Many factors impinge on student decisions whether to complete a program. Most speakers at the workshop agreed, therefore, that the raw data on the extent and timing of attrition must be supplemented with more contextual or qualitative data in order to analyze the causes of attrition. In addition, they emphasized the need to gather this qualitative data at the institutional level. The types of information they felt would be useful would include the financial resources available for students, selection and retention policies, and the culture of departmental relationships with graduate students.
Financial AidAlthough commonly believed to be a key factor in student completion, the speakers generally agreed that data on the amount of resources provided to individual students is extremely difficult to obtain. They also emphasized the importance of data on the timing and types of support available. One presenter emphasized that it was the type of support that students receive, not the amount, that is associated with differences in persistence outcomes. By becoming research or teaching assistants, she explained, students intensify their relationship with the department, which becomes a positive factor in their retention.
Illustrating the importance of timing, one participant cautioned that for students in some disciplines, a research assistantship in the early years can become a burden that substantially affects the time the student takes to complete the degree. Opinions varied at the workshop regarding the relative value of fellowships vs. teaching assistantships and research assistantships. Student responses to surveys and interviews indicate advantages and drawbacks to each. This value can also vary from discipline to discipline and from one stage of the graduate program to another.
Selection and Retention
Policies for admission and evaluating students for acceptance into doctoral candidacy vary not only from institution to institution but within departments and programs at an institution.
Departments within some universities take radically different approaches on retention. One participant contrasted two successful graduate programs at the same university. One has no attrition, but is almost impossible to get into. The other has a liberal admissions policy, then tests students through their first year; those who do not master the basic skills are cut.
Student perceptions of the graduate school experience are strongly influenced by their department's cultures. Unlike the case of undergraduate education, the graduate education experience is shaped by specific situations-the student's relationships with specific faculty, in some cases just one or two members of a department. Research designs, therefore, ideally should capture that complexity.
One survey (Educational Testing Service) elicited information on departmental culture in the space provided to students for open-ended comments. In other qualitative studies, specific questions have been asked about such factors as the degree of faculty mentoring, opportunities for participating in department events, student-to-student mentoring, relationships with research supervisors, and general social activities.