Economic Capital

Girves and Wemmerus (1988) suggest that reliance on one's own financial resources lengthens the time to degree completion and increases the attrition rate. Yet, surveys show women are getting a near equal share of financial support from their institutions as compared with male full time enrollees (National Science Foundation 1995).

Social Capital

Studies also find that the social environment of the department, the quality of faculty and peer mentoring, and the prevalence of informal networks affect career aspirations and graduate retention (Dickey 1996; Holland 1994; Lewis 1996; Astin 1985; Pascarella and Terizini 1991; Sinton 1996; Tinto 1987). Mentoring relationships for women graduate students are less established and less likely to be with same-sex mentors (Heinrich 1991).

Symbolic Capital

Prestigious awards and assignments represent positive symbolic capital. Moore and Keith (1992), in a study of sociology departments, found that “firstyear statuses” such as prestigious research or teaching assistantship assignments or prestigious mentoring relationships affected student career aspirations. Negative symbolic capital, on the other hand, is represented by sexual or ethnic harassment which denigrates a person and undermines self-efficacy and self-confidence. Harassment amounts to negative symbolic capital; it threatens the person's credibility and status in the education community and undermines her/his advancement.

Cultural Capital

Typically, departments have used undergraduate GPAs and standardized test scores as predictors of graduate student achievement. These measures, however, have been unreliable in consistently and validly predicting performance beyond the first year (Arrow 1993). With regard to attrition rates, to date no evidence exists that indicates systematic attrition from graduate programs is caused by lack of academic performance (National Research Council 1992; Hornig 1987; Zwick 1991).

The hypothesis that we derive from this review of previous research is that symbolic and social capital have a greater effect than economic and cultural capital on the disparity in retention and attrition rates among graduate students and on overall levels of satisfaction with the graduate student experience.

We undertook a survey in spring 1997 to determine ways to improve retention and success rates among graduate students. The study's target population was all active graduate students in Ph.D.-granting institutions in the Midwest region (Appendix A on the APSA web site). Among the questions asked in a mailed questionnaire,¹

Although we do not know whether those who report having seriously considered leaving graduate school will indeed leave at some point, we use this question as one of two indicators of expected attrition. Using this first indicator, we evaluated the responses to other items in our questionnaire to identify which factors are related to the serious consideration of leaving graduate school before completing the degree objective. In Table 2, we present the set of factors that best predicts the probability of giving serious thought to quitting graduate school.²

While good mentoring works against attrition, the most important factor that contributes to the thought of leaving graduate school is knowledge of inappropriate conduct (either in terms of physical behavior or verbal communication) by a department member. Respondents were asked whether they know of any person in their department who has been the recipient of inappropriate behavior. Given the inequality of power between graduate students and faculty members, such inappropriate behavior, especially when it occurs as part of a pattern, is considered harassment by most institutions. For this query, 50% of women and 36% of men said they had knowledge of such inappropriate behavior (Table 3).

Evaluations of the availability of sufficient mentoring also show gender differences, with males registering higher levels of satisfaction with the degree of encouragement and consultation (Table 3). Thus, the two strongest predictors of the possibility of leaving graduate school before degree completion are embedded with a gender component.

Within the questionnaire, students were also asked: “Since entering graduate school have you (yourself) ever been involved in an interaction with a department member which involved behavior you perceived as inappropriate either in terms of physical conduct or verbal communication?” While a majority of both sexes respond negatively to this query, a profound difference emerges between the experiences of men and women. Women were much more likely to report an experience with or personal knowledge of inappropriate behavior than were men. That is, women are significantly more likely to have experienced inappropriate behavior by a faculty member (28%) than men (18%). However, it is the query about whether the respondent knows of any person in his/her department who has experienced such behavior towards them that emerges as the significant predictor of the possibility of leaving graduate school before completing the degree objective. Knowledge of harassment has a statistically significant effect on expected attrition. Thus, a climate of inappropriate behavior can contribute to lower completion rates.

Another factor that is significantly related to the serious consideration of quitting graduate school is having come directly to graduate school from a baccalaureate program (Table 2). Those who either worked for a while or who were in a previous graduate program are less likely to think about leaving their current program. A fourth factor influencing expected attrition is whether a department offers an orientation program to incoming students. This finding provides additional evidence that what departments do to help their graduate students does matter. The provision of an orientation program significantly reduces the probability of seriously thinking about leaving. A fifth factor, however, is beyond the department's control—the number of children that a graduate student has. The more children one has, the less likely one is to consider leaving graduate school.³

The second best predictor of dissatisfaction in graduate school is a scale that measures whether the respondent thinks that incidents of sexual or racial harassment would be handled promptly and appropriately by the department. The scale consists of five questions about the expected response of the department to sexual and racial harassment. In essence, we find that perceptions about the degree to which the formal organizational structure would address inappropriate behavior are powerful predictors of satisfaction with the graduate student experience. A negative perception of departmental response has a significant deleterious effect on satisfaction with one's graduate program. Because women more frequently experience such inappropriate behavior, the argument may be posited that, for them, the effect is more onerous.

Related again to the findings in Table 2, the provision of an orientation program for incoming students helps reduce the average level of dissatisfaction (Table 4). A more sophisticated statistical analysis of these relationships (not presented here) reveals that the presence of an adequate orientation program increases satisfaction with the graduate experience among women more than men. This gender disparity may very well reflect the greater isolation that women face in informal network structures, which would make a comprehensive formal orientation program all the more important for women.

Returning again to Table 4, we see that negative assessments or experiences with methods and statistics course materials work against satisfaction. Satisfaction is also related to whether the decision to attend graduate school depended upon the reputation of the department. The respondent could select from among several alternative criteria used to choose their graduate program. We had expected that the student's own a priori expectations about the department would affect his or her current satisfaction levels; indeed, this is the case. Those who selected their current program on the basis of reputation are more likely to be satisfied with their graduate student experience. If one selected the department for reasons other than reputation, such as for cost or location, then, on average, one is more likely to be dissatisfied. This finding supports conventional wisdom that more committed students are more satisfied with their program.

The last important independent factor is gender; males tend to have higher levels of satisfaction and females tend to have lower levels of satisfaction. This finding becomes even more telling when we recognize that controls have been introduced for other factors that we expected to be relevant, such as the quality of the mentoring relationships. The data reveal that the graduate student experience is different in the eyes of men than the eyes of women, and that women are unhappier on average than men. Is it that women are more disgruntled than males, do they have different evaluative standards, or are they experiencing a different environment that causes them to evaluate their departments differently? Previous research suggests the last explanation and terms it a chilly or hostile climate.

The finding that the receipt of an assistantship or fellowship has no impact on satisfaction levels is surprising. The data indicate that greater financial support, i.e. economic capital, does not affect satisfaction levels or a considered decision to leave a program in the same way that it appears to affect completion rates in other (non-political science) graduate programs. Our findings also suggest that women and men do not differ significantly with regard to degree objectives. Seventy-eight percent of women and 82% of men planned to get a Ph.D. upon entering graduate school. Once the degree is completed, however, men do have somewhat higher career goals than women. We will come back to this and other findings in part 2 of this article. This first part has shown clearly that faculty mentoring and departmental orientation programs do contribute to a favorable experience for graduate students. In part 2, we will also discuss ways to make both mentoring and orientation experiences as valuable as possible for graduate students.