<p>PolarBear and Consolation, perhaps you misunderstand my argument. The SAT was once advertised as an aptitude test and as an indicator of freshman performance. It was also drummed into countless young people that “you can’t successfully study for it.”</p>
<p>Sure, reward the 16-year old who for whatever reason, has read the ‘Song of Roland’ or ‘Ivanhoe,’ but all other things being equal, should the kid sitting next to you at the exam lose points for never having read literature of that sort in the classroom?</p>
<p>Actingmt, care to clarify? I have no idea of what point you tried to make. Obviously kids whom are reared in less than luxurious circumstances read. Some, like myself 40 years ago, read a lot. But a kid who read everything assigned in school and did well shouldn’t be penalized in a standardized exam for being unable to define a word that seldom appears in any literary, scientific or historical context; the very disciplines which students will embark upon in college.</p>
<p>Okay. The point I was making was that recreational reading is available to everyone.</p>
<p>I think these tests are trying to measure much more than what’s assigned in school and that’s fine by me. BTW , I have a kid who was a voracious reader and scored very well, along with another who did what was assigned and did not do so great. Both are very smart. In fact, the poor scoring kid may be smarter, actually. But, I don’t mind a bit that the reader gets a higher score on the reading portion of the test because he reads.</p>
<p>Alright. Then is the exam truly an aptitude test or even an reliable for a smart but “poor scoring kid?” predictor of Freshman year performance. Often the answer is NOPE. So maybe then admissions should be based upon various measurements of achievement and ability. Just an idea.</p>
<p>Why don’t you look at it that the kid who has read widely is gaining points over some average score, and the kid who doesn’t read is simply not gaining any?</p>
<p>FWIW, I don’t think that the CR SAT can be studied for particularly well, precisely because it depends so much on what the kid was doing during the 12 preceding years.</p>
<p>Also, as it happens, I read both The Song of Roland and Ivanhoe while I was in HS/Jr High…on my <em>own</em> time using my library card. No one ever read anything like that in the classroom. I read all kids of things, because I was a voracious reader, and my V score reflected that. If I had been watching 5 hours of TV a day instead of reading, it would have reflected THAT instead. I see nothing unfair about it.</p>
I assume you are referring to the Geiser UC study. The study found that test scores only explained ~6 percentage points more of the variance in college GPA + income + API than without scores. The most influential scores for predicting college GPA were the writing score and the 3rd SAT II subject test. When adding controls for all variables, the correlation coefficients for SAT I V+M were only 0.08 and -0.02. The author of the study had a similar conclusion to the author of the article, as quoted below:</p>
<p>"The paper concludes with a discussion of the implications of these findings for admissions policy and argues for greater emphasis on the high-school record, and a corresponding de-emphasis on standardized tests, in college admissions. "</p>
<p>
Depending on which controls you use, it can be negligible and is much weaker than the correlation between SAT score and parents’ income.</p>
<p>So yes, GPA is a better predictor, but not that much better. Importantly, it clearly shows that the Bard Prez is statistically-challenged and/or just throwing out spin.</p>
<p>btw: for those in other states, UC admissions requires completion of a core set of courses, including math through Alg II and a course in visual and performing arts.</p>
<p>In your earlier post you wrote, “combined with GPA” , and the colleges we have been discussing do not look at SAT alone, so listing variance for SAT alone does not make sense in this context. SAT is correlated with HS GPA, so if you look at SAT alone you are looking at groups with notably different HS GPAs. Furthermore SAT has a negative correlation with improvement in grades from freshman to senior year, so looking at only first year GPA is misleading. SAT is also correlated with income, so you are also looking at groups with notably different average HS qualities and college preparedness via quality of college prep courses, which likely relates to the larger effect on 1st year GPA than overall college GPA. One study found that when adding controls for HS Quality and GPA, SAT score often had a negative correlation with college GPA, as quoted below (<a href=“http://www.nacacnet.org/research/research-data/nacac-research/Documents/DefiningPromise.pdf”>http://www.nacacnet.org/research/research-data/nacac-research/Documents/DefiningPromise.pdf</a> ):</p>
<p>“Crossing the Finish Line (p. 114) found that when controls were added for the quality of high schools, the predictive power of SAT/ACT testing disappeared, and often had a negative correlation with college performance.”</p>
<p>The Geiser study didn’t find a negative correlation, but it did find that the scores added little benefit in prediction beyond HS GPA + income + HS API.</p>
<p>The problem with the SAT is a better predictor of income or is better correlated with income ignores other factors that may be the real cause(s). It is not just one thing. Anyway, correlation is not causation. </p>
<p>How about these for thought: SAT is better correlated with vocabulary words used at the dinner table; SAT is better correlated with parents who institute serious study habits and rules; SAT is better correlated with parents who are involved in making sure homework is done; SAT is better correlated with kids who go to the library at least three times a week; SAT is better correlated with parents who institute school night curfews. </p>
<p>I do not ascribe that any of the above are accurate, but I bet that regardless of income, they do correlate well with the SAT. And, I bet, in many cases, they correspond to income as well. I just get this feeling that income is the catch-all phrase that people use when it really comes down to basic family values and behavior toward school and studying and how hard those values are enforced. It is one thing to tell your kid to study and tell everyone else you believe education is important, but without a school night curfew and checking of homework, it does not matter. </p>
<p>A more rigorous (and peer-reviewed) analysis can be found here: <a href=“http://gsppi.berkeley.edu/faculty/jrothstein/published/sat_may03_updated.pdf”>http://gsppi.berkeley.edu/faculty/jrothstein/published/sat_may03_updated.pdf</a> Rothstein argues that the SAT is not as good a predictor as previous researchers found, but he still finds that “an incremental validity of 0.044 may well be enough to justify use of the SAT” (p. 19). He also finds that “that admissions offices could admit better-prepared entering classes by giving explicit admissions preferences to high-SES students and to students from high-SES high schools. SAT scores would receive some weight in best predictor admissions rules, but considerably less than is indicated by sparse models” (p. 20). So, even with these SES variables, the SAT retains some predictive power. </p>
<p>I disagree. The Bard Prez made a blanket statement that the SAT does not… The Prez did not qualify his statement that is was based only on/for Test Optional schools.</p>
<p>But don’t forget, the TO schools are self-selecting. Students with extremely high GPA’s don’t need to be concerned about test prep or their score if they target a TO school. But students with low GPA’s need high test scores to get into the same TO school. Such students may have prepped hard, but their HS transcript is lacking. And given their work ethic in HS, the GPA in college may be a question. Of course, a study of TO schools will show this.</p>
<p>
</p>
<p>So you are complaining about the SAT about what it does not do and what it never claims to do? Huh?</p>
<p>
</p>
<p>But yet UC kept the test scores as an admissions requirement.</p>
<p>
</p>
<p>Well sure, I have no doubt that the bottom student at Andover or Exeter or TJ or Stuy would do well at any college. </p>
The students who improve the most tend to be the students who have the most room to improve. If you have a 4.0 in your freshman year, you have nowhere to go but down. </p>
<p>You can look up the Bates study on 20 years of test optional. Post acceptance, they collect scores from those who didn’t submit them for admissions.</p>
<p>All this talk about college grade performance skips that, with rare exceptions, college gpa isn’t what the colleges are most concerned with, other than it’s relation to graduating. </p>
<p>My kids both chose a TO, based on the school. And were challenged by grade-deflation in both stem and humanities classes. </p>
<p>I have always, from my own hs days to today, thought of the SAT as nothing more than a hoop to jump through. If anything, it’s that need to jump through either the SAT or ACT that is “standardized.” Nice to see some kids take it very seriously. But when you consider what’s included in a holistic application, as bcc noted, it is just one piece. </p>
If we are talking about predictive power of SAT alone without GPA, the relevant colleges would be the ones that admit based on SAT alone without considering HS grades/transcript (are there any?), not test optional schools. In any case, in the original post I quoted you said “combined with GPA,” which is the relevant case for college admissions and I expect what the Bard Prez meant.</p>
<p>
My point was you’ll often find SAT has a different degree of correlation between 1st year grades and overall college grades. Which measure do you think should be the more relevant metric for college admission… How the student does in their first year? Or how the student does overall across all years? In retrospect, this effect is quite small, perhaps too small to be worth the apparent confusion it is adding this discussion.</p>
<p>
The UC system added a top 4% admission rule in response to Geiser’s research (changed into the top 9% ELC since then), among other changes. His research, among others, has contributed to a growing number of colleges becoming test optional in recent years.</p>
<p>
Students who attended highly selective HSs only represent a small portion of the overall student body and likely a negligible portion of the groups in any of the discussed studies . As a more common example, many HSs in the United States offer very few AP classes. Some do not offer any AP/IB-level classes at all. Students of families of lower incomes are more likely to attend such schools than students of families with higher incomes. Students at the two schools may have the same GPA, but student who received the GPA while taking many AP/IB courses is more likely to have a higher GPA in college than the student who took few. The same effect also occurs for general teacher quality, which tends to be decrease in lower income public schools (I realize highly selective magnets in lower income areas are an exception). If college admissions were to look at HS GPA without considering course rigor or which HS the student attended, then SAT score would have more of an impact. However, colleges with holistic admissions generally do consider these factors. </p>
<p>Botstein is a leading musicologist. He discusses colleges through his Franconia and Bard highly clouded lens. His views about what colleges should do or be is pleasing to an extremely small portion of students and their family. He is entitled to his opinion, and the rest of us entitled to ignore the sounds of his keyboard. </p>
<p>That’s incorrect. If you go back and read the minutes [and politics] of the UC meetings when this proposal was adopted, you’ll quickly find that the ELC 4% rule was similar to UT’s Top 10% admissions rule. </p>
<p>ELC has never had anything to do with test scores. It had/has one objective only.</p>
<p>
</p>
<p>Regardless, the study you posted clearly shows submitters with a C average in HS. (not college ready, IMO.) Their high test scores got them into the colleges, but their lack of work ethic keeps their college grads down. Doh!</p>
<p>He also has been a college president virtually his entire adult life. That does gives him some standing to publicly express a viewpoint on entrance exams. His arguments should be addressed on their own merits.</p>
<p>What I would like to see is a study looking at family behavior and education values with respect to the SAT that includes low, middle and high income earners. That would shed much more focused light on the issue, as to what affects scores the most. Maybe that study exists, but I am not aware it.</p>
<p>Excellent point, Awcntdb! Let’s hear more about working class parents who sacrificed for their children, pushing them and instilling within them the motivation and for academic achievement. That fellow with the four daughters that became doctors and health professionals some years ago come to mind, as does heart surgeon Ben Carson’s mother.</p>
<p>"His work has contributed to the development of a number of new admissions policies, including UC’s policy on Eligibility in the Local Context, which guarantees admission to the top four percent of students from each California high school. "</p>
<p>
The study I linked to earlier included both less selective and more selective colleges. Less selective colleges a had lower average GPA, as one would expect. The C students who you claim were admitted to less selective colleges based on test scores had test scores ranging from 868 to 1113, depending on school. I wouldn’t exactly call these “high test scores,” nor would I assume that these scores are what got them admitted. If you ignore these less selective schools, and focus on scattergram’s for the UC colleges discussed in this thread, I doubt you’ll find any admits who were C students with high test scores (both due to admissions policies and the discussed correlation between HS GPA and test scores).</p>