Don't Put Up Barriers to Foreign Students
I am responding to the letters on foreign physics students published in the July 1996 issue of APS News. The free-marketeer's stance on the bumper crop of physicists is: if the U.S. ejects all its educated foreign nationals, or stops letting outsiders in at all, then instead of importing some employees, U.S. commerce will more vigorously export workplaces. But this will take a long time.
An argument against putting up barriers to foreign grad students that immediately benefits native-born students also exists. U.S. students benefit from the job-hunting advice of their foreign-born fellows. Speaking for myself, I found talking with newly-employed Ph.D.s whose temporary visas forced them to be extremely resourceful about looking for work to be helpful during my own recent job search. I suspect that there would be many fewer jobs available for physicists outside of physics right now were it not for the reputation that physicists (often foreign-born) have established outside of physics, often by obtaining a second degree in another field while earning a physics Ph.D.
Mr. Moodenbaugh's statement that U.S. nationals are "restricted" to the U.S. job market while foreigners are not is perplexing. First, Americans are not so restricted. Many European countries have special funds for foreign postdocs, and there are opportunities in Japan also. Second, a temporary visa imposes a disadvantage to U.S. job-hunting, a fact which may not be unrelated to the larger fraction of foreign nationals choosing postdocs over permanent work.
If there is a problem of oversupply of physicists the solution might be found by imposing some sort of accreditation program to limit the number of Ph.D. granting institutions. I don't know if the disease is serious enough to warrant so drastic a cure.
A Contrasting View of Standardized TestsThe article "Fighting the Gender Gap," in the July 1996 issue of APS News, is shamelessly one-sided. Arguments in favor of standardized tests are barely mentioned, but Pamela Zappardino, executive director of FairTest, a group that opposes standardized tests, is quoted extensively. Let me cite several distortions and instances of sloppy reasoning.
Ms. Zappardino states that the SAT accounts for at best 16 percent of the covariance in freshman college grades and is therefore a poor predictor. This is quite misleading, because of the "restriction in range" effect. A very selective institution admits students with almost uniformly high SAT scores. The fact that SAT scores correlate only modestly with grades in THAT group does NOT mean that students with much lower test scores would do well.
The article mentions that girls' scores improve when a test is not timed, while boys' scores change very little. So what? In the real world, people must solve problems under time constraints. There are few exams in college or graduate school where time is unlimited.
It is stated that foreign students do well on the GRE subject test, "although their performance in graduate school isn't any better or worse than their American colleagues." It is not stated how "performance in graduate school" is defined. At the University of Illinois at Urbana Champaign, where I am a graduate student in physics, foreign graduate students are much more likely than Americans to pass the qualifying exam at the "free shot" immediately upon entry. The qual exam is not multiple choice and allows ample time for completion, but the superior performance of foreigners on the timed, multiple-choice GRE seems to carry over.
A female physicist states that she needed to learn standardized test skills, not physics, in order to get an adequate score on the GRE physics exam. To prepare for the physics GRE subject exam, I simply reviewed the syllabus of the test and earned a very good score. This is just an anecdote, but so is her story.
Professor Georgi of Harvard says that standardized tests favor "idiot savants." He cites as evidence for this extravagant claim the cases of two female undergraduates he knew. He forgets that one should not generalize based on a sample size of two.
It is reported that Harvard's graduate admissions committee relies more heavily on letters of recommendation and the personal essay than the GRE in making decisions. Has the committee calculated the correlation coefficients of these measures with first-year graduate student grades or completion of the Ph.D. degree? Critics of standardized tests never bother to demonstrate with hard numbers that the admissions criteria they prefer do a better job. How can one compare students from different schools, much less different countries, without standardized tests?
The final quote from the anti-test lobbying group notes that the number of colleges not requiring standardized tests has grown. Admissions directors may decide to drop the SAT for reasons other than validity - it is good for one's ego to believe that one is making a decision that cannot reduced to "mere numbers".
The differences in boys' and girls' SAT scores is small, but the differences among ethnic groups are substantial. The consequences of this differential have been studied, but none of this research find its way into APS News.
The SAT does not underpredict the academic performance of blacks in college - in fact, it is a better predictor than high school grades (Thomas Sowell, Inside American Education, p. 128). Admitted black students to MIT have scored in the 90th percentile nationwide on the math SAT, but in only the 10th percentile among MIT students. Nearly 25 percent of these students failed to graduate (p. 144). At Berkeley, only 22 percent of Hispanic students preferentially admitted had graduated five years later, while more than half of those admitted under normal academic standards had graduated.
Before making self-righteous statements about the need for ethnic "diversity" in physics, one should estimate the pool of "under-represented minority" students qualified to major in physics in college. In 1985, the number of American Indian, Black, Mexican American, and Puerto Rican students scoring above 600 on the math SAT, was 3,929, compared to 11,903 Asian Americans (p. 142).
Physicists say that the critical thinking skills developed by training in physics will carry over to other fields. The essays published on "diversity" in APS News do not support this claim.
University of Illinois
GRE Scores Serve a PurposeWith regard to the article on gender bias in the GREs (APS News, July 1996), I agree with the opinion that the GRE should not be used to keep capable students out of graduate school. However, the GRE has a more positive function: It allows students from unknown schools to compete for positions in graduate school. The problem with letters of recommendation from unknown professors is calibration. If I say "This is the best student I've seen in ten years," it helps the student if the GRE scores verify and quantify that statement.
John W. Dooley
Straight (Back) TalkWe must bow to Francis Slakey ("Straight Talk," APS News, Aug./Sept. 1996), who could perfect his writing style by mere improvements in word choice and sentence length. The rest of us, brought up on television scripts, e-mail, and writings of the technically precocious but incompletely educated, have to start much further back. We must learn the many basics that were kept from us in the formative years of grades K-6, such as that split infinitives are not necessarily incorrect but are necessarily inelegant; that a gerund requires the possessive case; and that agenda, bacteria, criteria, and data ARE PLURAL. When the bricks and mortar of language are missing, the perfecting of its beautiful architecture will have to come later. Staffing technical publications with well-read, vigilant copy editors would be a big help.
Z's Fight BackThis letter is a comment on your "Straight Talk" article in the Aug/Sept issue of APS News. Although I found myself in agreement with most of it, I must say that, for obvious reasons, I am totally opposed to the proposal to drop the letter "Z" from the alphabet.In fact, I have believed for some time that the order of the alphabet should be revised, and should start "zabcde_".
An Age-Old Question...Your report on Don Vandenberg's new data on the age of globular clusters left me somewhat confused ("New Cluster Data Puts Universe at 13 Billion Years," APS News, July 1996). I do not understand why "the universe cannot be older than its older stars," as stated in the article. For instance, my Mom is older than I am. Likewise, it would appear to me that the universe actually MUST be older than its older stars, but it cannot be any younger.
However, in the same way as my Mom may not delight in sharing how much older than I she is, so also the Universe is unwilling to easily let us know how much older than Population II stars it is. If I missed the point altogether, I would appreciate any clarifications you could provide to me.
Portland State University
The Editor Responds:
As Professor Pinto noted in such a good humored way, we erred and the statement should have indicated the apparent paradox of the universe being younger than its oldest stars. This reminds me of the limerick: There once was a woman named Bright, whose speed was much faster than light; she eloped one day in a relative way, and conceived the previous night.
©1995 - 2014, AMERICAN PHYSICAL SOCIETY
APS encourages the redistribution of the materials included in this newspaper provided that attribution to the source is noted and the materials are not truncated or changed.