Ph. D. Completion Rates Data and Strategies for Improvement Completion rates



Download 422.24 Kb.
Page3/3
Date06.08.2017
Size422.24 Kb.
#27346
1   2   3

Lessons Learned

Visits to programs with low completion rates and review of other UA data as well as published research findings (especially work by Lewis Siegal, Vice Provost and Dean of the Graduate School at Duke University5) yield some lessons. Some interventions can be done without significant new resources. With regard to completion and time to degree, one size definitely does not fit all. Each program needs to do a careful analysis of their practices and determine what will improve their completion. Our visits to low-completing programs resulted in some useful brainstorming. Any subsequent changes in policies and practice need to be evaluated for effectiveness.


A laissez-faire attitude is tied to low completion and longer time to degree. Throughout our interviews with programs, we heard from some faculty that progress is really up to students. Yet we know that students are more likely to succeed with close monitoring, frequent feedback, clear expectations and structure. Changes such as setting dates for all students to take comprehensive examinations, rather than letting students each decide when they are ready, may keep students on track. Holding programs accountable for their students’ progress is an effective way to change attitudes.
In some programs, students can flounder in their first year or two, waiting until they have seized a dissertation topic before becoming closely involved with their research advisor. Early advising, exposure to projects and labs can improve retention.
It’s all about individuals. Not surprisingly, programs we visited pointed to individual students who left for personal reasons and remarked that results were not representative of their program. Perhaps not so obvious is the fact that for most programs, completion rates can be attributed to a relatively small number of faculty. In most programs a handful of faculty chair the majority of the Ph.D. committees. Two or three faculty who push students and work efficiently can contribute to a high completion rates while two or three faculty who are more laissez faire can cause a department to have a lower completion rate. Thus improving mentoring skills for those faculty who chair many committees and/or whose students have lower completion rates and longer times to degree will improve results. A more even distribution of workload can also improve completion rates; some faculty are simply overburdened with students.
In science programs, students often choose advisors based on funding. But in other areas, students may select advisors by following the crowd; because a certain faculty member has lots of students, more students may gravitate to her. In some programs, most chairing is done by full professors and in others, rarely by full professors. Patterns of chairing and serving on committees need to be re-examined by individual programs. Here as in other areas, a laissez faire attitude may deter student performance.
It takes a village. In addition to programs changing their policies and procedures, centrally we can do more to monitor and encourage progress. The Graduate College has changed procedures for advancing to candidacy and has instituted email follow-ups for students with outstanding incompletes. Plans are underway with the PeopleSoft installation to do more centrally to help departments track student progress.
The Director of Graduate Studies sets the expectations for the graduate program. DGSs can help create a sense of community and develop a clear path for students. Unfortunately some programs undervalue this crucial position. The Graduate College has begun regular meetings and training sessions for DGSs.
A good graduate coordinator is key. With budget cuts, many graduate coordinators have taken on additional duties and are not able to do the ‘herding’ that can keep students on track. In most departments, it is the graduate coordinators who remind and cajole students about deadlines and requirements. Here, too, the Graduate College is increasing training efforts.
Transparency helps. At Duke and other universities, graduate deans have found that publicly posting completion rates has improved performance.6 Some attrition results from students negatively assessing post-graduation opportunities. Departments should publish placement data, including non-academic career information.
Policies and procedures can be improved and clearly articulated. Student handbooks should clearly lay out expectations, milestones, and what needs to be done at each step. At the Graduate College, we are reviewing our catalog and website, and will also be checking links to all program handbooks. We are asking programs to update their Satisfactory Academic Progress guidelines.
Goal setting and self analysis helps. Programs need to determine what would be an acceptable completion rate. The rate will necessarily vary according to the discipline. As an example, our Computer Science program reported that they lose students to industry. We would expect, then, that completion rates for Computer Science programs nationally would be affected by competition from industry and may be lower than some other disciplines. Once the NRC data are available, the Graduate College will ask programs to set completion goals.

Comprehensive exams are one milestone where we lose students. One program we visited, for example, found they could expedite and simplify their exams. Another analyzed their high failure rate and changed the timing of the exams.


The first-year experience is crucial, yet some programs did not assign advisors until later and did little to integrate new students.
Evaluation efforts need to be coordinated. APRs, faculty evaluations, student evaluations, and surveys such as NRC all need to be used collectively to improve the Ph.D. experience. We reviewed APRs for programs with low completion rates and often found evidence for other related problems that may have contributed to low completion. Low completion does not happen in a vacuum.
Focus on the problem areas. We asked programs to identify patterns of withdrawal.

  1. Are students not completing because of academic problems? If so, carefully examine your admissions policies.

  2. Do your admission requirements include evidence for persistence, research ability, communication skills, and not just GPA and GRE? Do you interview prospective students? Do you arrange campus visits?

  3. What assistance do you have for students struggling academically?

  4. If students attrite early, re-consider your curriculum and course sequencing. Do you need to reconsider early coursework so that students have a better foundation?

  5. If you lose students during the dissertation stage, can you develop a course/workshop directed at dissertation completion?

  6. Do students find that their interests do not match faculty’s? If so, can you better inform potential students of faculty research areas?

Some programs have substantially lower completion rates among women or minority students. Looking at time to withdrawal can also pinpoint problems. Programs need to determine any patterns in their student attrition. Students who need extra mentoring and assistance are often left behind. Departments need strategies to identify struggling students early.


Tying funding to academic progress focuses students on the need to stay on track. One program we visited, for example, has formalized their evaluation process to include a broader set of data and tied their annual review process to their re-funding decisions. Students who delay their comprehensive exams or who fail them are given lower priority for funding.
More sophisticated recruitment and admissions practices can reduce attrition. Many students leave because they lack realistic expectations about graduate study or because of a mismatch of their interests with the department. Admission decisions need to be made on the basis of more than grades and GREs. While we want to encourage risk-taking rather than restricting our admissions to safe candidates, more communication and better information can help us determine an appropriate fit.
Find ways to shorten time to degree. One of the biggest complaints we hear from students is about delays (often semesters) because of slow feedback from advisors and difficulty in scheduling committee members. Timing of qualifying and comprehensive examinations can slow progress. Some programs reported that standardizing some of the content of comprehensive exams and formalizing the process for the dissertation prospectus decreased delays. Some programs have instituted a dissertation seminar and revised their prospectus process to move students more quickly.
Graduate assistant stipend levels need to be evaluated and calibrated. Most of the low-completing programs we visited argued that their stipends are not competitive which causes them us to lose some of their most-promising applicants. Programs also pointed to the uncertainties in TA funding and the frequent lateness of budget information. Here, too, programs will have NRC data from our peers and will be able to determine what stipend levels we should target. Duke now distributes GA budgets based on graduate education parameters rather than on service needs.
Teaching assistants’ workloads need to be evaluated. Not surprisingly, some of our programs with lower completion rates are those that have the largest undergraduate service loads. On the other hand, English, History, Communication, and Spanish have high TA workloads and all have above-average completion rates.
In summary, these data do not tell the whole story, but clearly our completion rates show room for improvement. The data should open the discussion and spur programs to set appropriate goals and develop strategies for improvement. The Graduate College looks forward to assisting in this process.



1 http://www7.nationalacademies.org/resdoc/index.html

2 Council of Graduate Schools, Analysis of Baseline Program Data from the Ph.D. Completion Project, 2008.

3 UA ten year completion rates for the cohort entering in Fall 1996 was 58% for men (N=160); 55% for women (N=105). There is considerable fluctuation, however. Among the Fall 1997 cohort, 56% of men and 56.1% of women had finished by Spring 2006.

4 Our NRC programs are all fulltime, so eliminating students who took a leave of absence or dropped to part time has a minimal effect. One NRC concern was to avoid penalizing departments where women students had longer time to degree because of child bearing or rearing.

5 http://www.phdcompletion.org/features/Siegel-DukeCompletion.pps#294,1,A

6 Ph.D. Completion---and Content, Inside Higher Ed, April 1, 2008. http://www.insidehighered.com/news/2008/04/01/graduate.


Download 422.24 Kb.

Share with your friends:
1   2   3




The database is protected by copyright ©ininet.org 2024
send message

    Main page