<p>"Predicting College Grades. The basic purpose of the SAT is to provide information to college admissions staff that will help them to evaluate a student's application. Results of research conducted using data from 26 colleges show that SAT test scores, in combination with a student's high school grade point average, predict college freshman GPA more accurately than either SAT scores or high school GPA alone.</p>
<p>For all college freshmen in the study (N=110,468), the predictive validity of an optimally weighted combination of SAT scores and high school record is .65. The correlation between the critical reading score and freshman GPA is .50; between the math score and freshman GPA is .52; between combined critical reading and math scores and freshman GPA is .55; and between high school GPA and freshman GPA is .58. Typically, the correlation for high school GPA is slightly higher than for combined SAT scores."</p>
<p>The nearly identical correlation of both parts of the test to freshmen gpa seem to indicate that the difference in scores of the average student can not be large. Even though the parts have little to so with each other. And since one is allowed to retake the test and count the best combined it would be unlikely that a student with a 770 math and a 690 reading would not retake. Economics come into play, one of the knocks on the test, with tutors or study courses to bone up on any weakness.</p>
<p>
[quote]
That does not mean that a child with a 1200 SAT (75th percentile) and few ap's could not get into a top school (the odds are small sadly) and maybe do ok with a less strenuous course load, but they would argue that the likelihood of that child doing well relative to children with higher credentials is quite low, lacking other significant factors that influenced those results.
[/quote]
My daughter had a 1200 SAT and at the end of her first year I decided to calculate her Columbia GPA separately from her Barnard GPA. I was prompted to do so by maternal curiosity and pride when she earned an A+ in a 5-point course at Columbia. The numbers: over 2 semesters, she earned 16 points at Barnard and 13 points at Columbia - the Columbia GPA was 4.04. [Barnard GPA somewhat lower (partly do to a Barnard prof who announced at the outset that "nobody" in the class would get an A) - but the net GPA still is higher than what I graduated with from a public U. -- and my SATs were a lot better than my daughter's.</p>
<p>You are absolutely right that a 1200 falls in the 75th percentile for SAT scores. You are sadly mistaken to think that any 75th-percentiler isn't up to the task for any elite or Ivy League school, with the likely exception of those pursuing math-heavy majors at colleges like MIT or Cal Tech. If anything, most of the Ivy League schools are known for grade inflation and my daughter has actually been surprised by the ease at which she has earned some of her grades. She did not start out with an easy courseload -- (she was quite surprised to see that A+) -- quite the contrary she signed up for a 3000-level course at Columbia and used AP credit to take an advanced level course at Barnard her first semester. </p>
<p>My guess it that you are the parent of a kid with really high SAT scores and very proud of it. I'm the parent of 2 kids, one with really high scores and one whose scores are merely better than 3/4ths to 4/5ths of the kids who test, depending on which tests and measurement scales you want to look at -- and the lower scoring kid is the better student and always has been. She's no drudge either -- she is an active, extroverted kid who is heavily involved in EC's, works 2 jobs to earn her own spending and food money, and maintains an active social life. If anything, I am sometimes frustrated by what I see as her tendency to give short shrift to her studies. </p>
<p>The elite colleges look for high scores because it keeps their rankings up and they need a quick and dirty way to weed through thousands of applications (and, as noted, are a good proxy for income level and protect their "need-blind" financial aid budgets) -- but I doubt seriously that there is any ad com anywhere who seriously believes that a kid with a 1200 can't do the work at their school, if that kid comes with a demonstrated track record of taking tough courses at their high school and doing well in them.</p>
<p>Windy, the problem with your post #101 is that you are extrapolating a predictive value when comparing high-end scores from a study that probably looked at full range scores. </p>
<p>If the average scores are around 500, then it is likely that studies of predictive validity are looking at below-average vs. above-average scores -- or wider score differentials. I certainly would agree that a student with a 700 CR score is likely to do a lot better in college than one with a 450 -- and the same observation would hold for math. </p>
<p>But that doesn't mean that you can assume the same predictive validity when you are trying to sort out the upper or lower end of the ranges. It's one thing to compare a 500 with a 750, quite another to try to draw the same distinction between a 600 and a 700 -- just as there probably isn't a dime's worth of difference between the kid with a 3.85 GPA who graduates 10th in his high school class and the kid with the 3.98 GPA who graduates 2nd -- but both are likely to do far better in college than a kid with a 3.2 high school GPA.</p>
<p>The one advantage of looking at GPA's over SATs is that they provide insight into a sustained level and pattern of performance -- that is insight that the ad coms get when they look at a transcript, but something that also is not going to be accounted for in a study looking at the raw GPA.</p>
<p>The elite colleges look for high scores because it keeps their rankings up and they need a quick and dirty way to weed through thousands of applications (and, as noted, are a good proxy for income level and protect their "need-blind" financial aid budgets) -- but I doubt seriously that there is any ad com anywhere who seriously believes that a kid with a 1200 can't do the work at their school, if that kid comes with a demonstrated track record of taking tough courses at their high school and doing well in them.</p>
<p>All students should be given the opportunity to be than they can be. In the real world lacking some kind of significant hook the student with the 1200 SAT will almost never get the chance to attend an elite school. At Amherst and Williams according to articles regarding athletic tips 1300 is a pretty hard floor, same with legacy. Rightly or wrongly the barriers are in place and that application doesn't get by. If the schools were as concerned as you think with nothing but ratings and budgets they can fill their incoming classes many times over with 1400+ sat's and full paying students. That is not what they choose to do. They system they use may be imperfect but is many times better than it used to be. Their hearts and there wallets are in the right place.</p>
<p>That's what they do, but it has nothing to do with academic ability. The point is that my 1200 scorer has no problem at all pulling A's in courses taken at an Ivy and I am sure that there are many, many more students who would be equally adept if given the chance. I'm also sure there are plenty of 1500+ scorers with weaker GPA's who would be likely to flunk out. One problem with SATs is that they are absolutely nothing like a college-level exam -- my d's AP exam scores (where she did uniformly well) are probably far more predictive, simply because the AP format is more like a college exam format. (Content based + requiring a written essay demonstrating some depth of thought and analysis). </p>
<p>The Ivies would increase minority enrollment as well as low-income enrollment if they would take a more holistic view. That is what the Bates data shows and there is other data as well to support that:</p>
<p>You have every right to be proud of your D...but she is the exception not the rule. The fact that Barnard students get to take Columbia courses is a wonderful but highly uncommon situation. Most, bordering on 99% of unhooked students with 1200 SAT's would never get the chance to see how they would do in that rarified air. Amherst and Williams have over 50% of the class of 2011 receiving financial aid (something like $25,000 average package)so they do not fill the school with full pays by any stretch of the imagination.</p>
<p>A $25000 package is what a middle class family gets, not what a low income student gets. It leaves an EFC of about $20K. If you look at the percentage of Pell grant recipients, the Ivies and top LAC's don't fare too well -- Amherst has less than 14% and Williams has only 11% of its students in that category. This means that a far greater percentage of low-income students are attending NYU (notorious for its stingy aid policies, but with almost 18% of its 20,000 students qualifying for Pell Grants). The LAC which serves the highest number of low income students is Smith (almost 26%). See: The</a> Chronicle: 5/12/2006: Elite Colleges Lag in Serving the Needy</p>
<p>And, assuming that the student is coming in with a strong high school GPA with a challenging course load, I stand on my assertion: any student with a 1200 SAT can do Ivy-level work. It's actually a lot harder for kids to make the grade at many of the larger publics which do not have the support systems in place for students. And I cited the stats that showed that. </p>
<p>My daughter is bright and capable and doesn't test well. Any other kid who has a high end high school GPA and 1200 SATs probably also is bright and capable and doesn't test well. </p>
<p>In that regard it is interesting to note that the discrepancies in SAT scores between white and African-Americans are reduced considerably if the tests are rescored using on the the hardest questions -- The</a> Bias Question</p>
<p>*If you look at the percentage of Pell grant recipients, the Ivies and top LAC's don't fare too well -- *</p>
<p>What makes you think it is the obligation of private colleges to offer free educations to low achieving socio-economically disadvantaged students to the exclusion of their traditional role? I choose to congratulate their efforts instead of criticizing. All schools operate under some kind of predetermined structural bias. That they have moved to welcome students who do not fit the traditional rich white kid stereotype is a good thing. Look if you will at Barnard. As an all women school it has chosen to exclude half the population of the planet. This in a time when similarly situated schools like Radcliffe have chosen a more mainstream posture. I am sure that the arguments that this form of bias is a redress for centuries of exclusion at all male schools, make some degree of sense, and I am sure there are other valid reasons. Either way it is a private school and it can do what it wants. As I have questioned before a 5% African American student population is clearly out of proportion with NYC (26%) and the nation at large. A form of bias, I doubt it. It is a private school it can do whatever it wants. Figures I have read in trying to learn more about the school also indicate a Jewish student population (ethnic designation not necessarily a religious one) of 45%. Again clearly disproportionate with national percentages. A form of bias, I doubt it. It is a private school and it can do what it wants. It is not necessary or even desirable that all schools have the same mission. There are many private schools that fulfill a myriad of roles, and of course a well established (and government regulated) public university system that does an outstanding job.</p>
<p>It isn't a matter of comparing one to the other. I stated a FACT: so called "need blind / 100% need colleges" use SATs as a rough proxy for income, though they know that SATs have little predictive value beyond high school GPA. Thus far only one example of a so-called need-blind, full need college which is also SAT optional has been presented -- Bowdoin -- which due to location is probably not deluged with low-income applicants. </p>
<p>I also pointed out the racially discriminatory aspect of SAT scores, particularly against African American students who score significantly lower even when the analysis controls for income disparities; Freedle's analysis (the one showing that the black students do better on the harder questions) would tend to indicate that this is not a function of innate ability but something related to test design. I raised that issue in part because you had raised the issue of racial disparity in another thread, and I was surprised that you would argue for preservation of a college admissions tool known to have a biased impact. Apparently you are the one who thinks that it is ok for some private schools, but not others, to maintain admissions standards that create additional barriers for low income and minorities: What makes you think it is the obligation of private colleges to offer free educations to low achieving socio-economically disadvantaged students to the exclusion of their traditional role? So you "congratulate" the schools you like on their efforts, while condemning the ones you don't like.</p>
<ol>
<li><p>Are you suggesting that highly selective schools intentionally use the SAT as a selection tool in order to maintain a wealthy student body (rather than continuing to use standardized test because of inertia or some other perceived benefit). If so, I would like to see the evidence.</p></li>
<li><p>The racial disparity in the tests is large measure counterbalanced by affirmative action policies that I believe are in place at all highly selective private schools.</p></li>
</ol>
<p>I am saying that they are not truly "need blind" and that they know that their admission standards will always result in a class skewed toward the upper middle class and wealthy. If that were not the case, how would it happen that year after year these "need-blind" schools end up with the same percentage of kids on financial aid?</p>
<p>I'd also point out that tuitions have been rising astronomically, whereas middle and lower class incomes have remained relatively stagnant over the last decade or so. That means more and more financial aid dollars are going to families who would not need aid but for the high cost of tuition.</p>