The Path to the Ph.D.: Measuring Graduate Attrition in the Sciences and Humanities (1996)

The most important variable in time-to-degree was the source of support during graduate school. "Own earnings" especially add to time-to-degree. Fifth in importance was field of study. Also included are data from the National Research Council's Doctorates Records File on 4,255 students receiving doctoral degrees at UCLA between 1976 and 1985.

Interviews with 21 students who did not complete the program are examined. The primary source of dissatisfaction expressed by students was their perception that faculty were not approachable.

Universities that have carried out comparative assessments of doctoral programs have found that departments with well-structured programs, clear expectations of graduate student performance and faculty responsibilities, and widely shared faculty commitments to encourage and facilitate students' progress have lower attrition rates and shorter times-to-degree than comparable departments whose programs lack those properties.

Questionnaires were distributed to approximately 1,200 university freshmen to develop and test a causal model of student attrition. The causal model synthesized research findings on turnover in work organizations with student attrition from institutions of higher education.

The study focused on approximately 4,300 students who entered doctoral programs at UCLA in 1969, 1970, and 1971. The status of students was examined at the end of the 1981 spring term.

The report discusses graduate education (master's and doctoral) in countries having membership in the Organization for Economic Cooperation and Development (OECD), with particular attention on the duration of research training and the problem of noncompletion. An important reason for not completing the dissertation was a poor working relationship with the advisor and/or committee (except for bioscientists). Data from telephone interviews with 25 former doctoral students in psychology, sociology, zoology, physics, electrical engineering, and biochemistry are included.

A theory of career instrumentality (the extent to which a doctoral student perceives that the completion of the degree is relevant to a career goal) is developed through a study of the literature on employee turnover and graduate attrition. Data from a survey of 670 Ph.D. students at the University of Maryland in 1986 is presented. Career motivation (a desire to enter the guild) is shown to affect student persistence in maintaining relationships with faculty members.

The study reviews data from ten leading research universities regarding the educational experiences of 35,000 graduate students in English, history, political science, economics, mathematics, and physics, as well as the graduate experiences of 13,000 winners of prestigious national fellowships. The authors conclude that there are opportunities to achieve significant improvements in the organization and functioning of graduate programs.

Tabulation of degrees awarded in chemistry in 1981-82.

Tabulation of degrees awarded in chemistry in 1982-83.

Tabulation of degrees awarded in chemistry in 1984-85.

The author notes that registered time-to-degree at his alma mater, the University of California at Berkeley, has increased from three and one-half years in the early 1960s to five years plus in the mid-1980s. He suggests a program design for degree completion in four years (emphasizing an early introduction to research and productivity), and the use of effective management practices, noting that there is extensive literature on the management of research. The article is based on the author's experience and offers a prescription for shorter time-to-degree in a graduate program in chemistry.

The article examines evidence for the convergent and discriminant validity of V. Tinto's student integration model and J. Beans's model of student departure by tracking student outcomes of an entering freshman class (fall 1988) at a large southwestern institution.

The relationship of research to teaching and learning is at the heart of core problems in higher education. This cross-national analysis of other national systems of higher education allows for isolating some of the similarities and crucial differences in the system of American higher education.

This study presents a tabulation of Ph.D. production by Roose-Andersen rankings, demonstrating that the best departments produce the most Ph.D.s.

This study presents the results of a survey of 266 doctoral students (145 Ph.D.s and 121 drop-outs) enrolled between 1970 and 1980 in the program of educational administration at Georgia State University. Significant differences are shown between Ph.D.s (positive experience) and dropouts (negative experience) in faculty relationships.

Data that were collected for a set of selected major universities during the 1970s and 1980s suggest that completion rates, while varying widely across fields and institutions, tend to lie in the 40 to 70 percent range.

This article proposes a model that links several variables to predict progress toward a graduate degree. Data were collected from a survey of 486 students entering at a major Midwestern university in the fall of 1977 and measured in 1984. Department characteristics were shown to be an important predictor of progress toward a degree, while the student's relationship with the advisor was found to be critical to graduate success.

The authors present an overview of a design for improving the quality of information available for the continuous operational study of student retention at the University of Hartford. They address concerns about maintaining the confidentiality of files, while providing information to professional administrators, counselors, and faculty advisors.

This discussion paper examines recent trends affecting academic research in the U.S.; considers the impact of trends on the current academic research enterprise (the group of American universities and colleges performing significant research in the sciences and engineering); identifies longer-term issues that will affect the enterprise in the future; and explores ways in which the enterprise can meet future challenges.

This article discusses the relationships in engineering over an unspecified three year period and provides a model for use in any engineering school.

The study presents data from the National Center for Education Statistics, the U.S. Census Bureau, the NRC Doctorate Records File, and various federal departments for finishing graduate students in year groups from 1965 to 1983. It points out significant differences by field of study and type of financial support for graduate education. The results are divided among physical sciences, life sciences, and engineering; social sciences and humanities; education and professional; and others.

The paper presents nine attrition studies, completed by members of the AGS, that indicate attrition is greater in the humanities and social sciences than in physical and biological sciences.

The report presents interviews with 149 current and former science, mathematics, and engineering (SME) students at seven institutions to elicit the range of factors which contribute to attrition from SME majors. Theories about the relative experiences of "switchers" and "non-switchers" are presented.

Tabulation of degrees awarded in chemistry 1987-88.

This article explores the circumstances of selecting a dissertation topic and suggests that the dissertation may be a significant factor in lengthening time-to-degree. Data are derived from a survey of 438 finishing doctoral

students at a major Midwestern research university in education, humanities, mathematics and physics, social and behavioral sciences, arts, engineering, business, agriculture, and biosciences from June 1986 to March 1987. Dissertation decisions affecting time-to-degree are shown to differ by field of study. The study also indicates that science students benefit from close faculty supervision by having a shorter time-to-degree than other students.

This study of former doctoral students in psychology, sociology, zoology, physics, electrical engineering, and biochemistry shows that the principal reasons for not completing their dissertation vary by field of study. The primary reason for failure to complete their program was a poor working relationship with the advisor and/or committee.

This study is based on a survey of 117 doctoral students (85 of whom earned the Ph.D.) at Temple University in 1986 and finds that that faculty management of student research is an important factor in degree completion or noncompletion.

This is an exploratory study of "persisters and nonpersisters" in a graduate level, nontraditional liberal education program. Five demographic factors proved to be significant: age; type of bachelor's degree; years between completion of the bachelor's degree and enrollment in an M.A. degree program; site; and the social science score on the Undergraduate Assessment Program Test.

This article considers doctoral programs in music in a general period of decline in job opportunities. It recommends a reduction in program size while working to improve program quality.

This paper describes efforts to build a data base from institutional files to analyze time-to-degree and Ph.D. completion rates in 32 graduate groups from students matriculating between the fall of 1973 and the spring of 1989.

This study presents an analysis of doctoral completion times since 1967. It provides a time-series data base for the period from 1967 to 1986 and develops a model that explains some of the factors that have been responsible for lengthening time-to-degree.

The study reviews data from the National Research Council, from interviews with approximately 300 doctoral students, and from selected campuses of the University of California System to examine whether students took longer to complete their doctorates in the 1980s than they did 20 years earlier at the University of California. Factors that might account for observed differences are examined.

The study compares average time-to-degree in Ph.D. programs and completion rates at the University of California at Berkeley to national trends and similar analyses at comparable institutions. It finds (1) a wide variance in time-to-degree and completion rates by field of study and (2) shorter time-to-degree for students in engineering and natural sciences than for students in the humanities and social sciences. A similar variance for completion rates was noted. Poor communication and lack of guidance from faculty members were found to contribute to longer time-to-degree. The authors describe various research activities to address these issues, including quantitative analyses supplemented with qualitative methods to develop a basis for designing recommendations and programmatic outreach activities.

Realizing the value of exchanging information on experience in the development of human resources for research and development, OECD organized a meeting of member countries to help them develop improved strategies and techniques for managing scientific and technological

development more efficiently. The various mechanisms which are involved in the education of research personnel are considered to form a dynamic system whose output is the flow of new personnel and whose behavior is affected by inputs, some of which are amenable to control.

Tabulation of research support in academic chemistry in 1985.

This study examines arguments of loss/gain by delay in achieving the doctorate, and it develops five groups of reasons for delay or noncompletion: (1) qualities of the student, (2) personal problems/accidents unrelated to their studies, (3) problems inherent in research, (4) personal academic problems, and (5) teaching and supervision. Recommendations pertaining to the organization of supervision are made. The data, measured as of 1966 on 1,008 students who entered doctoral study in 1957 and finished the doctorate in Britain, indicate that time-to-degree was shorter in science and technology groups than in social studies, languages, literature and area studies, and other arts groups.

The study looks at the problem of attrition and time-to-degree in Britain. Supervision of graduate work is identified as a major problem.

This study presents data from a survey of 190 doctoral students in Britain who completed their Ph.D.s in 1978-79 in engineering, science, and social science. It found that time-to-degree is higher in social science than in engineering and science, and almost half of the sample cited difficulties with faculty supervision of students' research.

The study discusses two findings: (1) that 288 undergraduate programs account for a majority of doctoral graduates in the natural sciences; and (2) that colleges are more productive than universities (private more productive than public), and that private colleges are more productive than private universities.

The author reviews the limits of the research and theory dealing with undergraduate student attrition.

Data from the literature are presented, showing that field of study has greater influence on doctoral persistence than the institution. The faculty-mentor relationship is also shown to be a major factor in degree completion.

The study discusses how time-to-degree has increased in 11 scientific and engineering fields, indicating the differences and increases and calling for further research on cause. Data from the NRC Doctorate Records File for 1967, 1977, and 1987 in physical sciences, biological sciences, and social sciences are presented, showing a greater increase in time-to-degree for social sciences than for physical and biological sciences.

This report examines the forces associated with elementary and secondary education that shape the talent pool; traces pathways to undergraduate and graduate education in science and engineering; and presents a discussion of policy areas for possible congressional action developed under two strategies labeled "retention" and "recruitment."

The paper looks at American colleges and universities as producers of a future work force and provides a perspective on fledgling scientists and engineers, focusing on the endpoint of educational preparation—undergraduate and graduate study—for science and engineering careers.

This paper is a study of completers and noncompleters of Ed.D. at West Virginia University and shows that recipients of the degree are more

likely to have positive relationships with faculty.

This narrative discusses problems experienced in time to doctoral degree in Sweden and steps taken. It found that long study periods and high dropout rates led to the improvement of student grants. Concomitantly, time limits were tightened up, and admission qualifications were made more stringent. Data are derived from Swedish literature. Disciplinary differences in the conduct of graduate study are noted. The study found there is less attrition and shorter time-to-degree in the natural sciences than in the humanities and social sciences and that strong faculty supervision results in lower attrition and time-to-degree in the sciences.

Data on 4,637 doctoral students entering at three major research universities from 1978 to 1985 in chemistry, English, history, mathematics, political science, psychology, economics, philosophy, physics, computer science, and sociology indicated that doctoral candidacy is achieved earliest in chemistry, mathematics, physics, and computer science; next in psychology, political science, sociology, and economics; latest in history, English, and philosophy.

Data measured in 1987 on 1,379 doctoral students entering Northwestern University from 1972 to 1978 in psychology, chemistry, English, history, mathematics, political science, chemical engineering, economics, philosophy, physics, sociology, theatre, and computer science show that attrition appears to be highest in English, history, political science, economics, and philosophy.