<p>
</p>
<p>Where is there published research on this point?</p>
<p>The most extensive data that I know of comes from Bates’s 20-year longitudinal study. The link provides a reasonably complete summary; I recommend dowloading the Powerpoint, which is loaded with data.</p>
<p><a href=“http://www.bates.edu/x58835.xml[/url]”>http://www.bates.edu/x58835.xml</a></p>
<p>Here’s the payoff line from the Dean of Admissions at Bates on their 20-yr. experience. He’s describing the difference in performance of non-submitters vs. submitters:</p>
<p>
</p>
<p>Nearly any study that measures the correlation between college achievement and SAT scores will be based on first-year GPA once you start to read the fine print. I hate to cite Fairtest.org as a source, because they’re hardly an uninterested party, but this is one of many links that will confirm that SATs were designed to predict 1st-year GPA.</p>
<p><a href=“http://www.fairtest.org/facts/satvalidity.html[/url]”>http://www.fairtest.org/facts/satvalidity.html</a></p>
<p>Here’s a better source for the reasons that researchers usually study the correlation between SAT and 1st-year gpa, from the other end of the spectrum, College Board itself. </p>
<p><a href=“College Board - SAT, AP, College Search and Admission Tools”>College Board - SAT, AP, College Search and Admission Tools;
<p>(third report from the bottom)</p>
<p>MarathonMan, do you trust the studies YOU are quoting and linking to?</p>
<p>If you do, I hope you’ll evaluate the validity of the studies of Wayne J. Camara and Gary Echternacht</p>
<p>
</p>
<p>Aren’t we back at square one, where people believe that the combination of the SAT and the GPA provide a better yarsdtick than one of them in isolation?</p>
<p>xiggi, would you please provide a citation for the Camara & Echternacht quote? Thanks!</p>
<p>xiggi,</p>
<p>I know a good bit of Camera’s work, but admit that I do not know this article; I will try to get hold of it. (It is worth noting that he works for the College Board or is routinely commissioned by them). </p>
<p>The part that of the statement that I think is misleading is this: “although high school grades typically are slightly better predictors of achievement, SAT scores add significantly to the prediction.” A number of studies, including the large UC study, suggest that the incremental predictive validity of the SAT is not particularly great. Fairtest cites as many studies showing low incremental predictive validity as College Board does showing high incremental validity.</p>
<p>I’m not sure that it’s all that crucial, really, because in practice, holistic admissions models code, either formally or informally, for a number of other variables besides just HS gpa: rigor of HS classes, number of AP or college classes, quality of the school itself, teacher recs., essays, interviews, etc. A lot of this is captured by the number that schools assign–whether it’s called readers rating, dean’s academic rating, or whatever. What’s always impressive is how good a predictor of college achievement the readers’ ratings–which represent something like the full admissions model-- are, regardless of whether the readers saw the SAT or not. Here are the results from the recent full Mt. Holyoke study cited by Asteriskea, above:</p>
<p>
</p>
<p>One more thing that I find noteworthy. In the studies at Mt. Holyoke and Bates, the only two somewhat selective schools that have made extensive data public, non-submitters and submitters had cumulative gpas that did not differ significantly, though submitters had slightly higher grades. But that’s not really a fair comparison, because the non-submitting group had SAT scores more than a standard deviation below those of submitters. That really calls into question the predictive power of the SAT for a very special group, the one that we should be interested in: those non-SAT-submitting students who were admitted.</p>
<p>I sound like a die-hard SAT-optional supporter, and I’m actually not. I think if you go into the process caring only about maximizing the college gpa of your admitted class, you use the SAT in a heartbeat. I also think the SAT is useful when you have to draw fine distinctions between many superbly qualified candidates. But I also think there are thoughtful people in the admissions business who truly believe that the emphasis on tests is out of balance and who think that they may draw a more diverse and interesting admissions pool by being SAT optional (and that the cost in the quality of their admitted class will be negligible). There really are some interesting and thoughtful questions about SAT-optional, it seems to me.</p>
<p>Love that last paragraph…Well said, MM88!</p>
<p>Am I on Candid Camera? </p>
<p>
</p>
<p>Seriously, here’s a link to the study: </p>
<p><a href=“College Board - SAT, AP, College Search and Admission Tools”>College Board - SAT, AP, College Search and Admission Tools;
<p>The title is “The SAT® I and High School Grades: Utility in Predicting Success in College” </p>
<p>Here’s the title of the “third report from the bottom:”
The SAT I and High School Grades: Utility in Predicting Success in College<br>
Author: Wayne J. Camara, Gary Echternacht - Published: 01/2000</p>
<p>Deja vu?</p>
<p>For some reason, I assumed one might have clicked to the QUOTED links. :)</p>
<p><a href=“cadbury:”>quote</a></p>
<p>The research on SATs is quite consistent. Studies regularly show that there IS a significant correlation between SAT scores and performance in college as measured by GPA. HOWEVER, results of multiple regression analyses that look at the contribution of more than one factor in predictivity of academic performance show that SAT scores contribute negligibly beyond the use of High School GPA alone.
[/quote]
</p>
<p>That’s incorrect. The recent UC study, which was anti-SAT, found the following (assuming you believe in regression models, which are the basis of the anti-SAT statistical argument): </p>
<ol>
<li><p>SAT-II scores alone are as good or better than GPA as predictors of everything (first-year GPA, graduation rates, etc), and are independent enough of SAT-I that using the whole SAT report (I + II) would certainly surpass GPA in predictive power. The authors calculated but did not publish the comparison of SAT I+II against GPA as predictors, since revealing the extent of test score superiority over GPA would be… inexpedient.</p></li>
<li><p>Regression results (multiple linear) showed a strong effect of course selection on grades. Math SAT subject test had a negative correlation with grades, due to the harsher grading in engineering and science. When separated out by major (but not individual courses) the predictivity of SAT increased. When accounting for course selection and performance within individual courses, SAT would look even better relative to GPA.</p></li>
</ol>
<p>
</p>
<p>They say “why bother with GPA” for the same reasons, but as one might expect, that point is never emphasized.</p>
<p><a href=“quoted%20from%20Mt%20Holyoke%20propaganda:”>quote</a></p>
<p>Interim results from that study show no meaningful difference in academic performance between students who did not submit scores and those who did. The study shows that there is a .1 difference between the aggregate grade point averages of submitters and non-submitters. The difference is equivalent to approximately one letter grade in one course over a year of study.
[/quote]
</p>
<p>In other words, a difference that is quite significant given the grade compression at schools such as Mt Holyoke, and big enough to show up even when measured using a yardstick tailor-made to suppress the effects of SAT.</p>
<p>Xiggi nailed it in post #57. Brilliant!</p>
<p>Okay, really now - a study put out by the CB is not “propaganda” while a sponsored study by the Andrew W. Mellon Foundation is - time for a reality check, at least for me. First off, I do want to make it clear that I am not a die-hard supporter of either ditching the SAT or proponent of a sea-change toward SAT optional admissions policy. In fact, I agree with Xiggi and Siserune on many fundamental points. As I have posted before, and even started that long thread way back when “Why not chuck the SAT”?", I fall squarely into the camp of SAT supporters, despite all of its imperfections. Yet, I do see the value and point of SAT optional policies - and I think MM88 nails it:</p>
<p>
</p>
<p>As Cadbury posted earlier on this thread, Weisbuch brings a varied fare to the table and much food for thought. Somehow, somewhere, product of an over-fertile imagination perhaps, I can see him and profs. like Barry Schwartz at Swarthmore, and Sternberg of Tufts and head of PACE, creator of the Rainbow project for admissions, cringe at much of this discussion. Instead of delving into all those thoughtful and interesting questions - both related and unrelated to the particulars of Dr. Weisbuch and Drew- it seems it is more interesting to parse the statistical validity of GPA vs. standardized testing or the happy combination of both used for admissions purposes. That said, I think it is stating the necessary obvious that SAT-optional policy is a strategy - both for student and colleges. Colleges want to both increase enrollment and, if they are lucky, attract the type of student they want to fill their seats. Students want to be seen in “their best light” and want to attend a college where they “fit” both academically and socially. Hence the multitude of SAT optional policies - some require graded papers, some interviews, some a mix or match of tests, and others nothing extra at all. For now, it does indeed seem that the SAT policy Weisbuch instituted at Drew is working well for Drew and I thank Xiggi for the link to the Drew publicity page which says just that. On this score, Weisbuch is, after all, a scholar and teacher at heart, as well as pragmatic-flamboyant college president. His years working for the WWFF, as regional director of the Andrew W. Mellon Fellowships in Humanistic Studies, and as head of the WWFF, testify to his dedication to teaching and scholarship and the creation of future teachers as well as to the fact that he is not a man given to half-measures. He thinks and acts on a grand-scale. To quote Wesbuch on his election to lead the WWFF:</p>
<p>
</p>
<p><a href=“U-M Web Hosting”>U-M Web Hosting;
<p>Food for thought indeed as is his- “President’s Annual Address to the University Community” on his ambitious vision for Drew 2020.</p>
<p><a href=“http://depts.drew.edu/pres/rweisbuch/talks/AnnualAddress2006.php[/url]”>http://depts.drew.edu/pres/rweisbuch/talks/AnnualAddress2006.php</a></p>
<p>Back to the questions related to SAT-optional - I am intrigued by Lafayette’s decision to rescind SAT-optional admissions policy put into effect in 1994 and the “whys” this strategy did not work there.</p>
<p><a href=“http://chronicle.com/subscribe/login?url=/weekly/v48/i09/09b01101.htm[/url]”>http://chronicle.com/subscribe/login?url=/weekly/v48/i09/09b01101.htm</a></p>
<p>“Nearly any study that measures the correlation between college achievement and SAT scores will be based on first-year GPA once you start to read the fine print.”</p>
<p>My d.'s school studied all four years. They had all “submitters”, and (until they changed the policy), SATs were used in admissions decisions. So what they studied were correlations between SAT scores of those they admitted and four-year performance. They obviously didn’t study those they hadn’t admitted.</p>
<p>
</p>
<p>But, if SAT is of no value, why use it at all, (which is the question I posed earlier to mini re: smith)? It either adds value to a specific college’s admission process or it doesn’t (Drew=“scam”). If a college does not think the SAT of any value, why not save the applicants some hard, cold cash and take the Sarah Lawrence route. But, if it does have some “minimal” value, perhaps in engineering, sciences, and, particularly pre-professional, as Drew ppt notes, why not require it of all? Or, if schools like the UC’s (and Smith) find that standardized testing hurts low income kids, why not come clean and publicize that they don’t want scores for anyone under, say $40k income?</p>
<p>MM what are the:
.</p>
<p>sorry, but I’m just a simple, boolean-logic type of guy, and don’t get it.</p>
<p>I don’t think that many people deny that the SAT adds <em>some</em> predictive value to HS grades in predicting college performance. But the SAT-optional LACs focus on the predictive value of the SAT for their admitted students who were non-SAT submitters and over the course of a college career, where the predictive value diminishes. (The College Board tends to emphasize the relationship between 1st-year grades and SAT scores across all students, where the value turns out to be higher.) </p>
<p>So in the semi-selective LAC context, it’s question of balancing a small increase in ability to predict college success against other factors. Here are a few of the questions of balance that I believe that such schools ask:</p>
<p>Would not requiring the SAT increase the size of our applicant pool and the number of interesting students in that pool? Does going SAT-optional produce a more diverse applicant pool? (Mt. Holyoke saw a dramatic increase in ALANA apps, over 50% increase for African-Americans, Hispanics, and Native Americans, the year they dropped SAT required, though no one attributes the whole increase to SAT-optional.) Does requiring the SAT discourage students from lower income brackets from applying? Do the well known correlations between income and race and test scores diminish the value of standardized tests in the admissions process? How do testing policies affect the institution’s image among guidance counselors, prospective students, and the academic community? Do we want to encourage our prospective students to devote time, energy, and money to prepping, coaching, and test-taking during their final years of high school?</p>
<p>And some schools doubtless ask less idealistic questions, particularly: Would we get a boost in the rankings?</p>
<p>When it comes to attracting, “interesting and diverse” students, SAT optional policy also has its place in international recruiting. Drew is indeed making great efforts to expand its international application pool to reach out to find academic talent overseas. The information sessions given by Ambassador Rick Haynes are a prime example and are given at quite a number of overseas high schools that do not even fall on most other colleges’ tour radar.</p>
<p>I can’t imagine a college doing research that didn’t end up reinforcing the wisdom of what it does.
And magically, what it does is what it perceives to be in its own best interests, given its market.
The only variable is the extent of the “our stuff doesn’t stink” pretentiousness.</p>
<p>With their yield I suspect Drew is reaching out for paying customers.</p>
<p>“I can’t imagine a college doing research that didn’t end up reinforcing the wisdom of what it does.”</p>
<p>Actually, in the case of my d’s school, it didn’t. And they changed policy as a result.</p>
<p><a href=“asteriskea:”>quote</a></p>
<p>Okay, really now - a study put out by the CB is not “propaganda” while a sponsored study by the Andrew W. Mellon Foundation is - time for a reality check, at least for me.
[/quote]
</p>
<p>You’re mixing and matching quotations from different people. </p>
<p>Sure, the College Board (Wayne Camara) study is not neutral, presumably defends a vested interest in SAT, and can be described as “propaganda”, though I haven’t seen any critic point out flaws in the analysis, either.</p>
<p>The word “propaganda” that you criticize was posted by someone who did not introduce the College Board study or rely on it. That already invalidates your charges of inconsistency — you treat different posters as one and the same. More importantly, the word “propaganda” as applied to Mt Holyoke’s claim was followed by a series of specific (or specifiable-as-needed) criticisms indicating WHY their statement is wrong. And Mt Holyoke has not, as far as I have seen, released the study or any data from which somebody other than their PR office can make an informed judgement of what they did and didn’t find.</p>
<p>Mini, did those measures entail putting a lower value on the SAT? </p>
<p>In one “research” document styled, “Re-imagining a Liberal Arts Education:
Linking Smith’s Past to Its Future in Bold New Ways” I noted a plan to expand MERIT aid to boost the number of high-ranking students. <a href=“http://www.smith.edu/planning/strategic/StrategicDirections.pdf[/url]”>http://www.smith.edu/planning/strategic/StrategicDirections.pdf</a></p>
<p>
</p>
<p>This seems to go hand-in-hand with the previoulsy released reports on Smith’s problem of having to offer one of the highest discount-to-tuition in the industry.</p>
<p>Unless I am mistaken, I did not see many mentions of lowering the standards by relaxing the reliance on standardized tests. </p>
<p>Potential Strategies Emerging from Campus and Alumnae Discussions:</p>
<p> Invest in admission marketing and outreach in the South and West to build awareness of Smith.
Build “feeder school” relationships in specific targeted areas, such as the arts, sciences and engineering, etc.
Improve Smith’s Web site and develop a virtual tour.
Reduce the traditional enrollment size of the college, striking a balance
among issues of diversity, access, and measures of quality.
Increase use of merit aid to yield top-ranking lower-need students.
Develop opportunities for the campus community to come together for conversations that explore critical intersections between excellence
and diversity.</p>