Reshaping the Graduate Education of Scientists and Engineers

America's system of graduate education in science and engineering has set the international standard, especially in preparing students to work successfully at the cutting edge of research, and it must continue to do so. Graduate schools have also, increasingly in recent years, contributed to filling the nation's growing need for advanced expertise in diverse nonresearch positions.

Nevertheless, the committee believes that there is room for substantial improvement in graduate education and that some immediate changes are needed in programs, information, and attitudes. These changes are recommended, in part, in response to contemporary stresses. In many important fields, employment in basic-research positions has not kept pace with expanding graduate enrollments, and this has led to unmet expectations among many graduates who have aspired to such positions. The available evidence on unemployment rates indicates that demand by less-traditional employers is growing fast enough to absorb most graduates. However, we note broad criticism from many such employers concerning graduates' immediate suitability for entry jobs--criticism that is often based on a belief that students are too specialized, in view of the variety of tasks that they will confront, and that it is hard for them to adapt to the demands of nonacademic work. With only one-third of new PhDs expected to enter the academic tenure system, the needs of these alternate employees should be given more attention.

There is also a broader concern: Although it is clear that human resources are the primary key to the nation's strength in science and technology, we have not, as a nation, paid adequate attention to the graduate schools as a system for meeting the full range of needs for advanced talent in science and engineering. That is perhaps seen most clearly in the fact that the United States has effectively lacked human-resources policy for advanced scientists and engineers. In effect, human resources have been taken for granted as a byproduct of our policies for the support of research. The simplifying assumption--both inside and outside the university community--has been that the dominant function of graduate programs is to produce the next generation of academic researchers. It is time for a fuller recognition, by academics and policy officials alike, of the changing way that graduate education in science and engineering contributes to the wide array of national needs. For many of these needs, it is a career in professional service, applied research, development, or consulting that graduates will find open to them.

The committee concludes that improvement of three kinds is needed. First, graduate programs should add emphasis on versatility; we need to make our students more adaptable to changing conditions. This is mainly a matter of local initiative by the universities themselves, but there is a supporting role for government, too. Second, much better information should be routinely provided to students and their advisers so that students can make more realistic career decisions than is now practical. Third, there needs to be a deliberate national reconsideration of graduate education so that the open policy questions, the current information gaps, and the contemporary stresses are systematically addressed by a suitable blend of university, industry, professional society, and government. Those improvements can be made without disruption of the traditional commitment to excellence in basic research that has been, and must continue to be, a hallmark of the US system of graduate education.

Although the universities are primarily responsible for implementing those changes, national and state government, industry, business, and others can help by providing opportunities to gain experience and exposure to a variety of occupations via internships, alternative certification programs, etc. We do not minimize the difficulty of effecting reform in a system as complex and diffuse as that of US universities. But we already have many relevant examples of the application of local imagination and initiative. We believe that most university leaders will find it in their own interest to reshape graduate education to meet students' career needs better and to ensure universities' vital role in the nation's steady progress toward a knowledge-based society.

5.1 NATIONAL OPTIONS

The committee arrived at its preferred national strategy--emphasizing versatility and information--after considering alternative approaches.

For example, it might seem tempting to remove any apparent imbalance between supply and demand by adjusting student enrollment. The reasons not to move toward anything like national enrollment quotas have been presented above (see Section 4.1). We found these arguments as persuasive when applied to discipline or fields as would be implied in suggestions to cut physics enrollments by X% or to increase the numbers of master's degrees in microbiology by Y%. Identifying the "right" number of graduates is chancy, to say nothing of administering nationwide compliance.

Another version of this suggestion is that we should set out to adjust the mix of master's degrees and PhD degrees that are awarded. Some question, for example, whether PhD-holders are not overeducated for the positions they fill--especially for nonresearch jobs--and whether a master's degree would suffice. But can one actually conclude that the PhD experience is unnecessary for such positions? From the information gathered by the committee (see Section 2.7 and Appendix F), the opposite seems to be true. Employers themselves appear to be seeking the intellectual standards, resourcefulness, and initiative that come with the successful completion of original research in a PhD program. The complexity and sophistication of more and more positions appear to require the qualities gained in the advanced coursework and original-problem formation of graduate programs.

Another possibility is the creation of a new form of degree--a "different doctorate," perhaps, or a degree that is intermediate between a master's and a doctorate. In theory, a new degree could be better tuned to the class of nontraditional jobs that PhDs are increasingly filling--for example, it might require less-intensive or different types of research and dissertation experience and as a consequence take less time to complete.

In practice, however, we are convinced that this approach would not work well. The proposal is reminiscent of the doctor of science (DSc) degree that some institutions have offered with the hope that it would catch on as the preferred degree for doctoral students who seek nontraditional careers. A key point is that employers report that they value the research experience required for the PhD degree. Without ready demand for a newly introduced degree, students risk investing substantial effort only to find that they receive a diploma regarded as inferior--one that critics might think of as "PhD-lite." It is more realistic, we conclude, to adapt the PhD degree than to try to invent and introduce a hybrid degree.

In opting for a strategy of making graduates much more versatile and informed, we believe we have a solution that allows the system to self-adjust continuously in a way that does not depend on the accuracy of an assessment of the number of graduates needed in the national aggregate or in particular fields. Thus, for example, if better-informed students conclude that the PhD is inappropriate or unnecessary for the jobs they want, enrollments will decline accordingly.

5.2 TOWARD GREATER VERSATILITY

Once enrolled, a graduate student might find many reasons to select a relatively narrow subject for intensive study. A student might be fascinated by a particular field of knowledge and see specialization as the surest route to a research position. If the selected field aligns with the research interests of a professor, the student might have an exciting and educationally enriching chance to work as an assistant on a path- breaking research team; this can enrich the student's educational experience immeasurably and can provide fresh ideas and energy to the research team as well.

The disadvantages of overspecialization in graduate school, although not immediately apparent, are real for both the student and the nation, whether or not the student becomes a researcher. Excessive concentration in a particular subfield can limit a person's later research contributions and affect later career choices. It is difficult to gauge whether a specialty chosen early in graduate school will be desirable in the job market or still be in the exciting forefront of research when the graduate years conclude. And midcareer changes might be desirable later, whether or not the person starts off in a research position; too narrow an educational experience makes later changes difficult, especially in the direction of nontraditional types of employment.

5.2.1. To produce scientists and engineers who are versatile, graduate programs should provide options that allow students to gain a wider variety of academic and other career skills.

Graduate programs should offer options that equip students for a wide array of eventual career opportunities. These options are beginning to appear on some campuses but need to be expanded to promote students' ability to adapt. Adaptability can be enhanced in two ways.

First, graduate students can benefit from a wider variety of academic preparation. For students who choose to enter a research career within the discipline studied in graduate school, it is important to have a grounding in the broad fundamentals of the field and some personal familiarity with several subfields--a breadth that might be much harder to gain after graduation. For such positions, if students become overspecialized in graduate school, they can later suffer from an inability to recognize and enter newly emerging kinds of research. For nonresearch positions, too, one's value to employers is likely to be enhanced by breadth in a related field, gained through coursework or, better, a minor. For example, a chemist might minor in computer science or a biologist in mathematics. Second, there is value in experiences that supply career skills beyond those gained in the laboratory and classroom. More students should, for example, have off-campus experiences to acquire the skills desired by an increasing number of employers, especially the ability to communicate complex ideas to nonspecialists and the ability to work in teams of interdependent workers.

The internship in off-campus settings is one option that needs to be expanded. Project-oriented teams in corporations provide potential opportunities for collaborative interactions and exposure to challenging practical problems. Joint industry-university projects should be explored as part of some students' preparation, with the possibility that students complete their dissertation work off campus. Such projects also acquaint faculty members with the needs and organizational cultures of nonacademic employers.

Graduate programs should also expand on-campus opportunities to allow their students to attain a broader range of career skills. Outstanding people in a large variety of careers can be brought to campuses for presentations and made available to students. And communication skills might be sharpened through organized presentations to people outside the discipline and through older students' mentorship of younger students, for example.

5.2.2. To foster versatility, government and other agents of financial assistance for graduate students should adjust their support mechanisms to include new "education/training grants" that resemble the training grants now available in some federal agencies.

The United States has a sound tradition of investing generously in the graduate education of scientists and engineers. Federal agencies, private foundations, industries, and other granting agencies can support the efforts of both students and their graduate programs to enhance the versatility of new graduates.

Most federal support for students is provided through research assistantships. Research assistantships have proved important for bringing graduate students into federally funded research projects, and they will continue to remain a major form of federal assistance.

Research assistantships can bring great educational benefits to students, but they are not specifically designed to enhance the versatility of graduate students. Assistantships are usually administered by the faculty researchers who receive the research grants, so the needs of the funded projects themselves are likely to be paramount in guiding students' work assignments.

We recommend increasing the relative emphasis on education/training grants, a concept adapted from the training grants[1] that are now awarded in selected agencies. We recognize that, in a period of constrained funds, increased emphasis on education grants could reduce the number of research assistantships that are available.

The essential features of education/training grants would be:

• The grants are awarded, on the basis of competitive proposals, to departments and programs, rather than to individual faculty members.

• Evaluation criteria feature the proposer's plan to improve the versatility of students, both through program and curriculum innovations and by upgrading faculty advice to acquaint students with the full range of future employment options.

The experience with training grants at the National Institutes of Health over the last few decades shows that this type of mechanism can be successful in establishing productive interdisciplinary programs and in encouraging students to enter emerging fields of research. We encourage the growth of this program in a way that further enhances students' command of subjects and skills needed by nontraditional employers.

The National Science Foundation (NSF) is reviving the training-grant program that it supported until the 1970s. In its initial stage, only a few hundred of the 20,000 students supported by NSF grants are receiving traineeships. By means of its training grants, NSF hopes to serve a variety of distinct purposes, including promotion of emerging fields, interdisciplinary programs, industry involvement, and the participation of women and minority-group members. We encourage deliberate expansion of this effort for the special purpose of fostering broader graduate experiences, which could well include industry involvement and emerging research that is particularly valued in expanding job markets. Some of the new education/training grants should be administered as demonstration grants for particularly innovative programs; if successful, demonstration grants should be expanded at NSF and replicated in other agencies.

Other federal agencies could use education/training grants effectively. The Atomic Energy Commission and the National Aeronautics and Space Administration once used what they called training grants to help to augment the national pool of nuclear and aerospace engineers; today, the Environmental Protection Agency and the Department of Energy, for example, could use this type of grant to induce more young researchers to address issues in environmental protection and remediation. Similarly, one goal of the recently launched technology-transfer programs--the Advanced Technology Program and the Technology Reinvestment Program--is to foster science-based technologies in industry. Those programs now operate by means of cost-shared research grants, but a worthy national objective could be the development of human resources as a component of technology transfer, and a portion of program funds could be devoted to education/training grants.

Education/training grants need not be restricted to federal agencies. For example, corporate sponsors could design grants to expose students to industrial research, development, and problem-solving. Foundations and state governments could fund graduate education/training grants with the aim of producing secondary-school teachers and science-curriculum specialists.

In summary, these suggestions are intended to encourage a better balance among the alternative types of grants: continuing fellowships for the top research-oriented students, expanding education/training grants to catalyze the development of innovative programs, continuing a substantial number of research assistantships, and continuing institution-supported teaching assistantships.

5.2.3. In implementing changes to promote versatility, care must be taken not to compromise other important objectives.

Introducing measures to enhance versatility will require care and imagination. They should be instituted in a way that allows universities to attain other ends--enumerated below--at the same time.

Maintaining Local Initiative. The changes that we recom-mend will likely come from local institutional initiatives and should show considerable local variation. We would not expect or want all universities to offer the same or similar options to their students. Programs should build on their own strengths and interests. Some universities and departments might want to focus on particular career paths (e.g., secondary-school teaching or subject fields of interest to local business and industry). Others might emphasize the development of particular career skills or cross-disciplinary combinations. Different university programs across the United States play different roles now, and that should continue.

Maintaining Excellence in Research. We are not recommending that all students be prepared for nonresearch careers. Opportunities appear to be growing in nonresearch jobs now; but we will continue to need many of the best students to dedicate themselves to research in academic and nonacademic settings, and they will need the depth and quality of graduate experience that basic researchers have long enjoyed.

Furthermore, we are not espousing what some call vocationalism. The idea is not to slot every student into a particular career path and then "train" him or her accordingly. Among other problems, that would bind students to jobs that can change or decline in number while they are in graduate school. What is needed is not additional specialization. We need a graduate system that is well tuned to the central feature of contemporary life: continuous change. Change is happening both within the research world and outside, and work in both spheres requires constant readiness to adapt. Our objective, therefore, is a breadth of experience so that graduates can keep career options open and have the capacity to switch career tracks both at the beginning of and throughout their professional lives.

Controlling Time to Degree. The recommended changes should not be construed as additional requirements that would in themselves extend a student's time in a graduate program. The steadily lengthening time to degree--and, more important, the time to first employment--is already too long, for whatever reasons. Many ways of fostering versatility, including several noted above, can easily be introduced within the time that graduate students now spend after registration. An industrial assignment, for example, might replace--and not supplement--an on-campus research assignment.

We are aware of some strain between broadening the graduate experience and controlling its duration. Both solutions are needed, even if considerable administrative energies are required. Although long average time to degree is often decried, faculty and administrators have not generally made the disciplined effort that is needed to tighten graduate programs.

Whatever the nature of a specific graduate program, it is crucial to establish the principle that each student is the responsibility of a department, not of a single faculty member. Thus, a small faculty group (including the adviser) should meet often with each student working for a PhD degree; this faculty group, not the student's faculty adviser acting alone, should determine when enough work has been accomplished for the PhD degree.

Some observers have suggested fixed limits--5 years, perhaps, which is about 2 years shorter than the current averages--for a doctoral-education career. In the abstract, it is not obvious why such a period, which would allow 2 years of coursework and 3 years for a dissertation, should not suffice for most full-time PhD candidates. However, we are not prepared to espouse strict limits, in part because today's more-diverse student population requires flexibility to accommodate family and other personal factors.

However, we do believe that the "Two Plus Three Plus X" model for doctoral education ought to be evaluated and debated within the academic community. The idea is that preparation for a career in research has three discrete phases. The first, which should require no more than 2 years (assuming adequate preparation and suitable adjustment for part-time students), is for developing a broad command of the field. The second, for which the norm might be 3 years, is for making an original contribution to research as reflected in the dissertation. The third, for refining research skills and specialized knowledge that might be required for a first research position, should be left to postdoctoral assignments. Our concern is that the second and third phases are often merged in current practice.

We urge institutions to set their own standards on time to degree. This could be done at the departmental or program level, and it could accommodate the features of individual disciplines and the character of the student body. The standards should be clearly communicated to students and advisers, and responsibility for enforcement should be accepted by university administrators.

Attracting Women and Minority-Group Members. It is essential that a fair share of the best students be attracted to each discipline in science and engineering. If it appears that the numbers of women and minority-group members are low in particular fields, an effort must be made to determine whether there are barriers to entry, including issues perceived as barriers by members of the group in question. If so, steps to encourage increased participation should be devised and implemented.

5.3 TOWARD BETTER CAREER INFORMATION AND GUIDANCE

The committee is concerned about the lack of organized and timely career information and guidance that is available to students and their advisers--especially about the absence of reliable information on the less-conventional career paths of scientists and engineers.

Faculty attitudes have sometimes favored academic research careers, and some students have come to feel that other career paths were less worthy.

During their graduate years, students by themselves have access to little more than anecdotal information about career options. Many proceed through these years presuming that research jobs will be available in sufficient numbers to allow them some freedom of choice. They might see no urgency to investigate alternative careers when actual job entry is several years away and few sources of information about such careers are available. Their faculty advisers, having spent most of their time interacting with other academic researchers, might have little personal knowledge about alternatives and thus no basis to advise students about them. Former students who have taken nonresearch jobs are often less visible to their graduate departments than former students in traditional positions and are too seldom available as career models for current students. Departments generally do not adequately track information on nonacademic nonresearch employment so that it will be available to potential and current students.

The lack of reliable and timely information impedes the adjustments of the supply of graduate scientists and engineers--both upward and downward--to the demands of the job market.

The committee stresses that departments should not assume that the burden of learning about realistic career options rests with students. They have an affirmative obligation both to know what the full range of options is and to impart that knowledge to students.

5.3.1. Graduate scientists and engineers and their advisers should receive more up-to-date, accurate, and accessible information to make informed decisions about professional careers. Broad electronic access to such information should be provided.

We recommend that a national database on employment options and trends be established. The database information, intended for both students and their advisers, should include, by field, data on career tracks, graduate programs (including financial aid), time to degree, and placement rates.

Given the diversity of the information for which there is a need, it is clear that the responsibility for providing data must be shared by all partners in the graduate-education enterprise, including the universities, federal and state agencies, and professional societies.

The rapid development of information networks--collectively called the Internet--makes it possible to organize employment and career information so that two important principles are maintained: the information made available in the information system retains decentralized "grass roots" and therefore more currency than information previously assembled into central compendia; and timely information is available where it is most needed--in the hands of the ultimate consumers, doctoral students, graduates, and their mentors and advisers. In the past (for example, as recently as the downturn in aerospace employment in the 1970s), it would not have been possible to construct an employment-information system that recognized those principles. The new technologies can and should be deployed to improve nationwide access to accurate, germane, and timely education and employment information.

The National Science Foundation should coordinate the federal participation needed to organize the database. However, it is preferable that the database be designed and managed within the research community itself so that it has accurate and timely information that is credible to students and other users, some of it collected from university departments and professional societies. A national organization that covers the many fields of science and engineering could be a catalyst in establishing the database.

5.3.2. Academic departments should provide employment information and career advice to prospective and current students in a timely manner and should help students see career choices as a series of branching decisions. Students should be encouraged to consider discrete alternative pathways when they have met their qualifying requirements.

Graduate students typically devote years of intense effort to their education, and they deserve thoughtful, individual advice about career options. Many faculty members find that advising and mentoring are among the most important and most rewarding of their responsibilities. But more can be done to make sure the advice that is given is both pertinent and complete.

Advice for students should not be limited to the personal knowledge of the faculty member who serves as a student's adviser. Departments should both understand and convey the employment prospects of their graduates. One way to start is to track--perhaps with the assistance of alumni affairs offices--their own past graduates systematically.

Use of information in the national database recommended above could help. We hope, in addition, that some of the demonstration effort funded under a program of education/training grants would allow departments to invent and try other novel means of improving the advice that students receive.

In the past, when most students were destined to become professors, graduate school was more accurately construed as a step on a simple career ladder. We are concerned that that perception is still held in some places. Departments should help students to conceive of their time in graduate school as a series of deliberate decision branches.

Academic departments can focus attention on the importance of career choice at two particular points. The first is the application stage. It would be helpful if more departments, in describing their programs to potential students, routinely provided more data relevant to career choice, such as location of job placements, salaries, and unemployment rates for the department and the discipline as a whole. Departments should report on the careers of all their graduates and provide the relevant information to prospective and current students. Such information could help to prevent unrealistic expectations among students.

The second point is the beginning of the research phase, which usually begins with the passage of the qualifying examination for doctoral students. That is when departmental advisers can help students to evaluate each of three distinct options:

• Some might conclude that a master's degree is sufficient, given their aspirations and the current employment demand for PhDs.

• Some might elect to proceed toward a PhD and try for a position in research.

• Students interested in nontraditional careers could design dissertations that meet high standards for originality but require less time than would be customary for a career in academic research.

The last of those options is often neglected. Implementing the first, which is typically undervalued, might require some reshaping of the master's program to ensure that those who switch from the doctoral program receive--and are perceived to receive--something more than a consolation award. Among other advantages, this counseling approach will require that the leading faculty members come to respect the alternative careers that are available to their students.

Professional societies are often in the best position to gather nationwide employment information on scientists and engineers by field. Some--the Graduate Student Packet of the American Physical Society and American Institute of Physics is a good example--have made impressive starts in this direction. University departments should help to communicate their results to students and advisers.

5.3.3. The National Science Foundation and National Research Council should continue to improve the coverage, timeliness, and analysis of data on the education and employment of scientists and engineers to support better national decision-making about human resources in science and technology.

5.4 TOWARD IMPLEMENTATION OF A NATIONAL POLICY

In preparing our last report, Science, Technology, and the Federal Government: National Goals for a New Era (COSEPUP, 1993), it became clear that no coherent national policy guides the education of advanced scientists and engineers, even though the nation depends heavily on them. That recognition was an important stimulus for the present report.

A casual observer might say that federal policy should simply be to fund the best research and that sound graduate education is an automatic byproduct. There is some validity to that view, but we believe that it is time to reconsider the stewardship of our human resources separately. The nation's graduate programs must prepare scientists and engineers for contributions not only to the nation's basic research, but also to a wide array of other national objectives. Simply to let the development of human resources be guided by the workings of the relevant labor markets is an inadequate policy, given the long lead times required to make career decisions.

At present, there is neither the conceptual clarity nor the factual basis for us to lay out a coherent policy. We are concerned that many prevailing views are obsolete or obsolescent.

5.4.1. A searching national discussion that includes representatives of government, universities, employers, and professional organizations should examine the goals, policies, conditions, and unresolved issues pertaining to graduate-level human resources.

Graduate education is the responsibility of private and state-supported universities; of the federal and state governments, which support many students; of the corporate sector, which increasingly employs those who complete it; and of public and private foundations, which support its conduct and study its workings. All those parties need to be involved in a continuing reconsideration of graduate education and its national purposes.

Three kinds of issues are suggested as worthy of a searching national discussion:

• National goals and policy options.

• System characterization.

• Contemporary issues.

5.5 CONCLUSION

In conclusion, the committee believes that science and engineering graduate programs will be improved if

• Science and engineering programs are made more flexible and provide more options for students so that they are more versatile.

• Graduate-student support is shifted to education/training grants.

• Time to degree is controlled.

• More women and minority-group members are attracted to them.

• Better and more-timely career information and guidance are provided while diversity and excellence in research are maintained.

NAS Home Page | NAP Home Page | Reading Room | Report Home Page