In the face of a chronic shortage of U.S. scientists and engineers, only one question really matters: Why is it so difficult to train scientists?
Proposals for addressing the shortage of scientists are plentiful and well known. A 2005 National Academies of Sciences and Engineering and the Institute of Medicine report, "Rising Above the Storm: Energizing and Employing America for a Brighter Economic Future," has had wide circulation. President Bush's Science and Mathematics Access to Retain Talent (SMART) grant program tops a set of new government efforts to encourage young people to study science. These initiatives, like most proposals for addressing the problem, concentrate on assisting large institutions to produce more science graduates.
Intuitively, it appears sensible to focus there, on the largest producers of both college graduates and PhDs. However, intuition can be misleading. In this case, quite counterintuitively, the most efficient producers of new scientists turn out to be small colleges and universities.
Why is it so difficult to train scientists? And why do small colleges do it better than the large universities?
It is difficult because the sciences are inherently cumulative fields. If a student cannot master the skills and knowledge in Chemistry 101, he or she cannot succeed in Chemistry 201. Small colleges do a better job of training scientists because their students are more likely to persist in comparison with science majors at large universities.
A giant lecture course can offer little help to a student who stumbles on one unit in the course. Some large universities have surmounted this problem: Virginia Tech's "Math Emporium," for example, shows how large institutions can successfully use technology to provide tutorials for students.
The National Center for Academic Transformation, led by IT in education expert Carol Twigg, has also helped universities use technology to increase students' grasp of course content and to reduce attrition in large, lower-level courses. A few highly selective, large universities consistently graduate large numbers of scientists, but too many large institutions have dismal attrition rates.
Meanwhile, at highly selective liberal arts colleges such as Oberlin (Ohio), Swarthmore (Pa.), Mount Holyoke (Mass.), and Williams (Mass.), many graduates-disproportionate numbers-major in the sciences and go on to earn doctoral degrees. These small colleges graduate a much smaller absolute number of students than large universities, but the percentages of their students who pursue and complete PhDs in the sciences is very high.
Critics dismiss what happens in elite small colleges as not relevant to national planning. They argue that the students at these colleges are among the nation's brightest and would do well anywhere, and that the small scale of these efforts will never meet a national need.
What critics fail to see is that many other, less affluent, less selective independent colleges also produce disproportionately large shares of PhD scientists. There is something about the format of small colleges and their approach to introductory science teaching that prevents the high attrition rates that are typical of larger institutions.
A promising approach to answering the question of why scientists are so difficult to train, therefore, may be to draw upon the many smaller, independent institutions with proven track records of successfully graduating science majors.
One recent CIC program illustrates how this strategy might work. Between 2001 and 2005, CIC conducted an annual competition that recognized outstanding achievement in undergraduate science education. Between 30 and 70 institutions each year competed to win a $10,000 prize. The program, funded by the Philadelphia-based Russell Pearce and Elizabeth Crimian Heuer Foundation, awarded two to four prizes each year. (The records of that program are available at www.cic.edu/projects_services/archives.)
CIC was delighted that the overall quality of the proposals was very good. In fact, it proved difficult to choose winners from many worthy proposals. And many of the colleges that won the awards are not well known.
To cite accomplishments of a few of the winners (figures are from the competition period, 2001-2005):
Juniata College (Pa.) tripled the number of chemistry majors in just a few years, with 60 percent going on to graduate school in chemistry. And 70 percent of all chemistry majors at Juniata are women.
Allegheny College (Pa.) created a neuroscience major in 1996. By 2003 it was attracting 35 majors per year. Since the program began, 47 students have entered graduate programs in neuroscience.
Hendrix College (Ark.) ranks 24th in producing PhDs in chemistry among all four-year colleges and universities in the United States, measured as the ratio of PhDs in chemistry to undergraduate enrollment. Almost all its chemistry majors have entered graduate school. Hendrix ranks 33rd in producing physics PhDs.
Roanoke College (Va.) experienced a growth in the number of chemistry majors from an average of 15 per year before 2002 to an average of 25 since then, 75 percent of whom go on to graduate school.
Small colleges are also having great success in preparing female scientists.
Juniata's impressive percentage of women chemistry majors is just one example. Other colleges with noteworthy records in this area include some women's institutions such as Mount Holyoke College, Cedar Crest College (Pa.), and Spelman College (Ga.). In fact, Mount Holyoke has been among the top colleges nationwide in graduating women who earn doctoral degrees in the physical sciences, geosciences, mathematics, and computer sciences. In biological sciences, Mount Holyoke is first in the nation.
These successes have come to small colleges because their full-time faculty members teach introductory courses. Those professors are more readily accessible to students than part-time teaching assistants or adjunct faculty at larger institutions. No matter who is teaching, the prevailing pedagogy at small colleges makes it much more likely that a student who is having difficulty will get help before the effects of not understanding lead to failure.
Unfortunately, the National Academies report is silent on these differences between small colleges and large universities in producing scientists. If scale matters, scaling up failure makes no sense. The fact is that graduating a half a dozen chemistry majors here and a half a dozen physics majors there will actually solve the problem.
Moreover, even some academically strong, big universities produce very few science majors who pursue PhDs. Here is one example: The University of Wisconsin-Madison has 28,000 undergraduates; Oberlin College is one-tenth the size. Yet those institutions graduate similar numbers of physics majors who later receive doctorates in physics: in 2001-Oberlin, 5 and UW Madison, 9; in 2002-2 and 2; in 2003-2 and 4; and in 2004-1 and 4.
There is an important efficiency and cost-effectiveness argument that is missing from most rhetoric of the public policy gurus about the problem of too few scientists. If state and federal investments in producing an undergraduate science major lead to fewer science PhDs per dollar when spent through a state university compared with a small college, the most cost-effective public policy for increasing the supply of U.S. scientists would be to rely more heavily on the small colleges and invest in them more.
The National Academies are not alone in overlooking the role of independent, smaller colleges and universities in preparing scientists.
The American Chemical Society (ACS) has been weighing a proposal to change its standards for the accreditation of college and university chemistry departments. ACS had proposed to raise the required number of full-time faculty members to a minimum of five. This change defies common sense: Some of the colleges with the best track records of graduating chemistry majors who earn PhDs have done so with departments of four faculty members or fewer.
Earlham College (Ind.) is a prime example of this. The ACS should look at results, not inputs. The backlash to the ACS proposal has been considerable, and properly so.
Why not offer awards through the National Science Foundation and private foundations to colleges and universities on the basis of their demonstrated success in graduating science majors and in the percentages of those majors who enter PhD programs in science?
Permit the money to be used for basic, top-priority campus needs such as new scientific equipment and generous scholarships for science students.
And keep the process of applying very simple. Making it easier for colleges and universities that have proven track records in producing career scientists to do even more is a highly promising approach to meeting the need for more U.S. scientists.
With better information, smarter policies, and wiser investments, it should be possible to change the question: Why was it so difficult to train scientists?
Richard Ekman is president of The Council of Independent Colleges, www.cic.edu.