Organizational practices and gender diversity: Factors contributing to differential participation in Computer Science and Engineering doctoral programs

The present study examines the sources of variation in gendered outcomes among Computer Science and Engineering (CSE) doctoral programs. The data was collected in 2003 for a project funded through the Computing Research Association by a grant from the National Science Foundation. Joanne M. Cohoon was Principal Investigator and contact. Survey data from a nationally representative sample of 48 CSE departments are examined to identify gender gaps in enrollment, thoughts of leaving the program, and reports of academic and social support from within the department. Responses from faculty are used to construct measures of department characteristics, and responses from graduate students and chairpersons are used to construct measures of outcomes. Factors tested in a logistic regression model for gendered effects in the admissions process are gender stereotyping, criteria representing a CSE occupational schema, formality of criteria, diversity practices, and attitudes towards increasing diversity. Results show that criteria representing a CSE occupational schema predict a lower enrollment of women, while consideration of minority status in admissions decisions predicts a higher enrollment of women. Preliminary evidence indicates that formalized criteria mediate unfavorable effects of gender stereotyping, and that attitudes supportive of increasing diversity mediate effects of diversity practices in expected directions. Factors tested for gendered effects in program participation using hierarchical linear modeling include gender stereotyping, diversity attitudes, diversity practices, faculty support of students, and faculty promotion of a competitive versus helping ethic towards students. Results show that research orientation (a control variable), diversity attitudes, a certain type of faculty-student support, and formalizing the flow of information predict gender differences in participation. The findings contribute to understanding how interactional and structural mechanisms operate together in organizations to influence differential outcomes. The discussion draws on the literature on gender and science, gender and occupations, and diversity and organizations.