Senior Vice-President for Research and Innovation, Professor of Theoretical Physics, Michigan State University

Thursday, December 15, 2011

Human capital, globalization and physics 101

In the past few years we have seen a large influx of undergraduate students from China. Since UO non-resident tuition is about \$20k per annum, these students must come from relatively affluent families there. The conventional wisdom among professors familiar with China is that most of these kids are slackers -- they didn't do well enough on the gaokao to be admitted to a top Chinese university. How good are "slackers" from China? Judge for yourself.

Below is the score distribution from the course I taught this fall, physics 101 for non-majors (about 200 students total). The black histogram is non-Chinese, the red is Chinese, most of whom, judging by their names, are from PRC. Why was this analysis necessary? Because I noticed the score distribution was very different from previous times I had taught the course. About 20-30 PRC kids scored higher than what is usually the highest score. (Click for larger version.)

Here are two exam problems.

An ant slowly pushes a box of mass .1 kg and coefficient of friction .1 a distance of 10m, moving at constant speed. Calculate the work done.

A satellite orbits the Earth at a distance of 3 Earth radii from the center. Compute its gravitational acceleration.

Jerry Lin said...

I wouldn't say they are slackers since there are lots of them in my university who are truly among the top academically. Given the rigorous curriculum in Chinese HS system, those students probably (definitely?) have learned most content in Physics 101 thus are better prepared for undergrad courses than most North American HS graduates.

Yan Shen said...

And I'm sure this pisses off the David Versaces of the world, who'll compensate by trying to assert that these Chinese aren't as alpha as him and his white investment banking buddies. ;)

sargento said...

Are the US kids that excelled in HS physics getting bumped up into higher level courses?  Your graph is pretty convincing evidence that something is going on, but it makes me wonder if the Chinese students that would have aced the AP Physics BC exam are starting to show up in your entry level physics class in increasing numbers while non-Chinese students are being more effectively sorted into higher-level classes.

steve hsu said...

This is more like Conceptual Physics, but I do make them solve problems involving algebra.

steve hsu said...

The PRC kids in my class are, by and large, not the academically strongest ones in their home country. Almost all of those kids would be going to upper tier universities in China, for a fraction of the tuition they are paying here.

bfgc said...

Ah. Thanks!

steve hsu said...

Because we got a big influx of PRC students who are qualitatively different from American students in their ability to do basic physics.

bfgc said...

Er, "Why was…taught the course." was me quoting you.

If you're comfortable editing the CSS on the page, you might want to add something like

.dsq-comment-text blockquote {
margin-left: 0;
margin-right: 0;
border-left: thick solid gray;
font-weight: normal;
}

to make blockquotes more obviously blockquotes.

David Stern said...

We have huge numbers of Chinese undergrads in Australia and our tuition is \$A22-26k at my uni depending on discipline (\$54k in medicine). Grad tuition is \$30k. I would think that some come in order to study in English and get an advantage over students from Chinese universities in that dimension. Who would you hire if you were a multinational firm? A grad of Beida or ANU?

Jerry Lin said...

My guess is that most American freshman who are majoring in non-STEM programs have only taken HS-level math, and we all know it's ridiculously easy. However, their Chinese and Asian counterparts have been through tougher curriculum in math/science while studying in their home country. Asian HS curriculum tend to place emphasis on math & science courses regardless of your choice of college majors. A PRC student majoring in Economics probably had gone through the same rigorous training in math/science courses as a PRC student majoring in engineering. (Correct me if I'm wrong!). Therefore, even though the humanities/social science kids from PRC are not the best among Chinese when it comes to math, they're still much better equipped than their American counterparts.

I've also seen this in engineering as well, although to a lesser extent because most incoming Canadian students have completed some AP/IB courses prior to university. Generally the gap disappears in sophomore year...not quite sure about other faculties though.

In all seriousness, you're comparing the performance of students from PRC who may very well be pursuing some STEM degree (if not, then what degree are they looking to get?) with the performance of students in a Physics for Poets class, and find it in any way remarkable that the students from PRC crushed the Poets?

Yan Shen said...

I noticed that Steve focused on Chinese Americans only, versus everyone else.

Are there any data on how Chinese Americans stack up against Korean and Japanese Americans? I've always had the sense that even when compared against other East Asians, Chinese Americans are distinctly stronger academically.

IIRC, on a per capita basis, Korean and Japanese Americans are virtually non-existent amongst Intel Science Talent Search and IPhO competitors compared to their Chinese American counterparts. Japanese Americans are non-existent in the USAMO/IMO. Koreans are over-represented amongst USAMO/IMO qualifiers compared to their US population percentage, but still trail Chinese Americans in terms of per capita representation, (probably by a factor of at least 2 or so). And while South Korea sends almost as many students to the US to study as does China, there has been only 1 Korean Putnam Fellow, while there have been numerous Chinese American Putnam Fellows.

Zhengzheng Zhou said...

With all due respect and I don't mean to imply anything, but those questions are exactly the type you would get in 9th grade in China 25 years ago.

steve hsu said...

Almost none of these PRCers are STEM majors. This course doesn't satisfy any STEM requirements -- for that they have to take a harder physics course. The two students who came regularly to my office hours this term were a business major and an architecture major.

What is surprising to me is that even though this is physics for poets, the top few kids in the class in a normal term are usually quite sharp (we're talking about 200 kids here!). But about half of the PRC students this term got *higher* scores than the usual top score in the class. No matter how much background you have in physics, that requires some g.

steve hsu said...

I'm not surprised that people in some countries might learn this material in high school. It's a scandal that in the US people can graduate from university without having learned basic physics. On the other hand, I doubt most adults in China or Korea or Japan could solve these problems -- they've pretty much forgotten it all.

bfgc said...

If you're on a STEM track, anything less than calculus-based physics isn't on your radar. If you're a PRC exchange student in a STEM major, the only reason I can think of to take UO's Physics 101 is to get familiar with most of the relevant vocabulary in a laughably easy course before taking Real (i.e. calculus-based) Physics sometime later.

Be that as it may, the typical student in a Physics for Poets class is, as you surely know, almost innumerate. Anyone with a modest talent for mathematics and science will beat them hollow. God only knows what selection factors are working with respect to the students from PRC who happen to choose U Oregon. It's just apples and oranges.

It would be a far fairer comparison to pit the better students at U Oregon in physics (or another STEM major) against their PRC student counterparts. While this too would be something of an apples to oranges comparison, at least there would be a clear selection for mathematical talent on both sides.

steve hsu said...

What do you mean by "modest talent"?

What fraction of the population could get the high score in this class of 200 kids (assuming no PRCers)? I would guess the math ability average in the class is the same as in the general US population. That would be slightly below the average for university students. The top kid is then probably about +2 SD in the general population. (Note, as I mention elsewhere, there are kids in the class who have had AP physics but are non-STEM majors trying to get an easy grade.) Nevertheless, *half* of the PRCers outscored this hypothetical +2 SD person.

As I note at the link below, the top 10 percent of Shanghai students scored above the US 99th percentile on the math portion of PISA. These physics 101 results suggest to me that the PISA results are not unrepresentative.

http://infoproc.blogspot.com/2010/12/shanghai-pisa-scores.html

ytrewq123 said...

hnhjk

ytrewq123 said...

David Versace (Davver from HS) brings no data to the table. His very point is made much better by Siserune on College Confidential, where a fair amount of evidence was put forth, IIRC, that Chinese-American over-representation diminishes (but doesn't vanish) the higher you go up the research ladder (measures by awards/faculty hires, etc). This is something I've seen in my field too, so Steve's remarks on faculty hires in theoretical physics will be an interesting data point.

Also note, many of the awards have low median ages for winners, so there's no serious time lag. Links will be posted if needed.

steve hsu said...

Sample size in theoretical physics is small, but I believe the answer is probably yes. It is also (I am told) yes in pure math and I am 100% sure the answer is yes in some fields like EECS. Note though this is including Asian immigrants from PRC, Korea, etc. who come here for grad school. If you just count A-A's then the numbers are perhaps too small to make sense of, but see the link below.

http://infoproc.blogspot.com/2010/10/g-math-ability-and-their-population.html

ytrewq123 said...

What's in the table from the link? Asian-Americans in one portion of the table (USAMO, etc) and Asia-born Asians + Asian Americans in another (faculty hires)? Trends would be nice too, but I'll go with your observation. I'll try to tabulate some evidence for EECS and post in some time.

4lpha toronto said...

"An ant slowly pushes a box of mass .1 kg and coefficient of friction .1 a distance of 10m, moving at constant speed. Calculate the work done.

A satellite orbits the Earth at a distance of 3 Earth radii from the center. Compute its gravitational acceleration."This seems a little too easy for an undergrad course. Surely success here is more closely tied to memorizing the relevant formulae than g.

steve hsu said...

I *give* them the formulas, but they have to understand the concepts to apply them. You haven't been around "normal" (non-STEM) students trying to do quantitative stuff much, I guess.

I find your argument quite unconvincing.

Look, there are at least the following major selection factors at work here:

1. the selection of PRC students who happen to choose U O
2. the selection of PRC students who go to U O and choose Physics for Poets
3. the selection of US students who go to U O
4. the selection of US students who go to U O and choose Physics for Poets

My overall point is, it's nearly impossible to reckon what the expected effects of such a set of selection factors might be on the distribution of mathematical talent in Physics for Poets.

In particular, I'd expect that the selection of US students who go to U of O would tend to exclude disproportionate numbers of students at the highest range of mathematical talent, because most such students would have more exclusive options. In addition, the selection of students in the Physics of Poets class would most definitely exclude disproportionate number of students at the highest range of mathematical talent, even within U O. What is the ultimate product of such selection factors? Who knows?

Now you may actually be in a position to settle some of this debate: you might have access to the math SAT scores of your students. If so, you might show us the distribution of those scores for US students (and for the PRC students?) so that a fairer comparison might be made. Perhaps then we could make some reasonable inferences.

Actually, if you could rummage up the math SAT scores for both the US students and the PRC students, it might make for an interesting cross cultural study: do the PRC students do better than expected given their SAT scores, or worse, or the same?

steve hsu said...

I'm not sure which argument you are talking about.

I'm not trying to make inferences about national populations as a whole. As you point out this depends on lots of selection effects. My main surprise in looking at this data was that these rich "slacker" kids are actually good students. Before I graded the final I hadn't looked carefully at how this subpopulation was doing in the class, but I didn't actually expect them to be doing well. Now I suspect that even the not so stellar kids (but admittedly from the upper class) in China are getting a pretty rigorous HS education, which is after all what PISA results say.

Perhaps you don't have a lot of experience dealing with data. I've looked at UO aggregate SAT data and SAT distributions by major. You get nice normal distributions with slightly shifted averages. You can consult the data mining paper I wrote on this (use google) to see the means and SDs. It is very unlikely that physics 101 students deviate much from the distributions in the non-STEM majors here. (Physics 101 is actually considered *harder* than the other intro science courses that fulfill the gen. ed. requirement here, such as Geology 101 or Astronomy 101.) So yes, I do have some idea about the distributions, and the +2 SD estimate I made elsewhere is almost certainly correct.

steve hsu said...

PRC students and other foreign students don't take the SAT, as far as I know. They are admitted through some different process at the moment.

This might explain my prior that these students would be not very good:

I'm sure that, for the most part, the distributions you would be seeing would appear roughly normal. But the particular situation you're describing here is, obviously, one that concerns behavior at the very extreme right tail of a distribution in a particular class. Rather than presuming that that tail behaves just as a normal curve would predict, you would do much better to produce if you can the actual distribution of, say, math SAT scores in that class. If the PRC students don't take the SAT, then that's unfortunate. But it would be very useful to see how those who do take the SAT perform in that class, and how those scores might be reflected in ultimate performance.

steve hsu said...

We're not only looking at SAT, but also "grit" (Big 5 Conscientiousness). We administered grit surveys at the beginning and end of the term.

FYI, the PRCers systematically *underestimate* their grit relative to Westerners. This has been reported before in other studies -- Asians work harder but still consider themselves lazy, Westerners exaggerate their work ethic. (An advantageous psychology if you want to climb your way to the top...)

But from previous work I can already tell you that SAT-grade correlation will be .3-.6 as it is in almost all UG courses (we did this in our earlier paper). It looks like grit will add a bit of predictive power.

Just as one example of how, say, the SAT distribution might seem normal given the selection factors, but might not reflect important underlying differences: it may be the case that the students at UO (from the US) who have very high SATs do NOT tend to have the same level of conscientiousness as in the larger population of students with equivalent SATs. If they did, they would be more likely to have gone to a more exclusive school.

It's in general a little hard to believe that there isn't a significant cutoff effect at universities like UO: why wouldn't one expect that a disproportionate number of the highest achieving students would be siphoned off by elite institutions? How do the overall numbers even add up if they aren't? Perhaps UO is an exception to this rule -- and I'll grant that a public university, because of its lower expense, is more likely to represent such an exception -- but I'd like to see hard evidence, by careful examination of the right tail at UO.

steve hsu said...

At +2 SD there are so many kids that the elite schools cannot even make a dent in the population. Do the math. At +3 SD you see a reduced number at state schools these days, due to the effects you mention.

http://infoproc.blogspot.com/2010/10/elite-universities-and-human-capital.html

steve hsu said...

I just checked for you. The top non-PRC kid is about +1.5 SD (top 6 percent or so) in the SAT-taking population, which is probably just a bit less than +2 SD in the general population (say, top 3-4 percent).

The SAT-M correlation with score in the class is about .55; Grit and score correlated at around .3

Grit and SAT-M are almost uncorrelated. So some linear combination of the two will give the best predictor. I'll post these results "officially" on the blog at some point :-)

George Shen said...

Calling these student slackers may be harsh.  They may not be the most capable Chinese students academically speaking but coming to the U.S. is still very big deal.  Most of them should have the typical Asian work ethic.  I expect most of these student work harder than their American classmates.  Another point is a small portion of them may even be ivy caliber.  If a Chinese student has the choice of going to Harvard without any assistantship/scholarship vs going to a state university with full assistantship/scholarship, I think the student would very likely choose the state school (unless he/she is from a very rich family where money is no object.)  Relating to my personal experience, it is entirely possible (i gave up the chance to go to U. Penn because my family can't afford the tuition and instead went to a state college while my roommate gave up Columbia because the state school gave him full assistantship).  Note that international students are not qualified for any student loan or similar type of financial aid. \$40~50K/year tuition plus books, board and other living costs is a lot of money even for very affluent Chinese families.  20 years ago when my generation came to the States for education (mostly for grad or post-grad study), undergrads were very rare.  And back then the Chinese student body tended to be very uniform - 99% STEM majors (and with some college education in China already).  Nowadays, more and more are going for finance, business, accounting, and law in addition to STEM majors.  Here in Northeast, I am seeing the trend of younger and younger Chinese students coming to the U.S. for pre-college education. Some of them are still in elementary school, which is quite shocking to me....

Alex Lamb said...

In general, are the students from China who go to elite American universities (i.e. top 15-20 or so) stronger academically than the students who attend top Chinese universities?  I go to Johns Hopkins, and I think that students from China are generally quite good, but for classes that aren't focused on memorization they are at or near the class average.

Anonocorp said...

Steve is the answer to the first one .981 j ?

Anonocorp said...

Steve is the answer to the first one . 981 j ?

davidwbudd said...

Steve:

a couple of questions on your plot.

1) Is it possible to disentangle who in the red part of the histogram are "Chinese" as opposed to say "Chinese-Americans?"  Your comment about "judging from the names, are from PRC" leads me to believe that you are sorting based on upon blunt criteria such as surnames (for example, distinguishing "Hsu" from "Xu").  It's a small point, but it does introduce an additional source of variance into your crude model

2) The phenomenon on display here may have nothing to do with who is a "slacker" (dubious to begin with, given the political choices surrounding whether to send one's children abroad from China to the US for education), and have everything to do with artefacts of the distribution.  Here, there is an obvious right-censoring, in that students cannot exceed 100, and in fact, most will "pass" your course.  The data are not symmetric.  When you throw into the mix the obvious fact that the biggest difference between the "Chinese" (however you construct them) cohort and the "non-Chinese" cohort has to do with variance, I am not sure what sociological conclusions are actually reasonably possible.

I am virtually certain you've seen this argument in re: why men dominate in mathematics (a result of the higher variance among male vs. female scores), but here, I wonder if the selection bias of having no "dummies" coming from China to a public American university is worth considering?  I presume that the University of Oregon, a public school, has somewhat 'relaxed' criteria for applicants from Portland, when compared to applicants from Peking, and thus, the dim bulbs from China are not coming in the same numbers as the dim bulbs from within Oregon.

I would say in short, that the "slackers" from China are largely driving the cabs or cleaning the subway stations in Shanghai, not attending American colleges.

davidwbudd said...

Steve, let's be serious here.

The claim that "most adults in China" could not solve these problems because "they've pretty much forgotten it all" seems laughable, when you consider the nature of public education in China.
In a nation with 1.4 billion (or whatever the current datum is now, my figures are out of date), I think it's a stretch to assume that people who are preparing for the Gaokao represent "most" of the school-age population.

Americans have, in my experience, a very, very distorted view of what life is like for most Chinese who are not in the eastern, urban cities....

davidwbudd said...

Is this really so, Professor Hsu?

I honestly don't know.  How many slots are there in the entering class at Bei Da (Peking University?)  How many kids apply?  China is a nation with nearly a billion and a half people.  If one or two or even 10 per cent (a dubious figure, really) are admitted to the "upper tier universities in China," that still leaves several millions without a seat.

It's certainly possible, based on sheer volume of numbers, that those students without a seat at, say Bei Da or Qinghua University or Shanghai Jiao Tong, could very well be among the strongest ones at home, but just due to the numbers game, unable to get in to the "right" university.  The competition for a coveted seat at Harvard (acceptance rate around 10 per cent) is nothing compared to how difficult it is rumoured to be to gain entry to Peking U.

The idea that these kids are "slackers" seems ridiculous to me.