Three Magic Letters: Getting to Ph.D. (review)

In Three Magic Letters: Getting to Ph.D., Michael T. Nettles and Catherine M. Millett investigate a set of measures of doctoral student progress and productivity that they hypothesize serve as proxies for student learning and development. Their conceptual model proposes a set of influences on those outcome measures. This book is not a direct investigation of pedagogy or learning, but rather an exposition of doctoral student experiences and resources assumed to promote learning and how these experiences differ by discipline, gender, and ethnicity.

The book is a lengthy and detailed research report, and so describing the study is the best starting point. The study is a survey of doctoral students (supplemented with data on Ph.D. completion by 2001) enrolled in the fall of 1996 at one of 21 major doctoral-granting universities in one of 16 distinct disciplines that are collapsed into five fields of study for the analysis: education, engineering, humanities, science and math (including physical and biological sciences), and social sciences.

Strategic and stratified sampling provided sufficient numbers of respondents to make statistically significant claims about the differences among five racial-ethnic identity groups (African American, Hispanic, Asian American, White, and international). An impressive 70% response rate yielded a total sample of 9,036 respondents. "The Survey of Doctoral Student Finances, Experiences, and Achievements" asks questions that cover a wide range, including such important but rarely investigated features of doctoral student life as office space, stopping out of graduate school, and access to a mentor as distinct from an advisor.

Because of the sampling strategies and the scope of the questions on the survey, this survey will probably stand as the definitive dataset on the doctoral student experience for some time to come.

Two strengths of the book are its clear organization and its nearly encyclopedic exposition of the literature, particularly prior research studies. After devoting three chapters to establish the context and methods of the study, Chapters 4–10 describe the survey findings on a particular topic: educational paths prior to doctoral study and choice of doctoral program; financing doctoral studies, including the prevalence of fellowships, research assistantships, teaching assistantships, and debt; social and academic interactions with faculty members and other students (which the authors call "socialization"); research productivity; satisfaction; rate and continuity of progress, time to degree, and completion rates.

Each chapter describes the response patterns for a number of variables. They are reported as simple descriptive statistics disaggregated by the five areas of study, by gender, and by racial/ethnic identity group. The robustness of the sample and the range of findings (49 are listed in the "Summary of Highlights") make the book compelling. Over and over, for example, the evidence shows that African American students are disadvantaged in doctoral studies.

After this avalanche of descriptive data, Chapter 11 uses logistic regression to determine the predictors of 15 experiences that the authors identified as both measures and predictors of student success. Four are financial, such as ever having a fellowship or ever having a research assistantship. Six are about personal relationships with peers and mentors. Five are outcome measures such as overall satisfaction with the program, stopping out during the program, and time to degree.

We learn, for example, that younger students in engineering, science, and the social sciences are more likely to receive a fellowship offer upon entering the doctoral program. Specifically, engineering students entering at age 25 are 4.3 times more likely to receive a fellowship offer than those who enter at age 45.

The book is a gold mine of such pieces of information. And this is simultaneously a strength and weakness. In most cases, the data are offered without explanation or discussion, although the final two chapters show the complex interactions among variables, interlaced with some interpretation. To return to the example about fellowships, one might ask: Are fellowships used to lure students into programs and, hence, are more likely to be given to younger...