For decades, graduate deans have been troubled by low completion rates and long times-to-degree in PhD programmes in the humanities. Now federal and state officials are talking about the need to improve completion rates in undergraduate programmes.

As university budgets remain tight, doctoral training may be subjected to similar demands. One indication that this issue is gaining traction is that measures of student outcomes will be part of the evaluation metric used in the National Research Council's new rankings.

In this environment, graduate schools are likely to be pressured to improve their completion rates and times-to-degree. In Educating Scholars: Doctoral education in the humanities (Princeton University Press, 2009), we report on the Graduate Education Initiative (henceforth GEI), the 18-year project funded by the Andrew W Mellon Foundation.

The GEI was intended to help 54 departments in the humanities at 10 major research universities improve their doctoral programmes, increase their completion rates, and decrease their times-to degree. Departments were to examine their existing programmes and identify needed improvements while the Foundation provided funds for student aid and departmental redesigns.

From the GEI's beginning, graduate deans at the 10 institutions were central participants. The deans used their policy-making, budgetary, reviewing, gate-keeping, and record maintaining powers to influence, advocate, and advance the project.

Their evaluations and suggestions for revising the programme helped keep it on track as did their effective wielding of proverbial carrots and sticks. The foundation expected that PhD programmes would be improved in three ways: 1) impediments to students' progress would be identified and changes made to facilitate their transiting from one stage to the next; 2) financial aid would be increased and made conditional upon students' timely progress; and 3) students judged unlikely to succeed would be identified early and encouraged to leave.

To evaluate the effectiveness of changes departments made, the foundation collected quantitative data from the institutions and received departmental reports on their progress and problems annually. Our book reports findings from analyses of these datasets and the responses of over 18,000 PhD students to a lengthy retrospective survey.

Taken together, these data indicate that graduate students' careers are complicated. Given the pivotal role that graduate deans played, we hope our findings will be of interest to them, their staffs, and to faculty members as they consider how to design more effective graduate programmes.

How can graduate deans help? Deans can use their budgetary powers (however limited) to assist students in the humanities while at the same time requiring that their progress be rigorously reviewed. Just as important is the need to persuade faculty members that it is possible to improve graduate programmes and student outcomes simultaneously.

As deans know, faculty members are central to the operation of graduate programmes. Indeed, we found that departments in which the faculty supported the GEI's goals increased their completion rates during the GEI more than departments in which faculty did not. Goals stated too broadly led to misunderstanding and resistance.

The call to 'shorten time-to-degree' is an apt example. One persisting theme in GEI faculty reports was scepticism regarding the value of reducing time-to-degree, based on doubt that high-quality dissertations could be done quickly. Measuring the quality of dissertations is daunting. We can, however, shed some light on the relationship between time-to-degree and 'quality' by treating job placement as a rough proxy for the quality of work done by doctorate recipients in our study.

Our finding that students who completed their PhDs in six years or less were no more likely to get tenure-track positions than students who finished in seven years seems consistent with some faculty members' reluctance to abbreviate time-to-degree. However, we also found that students who took longer than seven years to finish were much less likely to get tenure-track positions than students who finished in seven years.

Given this pair of findings, perhaps even reluctant faculty members might support a goal of reducing time-to-degree from nine or 10 years (or more) to seven years or less. The goal of 'reducing attrition' was also misunderstood.

Almost all departments participating in our study reduced attrition during students' first four years. But later, higher rates of attrition among advanced students occurred in these same departments. Reducing late attrition calls for being willing to identify unpromising students early and to counsel them to leave graduate school. Leaving does not end students' careers.

We found that almost 12% of the students in our study who left their initial PhD programmes went on to earn PhDs elsewhere and another 18% earned professional degrees in other fields. Such findings put attrition and its outcomes into a different light from what is commonly supposed.

When goals are clearly stated and agreed upon, deans can foster behaviour consistent with them. For example, deans oversee financial aid. We found that having a fellowship improves the probability that students will complete their degrees; but having a fellowship does not reduce time-to degree much more than being supported by a teaching assistantship.

Summer support, however, is especially advantageous. Other things being equal, it is positively associated with timely completion of the degree. Reallocating some fellowship money to summer support may reduce attrition and increase completion.

Multi-year financial-aid packages, now common at the universities that we studied, are not unalloyed goods. Students with guaranteed support have little financial incentive to make timely progress as long as funding in future years is assured. Institutions should be cautious about extending guaranteed support and should monitor the progress of students who receive it given the associations we found between late attrition and very late completion and guaranteed support.

Our study also earmarked several procedures that are effective and relatively cost-free. For example, students whose departments furnished them with explicit (and enforced) timetables for meeting departmental requirements were more likely to make timely progress and to complete degrees.

Similarly, the absence of faculty supervision is associated with late attrition. Students who reported receiving little or no faculty supervision were more likely than others to drop out late. Although periodic reviews of students' progress consume faculty time, they may save faculty and student time in the long run.

Although the GEI departments instituted additional monitoring and financial incentives (such as completion fellowships) to encourage students to complete dissertations promptly, our data showed little decrease in the time students spent in this stage. This surprising and frustrating result was often attributed to students' delaying graduation in response to the scarcity of jobs - often on the recommendation of faculty advisers.

But faculty and deans also noted that faculty incentives for advising numerous dissertations could prove perverse and should be reviewed. The loss of student benefits such as housing, health insurance, and teaching opportunities in the absence of future employment were also sources of delay.

In response, several deans imposed time limits on these benefits and some encouraged degree completion by increasing tuition charges for delayed completers. By monitoring the effects of such changes, deans and faculty can assess their outcomes.

As gate-keepers and maintainers of records, deans can also lend support to departments that try to improve their programmes. Departments often do not know how many students leave or how long it takes those who actually finish. By supplying data that measure students' progress and financial aid, deans can help departments assess their policies.

For example, our data showed that single men and single women had similar completion rates; likewise, women's completion rates were not adversely affected by their being married or having children upon entry to graduate school. Thus, there is no empirical support for taking marital or parental status into account in graduate admissions or financial aid.

Our book also describes a series of measures that deans may use in gauging departmental performance. For example, 'student-year cost', the collective time a department has expended on producing each new PhD, has already been used in at least one university. This measure of how well departments curbed attrition and fostered timely degree completion is also relevant for determining the costs of incoming cohorts of varying sizes.

Since the GEI began in 1991, awareness has increased about the extent of attrition and length of times-to-degree. This awareness has stimulated thoughtful review of programmes and sensible decisions to change them. The recent reports from the CGS PhD Completion Project and the forthcoming NRC rankings should motivate faculty members to reconsider the shape of PhD programmes and lead prospective students to demand information about programme length and completion rates.

These faculty and student reactions will provide the opportunity for deans to review policies and incentives, gather and use data more effectively, and re-invigorate enforcement of goal-enabling policies. The resulting programmes should save time and resources previously lost by students, faculty, institutions, and those who pay for them.

We hope that the lessons from our study help identify ways to reinforce these efforts.

* Ronald G. Ehrenberg is Irving M Ives Professor of Industrial and Labor Relations and Economics, Cornell University and Director of the Cornell Higher Education Research Institute; Harriet Zuckerman is Senior Vice President of The Andrew W Mellon Foundation and Professor Emerita of Sociology, Columbia University; Jeffrey A Groen is Research Economist, US Bureau of Labor Statistics; and Sharon M Brucker is Project Coordinator, the Survey Research Center, Princeton University.

* This article is from The Communicator, the US Council of Graduate Schools' latest newsletter and is republished with permission.

Printable version
Email to a friend
Comment on this article

Disclaimer: All reader responses posted on this site are those of the reader ONLY and NOT those of University World News or Higher Education Web Publishing, their associated trademarks, websites and services. University World News or Higher Education Web Publishing does not necessarily endorse, support, sanction, encourage, verify or agree with any comments, opinions or statements or other content provided by readers.