Nonlinear Psychometric Thresholds for Physics and Mathematics

<p><a href=“http://arxiv.org/pdf/1011.0663[/url]”>http://arxiv.org/pdf/1011.0663&lt;/a&gt;&lt;/p&gt;

<p>The authors, Stephen D.H. Hsu and James Schombert of the Department of Physics at the University of Oregon, studied SAT scores and academic performance.</p>

<p>Basically, they found that no student entering with SAT Math < 600 was able to make a 3.5 GPA majoring in physics or pure math, and that the “fraction of students with high upper GPA increases monotonically with SAT-M”. This was different from other majors, where some students with low SAT scores still achieved highly.</p>

<p>Although engineering students were not studied (University of Oregon does not have engineering), the authors “expect that similar results also apply to highly mathematical fields of study such as some areas [of] engineering or informatics.”</p>

<p>That is interesting since the SAT uses fairly rudimentary math. Perhaps it shows whether the math basics are strong enough to do well?</p>

<p>I think Erin’s Dad probably hit the nail on the head. If your basic math skills are not strong enough to score at least 600 then you aren’t going to do well in math, physics, chem or engineering.</p>

<p>I agree, from my experience as a physics professor at a Tech school, I know that the Math ACT score is the best predictor of success in engineering and physics. The rough equivalent of the 600 SAT is a 26 in the ACT and we find that those students with scores lower than this generally struggle and change majors to something not as mathematical.</p>

<p>The Math SAT 1 test is not so much a test of what you have learned in Math but instead evaluates your level of Math reasoning ability which is probably something you are born with and is a predictor of not only how well a student will learn high school Algebra but also advanced Mathematics found in upper division college courses.</p>

<p>Just by memorizing facts such as the Pythagorean Theorem, Quadratic Formula and formulas for the areas of various geometric shapes one can probably get a 500 on the test. To get over 600 though requires understanding of Math concepts and to surpass 700 probably requires demands a student to show real insight into Math concepts.</p>

<p>On the other hand, while being able to get a score of 600 or 700 on the SAT-M may be a necessary condition for success as a math or physics major (if the paper is correct in general), it is unlikely that it is a sufficient condition for success as a math or physics major, since ability to handle junior/senior/graduate level math courses with proofs (or perhaps even freshman calculus) may not be measurable with type of basic high school math questions found on the SAT-M.</p>

<p>Also, coaching may increase someone’s SAT-M scores beyond what s/he would otherwise get, including by use of techniques specific to the SAT but not applicable outside of standardized tests. (Perhaps those with better math skills would have been able to derive these techniques by themselves without external coaching.)</p>

<p>But the observation that the authors make does appear to be consistent with two things:</p>

<p>a. Low selectivity four year schools with relatively narrow ranges of student abilities are unlikely to offer engineering degree programs, perhaps because few students there are interested in or are able to handle the course work.</p>

<p>b. The [threatened</a> shutdown of physics majors at some low selectivity four year schools](<a href=“Texas Threatens Shutdown of College Physics Programs for Low Graduation Rates - Scientific American”>Texas Threatens Shutdown of College Physics Programs for Low Graduation Rates - Scientific American) may be a result of similar lack of student interest and ability to handle the course work (in the case of Texas, the threshold is 25 graduates over 5 years).</p>

<p>Great article & link – but I think you’ve missed the real story by focusing on the circumstance of the exception rather than the rule. From the article:</p>

<p>

Source: <a href=“http://arxiv.org/PS_cache/arxiv/pdf/1011/1011.0663v1.pdf[/url]”>http://arxiv.org/PS_cache/arxiv/pdf/1011/1011.0663v1.pdf&lt;/a&gt;&lt;/p&gt;

<p>Physic and “pure” mathematics were the ONLY areas where there seemed to be a consistent correlation between SAT scores and achievement. This suggests to me that colleges would do well to dispense with the SAT entirely except for math & physics majors – or certainly to reduce reliance on the SAT for other majors.</p>

<p>Or perhaps that suggests that non-math majors at the University of Oregon aren’t very rigorous…</p>

<p>Most of the paper was discussing the “exception” (physics and pure math majors), which was the actual focus of the paper.</p>

<p>As far as reducing reliance on the SAT goes, it would be ideal if it were not needed. The problem is that there is some need for a common measure across widely varying high schools, both for actual comparison purposes and for reducing the incentive for high schools to have more grade inflation. A nationwide set of curriculum standards like Canada has would eliminate the need for standardized testing (including AP and IB – if courses like calculus offered in high schools were standardized and graded consistently, then their course grades could be used by universities for placement into more advanced courses). But getting that implemented would be politically impossible.</p>

<p>" Low selectivity four year schools with relatively narrow ranges of student abilities are unlikely to offer engineering degree programs, perhaps because few students there are interested in or are able to handle the course work."</p>

<p>When it comes to Engineering, the state of Oregon is sort of a special case. Since their establishments in the 19th century there has been a policy that the University of Oregon in Eugene would be the primary campus for Liberal Arts and the Natural Sciences such as Physics while all Engineering programs, as well as Agriculture, Forrestry and Mining, are offered by Oregon State University in Corvallis.</p>

<p>My concern would be that some might take the results to indicate that the student with a sub-600 SAT-M score should be discouraged from even making an attempt at a STEM field, rather than encouraged to build their skills with some additional pre- courses before jumping in at the level of their classmates.</p>

<p>

</p>

<p>I was not referring to University of Oregon with the previous comment. I was more observing that, in looking at the ABET accreditation lists, the availability of engineering degree programs tends to drop off and disappear as one goes lower down the selectivity scale of the CSUs. Also, small, not very selective private schools tend not to have engineering degree programs.</p>

<p>Of course, this would not apply to broad range schools, like public schools in low population states that have only a few schools. There, while the minimum standard for entry may be low, but because the school serves every level of in-state student (i.e. a broad range of student abilities), it has enough students to fill engineering degree programs.</p>

<p>Canada does not have a national standard but a provincial one. Two western provinces, British Columbia and Alberta, I believe, have province-wide exit exams, and thus probably the most vigorous standards in the country. In Ontario where I grew up, exit exams for university applicants were in effect until I reached senior high (late 60s). It was abolished but the universities could make up their own if they so choose. None did because they all became intoxicated by governments grants which were based on number of students enrolled.</p>

<p>A professor told a friend of mine that senior calculus, one of three math courses on offer, is used as an “invisible sieve” by Ontario universities. I find this to be true; it seems to be a requirement for all competitive programs that I know (our universities are not competitive, but certain programs in certain universities are). One university, Waterloo, went a step further and asks all applicants to their most selective programs to do the Fermat math competition. In effect, it functions as an additional senior course. The school is also known to keep a list of high schools, showing how the students from that school do when they get to Waterloo.</p>

<p>We do not take standardized exams for college, but the standard seems quite consistent. What is not consistent is college grading. Because we use a curve to grade students, where you stand relative to your classmates is a function of the quality of your classmates to a large extend.</p>

<p>Canuckguy does the system of community colleges work to allow students who don’t have the requirements for university coming out of high school to reach the level they need to be at to compete for programs at those universities?</p>

<p>Although it seems quite logical to me that a student with a low SAT/ACT math score would not be able to succeed in a mathematically oriented major, I am extremely reluctant to go one step further and claim that the low score and the difficulty with the major are due to a lack of aptitude for math.</p>

<p>I think that in some cases – perhaps a substantial proportion of the total – the problem with math is due to inadequate learning of basic concepts. Often, students who did not fully grasp the content of their basic high school algebra courses, for whatever reason (not keeping up with the homework, poor teaching, too many absences, or many other possibilities, not all of which are related to a lack of aptitude for math) manage to complete high school math with passing grades (sometimes even As), despite the deficiencies in their understanding. But at some point, often in college courses that require math, the deficiencies catch up with them.</p>

<p>It is certainly possible that there are some students who do poorly on math measures in high school (grades and SAT-M scores) because of poor preparation, not because of poor ability.</p>

<p>Such students, if interested in studying something math-intensive, may find that going to community college the best option, where they may be able to first take the remedial courses that will give them a solid foundation for college level calculus, rather than going to a four year school as a freshman and bombing out of freshman calculus and/or being “behind” normal progress in their majors.</p>

<p>

</p>

<p>In Ontario, students can transfer courses taken in community colleges into a degree program. In general, they can transfer a third to a half of all courses taken provided they are academic in nature and completed with a certain grade point average. Universities generally do not accept any credit transfer until the student has completed at least one full year of community college though.</p>

<p>There also exist specific agreements between certain colleges and particular universities that would allow a student to transfer as much as “all” courses taken from one institution to another. If this is not confusing enough, some community colleges also offer post-degree certifications for university graduates who are looking for career-related training in a given field.</p>

<p>Community colleges are not the same as junior colleges. They are called colleges of “applied arts and technology”, with the specific mandate of training students for a given career or occupation.</p>

<p>I really believe that any student can learn anything. But not every student can complete a degree in a field such as math or physics within a four-year time frame, starting from his/her entry point. </p>

<p>I agree with both sylvan8798 and Marian. It would be great if some formal mechanism could be created to allow students to “come up to speed,” before jumping into the courses for the major full-throttle–even if by “courses for the major,” we are just talking about the introductory calculus-based physics course.</p>

<p>ucbalumnus,</p>

<p>I found it interesting that Texas might phase out low performing physics programs. The same thing may be happening here in California.</p>

<p>My son, a high school senior, tutored a student in physics 400A (the hardest upper division undergraduate class-classical electromagnetism- not to be confused with AP Physics E & M) at San Diego State University this semester. They met once or twice a week, an hour at a time (My son took the class last year).</p>

<p>During the few weeks they didn’t meet, she scored a 19 on a test. She was able to pull up her next test to something in the 80s, so hopefully, finished the class with a B or C. I’m not sure how many were in the class but it must have been a low number. </p>

<p>The second semester class, 400B isn’t even being offered. I have no clue what these students are supposed to do but it’s clear that there are very few physics majors at SDSU, and probably fewer that graduate with decent gpa’s.</p>

<p>Physics is just plain difficult and it does take a certain aptitude and interest to persevere, I guess.</p>

<p>Very interesting, thanks for posting. From their previous study I found this comment that might help to explain the discrepancy between SAT/GPA in Math/Physics vs. SAT/GPA in other subjects:</p>

<p>

</p>

<p><a href=“http://arxiv.org/pdf/1004.2731v1[/url]”>http://arxiv.org/pdf/1004.2731v1&lt;/a&gt;&lt;/p&gt;

<p>It is odd that they make no mention of the gender differential for “overachievers” in their discussion of threshold SAT (M) scores for success in Math and Physics. At most campuses a significant majority of students majoring in math or physics are male. It is possible that the lack of correlation of SAT to GPA in other majors is due to a higher proportion of female students in these majors.</p>