I agree that a standardized curriculum with some sort of standardized grading scheme would make the SAT obsolete. That’s how K-12 education is happening in most other countries already.</p>
<p>However, being from one of those countries myself, I very much appreciate the flexibility that the individualized American system provides. For example, take a student who’s strong in everything except math and science. In the US, such a student could take honors courses in everything except for math or science. In other countries, the same student would have to take all courses at a uniform level: either take everything at the highest level (and maybe repeat entire grades because you failed math or calculus-based physics) or take everything at a lower level and not qualify to attend an academic 4-year university. </p>
<p>Uniform grading is difficult to achieve as well. Either you need to tell teachers <em>exactly</em> what their assessments need to look like (even if half of the class fails as a result, which has happened at my German high school), or you need standardized external exams (like British A-levels) that replace grades given by teachers. </p>
<p>The two big disadvantages of the American education system, from my point of view, are:
that degrees by themselves are pretty meaningless (what exactly does a high school or college graduate know?), which in practice often means that the institution awarding your degree is more important than the degree itself
that the transition from one stage of education to the next is a mess and often disadvantages students from a lower socio-economic tier. I grew up with equal access to education - including higher education - being a basic right. In the US, it’s considered a privilege.</p>
<p>In theory, uniformity can be done at the course level without locking students into entire sets of courses (except perhaps closely related courses like math and physics). For example, a student can take advanced / honors math but regular other courses – but “advanced / honors” and “regular” have consistent definitions over all high schools so that universities and others could have reasonable expectations about what the student has learned if his/her transcript lists a particular course, eliminating the need for standardized testing for admissions or placement in subjects typically learned in school (though some placement testing would be needed for those entering the system from the outside).</p>
That’s an interesting idea. I am curious how you would propose to achieve a uniformation of grades without negatively impacting student learning though. True uniformation would require that all grades (that matter) are assigned through standardized external exams (as is done in some countries), which in particular means that homework or participation are no longer directly rewarded… I just can’t picture that working very well.</p>
OK, NewHope, you threw down the gauntlet, here is a response:</p>
<p>I wrote: (post 32)
</p>
<p>Vonlost wrote in reply: (post 34)
</p>
<p>So Vonlost, are you saying that as long as the adcoms know how to adjust for racial or class bias in test scores, that racial or class bias in the test itself is OK, not really a problem? Did you really think about what you were saying when you wrote that? Do you not care about how the existence of race and/or class bias in a test as widespread and important as the SAT affects disadvantaged groups in particular, and our whole society in general? </p>
<p>Food for thought: read this article about racial stereotype threat, and then see if you can continue to feel comfortable with the notion that the evidence of racial bias on the SAT isnt really a problem.</p>
<p>I agree with whoever said the tests “weed out the men from the boys.” You really want an intelligent kid? Pick the one with the 4.8 and 34 over the 4.8 and 27. You can see who got through school merely by studying like crazy and working for tests just by looking at standardized test scores, and they’re much better indicators of potential.</p>
<p>True, that could lead to some need for standardized testing in the classes to ensure uniformity. It is not an easy problem to solve without creating other problems. Alternatively, there could be some other means of checking course content and grading standards, but that could be even more difficult to manage than standardized testing.</p>
<p>However, it does seem that, in the absence of anything of this nature, AP (and IB to a lesser extent) are taking this role by default, at least for the higher achieving students. I.e. they are providing a known consistent external evaluation of what the student has learned (though obviously not a complete one, in that it is just a single “final exam”, while lab, project, and class discussion work is not counted), and schools are aligning courses to those syllabi (unfortunately, the other courses continue to be neglected as usual). This is not necessarily an ideal phenomenon.</p>
<p>That appears to be more about the phenomenon about how believed stereotypes affect performance than it is about whether the test itself has any bias.</p>
<p>I guess an AP-type model for courses at other difficulty levels could work: grades capture effort and performance inside of the classroom, while external exams measure mastery of the material at a uniform standard. Now you only need to convince politicians that that’s a good idea :)</p>
<p>I think such neutral tests cannot be constucted (or they would have been by now), so we recognize the bias and account for it. The situation is not ideal; little is.</p>
<p>UCBalum: Call me confused, but I don’t understand why you don’t appear to get why the larger issues discussed in the above-referenced article are relevant to this discussion. Are you saying that because this article doesn’t technically focus on test bias itself, the article is irrelevant or doesn’t matter? A test that discriminates based on race and class is not necessarily a valid test of ability – but it is likely a valid test of race and class. Until it is definitively proven that the SAT does not discriminate thusly, it is potentially quite harmful to use a biased test that does not dispel the notion that certain racial or class groups are more or less able than others. The point of the article is, students who believe the stereotype that they will fall short on tests of ability do less well than students who don’t have any notion of such a stereotype, when in fact both groups are equally matched in ability prior to testing. Stereotype threat and test bias can be at play at the same time and are not mutually exclusive. Don’t you think it might be emotionally, socially, academically or career damaging if you were of a minority group and statistics said you’d not likely do as well as the majority on the test? </p>
<p>There are, unfortunately, ill-informed people who really believe that some races are “smarter” than others, and that tests, including the SAT, are proof positive. Are you okay with this, UCBalum? You don’t think it matters that the well-publicized disparities in SAT scores between races might significantly contribute to the stereotype that some races aren’t as “smart” as others? You don’t think it matters that the pain this might cause minority students might then cause test anxiety and affect test performance, not to mention affecting overall experience of inclusion in schools and in the larger society?</p>
<p>
</p>
<p>Forgive me Vonlost, but your response seems to me to be an evasion of an uncomfortable truth. Little is ideal, so therefore why try to change anything??? Your statement may be the unfortunate reality; however, it is also quite possible that there has never been the political will to construct better tests. Or also possible that there is just too darn much money involved in the current testing system, too many people and institutions co-opted by the business of testing, and too much potential egg on people’s faces if the reality of how bad the tests really are at accurately gauging ability and predicting college success were to be exposed. A different approach could be: the SAT test is not ideal, so it should not be used – there are other well-documented effective means for students to demonstrate satisfactory levels of qualification for college level work without standardized testing.
The colleges that are SAT optional have figured out how to do without the tests for quite some time, with apparently good results.</p>
<p>Can someone tell me how does SAT discriminate against a particular race? This test is given around the world, and some of the best scores have come from outside of the US. </p>
<p>I was told by kids in my school, “go home Chink.” I don’t remember it caused any test anxiety or low performance in me. As far as “inclusion in school,” I was not asked out in school because back then most guys didn’t want to be seen with anyone different. Didn’t make me feel good, but it didn’t give me low self esteem either. </p>
<p>Ladybug- I don’t wholly agree with your assessment.</p>
<p>LB - It’s well-proved that SAT test prep works to increase scores. And it’s well-established that many/most in the majority group take the SAT just once, and without prep … in other words, a static target. At least two minority groups have used this static target to great benefit. This begs a question so obvious that I won’t even ask it.</p>
<p>The article does not claim that the test is biased; it claims that the racial stereotypes around “testing of ability” can have effects on people’s performance on tests. See the examples of where they were able to cause racially different results with the same test by just having pre-test questions or announcements that caused the test takers to think about the stereotypes, while the racial difference disappeared when non-stereotype-inducing pre-test questions were used.</p>
<p>The problem of stereotype threat is not the same as whether the test is actually biased; the test can be completely unbiased but the results affected by stereotype threat, or the test could be biased even if stereotype threat is eliminated from the test administration.</p>
<p>
</p>
<p>Except that in the complete absence of standardized tests, there will be more incentive for high schools to inflate grades and reduce course rigor without fear of being exposed by the standardized tests. While the current standardized tests have plenty of problems (including being relatively weak predictors of individual student performance in most cases), some means to control this problem needs to be found if they are to be eliminated. (The few schools that don’t use standardized tests are free-riding on the existence of such in terms of controlling this problem.)</p>
<p>“there has never been the political will to construct better tests”</p>
<p>Perhaps, but then we’re talking about decades, with the prospects for change as low as ever. </p>
<p>I, for one, am not in favor of all U.S. colleges becoming alike, nor for their application and entrance requirements to do so, so it’s fine to have both test-optional and test-required schools. Kids shine in different ways, and schools develop different cultures.</p>
<p>I guess I don’t really see a ‘need’ to change the current system.The SAT is simply one of several important information points presented about a student. Grades are usually dependent upon the rigors, grading policy and peer group of any particular high school. An A in school#1 might be equivalent to a B in school #2. Adcoms know this and that is why they evaluate a student’s grades and course load in the context of their particular school. ECs are a way for a student to stand out and have a lot more to do with a student’s individual motivation, talents and interests than anything else, IMO. </p>
<p>The SAT/ACT is just one more piece of information. High score on SAT with middling grades can mean one of several things - a talented but unmotivated kid or maybe a bright kid at a very rigorous school. On the other hand, a lower SAT score with super high grades could mean that the kid was at a school with a not-so-strong curriculum or there was grade inflation going on. If I were an Adcom, I would want to see the complete picture of each applicant. Until we have a national curriculum standards, there really is no other way to know what is taking place within each high school. It’s sort of like diagnosing an illness - you often look at both the subjective and objective data in order to arrive at a diagnosis. There’s not always a perfect answer, so you do the best you can given the information available. Not sure why having less information would be helpful.</p>
<p>The OP’s question #1 was “Should colleges be SAT optional?” By this I’m almost certain the OP meant “all colleges.”</p>
<p>The discussion has morphed a bit, now positing that the SAT is an unreliable indicator for many students and therefore should be done away with. Many posters feel that this would be gross overkill, given the many college options that remain for students who (um, let’s say) score below their potential on the SAT.</p>
<p>If I’ve summarized the argument correctly (that any data instrument with provable bias should be eliminated from the college application process), then GPA, class rank, private/public HS status, teacher recommendations, family education history, social status, economic status, residence data, and hooks should “obviously” be excluded because they too are biased in favor of certain student groups.</p>
<p>UCBalum, you didnt answer my question. Perhaps the point of your replies to me is to detract from engaging in an honest debate of the issues?</p>
<p>
</p>
<p>A quote from the conclusion chapter of the book SAT Wars: The Case for Test-Optional College Admissions, edited by Joseph Soares:</p>
<p>Jay Rosner, Executive Director of the Princeton Review Foundation, found serious racial and gender biases in the question selection process for the SAT, and it remains to be seen whether or not the Educational Testing Service (ETS) will open its records to independent scholars to evaluate that damning charge.</p>
<p>The track record of the ETS with respect to sharing data on its technical internal procedures is not encouraging. It took pressure from the University of California, and 2 years of in-house dithering, for the ETS to release data relevant to a previous charge of racial bias. Roy Freedle, a retired researcher with 30 years of employment at the ETS to draw on, found racial and social-class biases in the ETSs scoring of vocabulary questions. The ETS reluctantly provided data to the University of California to test Freedles case, and the work was done by two independent scholars, Marla Veronica Santelices, Catholic University of Chile, and Mark Wilson, Professor in the Graduate School of Education, UC Berkeley. They substantiated Freedles conclusion that the test has a racial bias, reporting that Freedles findings hold for current test forms the confirmation of unfair test results throws into question the validity of the test and, consequently, all decisions based on its results It is biased against the African Americans and could be exposed to legal challenge. (Santelices & Wilson, 2010, p.126) Naturally, the ETS rejects these findings and continues to resist independent evaluation of its data. Nonetheless, Freedles case was partially made, and that lends support to the notion that Rosners charges of bias are also true. P. 202</p>
<p>Here is the link to the book again in case youre interested in reading more:</p>
<p>It is not very honest to claim that stereotype effects on test performance are the same thing as the test itself being biased.</p>
<p>In any case, simply because there may be a racial difference in the results (that some racists may try to use as evidence for their claims) does not mean that the test is necessarily biased (nor does it mean that the racists are correct if it is not). What you are proposing is shooting the messenger (the test) and ignoring the real problems that it exposes – including the poor quality of K-12 education that disproportionately afflicts members of some ethnic groups. Hiding the evidence of such poor quality K-12 education does no one any good.</p>
<p>Of course, given how inflammatory anything race-related can be, that tends to obscure something much bigger, but is widely ignored: the SAT really should not be that difficult for a normal level (not advanced or taking a dozen APs) late 11th grade or early 12th grade student. That students in the US on average (regardless of race or ethnicity) do as poorly as they do on the SAT does not say very complimentary things about the K-12 educational system in the US.</p>