1
Introduction
In the twentieth century, U.S. graduate schools have changed from institutions for scholarly inquiry, largely separate from the rest of the university, to internationally preeminent institutions that are an inherent part of a university's teaching and research mission. They provide students from the United States and abroad the opportunity for advanced study in a wide variety of fields through carefully structured programs leading to the Doctor of Philosophy (Ph.D.)1 or related research degree.
As U.S. graduate education has ascended to international preeminence, its structure and purpose have changed. Some doctoral institutions are reporting surprisingly high levels of attrition from Ph.D. programs. Attrition, though some is to be expected, has a high cost in time and resources for both students and faculty. To the extent that avoidable causes of attrition can be identified and ameliorated, graduate education can be improved.
Changes In The Academic Environment
The graduate school is no longer an independent institution for scholarly inquiry detached from undergraduate education as it was when doctoral programs were first established over 100 years ago at The Johns Hopkins University and Clark University (Rudolph 1962; Veysey 1970). Graduate programs are much more closely linked today to the greater utilitarian concerns that characterize contemporary university-industry-government relations (see, for example, GUIRR 1992). Significant emphasis is placed on research productivity, and faculty are often heavily engaged in research activities, sometimes at the expense of being available to graduate students (AAU 1990). The functions of modern graduate study may thus be said to combine several goals, one of which is regulating the entry of individuals into a field while socializing them into a profession (Merton, Reader, and Kendall 1957; Wiebe 1967).
Graduate education also serves the university by educating research scholars and scientists who, in turn, teach and produce research, often in collaboration with faculty (Hagstrom 1965; Breneman 1970; Snyder 1981). Graduate students also supplement faculty as teachers of undergraduates (see Text Box 1).
TEXT BOX 1: REQUIREMENTS AND SPECIFIC ASPECTS OF THE DOCTORAL PROGRAM A doctoral program is an apprenticeship that consists of lecture or laboratory courses, seminars, examinations, discussions, independent study, research, and, in many instances, teaching, designed to help students make significant contributions to knowledge in a reasonable period of time. The first year or two of study is normally a probationary period, during which most of the effort of doctoral students will be devoted to acquiring a working knowledge of the field through study of the literature, taking formal courses and seminars, learning research and experimental techniques, problem-solving, and beginning to teach and do research. After being admitted to candidacy, students devote essentially full-time to completing the dissertation research planned with the major adviser, and the dissertation committee. Preparation of the dissertation usually occupies one to three years, depending on the field. An oral defense of the research and dissertation by the candidate before a graduate committee and sometimes other persons invited to attend constitutes the final examination. All requirements for the degree should be available to the student in written form (Council of Graduate Schools 1990, 14). A doctoral program is an apprenticeship that consists of lecture or laboratory courses, seminars, examinations, discussions, independent study, research, and, in many instances, teaching, designed to help students make significant contributions to knowledge in a reasonable period of time. The first year or two of study is normally a probationary period, during which most of the effort of doctoral students will be devoted to acquiring a working knowledge of the field through study of the literature, taking formal courses and seminars, learning research and experimental techniques, problem-solving, and beginning to teach and do research. After being admitted to candidacy, students devote essentially full-time to completing the dissertation research planned with the major adviser, and the dissertation committee. Preparation of the dissertation usually occupies one to three years, depending on the field. An oral defense of the research and dissertation by the candidate before a graduate committee and sometimes other persons invited to attend constitutes the final examination. All requirements for the degree should be available to the student in written form (Council of Graduate Schools 1990, 14). |
Transformations In Size And Organization
In addition to a shift in purpose, graduate education has undergone a dramatic transformation in terms of its size and organization. In 1978 the National Research Council's Board on Human Resource Data and Analyses reported that the number of Ph.D.s awarded in the United States essentially doubled in each decade over the past century. ''Quarter-century landmarks show that in 1900 the annual output was about 300; in 1925, about 1,200; in 1950, about 6,000; and in 1974, about 33,000" (NRC 1978). After a period of contraction in the late 1970s and early 1980s, Ph.D. production returned in the late 1980s to those levels recorded in the early 1970s. In the 1990s, degree production has continued to grow (see Figure 1-1). Data from the 1995 Survey of Earned Doctorates revealed that U.S. universities awarded over 41,000 doctorates (NRC 1996).
The transformation that has occurred in graduate education was made possible by a substantial increase in the overall number of institutions offering doctoral programs, especially small programs. The entry of new degree-granting institutions is summarized in Table 1, which shows the proportion of doctoral degrees granted, by year, between 1920 and 1992 in which the awarding institution first granted a Ph.D. As can be seen, most graduates earn their degrees at institutions that first awarded doctorates before 1930. However, since the 1970s the number of doctorates conferred by newer institutions has gradually expanded. Between 1990 and 1992, 8.2 percent of the doctorates conferred were awarded by institutions that recorded the first doctorate after 1970 (NRC 1995a).
The emergence of these newer institutions is also recorded in data from the Carnegie Foundation's A Classification of Institutions of Higher Education (1994), which indicate that the number of institutions awarding doctorates grew from 173 in 1970 to 184 in 1976, to 213 in 1987, to 236 in 1994 (see Table 2). Much of the growth occurred in "Doctorate-granting Universities II" whose
TABLE 1-1 Number of Ph.D.s Awarded by U.S. Universities (1920-1992) by Reporting Period and Year in which the Doctorate Records File Recorded the First Ph.D. Awarded by that Institution
TABLE 1-2 Number of Institutions Awarding Doctorates
|
1970 |
1976 |
1987 |
1994 |
Research Universities I |
52 |
51 |
70 |
88 |
Research Universities II |
40 |
47 |
34 |
37 |
Doctorate-Granting I |
53 |
56 |
51 |
51 |
Doctorate-Granting II |
28 |
30 |
58 |
60 |
numbers more than doubled between 1970 and 1994.2
As Bowen and Rudenstine observed:
In effect, a new structure has been created. This new structure reflects dramatic increases in the overall number of degree-granting programs, major reductions in the proportion of degrees conferred by highly rated programs, and especially rapid increase in the number of small programs.
From a national point of view, this transformation—in all its dimensions—raises inescapable questions concerning the nature of the system as a whole (p. 56).
Concerns About Rates Of Attrition And Patterns Of Degree Attainment
From the outset, the "wasteful" consequences of attrition have been a concern of serious educational analysts. One of the few studies of graduate attrition to generate national estimates came from work by Berelson (1960). Berelson's study involved a special-purpose survey of faculty and deans at universities asking them to "estimate the magnitude of attrition" from the cohort of students who had entered graduate school in 1950 (seven years earlier). This "impressionistic" study is one of the few of its type (see Text Box 2). Berelson observed that the attrition rate in graduate school is not high when compared with the undergraduate levels "where it is also about 40 percent. But the matter is perhaps more serious for the graduate school because its selection is supposed to be better; its type of education is much more expensive; and, as with law, its drop-outs stay around longer than the undergraduate drop-outs, half of whom leave in the first year" (Berelson 1960).
Berelson explored the phenomenon of attrition in some detail in his 1960 study. He invited deans, graduate faculty, and recent doctoral degree recipients to give the reasons for attrition at the doctoral level. He states that when he asked the deans and graduate faculty whether it was more a problem of selection or of the programs, they replied in effect that is was "both," but, as Table 3 indicates, "they do not consider [dropping out] the fault of the graduate school. It is either a matter of money or of the student's capability."
TEXT BOX 2: BERELSON'S ACCOUNT OF GRADUATE EDUCATION IN THE UNITED STATES CIRCA 1960 In his seminal work, Graduate Education in the United States (1960), Bernard Berelson asks, "Is attrition a major problem of doctoral study, wasting human and institutional resources?" He continues: There are certainly those who think so. President [Benjamin] Wright speaks of the 'inordinately high' attrition rate in graduate school and concludes that 'we can save a very considerable number of those who now drop out of graduate school.' While graduate dean at Columbia, Jacques Barzun deplored 'the appalling waste on both sides—of student energy, hope and money, and of faculty time and effort.' Citing data from questionnaires distributed to graduate deans and graduate faculty members at a number of major research universities, Berelson reports the following estimates from graduate deans: "Of the students who start work toward a doctorate at your institution, about what percentage never finish? Median Institution: 35-40 percent." When asked, graduate faculty placed attrition at a much lower level—around 20 percent—which Berelson attributes to more "selective attention" on the part of the faculty than the deans who actually maintain the official records. Berelson admits that 40 percent attrition is high when compared to medical school (at 10 percent) and "the better law schools," but points out that the professional schools typically screen their students more intensively before admission. Nonetheless, Berelson reports that, ''Graduate deans, being closer to the situation administratively, are … concerned. Half of them consider [attrition] an important problem for their institutions." |
TABLE 1-3 Reasons for Attrition at the Doctoral Level
|
percent |
|
|
|
Graduate Deans |
Graduate Faculty |
Recent Recipients |
Lack financial resources |
69 |
29 |
25 |
Lack intellectual ability to do the work |
50 |
64 |
52 |
Lack proper motivation |
38 |
45 |
47 |
Lack necessary physical or emotional stamina |
33 |
33 |
49 |
Found degree wasn't necessary for what they wanted to do |
19 |
10 |
12 |
Disappointed in graduate study and quit |
1 |
12 |
21 |
Berelson concludes, however, that the recent recipients are probably "closer to the mark than either the deans or the faculty in their estimates of the disappointment of the doctoral candidates when they leave. … The farther from the dean's authority, the more the inclination to blame the graduate school."
Berelson placed graduate attrition somewhere between 20 percent (according to faculty) and 35-40 percent (according to deans). It is important to note, however, that these are impressionistic numbers, based on anecdotal evidence rather than on quantitative analysis, and are likely to be low. Not since Berelson's study have national estimates of attrition from doctoral programs been generated. Instead, analysts turned their attention to a study of factors thought to influence attrition, including lengthy time-to-degree3 and the social-psychological effects of changes in the academic environment on student outcome.4
Origins And Scope Of The Study
This report responds to a request from the Andrew W. Mellon Foundation for a
summary of data sets that could be used to monitor trends in graduate attrition and degree attainment in the sciences and humanities at U.S. Universities.5 The study was conducted under the auspices of the National Research Council's Office of Scientific and Engineering Personnel (OSEP), which invited two members of the OSEP Advisory Committee on Studies and Analyses to serve as an ad hoc panel to oversee the project.
The ad hoc panel began its study with a review of systematic information about graduate attrition and degree attainment in the sciences and humanities. The panel commissioned summaries of known national data bases that have been or could be used to analyze patterns of graduate attrition and degree attainment. These are listed in Appendix A. In addition, the panel commissioned two papers to identify and summarize studies published in the past decade that treated the subject of attrition and degree completion. The results of that review are included in the Annotated Bibliography. The commissioned reviews revealed that there is a distinct and growing body of knowledge about patterns of student departure from doctoral programs in the sciences and humanities. However, most estimates of the degree of graduate attrition have been limited to special-purpose studies developed to meet institutional demands for information.
Finally, the panel briefly considered the feasibility of coordinating current efforts to collect information about graduate attrition and degree attainment in order to generate national estimates. There is no question that national data would be desirable, if possible.
Reliable estimates of graduate attrition are in the national interest because of their potential to reduce the waste inherent in the premature departure of talented individuals from advanced preparation in the sciences and humanities, but they did not find such an effort to be feasible within the country's existing diverse system of graduate education. The panel did consider what would be needed for individual institutions to establish their own longitudinal tracking systems, including a list of relevant data elements. These are presented in Appendix B.
Organization Of This Report
This report analyzes the data sources currently available on student attrition from doctoral programs in U.S. universities. Chapter 2 reviews studies that have been conducted in the last few years regarding patterns of attrition and degree attainment. These are studies that have generated precise estimates of attrition from doctoral programs in selected fields at the university level. Many of these studies involve the analysis of university records that also contribute to our growing understanding of when students are likely to "drop out." A second type of study, reviewed in Chapter 3, involves special-purpose surveys and analyses that reveal why students leave, pointing to certain features of the academic environment that contribute to the decision to suspend pursuit of the doctorate. Chapter 4 briefly discusses the viability of generating national estimates of graduate attrition and provides conclusions.
References
Abedi, Jamal, and Ellen Benkin 1987 "The Effects of Students' Academic, Financial, and Demographic Variables on Time to the Doctorate." Research in Higher Education 27 (March 14):3-14.
Association of American Universities (AAU) 1990 "Institutional Policies to Improve Doctoral Education." A Policy Statement of the Association of American Universities and the Association of Graduate Schools in the AAU. Washington, DC: AAU.
Association of American Universities/Association of Graduate Schools Project for Research on Doctoral Education (AAU/AGS) 1993 "Participation in Doctoral Education at Major Research Universities by U.S. Citizens, Women and Underrepresented Minorities." Program Profiles 1 (April).
Berelson, Bernard 1960 Graduate Education in the United States. New York: McGraw-Hill.
Bowen, William G., G. Lord, and J. A. Sosa 1991 "Measuring Time to the Doctorate." Proceedings of the National Academy of Sciences 88 (3):713-17.
Bowen, William G., and Neil L. Rudenstine 1992 In Pursuit of the Ph.D. Princeton, NJ: Princeton University Press.
Breneman, David W. 1970 "The Ph.D. Production Process: A Study of Departmental Behavior." Ph.D. diss., University of California, Berkeley.
Carnegie Foundation for the Advancement of Teaching 1994 A Classification of Institutions of Higher Education. Princeton, NJ: Princeton University Press.
Council of Graduate Schools (CGS) 1990 The Doctor of Philosophy Degree: A Policy Statement. Washington, DC: CGS.
Government-University-Industry Research Roundtable (GUIRR) 1992 Fateful Choices. The Future of the U.S. Academic Research Enterprise. A Discussion Paper. Washington, DC: National Academy Press.
Hagstrom, Warren O. 1965 The Scientific Community. New York: Basic Books.
Hartnett, Rodney T. 1985 "Trends in Student Quality in Doctoral and Professional Education." Project on Trends in Academic Talent (March).
Hauptman, Arthur M. 1986 Students in Graduate and Professional Education: What We Know and Need to Know. Washington, DC: AAU.
Merton, R., G. Reader, and P. Kendall 1957 The Student Physician. Cambridge, MA: Harvard University Press.
National Research Council (NRC), Board on Human Resource Data and Analyses 1978 A century of doctorates: data analyses of growth and change: U.S. Ph.D.s—their numbers, origins, characteristics, and the institutions from which they come: a report to the NSF, NEH, and U.S. Office of Education. Washington, DC: National Academy of Science.
National Research Council (NRC) 1995a Research-Doctorate Programs in the United States. Continuity and Change. Washington, DC: National Academy Press.
1995b Summary Report 1994. Doctorate Recipients from United States Universities. Washington, DC: National Academy Press.
1996 1995 Summary Report, Preliminary Tables. Washington, DC: NRC.
Rudolph, Frederick 1962 The University: The Autonomy of Academe. New York: McGraw-Hill.
Snyder, Robert G. 1981 "Federal Support of Graduate Education in the National Sciences: An Inquiry into the Social Impact of Public Policy." Ph.D. diss., Syracuse University.
Tuckman, Howard P., Susan Coyle, and Yupin Bae 1989 "The Lengthening of Time to Completion of the Doctorate Degree." Research in Higher Education 30:503-16.
Veysey, Laurence 1970 The Emergence of the American University. Chicago: University of Chicago Press.
Wiebe, R. H. 1967 The Search for Order, 1877-1970. New York: Hill and Wang.