In this issue:

Minority Ph.D. Production in SME Fields: Distributing the Work?

Context and Attrition

An Interview with Dr.Mary Louise Soffa

A Profile of an AGEP Institution: University of Puerto Rico

From the editors
 

Managing Editor:Yolanda S. George
Editor: 
Virginia Van Horne
Art Director:
Ann Williams

Making Strides is a free, quarterly (April, July, October, and January) research newsletter published by the American Association for the Advancement of Science, Directorate for Education and Human Resources Program. Its purpose is to share information about minority graduate education in the fields of science, mathematics, and engineering. It is available in print and electronic format. Inquiries, information related to AGEP, and all correspondence should be sent to the editor. 

By Dr. Barbara E. Lovitts, Senior Research Analyst, American Institutes for Research

Despite increased attention to doctoral student retention, graduate schools have faced persistently high attrition rates. Studies on persistence going back to 1950 (Tucker, 1964) show a consistent pattern of attrition by discipline (Benkin, 1984; Berelson, 1960; Bowen & Rudenstine, 1992; and Tucker, 1960). The sciences exhibit the lowest attrition rates (roughly 30 to 50%), the humanities the highest (roughly 50 to 70%), with the social sciences falling somewhere in between (roughly 40 to 65%).  Please see Figure 1. The variability in the attrition rates within discipline reflect not only variation in data collection methods, but, more importantly, variation among the departments from which the data were collected. A twofold problem thus arises: how to explain the consistent pattern of attrition exhibited by the disciplines while at the same time explaining variation in rates of attrition across departments within a discipline.  This paper presents a theoretical explanation for the stable rates of doctoral student attrition in terms of integration and assesses empirically how differences in structures and opportunities for integration within departments lead to observed differences in attrition rates across departments within disciplines.

The Relationship Between Attrition Rates and Disciplinary Structures

Standard rates across time reflect social structures and social forces that remain relatively unchanged from year to year (Durkheim, 1897/1951). These forces must be independent of individuals because the force acts with the same intensity, achieving the same end in the same numbers, on individuals who do not form a natural group and who are not in communication. Similarly, to paraphrase Durkheim, the regular recurrence of identical events in proportions constant within the same discipline but very inconstant from one discipline to another would be inexplicable if each discipline did not have a similar structure which affected its members with a similar force. Thus, one could expect standard patterns of variation of attrition across disciplines because of systematic differences in their intellectual and social structures. 

A discipline’s social structure is in large part determined by its intellectual structure. Intellectually, the sciences are highly structured disciplines. Their subject matter is vertically integrated and graduate students focus on mastering one or a few contemporary theories. Doctoral students in the sciences often chose, or are chosen by, an advisor by the end of their first year, and begin research on projects that will serve as the basis of their dissertations. Much of this research is done in teams, which ensures that doctoral students in the sciences are in frequent academic and social contact with faculty and fellow graduate students. (Bowen & Rudenstine, 1992; Wilson, 1965).

The humanities and many of the social sciences, by comparison, are more loosely structured. Their subject matter is horizontally integrated. Students in these disciplines frequently do not select an advisor or commence dissertation-related re-search until they have passed their qualifying examinations; and their research is often conducted in isolation in libraries, archives, or in the field. Consequently, doctoral students in these disciplines do not receive the same amount of academic and social support as their counterparts in the sciences. (Bowen & Rudenstine, 1992; Wilson, 1965). 

The description of doctoral education in the disciplines above only characterizes their intellectual and social structural differences in broad strokes. It does not explain how the academic and social interactions that are embedded in and develop out of these structures contribute to attrition. The next section develops the theory of integration and shows how the mechanisms that lead to integration are allocated differently in the different disciplines. 

The Mechanisms of Integration

The concepts of academic and social integration (Tinto, 1987, 1993) have been used extensively to explain both undergraduate and graduate student attrition. Lovitts (1997) elaborates on the mechanisms which contribute to each type of integration. Academic integration develops through formal interactions between and among graduate students and faculty as they work together on common tasks to achieve the primary goals of graduate education: intellectual and professional development. Social integration develops through informal, casual interactions between and among graduate students and faculty outside the classroom. The programmatic, social, and even physical structures of a department can facilitate or impede academic and social integration. 

The programmatic structure of a graduate department is related to the structure of the discipline. Some disciplines, like the sciences and laboratory-based social sciences, are structured around research teams. Students and faculty are in almost constant interaction all day long and sometimes late into the night. In such situations, the line between formal academic interaction and informal social interaction becomes blurred. Other disciplines, such as the humanities and nonlaboratory-based social sciences, are structured around individualized research that takes place in isolation in libraries, archives, and the field. Opportunities for social interaction to develop as a consequence of academic tasks is reduced. The differential attrition rate between students in the sciences and students in the humanities is consistent with this contention.

In the graduate school environment, academic and social integration are closely intertwined. Events such as weekly colloquia and brown bag lunches, on- or off-campus social hours, and departmental recreational activities that bring students and faculty into regular, informal contact foster an esprit de corps. Departmental traditions such as holiday parties, picnics, and the like heighten socioemotional integration between and among the graduate students and faculty who participate.  How a department’s physical space is structured has implications for both types of integration. Graduate lounges provide opportunities for students, as well as faculty who use them, to meet and interact informally. Group, as opposed to individual, offices for graduate students bring students into prolonged and informal contact. The frequent, and, sometimes chance, meetings and social exchanges that take place in hallways, and around mailboxes and coffee pots lead students to develop a sense of community membership. 

The extent to which departments provide structures and opportunities for integration and the relationship between department environments for integration, student integration in those environments, and attrition was tested empirically as described below.

Sample and Methods

The sample consisted of nine departments, three in each of the three major disciplines (sciences: mathematics, chemistry, biology; social sciences: economics, psychology, sociology; humanities: history, English, music) at two universities, Rural and Urban. 

Data on the departments’ structure and opportunities for integration were obtained through telephone interviews with the Directors of Graduate Study (DGS) and site visits to each department. The interviews explored the existence and nature of the departments’ formal and informal academic and social structures and activities for graduate students and for their professional development (see Table 1). During the visits, field notes were taken on the departments’ integrative environments. Special efforts were made to observe graduate student offices, to note student-student and faculty-student interactions, and to spend time in graduate student lounges.  Data on students’ actual integration in their departments came from the survey responses of 816 former doctoral students (511 completers, 305 noncompleters) who were members of the 1982-84 entering cohorts in the departments and universities noted above. This sample was 88% white. No analyses could be done by racial/ethnic group. Rather than asking students if the structure or activity was present in the department, the students were asked if they received or how frequently they participated in the structure or activity.

Table 2 presents this studies’ departments’ attrition rates by university. Note the large differences in attrition rates between Rural and Urban’s economics and sociology departments. The disimilarity indicates that the cause is not inherent in the discipline. 

Department Environments for Integration and Attrition

Using information reported by the DGSs or observed during the site visits, an overall integration score was calculated for each department. The overall scores were subdivided into academic and social integration scores based on the factors depicted in Table 1. When the departments’ three integration scores were compared with their attrition rates, two of the correlations achieved significance: overall integration and attrition (R = -.41, p = .044) and academic integration and attrition (R = -.54, p = .011). These results suggest that the more conducive the department’s environment for integration, academic integration in particular, the lower the department’s student attrition rate. 

Student Integration in the Department and Attrition

A score for students’ integration in each department was calculated by summing individual student responses to the survey questions and obtaining averages for each department. The overall student integration score was subdivided into an academic and social component.

The three student integration scores were compared with their departments’ attrition rates.  Although none of the correlations achieved significance, the signs of all the correlations were negative, as would be expected if the hypothesis, “The more integrated students are in the department, the lower the department’s student attrition rate,” were true.

The Relationship Between Department Environments 

To assess the relationship between the department environments for integration and actual student integration in the department, the department integration scores were correlated with the student integration scores. The correlation between the academic integration scores achieved significance in the predicted direction (R = .48, p = .021), indicating that the more opportunities a department has for academic integration, the more academically integrated students become. The correlation between the social integration scores also achieved significance, but not in the predicted direction (R = -.42, p = 0.42), suggesting that the more opportunities a department has for social integration, the less socially integrated students become. 

Discussion and Conclusion

From the evidence provided above, it appears that the intellectual structure of the discipline shapes opportunities for academic and social integration across departments within that discipline by structuring the nature of academic tasks and the frequency of academic interactions as well as the social relationships that develop out of task-related interactions. These differences in structures and opportunities for integration influence the characteristic and stable patterns of attrition that have been observed across disciplines for several decades. It also appears that different departments provide different structures or opportunities for integration and that these structures and opportunities affect student integration and persistence outcomes in a manner that is independent of the parent discipline. These differences in the structures and opportunities for integration help explain the variability in attrition rates among departments within a discipline. The evidence also indicates that persistence outcomes are affected more by opportunities for academic integration than by opportunities for social integration, as one would predict, because graduate students attend graduate school for academic not social reasons, and, consequently, their degree of integration into the academic systems of the department should, and does, matter more for persistence than their integration into the social systems.  

A more extensive elaboration can be found in Leaving the Ivory Tower: The Causes and Consequences of Departures in Doctoral Study. Rowman & Little-field, Boulder: Colorado (in press).

References

Benkin, E.M. (1984). Where have all the doctoral students gone? A study of doctoral student attrition at UCLA. (Doctoral dissertation, University of California at Los Angeles, 1984). Dissertation Abstracts International, 45A, 2770.

Berelson, B. (1960). Graduate education in the United States. New York: McGraw-Hill Book Company, Inc.

Bowen, W.G., & Rudenstine, N.L. (1992). In pursuit of the Ph.D. Princeton, NJ: Princeton University Press.

Durkheim, E. (1951). Suicide. (J.A. Spaulding & G. Simpson. Trans.). Glencoe, IL: The Free Press. (Original work published in 1897)

Lovitts, B. E. (1997). Leaving the ivory tower: A sociological analysis of the causes of departure from doctoral study. (Doctoral disseration, University of Maryland at College Park, 1996). Dissertation Abstracts International, 57A. (University Microfilms International, 9718277).

Tinto, V. (1987). Leaving college: Rethinking the causes and cures of student attrition. Chicago: University of Chicago Press.

Tinto, V. (1993). Leaving college: Rethinking the causes and cures of student attrition (2nd ed.) .Chicago: University of Chicago Press.

Tucker, A. (1964). Factors related to attrition among doctoral students (Cooperative Research Project No. 1146). Washington, DC: U.S. Office of Education.

Wilson, K.M. (1965). Of time and the doctorate: Report of an inquiry into the duration of doctoral study. Atlanta, GA: Southern Regional Education Board.