<p>
</p>
<p>I haven’t seen a survey or other data to support that assumption. Things are still not that different from bovertine’s day. Over the past 10 years, about 40% to 50% of physics majors have gone directly into industry. As I said before, only 15% of graduating physics majors get a Ph.D. If going directly into industry is “plan B” then we are misleading students as to the value of a physics bachelor’s degree.</p>
<p>A lot more students have aspirations of graduate or professional school than actually go to graduate or professional school.</p>
<p>People don’t usually consider art history, sociology, or comparative literature to be pre-professional majors, but it is likely that most graduates do not go to graduate school.</p>
<p>Momsquad, it was presented by gender. Look on page 19; overachievers and underachievers by major by gender.</p>
<p>@C Fang,</p>
<p>What is missing is the total number of male & female students for each subject. The tables show that no students, male or female, were categorized as overachievers in math and physics. Eleven males and 1 female were labeled as underachievers in these subjects. The authors claim that female students are more likely to be overachievers than male students. So the missing information is how many female students with SAT (M) scores of less than 600 were enrolled in upper level math or physics? My guess is zero, and so the population that has been shown by the authors to “overcome” low SAT scores is missing from the data.</p>
<p>I am leery of any study that uses gpa as an outcome without addressing the wide differences in rigor at various schools. I wonder if the finding was limited to math/physics because those disciplines are sort of “fixed” in how work will be evaluated – either the solution is correct or not correct, whereas in English and social sciences and humanities there is a great deal of subjectivity that enters into grading and, of course, the rigor with which writing is graded varies wildly.</p>
<p>I guess I’d be more inclined to look at GRE, GMAT, LSAT and MCAT scores after four years of college and see if those are strongly correlated to high school SATs. Grades are too squishy.</p>
<p>^^^
They do sort of mention something about GRE as an aside</p>
<p>
</p>
<p>Assuming your goal is to go to grad school, from what I’ve seen, you pretty much need an 800 Q (old GRE) or very close to get into a decent grad program in Physics. So this GRE information is more convincing to me as far as students who want to attend grad school.</p>
<p>Anyway, the study seems a bit limiited in scope to me. If the conclusion is “You likely won’t do very well as a Physics major at UO without an SAT M above 700”, well I think the study is pretty good evidence of that. I suppose you can generalize a bit beyond that, but how much exactly I don’t know.</p>
<p>
Even that is going to be squishy, since one school might assign the easier problems and another might assign mostly the challenging ones, and experienced professors can make the exams as easy or difficult as they choose. In some schools physics is watered down dreadfully to account for the general inability of the students to achieve more than the mere basics.</p>
<p>
</p>
<p>Actually, I think that math/physics exams are often graded on a curve. So a lower ability grouping would tend to have a lower curve. </p>
<p>I think it is just that the SAT is basically a test of basic mathematics, vocabulary, and English grammatical conventions. Students who have a hard time with the basic math tested in the SAT will probably have a hard time with math in their college courses. Maybe physics is also very math-intensive & math-reliant, while perhaps other STEM majors such as chemistry provide more leeway for the math-challenged student. </p>
<p>In other words, there is only one thing that the SAT tests directly that also has a direct bearing on college performance – math – and that only has a direct bearing on the most math-intensive majors. A student can’t succeed in a math major without being math-able, but a student can compensate for for SAT-deficiencies, including math, in other majors where students will be assessed based on the knowledge gained from reading, lectures & labs.</p>
<p>
</p>
<p>I think the study would disagree with that claim. The study showed that for all majors, people with higher SATs tended to do better, regardless of the their high school GPA. That is, for all majors, people with higher SATs tended to have higher GPAs in their college majors.</p>
<p>
</p>
<p>The University of Oregon is a typical state university and therefore fairly representative of the schools attended by a large number of students on a national basis. Results there can probably be generalized to a very large number of schools.</p>
<p>
</p>
<p>Yeah, maybe, maybe not. I’d like to see some sort of evidence or at least a cogent explanation for why it is considered representative.</p>
<p>Generally when people claim a study like this has universal application, they use more than one school in their sample. It seems like it would be valid, but at least what I read here is pretty skimpy.</p>
<p>Oregon is probably mostly typical of a state university in a smaller population state, so it takes students with a wide range of entrance characteristics. However, it is atypical in being a flagship without engineering.</p>
<p>In terms of student entrance characteristics, state universities in states with large populations may be more numerous, with each taking students with a narrower range of student entrance characteristics. For example, there may be little overlap in the student entrance characteristics at Berkeley and East Bay.</p>
<p>
</p>
<p>Er, no…</p>
<p>The study explicitly found:
</p>
<p>The only majors in which low SAT students did not sometimes end up becoming high-achieving students was pure math & physics. </p>
<p>The study did show a weak to moderate correlation between SAT scores & long term performance in other majors (0.35 - 0.5) – but my point was that some students in all majors other than math//physics seem to be able to compensate well for weaker math proficiency (as tested by the SAT). </p>
<p>Statistical note: a correlation coefficient of 0.2 to 0.4 is considered to be weak or low (not very significant); a correlation coefficient of 0.4 to 0.7 is considered moderate. If SAT scores were highly predictive of performance, then you would expect to see a correlation coefficient of 0.7+. I personally would not be comfortable saying students with factor X “tended” to do better/worse unless there was a high correlation.</p>
<p>Calmom, you’re hypothesizing the existence of these women who do badly on their math SATs (and let’s not mince words here, a math SAT of less than 600, for a physics or math major, is <em>terrible</em>), but nevertheless manage to get B+ averages in upper division courses in math and physics. Who are these women? None, as you note, were found at the University of Oregon. Where would they be found? </p>
<p>Twenty-one percent of “overachievers,” across all subjects, were male. Yet there were no male “overachievers” in math and physics. I’m inclined to agree with the authors that the math SAT actually measures a skill necessary for physics and math success.</p>
<p>As to the semantics of tended, we simply disagree about the meaning of the word. I’m perfectly happy saying that students with factor X tend to do better at Z, if the correlation is .4-- factor X is explaining 16% of the variation, and that’s a lot. If the correlation were .7+, I’d use a stronger word than tend.</p>
<p>
</p>
<p>That is not really surprising, since most other majors do not require more math than freshman calculus, if that (note: University of Oregon does not have engineering, so engineering would not be included in “other majors” in this study).</p>
<p>Indeed, [UC’s</a> study of the SAT](<a href=“Publications | Center for Studies in Higher Education”>Publications | Center for Studies in Higher Education) found that the math sections had close to zero predictive value for college grades. It would not be surprising if this was because students self-sorted into majors that used more or less math in correlation to their math section scores (students who did the worst on the math sections would be the least likely to choose math-heavy majors).</p>
<p>But it would also not be surprising if someone who could not handle the relatively basic math found on SAT math sections would have little chance of success in a math-heavy major like math or physics, as the study that is the topic of this thread indicates.</p>
<p>
</p>
<p>Corvallis is only about an hour up I5 from Eugene and is the only other flagship in the state of Oregon. A study that included both the UO and OSU and its school of Engineering would have been more informative and more generally applicable to other state university systems.</p>
<p>
</p>
<p>I said nothing of the sort. I never referred to “women” at all, nor did I say anything about students performing better than their SAT’s would predict in upper level math or physics classes. I said that the Univ. of Oregon study found that many students in ALL OTHER majors performed much better than their SAT scores would predict, including students with GPAs of 3.5-4.0. READ the study! (And READ my posts – you are pulling an invention out of thin air that isn’t even remotely related to anything I’ve posted).</p>
<p>
I’d just point out that a 550 math score on the SAT is 66% percentile for women, 53% for men; 600 is 79% for women, 68% for men. So it’s not as if students who score below 600 are idiots – among women especially, the 550+ student are still in the upper third of test takers.</p>
<p>It’s not necessarily math ability, either – there is some suggestion from studies that males are more willing to guess in a multiple choice test setting, and that may give them an advantage, especially in math, in a timed-test setting. In other words, there may be many women who are scoring below 600 not because they can’t do the math, but because they are in fact doing the math, but not able to finish all the questions. (See [Freakonomics</a> SAT Strategy by Gender: Men Guess, Women Leave it Blank](<a href=“Freakonomics - The hidden side of everything”>SAT Strategy by Gender: Men Guess, Women Leave it Blank - Freakonomics) )</p>
<p>So part of the math/physics issue might be more of a self-fulfilling prophesy issue – those who score under 600 on the math perceive themselves as being weak in math because of jerks who characterize above-average scores as “terrible” – and may be reluctant to enroll in math intensive courses. We don’t know from the data what level of support the Univ. of Oregon offers to students who need it.</p>
<p>
Gosh darn, and I have a BS in Civil Engineering, an MA in mathematics, and a phd in physics, all after getting a paultry 600 on my math SAT. I guess no one told me I was destined to fail :). Feminine failure BTW.</p>
<p>calmom, a hundred thousand apologies. I mixed up your comments with those of Momsquad, producing a jumble of my own invention.</p>
<p>Let’s look at those overachievers. The researchers included 4,084 students. They identified 96 overachievers, mostly clustered in Sociology (38) and Spanish (29). Of the Biology (521 students), Chemistry(84), Math (60) and Physics (77) students, exactly one was classified as an overachiever. In other words, the overachievers weren’t in science and math.</p>
<p>It’s debatable whether those Spanish students were actually overachievers-- the Spanish department had a high average GPA, so they apparently were handing out As more liberally than other departments.</p>
<p>It doesn’t matter whether a 550 Math SAT is the 66th percentile for women. If you can’t do math any better than that, you’re not ready to be a physics or math major. That’s not to say we can’t improve opportunities for young women to become math majors and physics majors-- but throwing unprepared young women into math classes they can’t handle is not the way to do it.</p>