Font Size: 
Indicator II-17 Attrition in Doctorate Programs
NOTE TO READERS: Please include the following reference when citing data from this page: "American Academy of Arts and Sciences, Humanities Indicators,"
Back to Section II-B

Updated (3/25/2011)

Attrition in humanities doctorate programs is a topic of considerable interest to higher education researchers and administrators, but data that could be used to systematically assess the extent of attrition have been scarce. Information compiled by individual universities and programs suggests that attrition rates are substantial, but just how many people begin work toward a humanities Ph.D. and then drop out—and, more important, why they drop out—are significant questions that have long gone unanswered.

Fortunately, three recent studies enhance our understanding of graduate attrition. Findings from the first of these, the Council of Graduate Schools' Ph.D. Completion Project, were published in the autumn of 2007. Supported by funding from the Ford Foundation and Pfizer, Inc., the project involved 29 U.S. and Canadian research universities in collecting data on doctorate completion rates, as well as on interventions designed to raise these rates.

The Andrew W. Mellon Foundation–funded Graduate Education Initiative, another new source of data on attrition, involved both the implementation of a set of interventions designed to improve graduate education in 54 humanities departments in ten major universities and an evaluation of the ten-year project’s outcomes. The findings of the evaluation are described in Educating Scholars: Doctoral Education in the Humanities (Princeton University Press, 2009).

The attrition data presented by the Humanities Indicators (HI) are from a third recent study, a comprehensive assessment of U.S. research-doctorate programs administered by the National Research Council (NRC) and funded by the National Institutes of Health, the National Science Foundation, the Alfred P. Sloan Foundation, the Andrew W. Mellon Foundation, and participating universities. The assessment involved the collection of a variety of data on doctorate programs, including Ph.D. completion rates. These data were then used to develop multidimensional ratings of programs at approximately 200 institutions of higher learning (see for more information about the project). While each of the three studies mentioned here is a source of important insight regarding attrition in humanities doctorate programs, the HI has used data from the NRC assessment because this study included the largest number of programs in the greatest variety of humanities disciplines.

Doctorate completion was defined by the NRC as obtaining a degree within eight years of entering a Ph.D. program for students of the humanities, and within six years for students in other fields. A completion rate—essentially the proportion of all students entering a doctorate program who completed their Ph.D.’s within the specified number of years—was computed for every research doctorate program at the participating institutions.1 Students who finished their doctorates but not in the specified number of years were not counted as completers. The completion rates presented here are thus conservative measures of doctorate completion. The figures are also for a group of doctorate programs in a given field (or discipline), not the student population in that field. The determination of the share of all doctoral students in a given field who ultimately obtain their Ph.D.’s will require detailed analysis of individual programs’ responses to the NRC survey.

Figure II-17a depicts the interquartile range (IQR) for doctorate program completion rates in the humanities and several other fields. The IQR is widely used as a means of describing the “typical” or “usual” values exhibited by a group of persons or objects and involves excluding the most extreme values of a particular variable (in this case, doctorate program completion rate). Quartiles are statistics that divide the observations in a batch of numeric data into several groups, each of which contains 25% of the data. The lower, middle, and upper quartiles are computed by ordering the values for a particular variable from smallest to largest and then finding the values below which fall 25%, 50%, and 75% of the data. The lower quartile and the upper quartile are the two values that define the interquartile range. The middle quartile is also known as the median.

Figure II-17a, Full Size
Supporting Data Supporting Data

The doctorate program assessment data reveal that the completion rate in the humanities is similar to that in the mathematical and physical sciences field. In both fields, the middle half of the programs graduated from slightly more than a quarter up to 55% of their students within the specified number of years (i.e., eight years for students of the humanities, six years for science students). The median program completion rate for both fields was 42%. The engineering and biological and health science fields had the highest median completion rates, 50%. The field with the lowest median completion rate, 35%, was the behavioral and social sciences. This field also had a somewhat greater range of “typical” completion rates (IQR) than others.

Median completion rates among the humanities disciplines showed considerable variation (Figure II-17b). With 56% of their students completing their Ph.D.’s within eight years, theater and performance studies programs had the highest median completion rate. Languages, societies, and cultures programs had a median completion rate of 33%, the lowest recorded within the humanities field. Programs in two of the most populous disciplines, history and English language and literature, had rates of 42% and 46%. The span of IQRs among the disciplines was similarly broad. While the completion rates of German programs were clustered relatively tightly around the median (for an IQR of 20, the lowest of all the disciplines), typical completion rates for French programs, those with the highest IQR, ranged from 17% to 64%.

Figure II-17b, Full Size


1 Doctorate programs at participating institutions were asked by the NRC to report the number of “graduate students who entered in different cohorts from 1996–1997 to 2005–2006 and the number in each cohort who completed in 3 years or less, in their 4th, 5th, 6th, 7th, 8th, 9th years, and in 10 or more years. To compute the completion percentage, the number of doctoral students for a given entering cohort who completed their doctorate in 3 years or less and in their 4th, 5th, 6th years were totaled and the total was divided by the entering students in that cohort. This computation was made for each cohort that entered from 1996–1997 to 1998–1999 for the humanities and 1996–1997 to 2000–2001 for the other fields. Cohorts beyond these years were not considered, since the students could complete in a year that was after the final year 2005–2006 for which data were collected. To compute the average completion percentage, an average was taken over 3 cohorts for the humanities and over 5 cohorts for other fields” (National Research Council, Committee to Assess Research-Doctorate Programs, “A Data-Based Assessment of Research-Doctorate Programs in the United States: Data Table in Excel (2010),”, under “Guide” tab in Excel workbook).

Back to Top
Skip Navigation Links.  

View figures and graphics: