Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
6 The Doctor's Degree The training of the Ph.D. engineer has been the subject of debate for many decades ~d is of special current interest because of the shortage of Ph.D.s in engineering for faculty positions. While there is a great need for Ph.D. engineers in academia, there are not equally well defined needs for large numbers of Ph.D.s in all industries. The Ph.D. engineer is generally utilized effectively by industry in research or development; however, that same degree and type of training are much less needed in operational divisions. While having a Ph.D. is advantageous in busi- ness, it is by no means a requisite for success. Industrial research is being driven by major advances taking place today in science, in analytical tools, and in computing capability as well as in market sophistication. Highly trained doctoral-level engi- neers are needed to tackle these complex research and development challenges. These experts command high starting salaries in industry relative to academia and are often given challenging assignments with access to the most modern facilities and equipment. Regarding the nature and purpose of a doctoral program, the Council of Graduate Schools in the United States has stated: s3 The doctoral program is designed to prepare a student for a lifetime of intel- lectual inquiry that manifests itself in creative scholarship and research, often leading to careers in social, governmental, business, and industrial organiza- tions as well as the more traditional careers in university and college teaching. The program emphasizes freedom of inquiry and expression and development 83
84 ENGINEERING GRADUATE EDUCATION AND RESEARCH of the student's capacity to make significant contributions to knowledge. An essential element Is the development of the ability to understand and evaluate critically the literature of the field and to apply appropriate principles and procedures to the recognition, evaluation, interpretation, and understanding of issues and problems at the frontiers of knowledge. A dissertation is universally required in U.S. universities for the doctor's degree. The purpose of the dissertation is twofold: {1) to develop ~ the candidate the independent ability to carry out a scholarly investigation of a challenging topic at a high level of professional abil- ity, and {2) to provide for an original contribution to knowledge in the field. Generally, the candidate is expected to defend the dissertation in a final examination; sometimes such examinations are open to the public. A nearly universal doctoral requirement is a comprehensive examination consisting of written and oral parts, generally imposed just before the candidate begins work on the dissertation. The purpose of the examination is to demonstrate an adequate knowledge of the field and an ability to use academic resources. If the candidate passes the examination, it is considered likely that he or she will successfully complete the dissertation. Many schools additionally impose yet another examination, given early in the student's program, to deter- mine fitness for doctoral work. The foreign language requirement tends to vary from school to school and frequently from department to department within a given school. Some kind of foreign language reading ability at one time was a nearly universal requirement. Now many departments have no foreign language requirement, apparently in the belief that such requirements generally produce little utility in reading articles published in foreign languages. The policy statement on the Ph.D. degree by the Council of Graduate Schools does not specify a foreign language requirement, and in fact scarcely mentions the topic.s3 Graduate schools generally require two years in residence for doc- toral programs, in order to provide an appropriate degree of student- faculty interaction and supervision of the thesis research. The result of the residence requirement is usually that a candidate must forgo full- time employment and become a full-time student. However, fully employed individuals have sometimes been able to complete doctoral programs without giving up their employment if they are close to a university campus and if their job assignments are flexible enough to permit extensive student-faculty interaction and faculty supervision of the thesis work. Since Ph.D. programs generally require a minimum of two years in residence and master's programs require a minimum of one year, the
THE DOCTOR'S DEGREE 85 minimum elapsed time between the B.S. and Ph.D. degrees for a stu- dent enrolled in full-time study would be three years. However, data collected by the National Research Council show that the average "reg- istered" time between these two degrees for engineers is 5.8 years, and the average total elapsed time is 7.5 to 7.9 years;~4 i5 These periods of time are greater than the minimum time, because few graduate stu- dents proceed directly from the baccalaureate to the doctor's degree with unbroken full-time study. There is often a period of employment before graduate school; also, doctoral students are frequently employed on a part-time basis by their universities as postgraduate researchers or as teaching assistants; furthermore, students often are required to take course work beyond the minimum in order to fill in gaps in their back- gro~ds; and, finally, there is the classic reason, which is that the thesis takes longer than expected. According to Engineering College Research and Gradinate Study, there were 149 U.S. universities offering doctoral engineering pro- grams ~ 1983.47 Of these, 137 reported that they awarded one or more doctor's degrees in 1982-1983.48 Nearly half (46 percent) reported fewer than 10 doctorates each, and 26 percent reported fewer than 5. The 30 largest engineering doctorate producers are listed ~ Table 23. In 1982-1983, these 30 institutions produced 62 percent of the engineer- ing doctorates in the country. Table 23 also shows that the ratio of doctorates per faculty ranged from 0.12 to 0.75 in 1982-1983, with an average of 0.32. It is likely that many of these schools could increase their Ph.D. output, although in order to do so they would probably have to reduce their undergraduate loads or else expand their resources, because most schools already are overloaded. It would seem reasonable that many of the schools which presently have small Ph.D. outputs could expand their production, and that they have an adequate quality base from which to do so. The conclusion, then, is that our existing system of engineering graduate schools is capable of expanding its pro- duction to the needed levels and that the startup of additional Ph.D. programs should not be encouraged. Expanding on the present base will require additional faculty and other resources, but is less expensive than starting new programs. Doctoral programs must have a strong base of funded research. Table 24 provides some insight to the magnitude of the cost. The total 1982- 1983 research expenditures for the 30 institutions listed in Table 23 are shown in Table 24, and are broken down into the following categories: { 1 J federal government, t2J state and local government, t3J business and industry, and {4) an "other" category, which includes private nonprofit organizations and institutional support. Not all universities reported
86 ENGINEERING GRADUATE EDUCATION AND RESEARCH TABLE 23 Thirty Largest Engineering Doctorate Producers {based On average of 4 years, 1980-1983) 1982- 1983 171 145 170 127 88 79 50 63 64 58 64 55 SS 51 63 44 41 30 50 37 51 41 38 40 35 37 29 36 26 33 1,871 School MIT I 11 i n o i s - U r b a n a / C h a m p a i g n Calif.-Berkeley Stanford Purdue Comell Calif.-UCLA Michigan Northwestern Wisconsin Ohio State Southern Cal. Texas Cal. Tech. Rensselaer Calif.-Davis Iowa State Harvard Virginia Tech. Camegie-Mellon Georgia Tech. Penn State Princeton Texas A&M Columbia Minnesota Univ. of Pa. Case Western Reserve Polytechnic Inst. of N.Y. Colorado State Degrees per Year {4-yr alga 166 156 144 124 94 74 66 66 62 57 ~6 52 47 45 44 42 38 37 37 36 36 36 36 35 34 34 33 32 32 31 Degrees/ Degrees Faculty Faculty 0.46 0.37 0.74 0.75 0.31 0.38 0.36 0.25 0.50 0.30 0.21 0.39 0.31 0.64 0.39 0.40 0.13 0.59 0.19 0.35 0.19 0.15 0.51 0.12 0.32 0.20 0.29 0.32 0.20 0.32 0.32 373 392 231 169 280 208 139 257 127 192 309 142 176 80 163 110 309 51 265 105 271 282 75 334 108 183 99 111 132 113 5,786 SOURCES: Engineenng and Technology Degrees 1New York: Engineering Manpower Commission, 1979, 1980, 1981, 1982, 1983~. Engineenng College Research and Grad- uate Study, Engineenng Education, March 1984. data in these categories, and in other cases the data did not match the totals given for the entire engineering school; in such cases, N/A {"not available" ~ is shown. The total 1982-1983 research expenditure for each school has been divided by the average number of doctor's degrees granted per year in the 1980-1983 period to give a general idea of the research base support
THE DOCTOR'S DEGREE 87 ing the doctoral enterprise. These ratios are shown in Table 24, with public and private institutions listed in different columns. These "dol- lars per degree" numbers shouldnot be considered as the cost per degree and should not even be taken too literally. There are too many differ- ences in the modes of operation of the different schools to permit such a simple interpretation. Engineering schools have, for example, varying kinds of organized research units with various organizational relation- ships to the schools. Some may be wholly within the schools; others may be partially or wholly outside the engineering school organiza- tions, and even though the research of the doctoral students may be heavily supported by these research units, the schools may report the dollar amounts in different ways. Furthermore, for some schools a significant portion of their research contracts may have little involve- ment on the part of doctoral students. Hence, the figures shown in Table 24 should be taken as the most general of guides concerning the dollar magnitude of research required to support a major increase in doctoral output. The columns headed "$ per degree" in Table 24 show an average of $241,000 per doctoral degree for public institutions and $317,000 per degree for private universities. Thus, if we conservatively use a figure of $200,000 per doctoral degree, we could anticipate that an increase in funded research for engineering on the order of $200 million per year would be required to support an increase from the present level of 3,000 Ph.D.s per year to the projected level of 4,000 or so per year by 1988. The last column of Table 24 shows that only a small fraction of the total research funding on the order of 15 percent has historically been provided by industry. Thus, unless there is a very large increase in research funding by industry, most of the increase will have to be pro- vided by government, and this means the federal government primar- ily, if past patterns prevail. Provision will need to be made also for major upgrading in research equipment, coming in part from the research contracts themselves. A different approach to this same topic may be taken by examining research support data for the Engineering Directorate of the National Science Foundation. In Fiscal Year 1983, NSF supported 1,795 graduate students for a research dollar total of approximately $101 million. Thus, there was an average of approximately $56,300 of support per graduate student. Of this, only about $9, 700 went directly to the salary for each graduate student, on the average, with the rest in salaries for faculty, postdoctoral students, undergraduate students, secretaries, technicians, and costs for equipment, travel, computer time, supplies, and general overhead. The 1,795 graduate students consist of a mix of
88 C~ a, .= V) ~ - CO ._ ._ .> C~CC ~.° ~ _ O- {5 , ~ ~ Z o G~ 1 C~ cn - o o v cn ~4 ._ C) ._ ~4 ._ CO ._ Q) P~ x C~ CD F~ ¢ :` ~ ~a ~> CJ ~0 C~ C~ _ O u) J 4 - Ct ~ C> ,. O - ee _ ~o e~ U~ ~ O ~ ~4 ~ tC O C~ 1 a a ~ O O O O O O O ~ O ~ _ O O O <S ~ _ Z - C~ 6a O C) g O z ~ O O O O 0\ U~ O C~ _ O O O O O O O O O O U ~O S0 U~ U) o o o - 00 C~ O O O O O O O _ C~ O O O O O O O O O O ~D ~ o ~ - - - o o o o o o o o o - - - o ~ oo o ~ ~ S o o oo o o oo o o oo o o oo o o oo o o oo - - -- - ~ -- ~ ~ o~ c~ u~ -- ~ ~ ~ ~ ~ - o o o - u~ o o o o o o - - - ~ ~ u~ o o o o o o o o o - - - ~ ~ c~ oo ~ ~ o o o o o o o o o - - ~ o ~ ~o ¢\ - - - ~ - - o o o o o o - - o o o o o o o o o o o o o o o - - - - - - ~ ~ r~ oo ~ oo 0\ - - ~ - d. ~ ~ c~ o z o ~ ~ o ~o - c~ - - - o o o o o o o o o o o o ~ - ~ ~ ~ o x ~ O oo - - - - ~ ~ ~D - - o o o o o o o o o - - - x ~ O`- ~o ~ o o o o o o o o o o o o - - - - ~ ~ - ~o o ~ ~ ~- 1 o o o o o o oo ~ ; o o o o o o o o o - - - ~ ~ c~ ~ c~ ~ oo eg oo - - - - ~ ~ o o o - o - c ~- o o o o o o o o o o o o o o o o o o - - - - - - ~ ~ - ~ ~ u~ x os ~ o ~ ~ ~ - - - - - - ox ~ ~ ox o o o o o o o o o - - - x ~ ~ oo c~ ~ - - - x ~ ~ ~ - ~ - o o o o o o - - - - e~ ~ - - - ~ c~ ~ - - ~D ~ ~ \0 C~ u~ ~ ~ 0\ ~ ~ ~D ~D ~ ~ u~ q" - - - - c~ ~ ~ ~ c) - ~ ~ I ~ m ~:5 ~ _ _ C) q ~ c) ~: - ~ "c o "c ~ cO ~ - ~ o -~ o o c) ~ c) c) u) ~ . cn x - ~ ~ - a 1 - C)
89 o o oo o o o o o o oo o o o o o o oo o o o o ~l ~ ~ X~ ~ ~ _ l CO U~ _ ~ ~ ~ C ~_ _ ~ _ o o o o o o o o o . o o o ~o C ~ o o o o o o o o o o o o o o o C o o o o o o o o oo _ ~ ~oo _ o~ _ - - o o o o o o o o o o o o o o o o o o o o o o o o - ~- ~- ' - - 'S - - I_ o o~ o ~ C~ - ~ ~ ~ z ~ _ _ _ o o o o o o o ~ o o o o o o o o o o oo 2 Z '~ o o o o o o o o o o o o o o C o o o o o o o o o o o o o o o o o o ~ ~ ~ ~_~ ~ ~ ~ ~ o o c~ ax - e~ ~7 ~ _ ox I_ ~ oo L~ _ U~ ~ ~ Z ~ ~ ~ ~ Z Z ~` - ~ - ~c~ e~ - ~ - oo o o oo o o oo o o oo o o oo o o oo o o - Inc~ 0 ~ ~ ~ ~ _ 1 X1 ° °\ - ~1 ° ~ - - oo C ~o ~ Z ~ ~Z Z o _ ~C~ o o o o o o o o o o o o o o o _ CN ~ ~ oo o ~ ~ _ ~ CO \c, _ _ ~ ~ Z o oo _ ~ o o o o o o o o o o o o o o o oo o - C~ o ~ o oo ~ o - o o o o o o o o o o o oo o o o o o o o o o o o o oo o o o o o o o o o o o o oo o o ~ ~ ~ o ~ ~ o~ X o C~ ~ C~ U~ _ ~ O4 ~ I o~ o ~ - ~ oo ~ o ~ - - _- _ ~ ~ _ _ __ X l I ~ ~ ~ ~o ~ ~ ~ ~ C ~O4 - o 9 E `, E e ' ° Io o ~ 4 - oo o o o o o o C~ 6e C~ ~4 c) CC CO 5 e~ 5 ~0 ~e C) C~ ._ O ._ 4- C~ 5: C~ ~ C) V 4- O ~ o O~ ~ D u, C > ~ e~ ~ _ X C) ._ ~ O e-4 ~ == 11 C) <: tt Z-O,O O ~ 0 4D o~ 3 .~ oc oo o~ - o o 0 ~ oo o~ o ~_ e~ o o - oo - - ~r - oo - - o oo ¢\ - - - o ._ co a, ._ o 3 . O ~ ~ ~o c~ b4 ._ ~ ~4 .c ~ - ._ 0 =, ._. .. :s ,!34' ~ 4 3 ~ 7 ~ . C~ oc t0 ~:S :: oo O _ ~ 0 q5 C) ~ ~ ~S b0= oo =4' °O .. _ U) _ O' ~ oo O ~
90 ENGINEERING GRADUATE EDUCATION AND RESEARCH master's and doctor's students, but a comparison with the figure derived earlier can be made as follows. Let us suppose all of the 1,795 students are doctoral students, and that the students are evenly distrib- uted throughout the four or five years that are required to complete a doctoral program. We can then estimate the number of doctor's degrees per year et one-fourth or one-fifth of the total in the program. If we take the more conservative figure of one-fourth, then a program with 1,795 doctoral students would produce' about 450 Ph.D.s per year. Dividing 450 degrees per year into the annual expenditure of $101 million gives a figure of $224,000 of support per Ph.D. degree per year, which corre- lates well with our earlier figure. Using the assumption of five years to the degree would give a figure of $280, 500 per Ph.D. degree per year. The available base of facilities and equipment has fallen well behind the needs ~ the face of overall enrollment growth. A specific statement concerning the total need is not possible in the absence of a detailed nationwide inventory of needs and resources. However, it should be noted that space needs per student are greater for graduate than for undergraduate students. The principal need in the case of undergradu- ates is for classrooms and class laboratories, with related support needs such as computer facilities and shops. For graduate students, in addi- tion to classrooms and support facilities there is a substantial need for research laboratories. For one research university, the University of California, the space standards provide for 200 square feet of laboratory space per graduate student, and 300 square feet of lab space per faculty member, plus another 220 square feet of office and clerical support space per faculty member. If we made the assumption that all available space nationally is already utilized and that there is no surplus avail- able, then the current projection of growth in Ph.D. production by 1,000 per year tunplying 4,000 additional students registered), plus the space needed to meet the full shortfall of 6,700 faculty, would require approximately 4.5 million square feet of new space as an upper limit. If space costs $100 to $250 per square foot, depending upon the sophisti- cation of the laboratories, then the estimated cost ranges from $450 million to $1 billion {1983 dollars) nationwide. For the states with the largest engineering enrollments {Califomia, Texas, and New York) on a proportional basis, this could require an investment on the order of $60 minion to $80 million per state for expansion of facilities and basic equipment. There appears to be no definitive data base regarding the true magni- tude of the need for research equipment in engineering schools. In Fiscal Year 1984, the National Science Foundation budgeted approxi- mately $18 million for equipment, out of its total engineering budges of
THE DOCTORS DEGREE 91 $123 million.62 NSF accounted for 38 percent of all federal obligations to Diversities and colleges for basic research in engineering in Fiscal Year 1983,63 so the total federal support for engineering research equip- ment probably lies in the range of $50 million to $75 million per year. NSF has estimated that it has been able to service only a small fraction of the total need, as evidenced by equipment requests in proposals. A' part of the need can be met by the $200-million increase in funded research recommended earlier. This could amount to $35 million to $40 million per year if past patterns prevail. An additional portion will be met by the $450-million to $1-billion construction estimate set forth in the preceding paragraph, since construction budgets typically include some provision for built-in equipment. The unmet equipment need could easily be on the order of $100 million, but this figure cannot be supported by any straightforward analysis. Since much of the need would be met by the construction budget discussed earlier and by an augmentation of engineering research of $200 million per year, no sepa- rate dollar recommendation for this element is included here. The major increases in federal funding needed to support increased doctoral programs would no doubt come from a variety of government agencies. The-traditions of the National Science Foundation have been productive for the sponsorship of academic research in engineering, although other government agencies have also found effective ways to sponsor academic research within their operating guidelines. Increased emphasis on engineering research within the National Science Founda- tion, a process that is already occurring, would be a strong stimulus toward the objectives outlined in this report. An important issue for engineering schools is the fact that research leadership in some fields has shifted substantially from academia to industry, as in the cases of VLSI and automated manufacturing based upon CAD/CAM. Such shifts are especially likely to occur in fields where the costs of laboratories are so great that few universities can afford them. The seriousness of this situation for American technical education is that graduate work should be couched in a research envi- ronment at the cutting edge of technology, and if the cutting edge is in industry, the educational experience will be less valuable than it should be. For a handful of universities, industrial funds have been brought together to establish major research facilities in specialized fields. Industry should be encouraged to provide such support to the maxi- mum extent feasible, and it should seek to support a mix of fields rather than a narrow selection. Government support has also been brought to bear on the problem of up-to-date research facilities in universities. All
92 ENGINEERING GRADUATE EDUCATION AND RESEARCH of these measures are beneficial in helping to establish cutting edge research environments in Universities, but they cannot cover all the needs. Additional measures that can be taken include cooperative industry-university research and consulting relationships for faculty. These have great potential benefits for academic institutions because they bring the academic environment closer to the moving frontier of industrial practice. Findings and Recommendations 1. The existing system of engineering graduate schools should be capable of expanding its doctoral production to the increased level that is needed, and the startup of additional Ph.D. programs should not be encouraged. 2. An increase of doctoral output will entail a corresponding increase in fended research. It is estimated that an increase on the order of $200 million of new funded research per year will be required, princi- pally from the federal government. 3. The available base of facilities and equipment has fallen well behind the needs for engineering education. Expansion of the Ph.D. Output, plus meeting the needs of the full 6,700 "shortfall" in faculty, would require space on the order of 4.5 million square feet of new space as an upper limit. The upper-limit cost of such space, depending upon the sophistication of the laboratories, would range from $450 million to $1 billion nationwide t1983 dollars). 4. The traditions of the National Science Foundation represent an excellent model for funding Diversity research. Increased emphasis OF engineering research within the National Science Foundation is strongly encouraged.
