• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

The determinants of gender inequality in higher education

Dissertation
Author: Stephanie L. Ewert
Abstract:
A dramatic reversal of gender inequality in education occurred when women reached parity with men in college graduation rates around 1982 and surpassed men since then. While scholars have documented this remarkable turnaround in the gender gap in college completion, few studies have offered explanations for why this reversal occurred or why women currently earn a larger percentage of bachelor's degrees than men. The focus of my dissertation is on illuminating the mechanisms of the contemporary gender gap in college completion that favors women. To more completely answer this question, I also examine gender differentials in the college experience that may ultimately affect graduation, including selection of college major and attendance patterns. In order to examine the effects of prior life experiences on college major, attendance, and graduation, I approach these questions through the lens of the life course paradigm and analyze data from the National Education Longitudinal Study of 1988. I find that academic performance and behavior in high school and college are critical factors in shaping men's more disrupted attendance patterns and lower likelihood of graduation relative to women. College experiences, particularly attendance patterns, social integration, and academic performance, do not merely mediate the effects of background characteristics and prior life course events on the likelihood of graduation but are independently consequential for the gender gap in graduation. Although I find that women's lower interest and self-confidence in math partially account for the overrepresentation of men in engineering and related fields, I do not find support for other leading explanations for gender segregation in college major

TABLE OF CONTENTS Page List of Figures ii List of Tables iii Chapter 1: Introduction '. 1 Chapter 2: Gender and Schooling 15 Chapter 3: Data and Methods 41 Chapter 4: Gender Differences in College Major 67 Chapter 5: Gendered Pathways through College I l l Chapter 6: Fewer Diplomas for Men 156 Chapter 7: Conclusion 186 References 202 Appendix A: NELS Survey Questions 217 Appendix B: Categorization of College Major 218 I

LIST OF FIGURES Figure Number Page 1.1. Bachelor's Degrees Awarded to Men and Women 12 1.2. Median Age at First Marriage 12 1.3. Average Age of Mother at First Birth 13 1.4. Fertility Rate for Women 13 1.5. Divorces per 1,000 People 14 4.1. Proportion of Selected Fields of Bachelor's Degrees 103 5.1. Multiple Disrupted Attendance by Race 148 5.2. Multiple Disrupted Attendance by GPA 149 5.3. Multiple Disrupted Attendance by SES 150 ii

LIST OF TABLES Table Number Page 3.1. Weighted Frequencies of Dependent Variables 65 3.2. Test of Dependence Between Outcome Variables 65 3.3. Frequency of Independent Variables 66 4.1. Proportion of Men and Women in College Majors 103 4.2. Determinants of College Major by Gender 104 4.3. Logistic Regression of Majoring in Engineering, Math 105 4.4. Logistic Regression of Majoring in Education, Health 106 4.5. Multinomial Logistic Regression of College Major 106 4.6. Regression of Majoring in Engineering, Math; 1996 Entrants 107 4.7. Regression of Majoring in Education, Health; 1996 Entrants 108 4.8. College Major by Gender, Race, and Mode of Entry 109 4.9. Logistic Regression of College Major by Racial Group 110 5.1. Attendance Patterns By Sex 150 5.2. Determinants of Attendance Patterns by Sex 151 5.3. Multinomial Logistic Regression of Attendance Patterns 152 5.4. Regression with Different Measures of GPA 155 6.1. Determinants of College Graduation by Sex 181 6.2. Logistic Regression of College Graduation 182 6.3. Regression Results for Entrants by 1996 183 6.4. Regression Results for Four-Year Entrants 184 6.5. Propensity-Score Adjusted Regression 185 iii

ACKNOWLEDGEMENTS I would like to thank several people for their unwavering support during the dissertation-writing process. I thank Charles Hirschman for guiding me through the research process. His knowledge, patience, and encouragement were invaluable to me. I would also like to thank Becky Pettit for her feedback and support throughout the dissertation-writing process. And finally, I offer a great thanks to my family and close friends who gently encouraged me to persevere during the difficult points and celebrated with me during the high points. iv

DEDICATION I dedicate this dissertation to my parents, who have been my biggest supporters throughout my life. Their emotional, intellectual, and monetary support developed me into the person I am today, instilled in me a value of education, and prepared me for graduate school. Their interest in my work and confidence in my ability sustained me through graduate school. v

CHAPTER 1: INTRODUCTION The nation's founders created the United States as a gender-stratified society, and for decades women have occupied lower positions than men within the occupational structure, home, and public sphere (Massey 2007). Such patterns of inequality in the United States are generally well-entrenched and slow to change. However, a dramatic reversal of gender inequality in education occurred when women reached parity with men in college graduation rates around 1982 and surpassed men since then. While scholars have documented this remarkable turnaround in the gender gap in college completion, few studies have offered explanations for why this reversal occurred or why women currently earn a larger percentage of bachelor's degrees than men (cf. Buchman and DiPrete 2006). The focus of my dissertation, then, is on illuminating the mechanisms of the contemporary gender gap in college completion that favors women. To more completely answer this question, I also examine gender differentials in the college experience that may ultimately affect graduation, including selection of college major and attendance patterns. The female advantage in college completion warrants the attention of sociologists of education since higher educational attainment of women relative to men is likely to have implications for marital formation, childbearing, labor market participation, and male-female relations more broadly. Historical Trends * 1

The reversal of gender inequality in college graduation that advantaged women is situated within a broader revolution of women's roles within society. As journalist Gail Collins reflected on the 1960's and 1970's, she noted, "The nation had, in just over two decades, accepted a radical new view of women's place: that it was everywhere" (2003; p. 447). Indeed, changing norms regarding women's family responsibilities, changing patterns of family formation, occupational restructuring, declining discrimination, and increasing opportunities for women in the labor market dramatically reshaped women's place in society. These trends likely contributed to the advances women made in educational attainment. In the 1950s and 1960s, men held an overall advantage in educational attainment and completed college at higher rates than women. For example, men received 65 percent of all bachelor's degrees in 1960 (Buchman and DiPrete, 2006). As Figure 1.1 shows, women made significant gains in college completion in the late 1970s and reached parity with men in college graduation around 1980. Since then, women have earned an increasingly greater percentage of bachelor's degrees than men. The higher percentage of bachelor's degrees awarded to women compared to men reflects two distinct trends—the college graduation rates of men and the rates of women. Over the last several decades the percentage of women who went to college continued to grow while the proportion of men graduating from college remained constant or even declined slightly. The combination of these trends resulted in the current gender gap in which the majority of bachelor's degrees are awarded to women. Women earned 56 percent of all 2

bachelor's degrees in 2000 and 58 percent by 2004 (DiPrete and Buchman 2006; Buchman and DiPrete 2006; Bae et al. 2000). This turnaround of the gender gap in college completion represents the reversal of a formerly persistent pattern of inequality which favored men. Besides the expansion of higher education that created available space for women to fill, scholars have hypothesized that declining discrimination, economic restructuring, and the transformation of women's roles within the home directly contributed to the advances women made in educational attainment. Declining discrimination against women contributed to women's advancement in college graduation in several ways (Buchman and DiPrete 2006). As women faced greater opportunities in society as a result of declining discrimination, girls received more encouragement and support from their parents to pursue higher education (Reynolds and Burge 2008; Dumais 2002). Daughters did not historically have equal access to parental resources, but as discrimination against women declined, parental investments in children changed and parents shifted family resources toward girls and now more equally invest in their daughters and sons (Buchman and DiPrete 2006). With a broader range of available opportunities and increased support and encouragement from family and friends that resulted from declining discrimination, women pursued higher education at greater rates than before. Changing economic opportunities also contributed to the advances women made in educational attainment. Economic restructuring, in combination with 3

declining discrimination, created opportunities for women to hold white collar jobs. Many available jobs required a college degree and so influenced women's decisions to pursue college (Goldin et al. 2006). Expanded opportunities to participate in the labor market through jobs that required college degrees necessitated that women pursue college degrees at higher rates. Although declining discrimination and occupational restructuring contributed to the rise in women's college completion rates, scholars have also hypothesized that social and cultural changes in family roles and responsibilities further facilitated this trend (Goldin 2006). Trends related to marital formation, childbearing, and divorce have influenced young women as they plan their lives. The median ages at marriage and childbearing rose during the same time that women's college completion rates increased. Figure 1.2 shows that in 1970 the median age at marriage for women was under 21 years old, but by 1982 the median age had risen to just over 22 years old, the age at which traditional college students graduated. By 2006, the median age of marriage for women was almost 26 years old, well beyond the age at which most students complete college. In 1960, just under 30 percent of women aged 20-24 had never married, but by 1998 a little over 70 percent of this age group remained unmarried (Casper and Bianchi 2002). As women married at older ages, they needed to be able to provide for themselves during the time between when they left their parent's houses and got married, thus encouraging educational attainment (DiPrete and Buchman 2006). Cohabitation rates increased as people married later and so 4

individuals still formed partnerships even though they delayed legal marriage. However, since cohabiting couples frequently maintain separate budgets, women in these relationships likely needed to provide for themselves. As women married at later ages, they bore their first children at later ages as well. Figure 1.3 shows that the average age of first birth was 21.4 years in 1970 but had risen to 25 years in 2006. Increased access to the oral contraceptive (the Pill) also contributed to delayed childbearing and increased the educational attainment of women (Edlund and Machado 2009). Not only did women have children at later ages, but they also had fewer children. Figure 1.4 shows consequential changes to the total fertility rate during the 1970s and 1980s. During the 1970s as women married later, the number of children women bore decreased. The fertility rate remained relatively constant during the late 1970s to the mid 1980s, and then increased modestly by the late 1980s when women had firmly established their advantage in college completion. Women who bore fewer children faced fewer family responsibilities and could expect to spend more years in the labor market. Therefore, more women went to college as women had fewer children in order to gain the skills needed to participate in the labor market. Furthermore, the rising divorce rate, in combination with the rising median age at marriage, led women to expect to provide for themselves for longer periods of time. Therefore, a college degree represented a form of insurance against poverty and so women also pursued higher education to ensure against income deprivation (DiPrete 5

and Buchman 2006). Figure 1.5 illustrates that the number of divorces increased by over 50 percent between 1970 and 1980. As the divorce rate rose, married women faced the possibility that they would not be able to depend on spousal income for financial support in the long term so needed to be able to support themselves should the marriage end. More women expected to be breadwinners as these marital trends changed, necessitating that women earn college degrees in higher numbers in order to find jobs with which they could support themselves. Scholars argue that these trends in educational attainment and family formation are linked. Women could expect to spend more years in the labor force as they married later, bore children later, had fewer children, and faced greater likelihoods of divorce. Consequently, women invested more in education (Goldin 2006). As expectations and norms regarding women's responsibility at home began to change, women came to understand that they could have both a family and a career and so pursued higher education at greater rates (Buchman et al. 2008; Goldin 2006; Walters 1986; Marini 1984). Women's educational attainment was responsive to cultural changes in the timing of family formation and so women pursued additional education as they married and bore children later. Although declining discrimination, occupational restructuring, changing norms and patterns of family formation, and the expansion of higher education may help to explain why women caught up to men in college graduation rates, these explanations cannot fully account for why women surpassed men. That is, why are women now 6

more likely than men to graduate? In order to shed light on this puzzling female advantage in college graduation, this study examines contemporary gender disparities in higher education. Instead of accounting for the well-documented trends in educational attainment over time, this dissertation provides a detailed study of a time point after the firm establishment of the female lead in college graduation. Through an in-depth study of one cohort, I explore the factors leading to the higher rate of college graduation among women than men. Gender and Education Although gender disparities have existed over time in schools, educational research has not traditionally focused on inequality between men and women. Jacobs (1996) observed, "I remain surprised at how much mainstream research in the sociology of education ignores women, and how much of the rest considers gender interactions rather than gender inequality. In other words, gender often becomes a matter of variations on the main theme of socioeconomic or racial inequality. My first recommendation for further research, then, is that gender deserves the attention of sociologists of education" (p. 177). Taking up Jacobs' call to action, I focus explicitly on contemporary gender inequality in higher education. My primary research question is why women are more likely than men to graduate from college. However, in order to more completely answer this question it is necessary to examine where men and women differ on other key aspects of the college experience that may ultimately affect graduation. A student's chosen field of 7

study is central to the college experience and may affect the likelihood of graduation depending on the rigor of the program and the number of required courses associated with a particular major. Therefore, I examine why women are underrepresented in math-based majors such as engineering and the physical sciences and overrepresented in "nurturing" majors such as education and health-related fields. College attendance patterns are also consequential since students who follow disrupted attendance patterns are less likely to graduate. Hence, I also explore why women are less likely than men to attend part time or discontinuously during college. I approach these questions through the lens of the life course paradigm, arguing that students' early educational experiences interact with societal forces and other domains such as family and work to shape their later educational trajectories and outcomes. In order to examine the effects of prior life experiences on college major, attendance patterns, and graduation, I analyze data from the National Education Longitudinal Study of 1988 (NELS), a longitudinal study that tracked a nationally representative sample of students from their eighth grade year in 1988 until 2000. These longitudinal data enable me to explore how high school performance and other background factors affect college majors and attendance patterns, as well as how these important undergraduate experiences influence eventual college graduation. These research agenda fall in the middle of heated public debate regarding the fate of men and women in education. Researchers have historically focused their attention on the problems of one sex or the other depending on society's opinions 8

about which sex is being better served by the education system at the time (Entwisle et al. 1997). There is little consensus on which sex is better served by the contemporary system. Citizens arguing for continued attention to women in education downplay the importance of the female lead in college graduation and stress that women remain disadvantaged in important realms of education including underrepresentation in math and science majors, poorer performance on standardized tests, and gender-biased interpersonal interaction with faculty (Sax 2007; AAUW 1995; Sadker and Sadker 1995). Others argue that efforts made by education officials to raise the performance of girls in school resulted in dramatic improvement for women at the expense of men, necessitating a new focus within education on improving the position of men (Mortenson 2008; Weaver-Hightower 2003; Gose 1999). Renewed interest in single- sex classes and schools developed out of this concern for the performance of boys in school (Younger and Warrington 2006). This research does not address these ideological dimensions of gender inequality in education, but recognizes these differing opinions regarding where the focus of research on gender and education should lie. I do not view education as a zero-sum game in which women achieve at the expense of men or vice versa. Instead, I pursue this research that studies why women are advantaged in some areas of higher education and disadvantaged in others in hopes that educators and policymakers can develop strategies to improve the education of all students. Past social movements 9

and policies successfully reshaped education, but only after understanding the location and magnitude of contemporary gender disparities in education will educators and policymakers be equipped to design programs to minimize these disparities. Dissertation Outline Six chapters comprise the remainder of this dissertation. Chapter two extensively reviews the literature on gender and educational attainment. I identify where the experiences of men and women diverge throughout the education trajectory, why these gender differences arise, and how they might affect the gender gap in college graduation. A comprehensive understanding of the theoretical and substantive findings of prior research guides the hypotheses tested in this study and the measurement of variables included in the analytic models. Chapter three provides an in depth discussion of the NELS data and a description of the measurement of key concepts. This chapter also presents an overview of the analytic strategy of the empirical chapters. Chapter four empirically tests the determinants of gender differences in college major, using logistic regression models to test leading hypotheses for why women and men select different majors. These explanations include: (1) men have greater confidence in their math ability and are mathematically better prepared for college than women, (2) women place greater value than men on intrinsic rather than extrinsic rewards of professions, (3) women perceive more limited career opportunities than men in the math and science fields, (4) women choose fields with more flexible 10

schedules and fewer penalties for time out of the labor force because of their anticipated childrearing activities, and (5) men and women engage in gendered extracurricular activities that develop particular skills that prepare them for different majors. Chapter five uses multinomial logistic regression to explore the underlying mechanisms behind gendered attendance patterns. I test several hypotheses that may explain gender gaps in attendance patterns, including that poorer academic performance for college leads to more disrupted attendance patterns for men and that institutional selectivity and family responsibilities suppress the extent of gender gaps in disrupted attendance patterns. Chapter six utilizes logistic regression to examine the determinants of gender inequality in college graduation, with a focus on how critical college experiences contribute to the contemporary gender gap. The concluding chapter summarizes and integrates the findings from the three empirical chapters. The conclusion further seeks to explore the status of gender and higher education more broadly and identify conceptual and empirical obstacles that researchers must confront to advance the field. 11

0.60 0.40 0.35 0.3 0 ~T 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 r o r»- 6961- CM 1^- cn cn CO CO o co cn CM co co CD oo CO CO o CD oo 00 cn 00 CO 00 cn o cn cn co cn Year CM cn T— cn cn cn CO cn cn to cn in cn cn oo cn cn cn o o O) cn cn CM o o o CM o CO o o CM CD O m o o CM Males Females Figure 1.1. Bachelor's Degrees Awarded to Men and Women in the United States, 1970-2006 Source: Data from the U.S. Department of Education. 2007. Digest of Education Statistics. Figure 1.2. Median Age at First Marriage in the United States, 1970-2006 Source: Data from the U.S. Census Bureau. 2007. Families and Living Arrangements. 12

25.0 1985 1990 -•? Year • - Figure 1.3. Average Age of Mother at First Birth in the United States, 1970-2006 Source: Mathews and Brady 2009 20051 2006 Figure 1.4. Fertility Rate for Women in the United States, 1970-2003 Source: Data from the U.S. National Center for Health Statistics, Vital Statistics of the U.S. 13

f Q. § 0) a. (0 o e o > 1,300 1,200 1,100 1,000 900 800 700 600 500 - i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — r o f o f <£* o f o f o^

CHAPTER 2: GENDER AND SCHOOLING Life Course Paradigm Education research on gender has historically highlighted the male advantage in educational attainment, but more recent trends suggest that women are now advantaged in overall educational attainment as well as in various educational experiences throughout the schooling years. The field of education studies lacks a comprehensive theoretical framework in which to understand the numerous gendered educational experiences. Although broad in scope, the life course paradigm provides a useful framework for studying education and gender differences. This paradigm asserts that early life course decisions, opportunities, and conditions affect later outcomes and that the various domains of peoples' lives such as family, work, and school interact with one another (Xie and Shauman 2003; Lucas 2001; Cavanagh et al. 2006). People live multidimensional lives with different areas, or trajectories, which require their time and energy, including family trajectories, work trajectories, and education trajectories (Xie and Shauman 2003). Events from one life course trajectory can interact with and shape outcomes in other trajectories (Elder 1977). For example, a significant event in the family trajectory, such as marriage, may affect college graduation in the educational trajectory. Furthermore, early events within a trajectory will influence later events in the same trajectory. Therefore, the life course perspective would suggest that not only background factors and high school 15

experiences, but also college experiences and other life course events during the college years, contribute to college completion. Although I primarily focus on events within the education trajectory, I consider how consequential factors in other trajectories such as the family trajectory relate to educational outcomes. The life course paradigm suggests that periods of life, such as childhood and old age, influence positions, roles, and rights in society, and are often based on culturally shared age definitions. Yet although students typically progress through the education trajectory in an age-graded fashion, the life course perspective also allows for variability. That is, age-graded norms are not deterministic and so there is individual variance. Ascribed characteristics such as race, class, and gender, and student characteristics such as personality, are key dimensions along which patterned differences in educational progression, dropout, performance, and attainment occur. Gender consequentially shapes various experiences and outcomes of students as they progress through the education system (Jacobs 1996). Therefore, in order to truly understand why women currently attend and graduate from college at higher rates than men, it is helpful to examine where the experiences of men and women diverge at each stage of the education system and how differences at earlier stages affect later outcomes. Highlighting gendered processes throughout the education trajectory, for example during high school, in college enrollment, and during college, will identify the mechanisms through which gender differences in the college outcomes of focus in this dissertation are structured. In the 16

remainder of this chapter, I identify consequential educational outcomes throughout the education trajectory, how these outcomes vary by gender, why they vary by gender, and how such differences may contribute to the contemporary gender gap in college graduation. The High School Years Education researchers frequently focus on the effects of family background and academic performance in high school on various educational outcomes to determine whether these outcomes are achieved or ascribed. Socioeconomic status undisputedly affects high school performance such that students from more privileged backgrounds earn higher grades, are more frequently placed in college preparatory classes, and are more likely to graduate from high school and enroll in college (Lareau 2003; Alexander 1974; Kane 1994; Buchman and DiPrete 2006). Students of high socioeconomic background benefit because their parents earn more so can pay for extra educational opportunities like summer camps and private schools, are more educated so are able to help their children with homework, and understand the education system so are able to intervene to ensure that their children get needed help and are placed in high track classes (Lareau 2003; Lucas 2001). Parents of higher socioeconomic status may provide more encouragement for their children to earn high grades in high school and pursue college than do parents in lower socioeconomic positions (Pascarella et al. 2004). 17

Although there are not gender differences in family background, the effects of high socioeconomic status may differ by gender if parents differentially allocate encouragement and resources to their male and female children (Buchman et al. 2008). Prior to the 1980's, parents were less likely to encourage their daughters than sons to excel in school and pursue college degrees since women had more limited educational and occupational opportunities at the time (Reynolds & Burge 2008). However, parental encouragement for girls increased as occupational opportunities expanded so that girls are now equally encouraged at home to pursue high levels of educational attainment (Reynolds & Burge 2008). Prior to the 1980's family background affected the academic attainment of girls more than boys and women were less likely than men to pursue higher education unless both parents had at least some college education (Alexander and Eckland 1974; Buchman and DiPrete 2006). Research among recent cohorts suggests that the effects of family background have changed and that mother's level of education is increasingly related to the educational attainment of daughters and father's education continues to be positively associated with the attainment of their sons (Buchman and DiPrete 2006). Scholars typically measure academic performance as high school grades, test scores, or enrollment in college preparatory curriculum instead of vocational or general education classes. Studies consistently find that academic performance has the largest effect of all covariates in models predicting a variety of educational outcomes (Buchman and DiPrete 2006; Mullen et al. 2003; Hearn 1991). Girls earn 18

better grades than boys in all subjects and earn higher test scores in English and writing, and scholars have documented this female advantage for decades (Bae et al. 2000, McCornack & McLeod 1988; AAUW 1995). Boys also have lower high school class rank than girls on average (Conger and Long 2010). In contrast, boys tend to earn higher test scores in math (Xie & Shauman 2003; Bae et al. 2000; Oakes 1990). Schools were less likely to place girls in the college track prior to the 1980's, but by 1982 girls' probability of placement in the college track was already four percentage points higher than boys' (Gamoran and Mare 1989). Although research consistently documents these sex differences in academic achievement, few scholars offer explanations that account for gendered patterns in both grades and test scores. Explanations for why girls score lower than boys on math tests by the end of high school include: teachers are less likely to encourage girls than boys to enter math and science fields, boys are more likely to take advanced math courses, girls perform more poorly on standardized, multiple choice, and timed tests, and girls experience more anxiety about math (AAUW 1995; Downey and Yuan 2005). According to the "stereotype threat hypothesis," negative stereotypes of women's math capabilities prevent girls from identifying with the math domain at school which depresses their math scores on standardized tests (Steele 1997). However, these explanations do not account for girls' higher math grades compared to boys or girls' higher scores on standardized reading and writing tests. 19

Full document contains 233 pages
Abstract: A dramatic reversal of gender inequality in education occurred when women reached parity with men in college graduation rates around 1982 and surpassed men since then. While scholars have documented this remarkable turnaround in the gender gap in college completion, few studies have offered explanations for why this reversal occurred or why women currently earn a larger percentage of bachelor's degrees than men. The focus of my dissertation is on illuminating the mechanisms of the contemporary gender gap in college completion that favors women. To more completely answer this question, I also examine gender differentials in the college experience that may ultimately affect graduation, including selection of college major and attendance patterns. In order to examine the effects of prior life experiences on college major, attendance, and graduation, I approach these questions through the lens of the life course paradigm and analyze data from the National Education Longitudinal Study of 1988. I find that academic performance and behavior in high school and college are critical factors in shaping men's more disrupted attendance patterns and lower likelihood of graduation relative to women. College experiences, particularly attendance patterns, social integration, and academic performance, do not merely mediate the effects of background characteristics and prior life course events on the likelihood of graduation but are independently consequential for the gender gap in graduation. Although I find that women's lower interest and self-confidence in math partially account for the overrepresentation of men in engineering and related fields, I do not find support for other leading explanations for gender segregation in college major