p>

<p>College Board also posts a link on the effects of retaking the tests. For a while the link they had on most of their summary Web pages went to an older version of this chart. Checking just now, it seems that the most current version of these data is once again posted for any member of the public to view. </p>

<p><a href="http://www.collegeboard.com/prod_downloads/highered/ra/sat/AverageScores.pdf%5B/url%5D"&gt;http://www.collegeboard.com/prod_downloads/highered/ra/sat/AverageScores.pdf&lt;/a&gt;&lt;/p>

473

<p>Around and around we go to explain lower SAT scores at Berkeley!</p>

<p>
[quote]
Hell, I think research at Berkeley showed that SAT II scores were more indicative than SAT I scores.

[/quote]
</p>

<p>Well, now that the universal clueless have helped The College Board pick up a lot of additional revenues at the expense of millions of test takers, the country is left to enjoy the results of the misleading and self-serving "research." ordered by Atkinson to satisfy his ego. In the meantime, the UC will simply look for yet another excuse to mask its growing ineptitude in maintaining a racial distribution that is remotely close to the population of California. </p>

<p>
[quote]
U. of California May Drop Requirement on SAT Subject Tests</p>

<p>The University of California should stop requiring in-state applicants to take SAT subject tests, according to changes in freshman-eligibility requirements proposed recently by the system’s Board of Admissions and Relations With Schools.</p>

<p>Applicants must now take two of the tests in subjects of their choosing.** In its proposal, however, the board says that the requirement “contributes very little to UC’s ability to predict which applicants will perform well initially at UC.” **The board also concludes that the requirement contributes to the underrepresentation of black and Hispanic students in the system’s applicant pool.</p>

<p>“The requirement isn’t helping us; instead it is differentially burdensome on certain groups,” Mark M. Rashid, the board’s chairman and a professor of civil and environmental engineering at the university’s Davis campus, told the Daily Bruin.</p>

<p>Before 2006, the university required three SAT subject tests — mathematics, writing, and one elective. But changes in the new SAT Reasoning test (including the addition of a writing section and elements of the math subject test) prompted the system to consider altering its policy on subject tests.

[/quote]
</p>

474

<p>I agree that SAT IIs are a waste of time. My understanding is that they are generally only required by top-tier schools, and most of those students applying will have already taken APs courses and the corresponding AP tests. I think our kids (in Cal. anyway) are too burdened with testing.</p>

475

<p>Abolish</a> the SAT — The American, A Magazine of Ideas</p>

<p>An interesting and relevant article.</p>

476

<p>
[quote]
Well, now that the universal clueless have helped The College Board pick up a lot of additional revenues at the expense of millions of test takers, the country is left to enjoy the results of the misleading and self-serving "research." ordered by Atkinson to satisfy his ego. In the meantime, the UC will simply look for yet another excuse to mask its growing ineptitude in maintaining a racial distribution that is remotely close to the population of California.

[/quote]
</p>

<p>I really have no idea how this is relevant. My point simply was (perhaps you didn't see it) that in a study--yes, an actual study, not "self-serving"--GPA, SAT, and SAT IIs were measured for their accuracy in predicting college success. The SAT came in last place. SAT IIs were slightly more effective, though still pretty ineffective. And GPA, of course, was the most indicative, very accurate in its predictions.</p>

<p>Whether racial diversity has a role in this is irrelevant. It's simply a fact. (Why you had to make an entire point on an aside of mine, I don't know.)</p>

477

<p>SATs are one of the best predictors of college success. Look at the US News data and the correlation between SAT and graduation rates.</p>

<p>Within a particular college the SATs may not seem to be as predictive of success because smarter students go into harder majors. It is like handicapping in golf.</p>

478

<p>
[quote]
SATs are one of the best predictors of college success. Look at the US News data and the correlation between SAT and graduation rates.

[/quote]
</p>

<p>Though I haven't run the numbers, I suspect there would be a similar degree of correlation between GPA and grad rates as well.</p>

479

<p>kyledavid80,</p>

<p>i didn't really read your long posts but if your point was to say superscoring makes a huge difference in SAT for the whole student body, then maybe you should look at the difference in ACT between Michigan and Northwestern. Both schools don't superscore ACT. Yet, Northwestern is supposed to be superscoring SAT like other top privates while Michigan doesn't. So if superscoring really makes a huge difference, we'd expect the gap in ACT to be significantly smaller than the corresponding gap in SAT. It turns out the difference in ACT (2-3 points) corresponds very well to the difference in SAT (~80 points). This applies to other schools as well, not just Michigan vs Northwestern. </p>

<p>Remember, a lot of kids improve in both sections and a lot of kids take it only once. For every one of these kids, you need a 120 point difference in superscoring to make the overall difference of 60 points. You can't project your own case and extrapolate it to everyone else. My guess is the difference is no more than 10-20 points once the whole student body is considered.</p>

480

<p>Tokenadult, those data from the College Board would be meaningful to me as well. </p>

<p>I am refamiliarizing myself with the college admissions world for the sake of my high school freshman. When I look at UC Berkeley and UCLA's incoming freshmen SAT scores, they are alarmingly lower than other schools of similar reputation. Now I learn that the published scores of UC schools and just about all the other top 50 are not directly comparable. I would be very interested to learn either what the UCB and UCLA superscored SAT results would be if published, or a correction factor to normalize those single sitting scores against superscored results.</p>

<p>Please post when you find this information. My gut says the correction would be in the neighborhood of 20-30 points per test section -- therefore 50 points on the two section reporting, and 75 points on the three section scoring. </p>

<p>Having attended both Stanford and UCLA, and having visited Berkeley innumerable times, I didn't find the top half of the UCLA or Berkeley students much different from those at Stanford.</p>