<p>^^clearly the colleges that have begun to require scoring from all testing dates “get this.” This is a trend I believe will continue. It is to the colleges benefit to cull all the “top scores” for their reporting but it is also in their admissions interest to “see” early test scores before the learning increase comes into play. The ability to test and retest and retest in and of itself blunts the SAT.</p>
<p>In my state at least, even kids who do not take formal “SAT” review spend a significant amount of school class time reviewing and prepping for similar multiple choice tests, much to the chagrin of some of my teacher friends.</p>
<p>Since the advent of NCLB they complain that large portions of class time are spent merely reviewing for state proficiency exams. My own kid started taking these tests in middle school, along with an exit exam for high school. I’ve seen the tests, and yes, I know the content is somewhat different, but I also believe all this test prep has had an effect.</p>
<p>In my time, we took a basic skills test (Iowa test, I think it was called) in the 8th grade, a PSAT in the 11th and usually the SAT once in the 12th.</p>
<p>maybe the difference in scores is logarithmic…</p>
<p>there IS a difference between a 2400 and 2390, but a very small and insignificant one.</p>
<p>calmom, the study on the effects of retesting that you cited, by F. Lievens, C. L. Reeve, and E. D. Heggestad, involves a set of 18 items that were repeated without change in test administrations about 2 months apart, on a medical/dental test in the Flemish part of Belgium. These items were embedded in a larger test, but apparently all 18 of the items (plus others) were contained in the first of four sections of a longer test.</p>
<p>In their section on “Implications for Practice and Research,” the authors note:</p>
<p>" . . . these two results do not provide support for the operational use of scores from repeated cognitive ability tests when the same items are used in both administrations. Further research should test whether the same conclusions are valid for use of parallel (instead of identical) tests in retesting."</p>
<p>The average number of correct answers given on these 18 repeated items by first-time testers was 9.29 (Group B). The average number of correct answers given by first-time testers who later retested (Group A1) was 6.77 and their average on retest (Group A2) was 9.86. So the second time around, they answered correctly 3 questions out of 13 they had originally missed (roughly, on average).</p>
<p>I don’t find this surprising. Are students encountering identical items on two SAT’s they take, two months apart? I’d hope not.</p>
<p>momofthree:</p>
<p>IMO, the SOLE reason some colleges are requiring all scores is to avoid cheating, which used be rampant when Subject Test had score choice. (google the topic – I was ‘shocked, shocked’ at what went on.)</p>
<p>bovertine: depends on your state. Our state tests are based solely on Calif curriculum, and have few “reasoning” questions, i.e., the “tricky” ones on the SAT. (A train leaves Chicago traveling towards Cleveland while at the same time…)</p>
<p>Smiles…bluebayou…
I had a friend who missed the ‘train problem’ because he said “Well… they will never meet because the conductors will stop the trains before they crash!”</p>
<p>
</p>
<p>Identical? Maybe the numbers change. Unless things have changed radically in the past 5 years, SAT MAth commonly repeat problems structures with little change - eg the train problem, the changing population over two time period problem, the proportion problem, the crossing lines forming triangles/supp/comp angles problems, the extremely simple two equation in two unknown problems, the recipe problems,there are many more I could list if I had my reference books with me. That’s why coaching is so effective. Multiple problemns where you can simply try answers or numbers. As a tutor at a large tutoring company I saw several official versions of the test and you would be shocked at the repetition.</p>
<p>Just a word to the wise (well, specifically silverturtle - which is not to say he is or is not wise…)
Asserting too strongly that SAT correlates in a meaningful way with IQ does suggest that you need to deal with the fact that blacks do less well on all parts and that girls are underrepresented at the high end of the math portion. I doubt that anyone here would agree that blacks are not as smart as whites or that girls are not as good at math as boys.
The key here is ‘too strongly’ and ‘meaningful’.</p>
<p>I’ll admit, I am surprised that there would be so much repetition. Never really looked through “Ten Real SAT’s” to see how much repetition there was in those tests. </p>
<p>Separately, I admire the use of “ceteris paribus” in ThisCouldBeHeavn’s post #256.</p>
<p>Perhaps we should return to the initial topic - we’ve ranged far afield here.
Is the SAT a blunt tool?
We should, properly , define blunt in this context.
I would definite blunt tool as “an instrument which is of use, but is imprecise”.
Given that…Have at it, droogs!</p>
<p>
</p>
<p>Well … it is commonplace to argue all sorts of things at which girls are better than boys, language, cooperative problem solving, etc.</p>
<p>is it a true statement that kids who score high on the SAT are smart?
is it a true statement that some kids who score low are smart?</p>
<p>I’d say yes to both. But if I was a college admissions officer I would try and take the high scoring kids because ALL are smart. with the low scoring kids I don’t know which one is smart.</p>
<p>that may be a blunt tool approach, but it works. I don’t think you need a finer tool to shovel the smarter students into the elite schools.</p>
<p>
</p>
<p>When I said “IQ,” I was using that as a short way of meaning “IQ scores.”</p>
<p>Re: posting 273
Ah. Well, IQ tests do measure IQ. Now how, precisely, IQ is related to intelligence is another question. But, clearly IQ is very strongly correlated to IQ scores, presuming a good IQ test, properly administered.
Still, I see no real reason to enter the very complex debate about these matters (IQ, IQ scores and general intelligence - also known as ‘g’) here.</p>
<p>
</p>
<p>But the colleges have additional information – they also know the kid’s GPA, class standing, and generally have some sort of information about the quality of the high school (school profile, etc.). And its not about high vs. “low” scoring applicants, its about high vs. moderate scoring applicants. And its not about test scores alone – its test scores along with a whole lot of other information about student achievement.</p>
<p>I mean… its a pretty fair assumption that the top 5% of graduates at a competitive high school are smart, no matter what their test scores. So when the colleges can see the kids GPA, the rigor of high school course work, and have some indication as to the relative quality of the high school – they know something that is much more important than how “smart” the kid is – they know how the student actually performs academically.</p>
<p>Ok calmom, fair enough. But if the other student with the higher score has everything else held equal, why not pick the one with higher score? Even if the SAT measure ability under pressure, that’s at least somewhat useful. </p>
<p>Also: let’s not bring race/gender into this. I have only ever seen it lead to ugly results.</p>
<p>All things being equal, why not accept the student with the lower score?</p>
<p>Because there is a limited number of spots. Besides the SAT, they are equal. With the SAT, they aren’t equal. For whatever reason, the student with the higher SAT has demonstrated higher potential, intelligence, whatever you want to call it. But the the point is that you couldn’t admit that student OVER the student with the higher score.</p>
<p>All normal criteria being equal, test scores, EC’s and so on, do you think better-looking applicants should be chosen? Or taller ones? Looks and height are correlated with future success, and a college wants successful alumni after all.</p>
<p>No, it is equally possible that the student with the higher SAT has merely demonstrated that his parents have greater resources to pay for test prep. Both students have demonstrated equal abilities through high school achievement, and the school can give its slots out to whoever they want.</p>