In Pursuit of the PhDby William G. Bowen, Neil Rudenstine
What percentage of graduate students entering PhD programs in the arts and sciences at leading universities actually complete their studies? How do completion rates vary by field of study, scale of graduate program, and type of financial support provided to students? Has the increasing reliance on Teaching Assistantships affected completion rates and time-to-degree? How successful have national fellowship programs been in encouraging students to finish their studies in reasonably short periods of time? What have been the effects of curricular developments and shifts in the state of the job market? How has the overall "system" of graduate education been affected by the expansion of the 1960s and the subsequent contraction in enrollments and degrees conferred? Is there "excess capacity" in the system at the present time? This major study seeks to answer fundamental questions of this kind. It is based on an exhaustive analysis of an unparalleled data set consisting of the experiences in graduate school of more than 35,000 students who entered programs in English, history, political science, economics, mathematics, and physics at ten leading universities between 1962 and 1986. In addition, new information has been obtained on the graduate student careers of more than 13,000 winners of prestigious national fellowships such as the Woodrow Wilson and the Danforth. It is the combination of these original data sets with other sources of national data that permits fresh insights into the processes and outcomes of graduate education. The authors conclude that opportunities to achieve significant improvements in the organization and functioning of graduate programs exist--especially in the humanities and related social sciences--and the final part of the book contains their policy recommendations. This will be the standard reference on graduate education for years to come, and it should be read and studied by everyone concerned with the future of graduate education in the United States.
- Princeton University Press
- Publication date:
- Product dimensions:
- 7.51(w) x 10.29(h) x 1.57(d)
Read an Excerpt
In Pursuit of the PhD
By William G. Bowen, Neil L. Rudenstine
PRINCETON UNIVERSITY PRESSCopyright © 1992 Princeton University Press
All rights reserved.
Introduction and Principal Findings
At the doctoral level, the preeminence of this country's programs of study, viewed in terms of both quality and scale, is widely accepted and perhaps even taken for granted. Incontrovertible evidence is provided by the large (and rising) number of students from other countries who elect to study for doctorates here. In 1989, 6,590 doctorates were awarded by U.S. universities to non-U.S. residents.
Still, all of its accomplishments notwithstanding, graduate education in the United States is far from any ideal state. Nor would anyone claim that its prospects and future role are well understood or assured. The recent Hearings on the Reauthorization of the Higher Education Act, reports issued by the Association of American Universities (AAU), and numerous commentaries attest to wide-ranging debates that continue. Similar issues are also being discussed in other countries.
While there have been a number of good studies of certain facets of graduate education (especially of the sciences, engineering, and professional education in fields such as medicine), it is surprising that so little systematic study has been devoted to doctoral education in general. Both its undisputed importance and its substantial cost would seem to justify considerably more attention. Not since the publication of Berelson's major study in 1960 has there been a comprehensive review of the overall system of graduate education. The humanities and the related social sciences have been especially neglected.
There has been considerable speculation about the reasons for the relative lack of scholarly interest in graduate education. Possible explanations include a general tendency among academics in several social science disciplines (especially economics) to prefer to study other areas, as well as the particularly daunting conceptual and empirical problems that bedevil study of graduate education. The specialized nature of fields of knowledge at the graduate level makes it unusually difficult to generalize, and the decentralized administration of graduate programs means that the problems of collecting even the most rudimentary data can be monumental. One clinical psychologist has suggested that the traumas associated with pursuit of the PhD may even have discouraged many scholars from returning to such a personally painful subject!
The Context and Main Questions
Concern for the future staffing of colleges and universities has been one stimulus to the research reported here. For reasons described in detail in Prospects for Faculty in the Arts and Sciences, we believe that many American colleges and universities are likely to face serious staffing problems by the end of the 1990s. Significant imbalances between demand and supply could result from a combination of expected retirement patterns, the demography of the college-age population, and declines over recent decades in the number of PhDs awarded to U.S. residents. A surprising (unanticipated) finding of the earlier research was that recruitment of faculty could prove to be as difficult in the humanities and related social sciences as in the sciences and engineering.
It is outside the scope of the present study to reexamine the projections made in 1989. We should reemphasize, however, the long-term, "other-things-equal" nature of those projections. Any projections based on simple, static assumptions will frequently be overtaken, at least in the short term, by external forces. At the present time (summer of 1991), the current recession and the accompanying fiscal problems of governments at all levels are creating difficult budgetary problems for colleges and universities. Many faculty vacancies are being left unfilled, and recruitment plans are being put on hold. If they continue for extended periods of time, financial constraints of this kind can have long-term effects on faculty staffing. But they are more likely to affect the timing of shifts in the pattern of faculty recruitment than eventual staffing levels.
Since the main underlying trends in the age distribution of the present faculty and in the size of the college-age population are unchanged, we see no reason to modify our sense that serious staffing problems should be anticipated by the late 1990s. We also continue to believe that policies affecting graduate education have more potential than any other set of actions to address potential staffing problems. Failing to make any provision for deeply rooted trends could be a most serious error—and an error very hard to correct subsequently because of the long time lags between changes in "inputs" and "outputs" that are inherent in graduate education.
If graduate education can be such an important lever of change, obvious questions arise: Is there sufficient capacity within existing graduate programs to meet the projected needs for additional faculty members? Should consideration be given to encouraging the creation of new programs? Can existing programs be expected to meet projected needs through the operation of more-or-less automatic adjustment mechanisms, or are some forms of external support/ stimulation likely to be required?
In evaluating whether proposals for additional external support ought to be encouraged, consideration must be given to the effectiveness of current programs and patterns of support. How efficient are they? Do graduate programs appear to have become more or less effective over time? Are there lessons to be learned from the differential outcomes achieved when students are grouped by field of study and scale of graduate program? Is it possible to find ways to reduce attrition and time-to-degree, while simultaneously sustaining, and perhaps improving, quality? How well structured and well organized are existing programs? Is it true that much greater reliance has been placed on teaching assistantships (TAs) in recent years, and if so, what have been the consequences? What has been the record of national fellowship programs, both governmentally and privately funded?
The more we thought about such questions, the more convinced we became that many of them were of such intrinsic importance that they deserve the most careful consideration, regardless of what one believes about the likelihood of possible shortages of faculty by the end of the 1990s. Doctoral education is, after all, the apex of this country's system of higher education in the arts and sciences. The effectiveness of undergraduate teaching as well as the quality of scholarship and research depend to some considerable extent on how well graduate programs function. Talented, well-motivated, and well-trained graduate students—both as students and later as professionals, working in or out of academia—contribute critically to our collective ability to generate ideas and educate new generations of students. They are assets of incalculable value.
Present concern with the costliness of higher education at all levels reinforces the conviction that graduate programs deserve careful analysis. It is incumbent on those of us in universities, and those associated with organizations that support universities, to improve the effectiveness with which we deploy whatever resources are at our disposal. This obvious proposition surely applies to graduate education no less than to undergraduate programs and the conduct of research.
Recognition of the extraordinary amounts of time and effort (as well as money) that many of the brightest students are expected to invest in graduate study heightens still more the feeling that it is wrong simply to accept current rates of attrition and present assumptions about how long it should take to earn a doctorate. We need first to understand the factors responsible for current norms. Then we need to see if there are ways to do better.
Scope of This Study
This book is intended to help fill part of what we perceive as a significant gap in knowledge. It is not, however, a comprehensive treatment of graduate education, and many important questions are not addressed. The scope of this study can be summarized as follows:
The focus is on PhD programs within the arts and sciences. We do not discuss master's programs or professional programs at the doctoral level.
Within the arts and sciences, we have elected to study most intensively a limited number of graduate programs in specific fields of study in particular universities. This "departmental approach" reflects our conviction that many of the most important issues facing graduate education can be understood and addressed most effectively at the level of the individual department or graduate program.
Many of the most important findings reported here pertain to two measurable outcomes of graduate education—namely, completion rates (or, conversely, attrition) and time-to-degree. We analyze trends in these two measures over time, differences in them across fields of study and types of universities, and the effects on them of different forms of financial aid, as well as departmental structures, requirements, expectations, and conventions ("cultures").
The attention given in this project to seemingly mundane numbers, such as the percentage of students in an entering graduate school cohort who complete their studies within various periods of time, does not reflect any lack of concern on our part for questions of quality, which are fundamental. Qualitative concerns are highly relevant both at the level of the individual graduate program and at the national level. But emphasizing the need to maintain and, if possible, improve quality, cannot be an excuse for failing to examine critically the more measurable outcomes of the process of graduate education that must be considered in assessing the effectiveness with which both time and financial resources are employed.
We are concerned mainly with the experiences of U.S. residents enrolled in these programs. We make no effort to analyze differences in outcomes for U.S. residents compared with non-U.S. residents.
While our primary interest is in outcomes for all U.S. residents enrolled in these graduate programs, we also examine differences in outcomes related to gender, as well as trends in the participation in graduate education of women and (to a lesser extent) members of ethnic and racial minorities.
Finally, while specific findings often relate most immediately to individual graduate programs and particular universities, we also analyze the results achieved by various national fellowship programs and, where possible, include references to other national data as well.
Data and Methods
It was obvious from the outset of this project that the data needed to analyze many of the central questions did not exist in any manageable form. The lack of a comprehensive set of reasonably "hard" data has shaped both our initial approach to these issues and the final presentation of results. It would be difficult to exaggerate the importance of the collection of new sets of data. Many of the most important findings depend directly on data heretofore unavailable.
One irony is that the lack of certain essential information has been accompanied by a veritable treasure-trove of other kinds of national data, which have proved to be extremely valuable. The basic source is the Doctorate Records File (DRF), maintained by the National Research Council (NRC). For about 70 years, records have been kept of the number of doctorate recipients each year, classified by field of study and by the institution granting the degree. From 1958 forward, far more detailed data are available.
This remarkable database (which is generated by reports on each individual receiving a doctorate, rather than by a sampling technique) is the source of the time series on degrees conferred and numbers and sizes of doctorate-granting programs that are presented and analyzed in Chapters Two, Three, and Four. It is also the source of estimates of median time-to-degree for specified populations of recipients that we present in later parts of the study.
From the standpoint of ability to answer many of the key questions posed above, the DRF has one fundamental limitation: It contains information only on those individuals who receive doctorates. It does not permit the calculation of completion rates because it is silent on the numbers and characteristics of students who enroll in PhD programs but do not earn degrees.
There is no substitute for longitudinal data that track each entering cohort or "class" of students as its members move (or fail to move) from one stage of graduate education to another. The only method that we could devise to assemble such data systematically, at the necessary level of detail, was by soliciting the cooperation of particular universities that had enrolled reasonably large entering cohorts over a number of years and were also willing and able to provide relevant statistics for individual students (without identifying them, of course).
Practical considerations—the time and expense involved in building such a database—dictated that we limit this data-collecting effort to a manageable number of fields of study and a reasonably small number of carefully selected graduate programs. Because we were interested in studying trends, we decided to request data going back to the early 1960s, even though we realized the formidable nature of such a data-gathering process.
Following considerable consultation with faculty members and administrators, we chose six fields of study for intensive analysis: English, history, political science, economics, mathematics, and physics. Each is a reasonably well-defined field that has enrolled significant numbers of students over many years. English and history represent important areas within the humanities, broadly defined; political science and economics illustrate somewhat different emphases within the social sciences; and mathematics and physics are central disciplines within the broad sphere of the physical sciences. These six fields together covered, we thought, a reasonable cross-section of those graduate programs in the arts and sciences that often lead to academic careers.
In much of our work, we have found it useful to group certain fields of study because of their common characteristics and the tendency for patterns of behavior to be quite consistent. For instance, English, history, and political science form a group referred to throughout this study as the "EHP" fields. Similarly, it is sometimes appropriate to group mathematics and physics (the "MP" fields), even though both basic characteristics and outcomes differ in some noteworthy respects between these two fields. Economics is in many respects a case all its own, but in terms of the measurable outcomes considered here, it behaves much more like the MP fields than the EHP fields.
We have not been even-handed in the attention we have given to these six fields. While the same statistical analysis of outcomes has been done for all fields (so that comparisons could be made), the EHP fields have been the focus of our detailed study of the content of graduate programs, their design, internal structure, and oversight. That part of the research required extensive interviews, an examination of institutional self-studies, and the perusal of related materials. It would have been impractical to extend such an analysis to more fields, and we felt that the most complicated (and troubling) issues facing graduate students and graduate programs were to be found in the humanities and related social sciences.
Choosing universities for concentrated analysis was even more difficult than choosing fields of study. We finally decided to seek permission to collect data at 10 major universities: the University of California at Berkeley, the University of Chicago, Columbia University, Cornell University, Harvard University, the University of Michigan, the University of North Carolina at Chapel Hill, Princeton University, Stanford University, and Yale University. All 10 agreed to participate in the project, and this is therefore referred to as the "Ten-University" data set. (The questions asked of these universities are described in detail in Appendix A, along with conventions adopted to achieve reasonably consistent reporting.) Understandably, not all 10 universities were able to provide the full array of information that was sought for all fields and time periods. Still, statistical records relating to over 36,000 individuals were collected, which gives some indication of why it was essential to limit the number of fields and universities included in this part of the study.
Excerpted from In Pursuit of the PhD by William G. Bowen, Neil L. Rudenstine. Copyright © 1992 Princeton University Press. Excerpted by permission of PRINCETON UNIVERSITY PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
and post it to your social network
Most Helpful Customer Reviews
See all customer reviews >