Best and Brightest Part Two: Are Sciences and Engineering Graduate Programs Still Attracting the Best Students?

Report to the

Alfred P. Sloan Foundation

Grant No. B-1998-64

presented by

Purpose - This is a report to the Sloan Foundation on a research project originally entitled, “Identifying and Understanding Certain Quality-Related Trends in Advanced Training in Science and Engineering,” supported by a grant from the Foundation to the University of Washington, over the period November 1, 1998 to September 30, 1999.  The University of Washington partnered with the Commission on Professionals in Science & Technology.   This project, a follow-on to an earlier Sloan project carried out by CPST,[1] was concerned broadly with investigating evidence that U.S. graduate programs in the natural sciences[2] and engineering (S/E) may be attracting less than their historic numbers or shares of the highest ability students, particularly those of U.S. origin.  More specifically, the first purpose of the project was to assess what is known or could be learned about trends in the quality of students bound for S/E graduate programs from the existing literature or relatively accessible sources.  The second purpose was to identify additional data sources and analyses from them that could be usefully carried out over a longer time frame to add significantly to policy-relevant knowledge about this issue.

 Project activities and major findings

 1- Review of the pertinent existing literature - This review was selective to avoid repetition of the related review completed for the original Best and Brightest report.  In short, although there is peripherally relevant literature, quality trends in the incoming graduate student pool in S/E have not been thoroughly examined empirically for many years.

2- Analysis of trends in destinations of GRE top-scorers- with the cooperation of the Educational Testing Service analyses of trends in Graduate Record Examination General Test scores were designed and executed.  These analyses covered selected years from 1988-89 to 1997-98 and focused on U.S. citizen examinees who indicated intent to pursue graduate studies in the natural sciences or engineering, compared to examinees headed for other fields.  Some data on trends in examinees and scores for the major professional fields that use other tests (medicine, law, and business) were also obtained.  The key findings from these analyses are:

 •  The number of U.S. citizens taking the GRE General Tests has declined significantly since the early 1990s (Figure 1).

 •  Since the benchmark year for most analyses herein, 1991-92, numbers of U.S. citizen examinees indicating intent to pursue graduate studies in natural sciences or engineering are down more sharply than numbers headed for other specified fields (-16% vs. –7%).  Declines range from –10% in physical sciences to –39% in mathematical sciences.  Only biological sciences among the S/E fields (+27%) is an exception to this pattern of declining interest.  Among non-science fields, the major gainer is the health fields (+29%) (Table6).

 •  With the exception of biological sciences, the S/E fields generally show small declines in mean GRE quantitative and analytical scores, proportion of high (>700) and very high (>750) scorers, and share of all such scorers expressing intent to enter the field (Table 1).

 •  Combining the trends in proportions of high scorers on the quantitative scale with the declines in numbers of examinees reveals that S/E graduate studies (except biological sciences) have lost significant absolute numbers of high scoring U.S. citizens while non-science fields as a whole have seen little change in the number of high scorers.  These findings are very similar whether a 700 or 750 cutoff score is used: numbers of high scorers headed for S/E fields are down 21-22% from 1991-92 to 1997-98 and numbers headed for non-science fields are essentially stable (Table7).  In sum, since 1991-92, percentage declines in numbers of high scorers headed for S/E are about one-third greater than decreases in total examinees headed for S/E and are nearly twice as large as the aggregate decline in GRE test-takers.

 •  The number of high scorers on the GRE quantitative scale intending to enter graduate study in the health professions has increased by 75-78%, depending on the cutoff score used (Table 8).  The vast majority of the gains in the health fields are in applied professional fields, led by speech/language pathology and physical therapy, not in basic science disciplines.  (Note that few prospective M.D.’s take the GRE.)

 •  Data on recent trends in graduate-level test-takers in other major professions (medicine, law, business) do not show clear signs of a “brain drain” toward these fields.  Law school applications and enrollments are down significantly in the 1990s.  Medical school applications (unduplicated) grew steadily through 1996-97 before falling off significantly in the last two years, but enrollments have been stable since 1993-94.  Mean MCAT scores gained slightly through 1997-98.  In business, GMAT examinees and mean scores have increased modestly in the 1990s but MBA awards have leveled off.  Absent a massive national effort to match early individual academic and testing records over a number of cohorts, it is not possible to determine definitively if the types of high ability students who once enrolled in S/E graduate programs are now enrolling in medical or business schools.  The test score trends and flattening graduate professional enrollments in these fields suggest it is unlikely there was a large-scale recent shift in this direction.

 •  The number of women U.S. citizens scoring above 700 on the GRE quantitative scale and intending graduate study in one of the natural sciences or engineering has increased modestly during the 1990s, although there are significant differences by area of S/E (Table9).  Overall, the gains in numbers of high scoring women are dwarfed by the decreases in the number of high scoring men headed for S/E fields.

 •  The numbers of African-Americans, Hispanic-Americans, and Asian-Americans with >700 GRE quantitative scores intending graduate studies in S/E have all grown by impressive percentages (in the 60-70% range) during the 1990s.  But in numerical terms their combined gains are considerably less than one-third of the decline in white U.S. citizens with such high scores headed for S/E graduate studies.

 3–Analysis of trends in S/E graduate enrollments by department quality and citizenship- Data from the National Science Foundation’s annual surveys of graduate students by field[3] for the years 1993-1997 were merged with the National Research Council’s latest published assessment of department quality[4] in order to examine whether enrollment declines were disproportionately concentrated in high- or low-rated departments.  (These years span the period from when S/E graduate enrollments peaked through the latest available NSF survey year, a period of considerable enrollment decline in engineering, physical sciences, mathematical sciences, and computer science.)   Disciplines studied were those where the NSF and NRC discipline categories could be closely matched: chemistry, physics, geosciences, mathematics, computer science, chemical engineering, civil engineering, electrical engineering, and mechanical engineering (Table 1).[5]  Key findings are:

 •  In the majority of disciplines examined no evidence of disproportionate declines in U.S. citizen students (including permanent residents) in top-rated departments as compared to lower-rated departments was found.  Exceptions were chemistry, chemical engineering, and electrical engineering where negative changes in U.S. citizen enrollments were 6-10 percentage points greater for top-rated departments than for all departments.

 •  Non-citizens’ (temporary residents) numbers generally fell along with those of U.S. citizens.  The only exceptions were in computer science and electrical engineering where non-citizens increased (though by only 1.6% in electrical engineering) while citizens declined.  (In top-rated departments only there were small gains in temporary resident students in civil and mechanical engineering as well as electrical engineering and computer science.)  Thus, with a few possible exceptions, there seem to be signs that American graduate programs in S/E can no longer draw on a “bottomless well” of international students.

 4– Obtained access to a data base on newly-enrolled graduate students in several S/E disciplines at AAU institutions – These data include citizenship and GRE scores of entering graduate student cohorts at 25-30 AAU universities over 1989-96 (in biochemistry, mathematics, and mechanical engineering), or 1992-96 (in chemical engineering and physics).  Analyses of changes over time in the distribution of GRE scores among these students should help confirm or disconfirm the suggestion of recent declining interest in S/E graduate studies among high ability students indicated by the GRE analyses (which are based on test-takers’ stated intentions about graduate study).

 5– Obtained access to a data file on postgraduation study plans of graduating seniors from COFHE institutions, 1982-1998– COFHE is the Consortium on Financing Higher Education, a group of highly selective private colleges and universities.  The COFHE-board approved analyses cover four senior cohorts from this period and six COFHE schools with survey response rates exceeding 50% in each year (about 4,000 respondents per year).  The analyses should give a good picture of changing patterns in the choices about postgraduate study and field of high ability students, including S/E majors and students with high grade averages, from these top schools.

 6– Reported on these activities and preliminary findings therefrom – These findings and plans were reported at a day-long workshop attended by a group of leading government and private researchers and analysts of S/E human resources issues held at the American Association for the Advancement of Science in Washington on August 11, 1999.  (List of attendees is attached.)  Valuable feedback was received at this session on the analyses and their presentation, as well as about dissemination strategies.

 7– Prepared and had accepted a conference paper on framing the policy implications of the apparent decline in interest in S/E graduate studies among high ability students– This paper, "The Best and Brightest for Science: Is There a Policy Problem Here?", will consider problem framing issues and policy options on the assumption that further analysis will show clear evidence of declining interest and participation in S/E by top students.  The paper will be presented at the annual research conference of the Association for Public Policy Analysis and Management (APPAM) in Washington in November 1999.  Comments received should be helpful in developing subsequent policy analytic papers for publication.

  Further Plans.  

 These are detailed in the accompanying proposal for further research and dissemination activities.   In short, the following tasks would be undertaken in the proposed new project:

 1– Update through 1998-99 the analyses of GRE score trends of high scoring examinees indicating intent for graduate study in S/E, compared to other fields.  More attention will be given to the GRE analytical scale scores, in addition to the quantitative scale scores.

2– Focusing only on the physical sciences, mathematics, computer science, and the engineering disciplines where the NRC and NSF discipline categories closely match, trends in first-year graduate enrollments by department quality and citizenship over the period of enrollment decline will be analyzed.   Also, the apparent negative patterns in enrollment shares of the highest-rated departments in chemistry, chemical engineering, and electrical engineering will be further investigated.

3– As described in the previous section, trends in the GRE score distributions of newly-enrolled graduate students in 25-30 AAU universities in several S/E disciplines will be analyzed to ensure that evidence from students’ intent at the time they register for the GRE is consistent with evidence from students who actually enrolled.

4– Also as described above, shifts over time (at least four cohorts over 1982-1998) in the postgraduate study plans of high ability S/E majors from very selective COFHE colleges and universities will be analyzed.

5– Once the findings from these analyses are clear, disseminating them via appropriate electronic means, publications, and presentations at conferences and forums will be a prime focus.