Reshaping the Graduate Education of Scientists and Engineers

The committee solicited comments from more than 1,000 persons: graduate students, postdoctoral students, professors, university administrators, industry scientists and executives, and representatives of scientific societies. More than 100 responses were received, of which about half came from industry.

The committee noted several trends in the responses. There were general support for the current concept of PhD education (with a variety of suggestions for improvement, as outlined below), support for closer ties between the universities and industry, support for improved student counseling, and opposition to artificial limits on enrollments.

Although the call for comments was not a formal survey, we thought it would be useful to summarize the information gathered to indicate the diversity of views held by those with a stake in the success of our system of graduate education.

Industry and academic administrators generally responded favorably to the current concept of PhD training. Most comments affirmed US superiority in graduate education, but with the observation that there is always room for improvement. No substantial dissatisfaction was described. The following statement typifies the general sentiment: "We may see some specific difficulties in the relationship between academe and the profession it is intended to serve, but the structure itself is generally sound."

However, some concerns were expressed about the level of additional education that is needed before recent graduates become fully participatory employees. Here is an example of a response from one major industrial employer that hires 150-400 advanced-degree people into its laboratories each year from many universities and in many disciplines:

Even "the best of the crop" take anywhere from 6 months to 2 years to become good, productive industrial researchers. Most recent graduates, particularly those who have not summer-interned, do not have the foggiest idea of what industrial research is all about. Some even think that using or developing technology to do something useful is not research and if it is a product that makes a profit, is even slightly dishonorable.

Almost everyone expressed support for better preparation of graduate students for teaching. Respondents generally cited numerous reasons for this improvement, including the following:

• Students pay high tuition for instruction, and they deserve better. Courses taught via recitation do not help students learn or graduate students teach.
• It is wrong to assume that anyone working on a PhD is automatically able to teach.
• Students aiming at careers in academe should take formal teacher- training courses to learn pedagogy as well as they learn research.

For example, the following comment is from a graduate dean and provost:

I have long been concerned about the teaching expectations of graduate students--all graduate students, not just in the sciences and engineering. How we can expect that an individual will intuit teaching skills is an amazement. While teaching is somewhat an art, there are many skills and techniques that need to be learned before an individual should be turned loose to teach a course. We do our graduate students no service, and certainly provide no service to the teachers, if we expect them to function in that capacity. ...They also need to be prepared to be academic advisors. It is not enough to walk into a class and conduct that experience. If graduate students are to be teachers, they need to know how to interact outside the classroom with undergraduate students, providing them the support that they should have during their undergraduate experience.

This is another:

The universities are not doing any better in training PhDs for academe either. Except for the recent initiatives taken by some universities in giving them pointers on effective teaching, generally their training is in a narrow area of research and they are faced with on-the-job training.

Expectations for those with master's degrees and PhDs are slightly different. Here is an overview from a major company:

In the case of PhDs we are looking for high intelligence and creativity, the ability to originate and conduct independent research, a research background involving at least a solid thesis research experience, and the potential breadth of talent to move from one research field to another. The flexibility required by the latter point is important to us because we cannot hire new talent every time we wish to enter new research fields.

We are also looking for excellent communication and interpersonal skills, so that with proper training they can develop into potential management candidates both in the research organization and in management positions in our operations. We have had a good track record in our research organization in supplying high-caliber talent to our operations.

In the case of MS candidates, we are looking for the same kind of talents, except we do not expect experience in conducting research.

Another consistent comment was on the changing environment--in both the industrial world and the academic world. The following comment is from the dean at a major graduate school:

Graduates are not necessarily being well trained to participate in much of our high educational system as faculty: facilities for front-line research in sciences are not likely to get less costly. Not many colleges and universities will be able to afford the kinds of equipment required for faculty to make significant contributions to science in many areas. If this is true, most academic PhD positions will be in institutions which do not have essential facilities for what is viewed by these fields as cutting-edge research. Either the faculty in such institutions will have to carve out areas of research which don't rely on expensive equipment, or they will have to change their expectations of being significant players on the national and international science scene. It may be that there should be some effort devoted to training PhDs for research appropriate to those other institutions, either for enhancing their instructional roles or for providing them with realistic lines of research.

This is an industry perspective:

In my judgment, educating and training students to do research as well as conducting basic research are still the primary objectives of graduate programs. However, it must be responsive to changing national policies and industrial needs.... I would agree that the American graduate system has been/is a great success. However, to ignore the indicators that show change is needed would be a mistake. Clearly, the challenge ahead is to retain the best of the system while making the changes that will strengthen the nation's outstanding research universities and make them more responsive to the nation's needs.

Yet another comment is the following:

The days when a person could do a PhD thesis in surface thermodynamics (as I did) and reasonably expect to work in the field for a career are over--and I think will never return. One must be ready with the skills to change one's area of focus several times over a career. Most PhD education is training people in the exact opposite direction, and I think this needs to be changed promptly.

Respondents indicated some general concerns about the level of specialized training some graduates receive:

Unfortunately, the training the graduates receive in universities is not directed to any specific career path. Most of the time, after some necessary training in their background, graduate students are pushed into narrow specialization. The consequence of such training is that many of them lack the breadth for work in industry. From what I have seen from the job offers received by our engineering students, they are successful with relatively less effort if their research topic and/or their assistantship experience is closely related to the prospective job description.

And they recommended a broader education for graduate students. One stated,

we may place a new employee in a position which exploits any special expertise he/she may have gained in order to provide "a soft landing," but [he or she] will eventually be called upon to handle a wide range of problems that go far beyond the training received during the completion of the PhD.

A vice president of an applied-research organization wrote, "Everything else being equal, individuals with graduate training cutting across areas of engineering, management and business will turn into better candidates for employment than more narrowly educated specialists."

But one industry respondent warned, "It's a terrible idea to turn [PhDs] into some kind of generalists who don't know anything deeply."

Nevertheless, here is a comment from an international corporation:

Why are industries such as ours not more accepting of PhDs with little or no experience? Because many fresh PhDs see their research area as their sole focus, at least for the immediate future. They generally tend to be very narrow. And, more important, they generally have no meaningful understanding of the business of business. Some might say that such understanding is the responsibility of business to provide. I say no. A highly trained scientist and engineer cannot be very effective if she/he has no knowledge at all of how a company is organized and why, lacks understanding about the principal staff and operating functions, is ignorant of the rudiments of accounting and finance, is unaware of product liability issues that directly affect product development, etc., etc. Industry cannot be expected to deliver such training and education in a short period of time. True, with years of experience working in industry such knowledge is slowly acquired--but it is an extremely inefficient transfer mechanism. Meanwhile, in the early years when the new technologist is working without awareness of these forces and boundary conditions, that person cannot be as effective as she/he otherwise might be. Careers are throttled.

And this is another:

Most of the new PhDs that we hire seem to be relatively well prepared for careers in our organization. I would urge, however, that rather than move towards increasing specialization, which occurs very early in their training, the students should be given a broad array of courses in related areas early in their training. I have the impression that, also from day one in their program, students are now put into laboratories and given a research project so that they can develop the knowledge and skills in their specific area of activity to allow them to complete for grants in the future. However, it has been my observation that this type of training limits their ability to participate in multidisciplinary teams that are often necessary in the industrial setting.

How should this change the current graduate educational system? Respondents agreed that the apprenticeship system of learning research should be preserved. At the same time, many in industry expressed a desire for mentors to be more open to the changing needs of industry. Some professors and administrators favoring apprenticeship thought that their programs already produced the kind of flexibility that industry desires. For example, a dean noted, "We must not change [apprenticeship]... however, ...[it] can also produce a very narrow specialist who is confined and limited by departmental or disciplinary perspectives, even though the obvious trend within the sciences is a breaking down of these barriers and a movement towards greater interdisciplinary perspective."

There was also a general concern that students need to expand the experiences they have during graduate school:

It is our general finding that US graduate schools successfully continue their tradition of producing well-educated scientists and engineers that are capable of making important contributions in their chosen fields. We also believe that the effectiveness of these graduates could be enhanced through practical ("hands-on") experiences/traineeships, functioning as a member of a (multidisciplinary) team, strengthened interpersonal skills, ability to communicate clearly the purpose (including the "strategic" value and relevance of the activity in question), and substantial knowledge of the business environment/culture (including project management fundamentals, time/effort/budget deliverables, sensitivity to human resource concerns, safety, intellectual property, etc.).

In general, employers do not feel that the current level of education is sufficient in providing the following skills and abilities for the people that they are interested in employing:

• Communication skills (including teaching and mentoring abilities for academic positions).
• Appreciation for applied problems (particularly in an industrial setting).
• Teamwork (especially in interdisciplinary settings).

For example:

We look for top-notch technical skills and some evidence of ability to "reduce to practice" the technologies the candidate has been involved in. If we look at new graduates, we look for curiosity about and an appreciation for practical applications of science. As we move away from independent, stand-alone research, and toward more team projects, we screen and hire candidates based on their ability to work in teams, to lead collaborations and teams in an effective way. Skills like project management, leadership, planning and organizing, interpersonal skills, adaptability, negotiation, written and oral communication and solid computer knowledge/utilization and critical for an industrial R&D scientist/engineer. If you walk on water technically but can't or won't explain or promote your ideas and your science, you won't get hired. If you do get hired, your career will stall.

Respondents were also asked about a number of key graduate-education issues.

Industry and administrators seemed to agreed that limiting enrollments was the job of market forces; professors disagreed. As a justification of limitations, they often cited an internal necessity, such as resource or space limitations, rather than a desire to affect the overall market. An industry respondent wrote, "Limiting enrollments is a drastic action to take since the law of supply and demand will usually bring about a correction, albeit several years out of phase. In a few particular disciplines, e.g., chemical engineering, limiting enrollment at the PhD level may need to be seriously considered." A dean wrote:

I would be hard pressed to argue that the world can ever have too many persons trained in the methods of inquiry.... The experience of students graduating with PhDs in the humanities in the 1970s showed us that good minds well trained will find a way to make a difference in places that didn't even know they needed or wanted PhD holders. Such may be the case with science and engineering....

This is from a graduate adviser:

I am skeptical about the utility of attempts to manage enrollment; it is simply too hard to predict what is appropriate, let alone optimal. We at [our university] have been asked to control graduate enrollment, but I suspect that the real controls are still market mechanisms.

There was nearly uniform agreement that the time to degree completion or initial employment is becoming longer. Many respondents favored shortening the time to degree, but others noted that adding teacher training, minor degrees, and other projects for graduate students could add to the time to degree and make it less easy to identify. A dean remarked, "It is not clear that 7 to 12 years of graduate work is either required or appropriate for most positions in business or industry."

A question about national goals for graduate education drew a common response: there are none. A common impression was that graduate education, though considered the "best in the world," is generally rudderless without the external stimuli of the type provided, for example, by Sputnik and the Cold War. A professor at the New Jersey Institute of Technology wrote, "The problem is one of national focus and goals and not the education to support them. There does not seem to be a unifying technological endeavor or an idealized goal with technological underpinnings, to inspire our students and engender popular support."

In conclusion, the themes of the anecdotal information collected via the committee's call for comments indicates that although employers are generally pleased with the result of US graduate education, they have some specific concerns as to the breadth, versatility, and skill development in that education. Furthermore, they are concerned that the graduate education system as it exists today--although acceptable for the past employment world--is less and less acceptable in today's more global world.

Many interesting ideas were suggested by the respondents, and they are summarized below. The committee did not have the resources to evaluate all suggestions fully, so it presents them here for further thought and discussion. They are divided into three categories: career preparation, information needs, and funding. Each is summarized as action items and advice for universities, industry, students, and the federal government.

After the responses were received, the committee decided to conduct a separate survey on information needs; a summary of the survey results is provided in Appendix G.

Universities should

• Initiate more collaborative arrangements with employers.
• Initiate more opportunities for research programs in association with industry.
• Eliminate tenure so that diverse teams of full-time and adjunct faculty can work together.
• Elevate manufacturing to the same level as the arts.
• Enrich the science and engineering curriculum to include training in interpersonal communication, technical writing, team skills, business-process management, accounting, and competitive assessment.
• Seek industrial appointments and sabbaticals for their faculty.
• Require that graduate students who teach be supervised on their first assignment by persons who can give them advice feedback on style and method.
• Help students to complete their education expeditiously.
  
• Continue to train research scientists of the highest caliber.
• Replace half the PhD candidates' course load for the first year with some form of apprenticeship.

Industry should

• Provide more internships in which academic fundamentals can be applied in real-world problems.
• Seek persons who have the ability to conceptualize, apply, develop, and continuously improve processes or products that can be sold at a profit.
• Support more "centers of excellence" to eliminate duplication of effort.

Industry is looking for

• Persons trained and specialized in the traditional scientific disciplines who can integrate skills with science.
• Persons with good writing skills.
• University-trained graduate students who are knowledgeable about safety, quality, statistics, and communication and who have good interpersonal skills.
• Highly technical employees and general managers.
• Persons who can negotiate with, work with, and lead others on a team (broadly trained scientists).
• Persons who can originate and conduct independent research.

Students should

• Take courses dealing with "applied interdisciplinarianism."
• Take a minor and more courses outside their specialties.
• Submit a proposal detailing the plans for their primary research efforts.

Universities should

• Collect benchmarking data to monitor the quality of approved programs and to try to satisfy hiring patterns of employers who regularly draw from particular program.
• Improve student counseling so that those planning PhD in the basic sciences understand, in advance, what the government gauges as appropriate for a PGY1 PhD salary.
• Assess the needs for professors and doctoral-level industrial personnel, translate the results of this assessment into a set of goals, and design strategies to achieve the goals.
• Offer students some comment about job prospects and frank cautions where job prospects are poor, as in physics. Disclosure of programs' completion rates and completion times would help prospective students to make informed choices.

Universities should

• Reduce their size by establishing quality criteria for doctoral programs.
• Refocus their funds through wider use of fellowships than of institutional grants.
• Provide adequate preparation of scientist by using methods that do not depend on external funding sources. Dependence of doctorates in the sciences on external (federal) funding makes programs vulnerable.
• Stop defending lower quality programs--some institutions are better prepared to deliver first-class PhD education.
• Bring the cost of graduate research assistants in line with the cost of laboratory technicians.
• Improve teaching by establishing a policy that senior graduate students can apply for federal grants to supplement their support and in exchange teach graduate courses.

The federal government should

• Grant more fellowships directly to students, which would separate financial support from the research process.
• Set guidelines so that no professor can be funded for more than five research assistantships; this would curtail empire building and encourage collaboration.
• Create a category of funding open to all faculty members, regardless of seniority.
• Restrict the number of foreign students on federal grants; this would force institutions to look inward for their supply of graduate students and make more of an effort to coach them.
• Levy a surtax on employment of foreign graduate students, i.e., require more money for scholarship funds for each foreign student used.
• Eliminate NSF graduate fellowships and redirect the money to research funding. Their fellowships are outdated and send the erroneous message to students that the nation needs more scientists.
• Provide, with state governments, financial support for advanced graduate students to visit public universities and liberal-arts colleges for one semester as professorial interns.

Industry should

• Help to change the science and engineering culture by setting aside a small amount of R&D money for internships for graduate students.
• Cooperate with funding agencies in preparing and reviewing requests for proposals without requiring rights to intellectual property as long as there is no direct monetary support.

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