What are the Top 10 Publics for OOS students?

College Search & Selection
101

<p>[The</a> SAT Warps Debate About Admission - The Daily Californian](<a href=“http://www.dailycal.org/article/13120/the_sat_warps_debate_about_admission]The”>http://www.dailycal.org/article/13120/the_sat_warps_debate_about_admission)</p>

<p>Hmmm…former presidents of Harvard, and Princeton…a Stanford psychologist…</p>

<p>"A 2001 UC Office of the President study of freshmen at Berkeley and six other UC campuses found that the SAT only explained 13 percent of the variance in freshmen grades, and that even this modest correlation was spuriously elevated by socioeconomic factors.</p>

<p>An earlier paper from Gregg Thomson, UC Berkeley’s Director of the Office of Student Research found that there was “zero correlation” between SAT scores and graduation rates for African Americans at Berkeley. The same report found that after separating out recruited athletes, Berkeley students with SAT scores in the 800s had six-year graduation rates of 75 percent, and those with scores in the 900s graduated 79 percent. This is nearly the same as the 82 percent graduation rate among Berkeley students with SATs in the 1500s, who on average are from far more affluent backgrounds.</p>

<p>In The Shape of the River, a mammoth study of elite universities, William Bowen and Derek Bok, former presidents of Princeton and Harvard respectively, found the SAT bore little relationship to graduation rates after controlling for socioeconomic status and other factors. Students with SATs under 1000 graduated 83 percent of the time, while those with SATs above 1300 graduated 87 percent of the time.</p>

<p>In short, these studies confirm Stanford psychologist Claude Steele’s expert report in the landmark University of Michigan affirmative action cases, where he explained that a gap of several hundred points on the SAT scores “can sound big, but actually represents a very small difference in skills critical to grade performance.”</p>

102

<p>“Either the SAT means something or nothing.”</p>

<p>Uhhhhh no.</p>

<p>SAT scores should be taken in context…they mean less than many posters think…but they don’t mean nothing.</p>

<p>Amherst looks at SAT scores and the student’s background.</p>

<p>Other schools do this too.</p>

<p>Obviously… UC Berkeley and the other UCs do this too.</p>

103

<p>^
In that case, they still mean something. There’s simply an additional factor.</p>

<p>The issue with that is some publics have a need-blind admissions process, and factoring in income/background for test scores would potentially get into questionable territory (of course, many schools already skirt this). Mini has noted many times that it’s reasonably easy to get a proxy for income through various other factors (e.g. zip code), but it’s still a very haphazard procedure, I think.</p>

104

<p>Well… Amherst does it anyway. Michigan does it. Smith does it too.</p>

<p>The UCs do this on a much bigger scale.</p>

<p>If the UCs look at SAT scores one way and another school looks at SAT scores another, like Wash U…why should the UCs be judged like Wash U?</p>

<p>Why should UC Berkeley be judged on SAT scores when they don’t use them the way other schools do?</p>

105

<p>RML,
I think you overrate how much value is to be found in things like GPA.</p>

<p>3.9 UC Berkeley
4.0 UCLA
4.0 UC San Diego
3.8 UC Davis
3.8 UC Santa Barbara
3.8 UC Irvine</p>

<p>I think that you overrate how much value is to be found in things like class rank:</p>

<p>98% UC Berkeley
97% UCLA
100% UC San Diego
98% UC Davis
96% UC Santa Barbara
96% UC Irvine</p>

<p>I think you underrate how much value there is to be found in standardized test scores:</p>

<p>1210-1470 UC Berkeley
1170-1410 UCLA
1140-1370 UC San Diego
1050-1300 UC Davis
1080-1320 UC Santa Barbara
1090-1300 UC Irvine</p>

<p>I don’t think that you understand the high correlation between SAT scores and achievement in other parts of one’s application and how this can impact applicant pools, selectivity, and acceptance rates. </p>

<p>21.6% UC Berkeley
22.8% UCLA
41.6% UC San Diego
52.6% UC Davis
49.2% UC Santa Barbara
48.7% UC Irvine</p>

<p>Finally, I think you overrate the degree to which privates use standardized test scores in the admit/deny decisions. For example, U Penn accepted only 22% of applicants who submitted SAT scores in the 750-800 range for Math and only 25% of those who did so for Critical Reading. If, as you claim, privates are so tricky in their admissions methods, they could easily fill their class with much higher scoring students. </p>

<p>BTW, none of this changes the fact that UC Berkeley is an excellent place and probably on everyone’s short list for colleges claiming title of Top Public in the USA.</p>

106

<p>I don’t underestimate the use of SAT scores. I listen to what the schools say and do. The schools use them differently.</p>

<p>The schools shouldn’t be judged like they are using them the same when they are not. Just common sense. Nothing complicated.</p>

<p>The more likely you are coming from privilege, the more important the SAT score…and the more your student body comes from wealth…the more likely SAT scores will be higher.</p>

<p>Common sense.</p>

<p>Just look at Wake Forest and what it is going to do.</p>

107

<p>Michigan tops web rankings.</p>

108

<p>And if there is a study about the SAT that differs from what the former presidents of Harvard, Princeton and UC Berkeley say…where is it?</p>

109

<p>dstark,
You can declaim all you like about the advisability of social engineering in the admissions practices on college campuses. Goodness knows that you’ll have almost unanimous support with that crowd for such a position.</p>

<p>But as independent observers looking at data and trying to make judgements, we are looking for data points that have the most indicative value. IMO, the standardized test scores have a lot of value in that regard and a 2nd grader could figure out that they certainly carry a lot more weight for Accept/Deny decisions than what is learned from things like class rank and GPA. </p>

<p>Look at the data above again in # 106. It is abundantly clear that the UCs, like the vast majority of colleges all across the USA, use standardized test scores as a key criterion in deciding who gets in and who does not.</p>

110

<p>The UCs use SAT scores. They don’t USE SAT scores.</p>

<p>It’s not that complicated. </p>

<p>They use them the way I said.</p>

<p>The way they said.</p>

<p>Other schools have different objectives. Different student bodies… different SAT scores.</p>

<p>The UCs educate more poor people and use SAT scores less…not zero… than other top schools. When looking at poor people, SAT scores don’t mean much. When looking at students from wealthier families…SAT scores matter more. SAT scores are important for a segment of the population. SAT scores are going to be more important if you live in Marin than if you live in Pinole.</p>

<p>If you want go to Wash U…SAT scores are going to be very important… for almost all students.</p>

<p>So it makes no sense to judge SAT scores at Berkeley and Wash U the same. At Wash U …SAT scores are important for more than 90% of the spots in the student body and at Berkeley…it is probably closer to 50%.</p>

<p>I look at the data too. And what I say is NOT contradicted by the data.</p>

111

<p>posted by dstark:

</p>

<p>hawkette, please answer this. </p>

<p>How is it going to be fair for Berkeley to be judged according to the way you do with the privates when Berkeley does have a different admission system/criteria from the privates? </p>

<p>Is it fair to penalize Berkeley for having a different admission system/criteria from, say, Washington U?</p>

112

<p>It seems to me that there would be very few ways to quantitatively rank schools if you had to base it on how each one decided their own admissions. Rather, you must decide on the best way to weight the values to determine which school is “best” by a certain measure. </p>

<p>For example, if school A based 90% of their admission decision on SAT and school B based 90% on GPA, it would be nearly impossible to rank them separately. Rather, we must say that “we believe the best undergraduate school is one where entering freshmen have high GPAs, SATs, Class Rank, etc… all weighted equally.” From there, you would crunch the numbers and get your answer. It may be that school A’s preference of SATs over GPA and other markers, along with numerous other factors (I doubt that it alone would explain all of the variance), leads it to have, using our multi-pronged metric, a higher rating than school B.</p>

113

<p>^</p>

<p>The criteria for selectivity would be:</p>

<p>40% - HS GPA
40% - SAT scores
10% - Class Rank
10% - Admission Rate</p>

114

<p>dstark,
Unless we actually work inside the admissions offices of UCB or Wash U, we’ll never really know how it works in practice at an individual college. Schools will say this and schools will say that, but at the end of the day, their actions don’t always match with their words (ie, their political propaganda). BTW, IMO the privates consider standardized test scores LESS in the overall Admit/Deny decision than do the publics, including UCB. </p>

<p>RML,
Your selectivity suggestion is interesting, but there are some significant drawbacks. First, there is HUGE variance among GPAs from high school to high school. How do you normalize for this? For example, the average GPA at U Pittsburgh is 4.0. At UCB, it’s 3.9. At Princeton, it’s 3.9. See a problem there?</p>

<p>Second, class rank is again a pathetic indicator for academic strength among high achieving students. Look at these California colleges and their Top 10% students:</p>

<p>98% UC Berkeley
97% UCLA
100% UC San Diego
98% UC Davis
96% UC Santa Barbara
96% UC Irvine</p>

<p>97% Caltech
87% USC
40% Pepperdine
95% Harvey Mudd
86% Pomona
85% Claremont McKenna
70% Scripps
65% Occidental</p>

<p>Forget UCB and UCLA for a minute and tell us if you think that the student bodies at the other UCs are equal to Caltech and Harvey Mudd and stronger to A LOT stronger than USC, Pepperdine, Pomona, Claremont McKenna, Scripps, and Occidental.</p>

<p>BTW, the average GPA at Scripps is 4.0. </p>

<p>Acceptance Rate is another measure that is generally difficult to gauge and even more so in California with their “check-the-box” approach to applying to various UC campuses. They’re probably a good reflection of the relative prestige of these colleges within the USA’s most populous state, but are they really an accurate reflection of their national selectivity? I doubt it. </p>

<p>Finally, re SATs, I know you like to portray my comments as saying this is all that I care about. Hopefully at some point, you’ll come to realize that that is NOT what I am saying here or anywhere else. What I am saying is that standardized test scores are the most reliable statistic that independent observers have for evaluating the strength of a student body and are far, far better than things like GPA and class rank. Underlying this is my belief that there is normally a large correlation between achievement on standardized tests and other parts of one’s application. I understand that there are exceptions, but IMO not enough to discredit the use of the standardized test as a very valid indicator of student achievement/preparedness.</p>

115

<p>Hawkette…Maybe UC Berkeley lies. </p>

<p>I know you would like to think Berkeley considers SAT scores more than the top privates.</p>

<p>[frontline:</a> secrets of the sat: history of admissions at uc berkeley](<a href=“http://www.pbs.org/wgbh/pages/frontline/shows/sats/etc/ucb.html]frontline:”>http://www.pbs.org/wgbh/pages/frontline/shows/sats/etc/ucb.html)</p>

<p>[frontline:</a> secrets of the sat: an update](<a href=“http://www.pbs.org/wgbh/pages/frontline/shows/sats/etc/update.html]frontline:”>http://www.pbs.org/wgbh/pages/frontline/shows/sats/etc/update.html)</p>

116

<p><a href=“http://www.equaljusticesociety.org/press_latimes_2003_10_25.html[/url]”>http://www.equaljusticesociety.org/press_latimes_2003_10_25.html&lt;/a&gt;&lt;/p&gt;

<p>I wonder what the fuss was about if UC Berkeley uses SAT scores for all students…</p>

117

<p>Maybe Berkeley lies.</p>

<p>[frontline:</a> secrets of the sat: interviews: bob laird](<a href=“http://www.pbs.org/wgbh/pages/frontline/shows/sats/interviews/laird.html]frontline:”>http://www.pbs.org/wgbh/pages/frontline/shows/sats/interviews/laird.html)</p>

<p>" There are some people who say that Berkeley relies very heavily on the numbers…</p>

<p>I really think this is widely misunderstood. what we do is look at an individual applicant against the high school from which she comes. We have gone through all of the course lists for every one of the high schools in California, to count the number of honors courses for example offered in those individual schools. So when we say that we use an uncapped GPA and give students the additional grade points, which is a UC system wide policy for honors level courses taken in high school, we don’t simply take the highest GPAs that we have in our applicant pool and admit those students. Because that would concentrate our admits in a small number of schools. We look at what a student has done within the curriculum that’s available to her. So if we have a student from a high school that offers only two semesters of honors courses for example, and that applicant has taken both of those courses and done well in them, that student can make an academic score of one in our comprehensive review. Just as the student from a high powered suburban high school, where there are 44 semesters of honors courses, and who has taken 24 semesters of those courses and done well–can make a one as well. </p>

<p>The same thing is true with test scores. We’re interested in test scores, but not in any mechanical way. We don’t add up the scores and make some judgement or some cut, in our applicant pool. We look at the SAT verbal score and the SAT math score. We look at the three required SAT two examination scores, as individual pieces of information. And we look at those numbers in relation to all the other academic information that we have on a student. And we look at those numbers in relation to the student’s own history. So that an SAT one verbal score of 600 doesn’t have a single meaning across our applicant pool. A score of 600 may mean one thing for a student who’s first language isn’t English, or who’s parents didn’t complete high school, or a student who comes from a low income family, where that same 600 may mean something entirely different for a student from an affluent background with educated parents. </p>

<p>So we’re looking at context as well as numeric achievement. And we’re matching those test score numbers against other things in the academic record. We might have a student with a verbal score of 500, let’s say. But when we look at the academic curricul we see that the student has done well in advanced level English classes. Maybe has taken AP English already. Maybe has a score of 4 on the AP English exam. And so that verbal score then is going to get weighed against another indicator of strong achievement in, English language and literature skills.</p>

<p>What about the notion of grade lock–such similar grades-- don’t you have to go the SAT to break the tie? </p>

<p>No. We don’t have to go to the SAT to break this grid lock among all of these applicants with very high GPAs. For one thing, it is true that among our 31,000 applicants this year, we have 14,800 students who report a grade point average of 4.0 or higher. Keep in mind that you can have above a 4.0 because of the university wide honors grade point policy. Nevertheless, all of those 4.0 GPAs, first of all, are not straight A’s. There’s sometimes a confusion in an observers mind–well, gee, 4.0 means straight A. But with the honors grade point policy, that’s not true. So, all of those GPAs are very definitely not equal. And second, the course work that students have taken in earning those GPAs will vary tremendously within our applicant pool. So it isn’t just the GPA that we care about. We want to see what the actual grades are that the student has made. We want to see what kinds of courses the student has taken. We want to see what kinds of courses the student has taken compared to what was available in that school. So we may have a student with a 4.0 GPA in a school with 24 honors courses. But that student has taken none of those honors courses. That 4.0 GPA is going to be different than a 4.0 GPA from, for another applicant from that school, where that student has taken 12 of those honors courses. </p>

<p>So we’re going to look at a whole range of academic measures as we assess what the level of academic achievement is. Not just that GPA. Same thing is true for the test scores. </p>

<p>Do you think using the SAT as the major basis for admissions could drive out college diversity?</p>

<p>My personal view is that that’s a very definite risk.</p>

<p>I think that’s a really critical public policy question for the entire state. I think we have to ask ourselves–what does it mean for the state if more than half of the school children in California look at the premier public university in the state and say I don’t have a chance to go there? I think we’re looking at extending the already growing inequality between the really affluent, folks in California and the increasing number of increasingly poor folks in California–and the correlation between race and ethnicity. And those income distribution patterns are clear and serious."</p>

118

<p>

</p>

<p>It’s easy, hawkette. You can normalize any imbalance by factoring in more criteria. It’s in the allocation of percentages that matters then. Let’s take UCLA vs USC as an example.</p>

<p>UCLA has 97% students from the top 10% but has a SAT score average of 1980. USC has only 87% students in the top 10% but has a SAT score average of 2080. By assigning values on both given factors, you’ll be able to normalize whatever you think there is a need to be normalized. Then you have % of students admitted and class rank, so you’ll be able to get a better result of selectivity. Of course, that isn’t a perfect formula. But I think it’s safe to assume that it is an acceptable one.</p>

<p>

That’s why I have been telling you - since the very beginning - that you don’t just rely on 1 single criterion in assessing school selectivity or intelligence. You were the one who kept on telling me that Emory students are smarter than those Berkeley L&S students because of one thing - SAT scores. I objected to that claim because you have no other sources/data and you have repeatedly failed to present to me other important measures of intelligence aside from the SATs. Like I said, you should look at these students’ other achievements too before you even make a conclusion. The SATs won’t singlehandedly measure intelligence and the Berkeley study has proven that to be true. Things like HS achievement ranks, HS GPA, HS academic involvements, competition per place, etc, are just as important measures of intelligence/school selectivity. The Berkeley study has shown that the SATs favor a certain race and/or socio-economic level of students. For example, if you would let all Cambridge and Oxford students take the said test, the average result would definitely be lower than many privates’ average scores. It would most likely be lower than the average score of WashingtonUSL’s. But are Oxbridge students inferior to WUSL’s? I very much doubt it. But not according to your logic. Now, notice how absurd that is? </p>

<p>

</p>

<p>That’s just simple math or statistics, hawkette. </p>

<p>The more applicants you compete with, the lesser your chances of getting admitted. But then again, I didn’t say the level of these applicants shouldn’t to be looked into. That is why this would only constitute a small portion of the criteria. (Noticed that I only assigned 10% value in my suggested method?) But do not tell me student-per-place is not a valid criterion for selectivity level. You are looking at your odds here. In statistics, odds matter. </p>

<p>

</p>

<p>That was actually what you have always implied since the very beginning of this conversation. Did you forget what you told me about Emory students being smarter than Berkeley L&S students? You told me that Emory students are smarter because they have higher SAT scores. That has what pushed me to react to your post. If you have made calculations before you made such a claim, I would have silenced and just analyzed your data. But you have no other data other than SATs, and again, SATs aren’t everything. </p>

<p>

</p>

<p>If you think SATs are really that huge issue for one’s intelligence, then tell me why there is still a huge number of students at Caltech and MIT with near perfect to perfect SAT scores doing considerably poor performance in the classrooms to the point that many of them are expelled from the school or don’t graduate from college anymore? </p>

<p>Again, SAT scores aren’t the sole measure of intelligence and if you rely on this alone because it is – for you – the most reliable statistic that independent observers have for evaluating the strength of a student body, then you’re just lazy, or you’re the type of person that only looks at the outer beauty of a person. You didn’t even check if it was just a thick make up and lipstick… In this type of conversation, you don’t make conclusions with incomplete, insufficient, inadequate data. That’s not a scholarly approach at all.</p>

<p>And lastly, before you make a comment about Cal, please read the articles which dstark has provided. It would help you understand more what we are saying.</p>

119

<p>RML,
I appreciate the effort you are making, but something is getting lost in the translation of the two dialects that we are writing. </p>

<p>There’s lots of tit-for-tat that I could pursue (eg, factors like GPA and class rank differ dramatically from school to school while standardized test scores are “standardized!”), but I doubt if we’d get anywhere. But one theme does strike me in your writings about UCB-somehow you seem to think that the UCB students are the only ones who have strong records away from standardized testing. Either that’s arrogance or that’s misunderstanding the applicant pool at dozens of top colleges around the country. I think it’s the latter, but it can come off as the former. The reality is that there are a lot of excellent students and colleges around the country, many with overall student bodies markedly stronger than what one would find at UCB. </p>

<p>Finally, I will respond to one statement you made about Caltech and MIT:</p>

<p>

</p>

<p>Do you realize what the respective graduation rates are for these colleges?</p>

<p>4-year graduation rate</p>

<p>81% Caltech
82% MIT
64% UC Berkeley (all students)</p>

<p>6-year graduation rate</p>

<p>88% Caltech
94% MIT
90% UC Berkeley (all students)</p>

<p>I note all students for UC Berkeley because engineers make up about 16-17% of the class. </p>

<p>Not sure what you meant in your comments about “huge numbers,” but the numbers would not back you up. </p>

<p>Oh…and I read the articles. Sorry, but it doesn’t change a thing in how I see the veracity of claiming standardized test data as the best single data point (by far) for judging a student body. I think you aren’t grasping the correlation effect.</p>

120

<p>“Sorry, but it doesn’t change a thing in how I see the veracity of claiming standardized test data as the best single data point (by far) for judging a student body”</p>

<p>That doesn’t mean schools agree with this… or use this metric in the same way.</p>

<p>“I see the veracity”</p>

<p>Well RML… some people see better than others. ;)</p>

<p>Hawkette, I am glad you read the articles.</p>