NYT: College Panel Calls for Less Focus on SATs

<p>I certainly agree with that, idad, and this access is one of the many reasons that so many foreigners love our college opportunities and praise its general quality. The U.S. college system spells opportunity, regardless of origin. Nevertheless, the elite colleges are fond of locating those of both humble origin and exceptional brainpower, measured somewhat by standardized tests but additionally by many other factors.</p>

<p>I just read the article for which Marathonman88 provided the link -</p>

<p>Dramatic</a> Challenge to SAT and ACT :: Inside Higher Ed :: Higher Education's Source for News, Views and Jobs</p>

<p>It has me very depressed. This time I really wonder if coreur is correct and this is all much ado about nothing. I cannot help but feel that turning away from the SAT and ACT - which meaure fundamental academic capabilities - will just lower our overall standards in education and worsen, not help the situtation.</p>

<p>My take based on Fitzsimmons' comments is that the SAT/ACT should be replaced by the Subject Tests/APs because if students are going to spend so much time and money prepping for a test, they ought to be learning something substantive, rather than mere test-taking skills, which is what SAT/ACT prep teaches.</p>

<p>Also, Fitzsimmons' acknowledges in essence that the SAT/ACT discriminates against URMs (although his words are "calcifies differences"), and should therefore be de-emphasized.</p>

<p>Yes, I know that's what Fitzsimmons said. Here's the thing though - the SAT subject tests still rely on reading comprehension and mathematical reasoning to do well on. Why will they lead to any difference in performance among various racial and income groups? Furthermore, prowess in a specific subject area is more dependent on quality of high school education (i.e., really good chemistry teachers, labs) and will most probably accentuate the "calcification" of income disparities.</p>

<p>How does the SAT discriminate against racial groups? I understand the income argument but not the racial one.</p>

<p>I believe that the comments on the linked article are more interesting than the article itself. I happen to believe that the story posted by this well-meaning commentator speaks volumes about the state of our education, and especially the state of our education of ... people who might be placed in a position of responsibility to educate the next generations:</p>

<p>
[quote]
Standardized Tests! Have I got Stories!
When I graduated from high school, I applied for admissions to a private liberal arts college. I was denied admittance based on a standardized test I had taken in the 8th Grade (an IQ test at that). So, I joined the Navy. After I completed my service,I enrolled in a junior college, earned an AA with a 3.67 GPA. I then transferred to Indiana University and earned a BA in History with a 3.47 GPA.</p>

<p>Next, I decided a Masters Degree would be nice. So I enrolled at Indiana State University. Had to take a GRE scored around 500. Earned a M.ED. with a 3.94 GPA. Then, I wanted a second Master Degree so I went back to ISU Earned a Masters of Science. It was supposed to be an MA. But at that time ISU had an interesting language requirement. You had to demonstrate the ability to read another language. The way this was accomplished was to take a class and then, you guessed it, pass a standardized test. I received an A in the class and believe me, I could read French. Took the standardized test 3 times; never did pass it. So I was awarded an M.S instead of an MA.</p>

<p>I just finished my Ph.D. at Indiana State. As an entrance requirement, I had to take the GRE again. An interesting thing happened; four degrees later and 20 years of teaching experience, I scored lower than ever on the GRE (450 on verbal). Fortunately because I had earned previous degrees at ISU the score did not keep me out. I completed all of my course work with “A"s and finished my dissertation in August. The point of this pedantic tirade is not to impress anyone with my academic record. Rather, my experience is just one of many examples that demonstrate how flawed standardized testing is. It is about time that the decision makers in higher education acknowledge what we have known for years. Standardized testing is a terrible, completely invalid and unreliable way to predict academic success. Of course, I am not surprised that the testing agencies disagree. They stand to lose millions of dollars if higher education decides to just say “no” to standardized testing.

[/quote]
</p>

<p>Draw your own conclusions!</p>

<p>How about GPA? Like they did research on how SAT can/cannot predict college success, has they done a same research for GPA?</p>

<p>
[quote]
the SAT subject tests still rely on reading comprehension and mathematical reasoning to do well on. Why will they lead to any difference in performance among various racial and income groups? . . .</p>

<p>How does the SAT discriminate against racial groups? I understand the income argument but not the racial one.

[/quote]
</p>

<p>My impression is that these are two separate issues: First, if colleges are going to required standardized tests, then they ought to encourage the student to learn substantive material, not test-taking skills. Second, the statistics show that on the whole, URMs perform worse than non-URMs on the SAT tests in current use. Ergo, emphasizing the use of a test that certain races have not proven to perform well on de facto discriminates against them. Therefore, the value of the test should be de-emphasized in admissions. That was my interpretation, anyway.</p>

<p>^^--^^</p>

<p>I believe that people love to overplay the element of test-taking skills in standardized tests. While it is true that EVERY test requires the takers to be familiar with its presentation and format, the basis of the test remains the ... material. In the same vein, detractors prefer to ignore the fact that a test such as the SAT attempts to measure math and verbal reasoning abilities. </p>

<p>As far as racial discrimination, the SAT itself does not discriminate; it simply returns a snapshot of the differences in the academic preparation of students. Those differences are created by SES conditions and differences in the attempts to educate different racial groups. </p>

<p>The question that begs for an answer remains ... what do we after we punt the SAT? Will URM and low SES be better off? Fwiw, NOTHING precludes any adcom in the country to simply ignore the test scores of special groups of admits. NOTHING!</p>

<p>Why don't Harvard and its peer institutions invest in k-12 education in this country by waiving tuition for their admitted students who commit to teaching in low-income high-URM communities following graduation? And by making educational research really cool so that it attracts top talent to trying to solve the persisting disparities in academice achievement among races and income groups?</p>

<p>I just don't think the admissions office is the place to solve these problems. Solving this issue lies in K-12 education. If we get that right, Harvard can use any darn test they want for admissions and the composition racially and economically of the top scorers will be just fine.</p>

<p>
[quote]
Second, the statistics show that on the whole, URMs perform worse than non-URMs on the SAT tests in current use.

[/quote]
</p>

<p>Do URMs do any better (relatively) on Subject Tests and AP tests?</p>

<p>^No, (in response to #170) that is why I stated that there seem to be 2 separate issues. It sounds to me as though Fitzsimmons would prefer the Subject tests/AP tests over the SAT/ACT, or better still, each college would have its own test. Also, the importance of all standardized test results would be de-emphasized in order to level the field for URMs and SES.</p>

<p>
[quote]
NOTHING precludes any adcom in the country to simply ignore the test scores of special groups of admits

[/quote]
</p>

<p>I agree with this and have no problem with it. Unfortunately, there is a huge contingent (some with hired lawyers) who believe otherwise.</p>

<p>
[quote]
Why don't Harvard and its peer institutions invest in k-12 education in this country by waiving tuition for their admitted students who commit to teaching in low-income high-URM communities following graduation? And by making educational research really cool so that it attracts top talent to trying to solve the persisting disparities in academice achievement among races and income groups?</p>

<p>I just don't think the admissions office is the place to solve these problems. Solving this issue lies in K-12 education. If we get that right, Harvard can use any darn test they want for admissions and the composition racially and economically of the top scorers will be just fine.

[/quote]
</p>

<p>Mammall, I believe that several leading colleges actually do what you are suggesting, at least to a certain extent. There is a growing number of faculty members who are trying to make educational research "really cool" at places such as Harvard, Stanford, Vanderbilt, or even at the University of Arkansas. Unfortunately, the power of interest groups that live only to reject any attempt at reforming the current system is still much stronger. </p>

<p>The good news is that the world is getting smaller every day and that the United States is slowly undertstanding that its education system is not *that *unique and that there is much to learn from countries that did not make the same mistakes as the United States in supporting a monopolistic system controlled by the service providers. </p>

<p>There is hope!</p>

<p>
[quote]
Do URMs do any better (relatively) on Subject Tests and AP tests?

[/quote]
</p>

<p>I think the score gap between "black" students and "white" students is smallest on some of the foreign language tests, e.g., German. The College Board has definitely done research on this.</p>

<p>
[quote]
The College Board has definitely done research on this.

[/quote]
</p>

<p>Of course they have, but I notice that they fail to publish it. :rolleyes:</p>

<p>CB does report SAT Reasoning Test scores by state, sex, race and ethnicity, not to mention how many years of math, english, foreign language, etc. There were 1.5 million students who took the SAT last year, and 273,000 Subject Test takers. Obviously, some of those Subject Test takers took the ACT and not SAT Reasoning Test. But regardless, CB has the sex/race/ethnicity data on every kid who took both the SAT Reasoning Test and the Subject Tests. So, CB could report Subject Tests by sex, race, ethnicity, & state if they so chose. Since they choose not to make it publicly available......hmmmmmm</p>

<p>And why won't they pulblish it? I don't get it.</p>

<p>Actually, the College Board DOES publish, each year, data about an aggregate mean score for each self-reported ethnic group for each SAT Subject Test (and for the AP tests), but I just don't have those links at hand. (They are not terribly difficult to look up, if you are really interested.) What College Board has NOT published since the mid-1990s, but surely still has raw data for, is mean scores on the SAT Reasoning Test by self-reported income-level groups, separated by self-reported ethnic groups. (It would be even better if College Board published income to SAT score correlations using bivariate plots, that is "scatterplots," but College Board seems never to do that.) The one time the ethnic groups were used to show income-to-score correlations, in a report that is summarized in the chart </p>

<p>Image:1995-SAT-Income2.png</a> - Wikipedia, the free encyclopedia </p>

<p>on Wikipedia, apparently some people were taken aback that ethnic groups could vary so much in what score level might be expected at what income level. </p>

<p>There are several possible critiques of this finding: </p>

<p>a) All the income data are self-reported by students, and may be wildly inaccurate. </p>

<p>b) All the ethnic categories are likewise self-reported by students, and in some years many students declined to self-report any category. </p>

<p>c) Correlational data of this kind doesn't prove causation--but that is equally true of some objections to standardized testing for college admission in general. </p>

<p>d) We aren't sure what the true causative relationships, if any, might be. </p>

<p>All of the points above caution us from leaping to conclusions about what a data report underlying the chart like the linked chart from Wikipedia might mean, but of course maybe the way to get to the facts about what is really going on is to gather more and better data, and report all the data currently known, so that scholars can argue about what the data mean and what other kinds of data should be gathered.</p>

<p>I was just at Harvard's regional information meeting in my town this afternoon. The admission officer present, who has great experience in the Harvard admission office, said that despite what you might read in press reports about the NACAC study, Harvard will not go in the direction of not relying at all on the SAT or ACT tests as one part of the admission process. The NACAC report, rather, would provide Harvard a rationale for requiring five rather than only three SAT Subject Tests, if any change to the admission criteria were to be made in the future.</p>

<p>I am not sure why we discuss a lot about SAT subject tests. The panel did not propose to substitute SAT-I with SAT subject tests. "The report suggests that what is needed is a new achievement test, pitched to a broad group of students, that would predict college grades as well as or better than available tests." They only said that tests based on subject contents like IB, AP, and SAT subject tests "are more closely linked to the high school curriculum than the SAT and ACT, and have little expensive test preparation associated with them."</p>

<p>Thank you Token Adult - for both the data link and for relating what the Harvard admissions rep had to say. Wow - five subject tests. I wonder if my ninth grader should take the SAT II subject test in math this spring - she's in precalc this year.</p>

<p>Regarding the chart you provided via Wikipedia - I wonder why CB cuts the income levels at 70k. I think on the questionnaire when they kids test, it goes up to at least 100k. And of course, colleges like Harvard are now distinguishing between incomes over 100k all the way to 180k or 200k when they distribute FA. I notice the black SAT scores start moving upwards more steeply at 70k and am curious what they do over that income level.</p>

<p>Realistically, 70k is a pretty mid-level income for a family in America right now, assuming two working parents.</p>