<p>S is an SAT tester. Did great - better than on ACT and does not like math or science. D did better on ACT and she is a biology major. So much depends on the individual.</p>
<p>“I think that a promising student in math, engineering, or the physical sciences ought to be able to achieve the SAT M score that corresponds to missing 2 questions (or fewer),” …"…varies from session to session, but somewhere around 730 to 740 is probably on the low side of scores for missing two questions…"…Adding to QM’s point, a study by Hsu and Schombert on University of Oregon students has shown that below SAT-M score of roughly 600, the probability of success in math/physics majors is very low."</p>
<p>Seems to me there are thousands of students scoring between 600 and 730. Wait; what are we talking about again?</p>
<p>sosomenza, I think at this point we’ll just have to agree to disagree. I agree that a differential of 200 points across three sections isn’t all that important compared to holistic factors.</p>
<p>I gave you an example of two students with a score differential of 400 points on ONE section, and you indicated that you wouldn’t take it terribly seriously. I’m not sure how anyone can honestly say that kind of disparity wouldn’t suggest some skill differences that should be relevant to a college admissions officer. I don’t care if it was a really good day for one kid and a really bad day for another. Anyone who is in the ballpark of a 400 verbal has some real reading weaknesses, and anyone who is in the ballpark of an 800 has at least reasonable facility with moderately difficult texts. </p>
<p>Of course there are no guarantees about which student’s general life outcome will be better. There are never guarantees. That doesn’t mean we can’t make educated guesses about which student is going to do better in an academic environment and, later, in a professional environments that depend on a high level of academic achievement.</p>
<p>“It varies from session to session, but somewhere around 730 to 740 is probably on the low side of scores for missing two questions.”</p>
<p>Oh good lord. You’re not “promising” unless you’re a 730-740? Might as well give it up and work at Starbucks? Do you guys have ANY idea how the real world works, where plenty of people are smart and successful and (Teh horrorz) didn’t get 730’s?</p>
<p>^The problem with these discussions is that people take the SAT to be a barometer of life success. </p>
<p>All it needs to be is a snapshot of where the student is now.</p>
<p>Right, I agree with collegealum314, post #385. I have tried to make it clear that a student could succeed in the more quantitatively oriented STEM majors with lower SAT scores than the “missing 2” level. I think that the faculty of many major universities will be teaching the introductory calc-based physics course so that an A level of mastery requires that the students be able to perform at that level. A student with lower scores does not require remedial course work, necessarily; but I think the student will be disadvantaged if he/she takes the viewpoint that SAT score are meaningless. A lower score is a sign that the student needs to improve his/her game–absent the various factors that could artificially lower a score (I forgot to include the possibility of food poisoning).</p>
<p>I am not writing about life success in general, Pizzagirl. I am just writing about the introductory calc-based physics course (and later mechanics, E&M and quantum mechanics). lookingforward specifically asked me to comment on the posts on physics. Otherwise, I probably would not have returned here after my first post on this thread.</p>
<p>Also, the threshold for success in the University of Oregon study was appreciably lower than I am suggesting.</p>
<p>
</p>
<p>I agree that acing the SAT by studying a lot is similar to acing a class by studying a lot. </p>
<p>To answer your question, it depends on what information you are trying to get out of the SAT and what you believe it measures.</p>
<p>In particular, it depends on whether:
- you are trying to measure skills (e.g., mathematical manipulation/basic skills, critical reading skills) or native intelligence.
- whether you think the skills which one can use to do well on the SAT are useful for school and/or in a future career. Some people think the SAT is pointless, and that the skills acquired to do well on it are useless.</p>
<p>I think it’s best to think about #1 in terms of measuring skills, while acknowledging that people with higher native intelligence will more easily acquire such skills. I generally don’t think such skills needed to do well on the SAT are pointless. That is, I think if one studies to become more adept at mathematical manipulation and/or to master basic algebra and geometry, and then performs better on the SAT math section, I don’t think this really subverts the point of the test. That person presumably can bring those skills to college (pt. #2). Same goes for the critical reading part. I am more skeptical on the writing sample, at least in terms of how it’s graded–but this was after my time so I don’t have personal experience with it.</p>
<p>
</p>
<p>Well, that depends on how much you value evidence of academic skills/abilities versus other activities and talents.</p>
<p>If performance on the SAT were largely correlated to the amount of time studying and prepping for the test, you would expect little correlation between PSAT scores and SAT scores, as few kids study for the PSAT (except perhaps those who think they have a shot at NMF). In fact, PSAT scores are highly predictive of future SAT scores:</p>
<p>
</p>
<p><a href=“http://research.collegeboard.org/sites/default/files/publications/2012/7/researchreport-2010-4-psat-nmsqt-indicators-college-readiness.pdf[/url]”>http://research.collegeboard.org/sites/default/files/publications/2012/7/researchreport-2010-4-psat-nmsqt-indicators-college-readiness.pdf</a></p>
<p>That’s not to say that some preparation isn’t helpful, but it’s unlikely someone is going to study their way to a substantially higher score.</p>
<p>When you get your SAT scores they also tell you how candidates with your scores on each section did if they took the SAT again. People who score 700 on the SAT math on average score slightly lower the next time they take it, presumably even with some further preparation.</p>
<p>
</p>
<p>The entire SAT prep books/classes/tutoring industry is banking on that not being the case…</p>
<p>Or perhaps that correlation takes into account the kids who do test prep of various kinds?</p>
<p>I suppose I am of the mind that a certain level of education needs to be successfully completed in order to score well (finish Alg 2, English, read a lot). After that I suspect a lot of gain is due to familiarity with the test and test-specific strategy.</p>
<p>
</p>
<p>Doesn’t the ETS send everyone a sample test? No excuse for not taking a look at it.</p>
<p>I’d guess that some people improve by learning actual math, expanding their vocabulary, or get somewhat better at reading passages and figuring out what they mean.</p>
<p>
Well, very few people do test-prep for the PSAT compared with the SAT.</p>
<p>I thought I simply said. “Did you see the physics comments?”.</p>
<p>Possibly my misunderstanding–I thought you wanted my opinion about the physics comments, lookingforward? Perhaps not, or perhaps I provided an opinion on the wrong set of comments?</p>
<p>(Your post #347 started out with “Needed you here, QM.” I took that as an invitation to comment.)</p>
<p>To provide some information specifically on the citizenship of physics Ph.D. students and Ph.D recipients, here is a report from the American Institute of Physics:</p>
<p><a href=“http://www.aip.org/statistics/trends/reports/EDphysgrad07.pdf[/url]”>http://www.aip.org/statistics/trends/reports/EDphysgrad07.pdf</a></p>
<p>Unfortunately, there is quite a time lag between data collection (2007) and publication (2010) and a similar lag between publication and the present. I may be able to find more recent data with just a little more digging.</p>
<p>What you will see from the report, though, is that in 2007, 55% of Ph.D.s in physics awarded by US universities went to non-US citizens. In about the year 1996, among first-year Ph.D. students in physics at US universities, the number of US citizens roughly equaled the number of non-citizens. Then, from 1996 to about 2001, the number of foreign first-year Ph.D. students in physics at US universities exceeded the number of American citizen first-year Ph.D. students. This is connected with the majority of the physics Ph.D.s awarded by US universities in 2007 going to non-citizens. After 2001-2, the situation shifted. By 2007, the first-year physics PhD students in American universities were 57% American citizens and 43% foreign citizens. (Incidentally, the first-year Ph.D. cohort was 19% female.)</p>
<p>I believe that the situation is similar in engineering, except that I would guess that a majority of Ph.D. degrees still go to foreign students. My explanation is that American students can find excellent employment in engineering with a BS or MS degree, and the additional qualification of a PhD is not really needed, unless the person wants to go into academia (which generally pays less than industry). In contrast, in physics, to be in charge of a research group, a person needs a Ph.D. in general. This would tend to depress the relative number of American citizens among engineering Ph.D. students.</p>
<p>What the causes of all of this are, I do not know. I can’t say whether holistic admission practices contribute to the situation or not. I don’t oppose holistic practices, generally speaking.</p>
<p>Here is a report from August 2012: </p>
<p><a href=“http://www.aip.org/statistics/trends/reports/1styeargrad.pdf[/url]”>http://www.aip.org/statistics/trends/reports/1styeargrad.pdf</a></p>
<p>They have combined the data on first-year physics Ph.D. students from 2007 and 2009. Summed over the two years, there were 929 first-year physics Ph.D. students in US universities who were American citizens and 920 who were foreign citizens.</p>
<p>Here are some numbers from 2012 (kind of interesting, if you are interested in physics–with apologies to Pizzagirl who is most likely <em>not interested!</em>). This just gives the total numbers of junior and senior physics majors, and grad students by type (Master’s or Ph.D.), with no demographic data, though.</p>
<p><a href=“http://www.aip.org/statistics/trends/reports/physrost.pdf[/url]”>http://www.aip.org/statistics/trends/reports/physrost.pdf</a></p>
<p>
</p>
<p>They use it because it is theoretically a common measure for all college applicants (perhaps not actually, since colleges allow the rather different ACT as a substitute), due to inconsistency in course rigor and grade inflation/deflation across high schools in the US (unlike in other countries like Canada and the UK, where high school grading is better controlled so that universities have no need for separate standardized tests for domestic applicants).</p>
<p>However, even the College Board acknowledges that the SAT-R is a weaker predictor of college grades than high school grades are, although the SAT-R does add some predictive value over high school grades alone. Note that Harvard’s admissions dean has said that AP/IB scores, followed by SAT-S scores and the writing* sections of the SAT-R and ACT, have more predictive power than SAT-R and ACT scores (non-writing sections): <a href=“Guidance Office: Answers From Harvard's Dean, Part 2 - The New York Times”>Guidance Office: Answers From Harvard's Dean, Part 2 - The New York Times; . UC has come to similar conclusions with respect to the relative predictive value of the different standardized tests. Perhaps that is not surprising that standardized tests that come closer to measuring what the student actually learned in high school are better predictors than a test that tries to measure “aptitude” or “reasoning”. But then that brings up the question of why the colleges in general do not just switch to the SAT-S tests. Perhaps it is due to inertia – they do not want to lose applicants who take the SAT-R or ACT for most colleges and not bother with the SAT-S for the few outliers that use the SAT-S instead.</p>
<p>*Apparently despite the “gaming” of the writing section taught by some test preparation schools.</p>
<p>
</p>
<p>Probably because a student who struggles on the relatively low level math on the SAT-R math section would struggle even more on the more advanced math needed for math and physics majors. Biology majors typically just need a semester or year of calculus and maybe introductory level statistics, so relatively weak math skills won’t be blocker for them like they would be for math and physics majors.</p>
<p>Here is the article: [[1011.0663</a>] Nonlinear Psychometric Thresholds for Physics and Mathematics](<a href=“http://arxiv.org/abs/1011.0663][1011.0663”>[1011.0663] Nonlinear Psychometric Thresholds for Physics and Mathematics)</p>