Programs for Undergraduate Women in Science and Engineering: Issues, Problems, and Solutions

Abstract

We analyze programs for undergraduate women in science and engineering as strategic research sites in the study of disparities between women and men in scientific fields within higher education. Based on responses to a survey of the directors of the universe of these programs in the United States, the findings reveal key patterns in the programs’ (1) definitions of the issues of women in science and engineering, (2) their solutions to address the issues, (3) their goals and perceived success with goals, and (4) their organizational characteristics and relationship to the larger institutional environments. The findings—which are conceptually grounded in the distinction between structural/institutional and individual issues facing women in science—have implications for understanding gender, science, and higher education, and for initiatives undertaken to improve the condition of women in scientific fields. The findings may also inform strategic efforts to reduce gender disparity in other organizational contexts.

Full text

http://gas.sagepub.com/
Gender & Society
http://gas.sagepub.com/content/25/5/589
The online version of this article can be found at:
DOI: 10.1177/0891243211416809
2011 25: 589Gender & Society
Mary Frank Fox, Gerhard Sonnert and Irina Nikiforova
Problems, and Solutions
Programs for Undergraduate Women in Science and Engineering : Issues,
Published by:
http://www.sagepublications.com
On behalf of:
Sociologists for Women in Society
can be found at:Gender & SocietyAdditional services and information for
http://gas.sagepub.com/cgi/alertsEmail Alerts:
http://gas.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
http://gas.sagepub.com/content/25/5/589.refs.htmlCitations:
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

PROGRAMS FOR UNDERGRADUATE
WOMEN IN SCIENCE AND
ENGINEERING
Issues, Problems, and Solutions
MARY FRANK FOX
Georgia Institute of Technology
GERHARD SONNERT
Harvard University
IRINA NIKIFOROVA
Georgia Institute of Technology
We analyze programs for undergraduate women in science and engineering as strategic
research sites in the study of disparities between women and men in scientific fields within
higher education. Based on responses to a survey of the directors of the universe of these
programs in the United States, the findings reveal key patterns in the programs’ (1) defini-
tions of the issues of women in science and engineering, (2) their solutions to address the
issues, (3) their goals and perceived success with goals, and (4) their organizational
characteristics and relationship to the larger institutional environments. The findings—
which are conceptually grounded in the distinction between structural/institutional and
individual issues facing women in science—have implications for understanding gender,
science, and higher education, and for initiatives undertaken to improve the condition of
women in scientific fields. The findings may also inform strategic efforts to reduce gender
disparity in other organizational contexts.
Keywords: engineering, gender, higher education, organizations, science
AUTHORS’ NOTE: The research reported here was supported in part by a grant from the
National Science Foundation (SES-0080638).
GENDER & SOCIETY, Vol. 25 No. 5, October 2011 589-615
DOI: 10.1177/0891243211416809
© 2011 by The Author(s)
589
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

590 GENDER & SOCIETY / October 2011
INTRODUCTION
S
egregation by field in higher education is a key mechanism in
maintaining occupational sex segregation. Women’s limited access to,
and participation in, particular educational specialties in higher education—
especially areas considered “elite fields”—produce and perpetuate the
existing segregation of women and men in occupations (Charles and
Bradley 2002). Occupational segregation of women and men is
important because it creates and sustains disparities in their earnings,
job autonomy, and opportunities for advancement in workplaces (Reskin
and Bielby 2005).
In this article, we focus on educational fields that are emblematic of
this segregation and disparity: science and engineering. We examine the
ways programs for undergraduate women in science and engineering in the
United States express issues, problems, and solutions for improving the
participation of women in science and engineering as “individual,” as
opposed to “structural/institutional,” concerns, and the ways programs
relate to the organizational settings of higher education in which they are
located. The findings reveal why programmatic efforts to improve the
status of women among undergraduates in science and engineering are
challenging, and why numerous organizational efforts fail.
Disparities in the number of undergraduate majors and degrees awarded
to women compared with men in science and engineering fields—and in
the quality of the experience while working toward those degrees—are
persistent concerns. Disparity persists particularly in computer and
information science, where women receive 21 percent of the undergraduate
degrees, and in engineering, where women receive only 19 percent of the
degrees. These percentages contrast markedly with women’s attainment
of undergraduate degrees in higher education broadly, where, across all
fields, 58 percent of undergraduate degree recipients in the United States
are women (CPST 2008, table 3-13).
Undergraduate education, particularly, is critical to understanding and
addressing disparities between women and men in science and engineering.
To pursue advanced study in science and engineering, one usually needs
to have an undergraduate degree in a scientific field. In scientific fields,
education progresses in a structured sequence of study: “Students master
a given unit and progress to the next, complete a given course and advance
to the following” (Fox and Stephan 2001, 109). If students are not tracked
into a scientific field by the time of undergraduate study, it is very difficult
to step into that sequence of study at a later stage. In fact, the undergraduate
degree level is acknowledged to be the latest point of standard entry into
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 591
scientific fields (Xie and Shauman 2003, 96). This contrasts with fields
outside of science that people can enter and pursue from a range of
educational backgrounds, and at later (and more variable) stages of the
life course.
Science and engineering are a revealing case for the study of gender in
society. First, scientific fields are influential and powerful. Scientific
fields embody “authoritative knowledge” (Derber, Schwartz, and Magrass
1990), and based on this authority, science and engineering define what is
taken for granted by literally billions of people (Cozzens and Woodhouse
1995). Scientific and technological advances and products play key roles
in shaping modern society. Hence, to be in control of scientific research
programs is to be involved in directing the future (Wajcman 1991). Science
and engineering have become gauges of national resourcefulness and
prestige (Montgomery 1994). Numerous American policy makers have
voiced concern that if the system of higher education fails to produce a
sufficiently large cadre of skilled scientists and engineers, the economic
and political stature of the country will suffer (National Academy of
Sciences 2003). Second, gender disparities in science and engineering are
substantial. Research has documented the lower levels of participation,
presence in prestigious locations, ranks attained, and salaries earned
among women, compared with men, in these fields (Fox 2001; Long 2001;
Long and Fox 1995; Sonnert and Holton 1995a, 1995b; Xie and Shauman
2003). Thus, because science and engineering are influential and powerful
in society and because gender divisions persist within these fields, this
means that gender stratification within science and engineering not only
exemplifies, but also legitimizes and supports, the hierarchical relations
of men and women in society at large (Fox 1999).
Programs for undergraduate women in science and engineering intend
to open pathways for women into scientific fields through sets of activities
thought to affect women students positively (Clewell and Ficklen 1986;
Matyas and Dix 1992). Such programs consist of initiatives that are
organized responses to issues of, and perceived barriers to, the representation,
participation, and performance of the targeted group (Clewell, Anderson,
and Thorpe 1992). For the study of women and men in higher education,
programs for women in science and engineering, in particular, are
strategic research sites (Merton 1973), that is, revealing sites for study.
This is because programs embody their directors’ (and sometimes, founders’)
conceptions of what is “wrong” or at issue for women in science and
engineering, and what can be done to improve women’s condition. In this
sense, programs for undergraduate women in science and engineering are
real-life organizational structures that reveal underlying beliefs about the
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

592 GENDER & SOCIETY / October 2011
crucial factors that affect the participation and performance of undergraduate
women in science and engineering.
In this article, we examine patterns of the programs’ (1) definitions of
the barriers facing women in science and engineering, (2) their solutions
to address these barriers, (3) their goals and perceived success with goals,
and (4) their organizational characteristics and relationship to the larger
institutional environment of the universities. Our aim is to provide an
understanding of programs as a key case in the study of gender, science,
and higher education.
Conceptualizations of the problem of participation and performance of
women in science and engineering—and solutions to the problem—have
tended to be either individual or structural. The former conceptualization
emphasizes the ways individual women’s attitudes, values, aptitudes,
experiences, and/or behaviors may operate as deficiencies (or as facilitators)
of their participation and performance. This conceptualization has been
applied also to undergraduate women as majors in science and engineering
(Astin and Sax 1996; Cronin and Roger 1999; Phipps 2008; Sonnert and
Holton 1995a, 1995b). From the individual perspective, undergraduate
women’s levels of self-confidence in mathematics, for example, and
internal sense of their potential for scientific achievement are barriers
to continuing participation in science (Astin and Sax 1996). Likewise,
women’s levels of motivation to pursue careers in science are supports
or barriers to participation in these fields.
The structural conceptualization, on the other hand, emphasizes
potential deficiencies in the characteristics of settings in which women
are educated (and work) in science and engineering. These characteristics
may include, for example, patterns of exclusion or inclusion in research
groups, access to human and material resources, and practices of evaluation
as they operate differently for women and men (Astin and Sax 1996; Bird
2010; Cronin and Roger 1999; Fox 1998, 2001, 2008; Phipps 2008;
Robinson and McIlwee 1989; Smith-Doerr 2004; Sonnert and Holton
1995a, 1995b). Structural factors that operate as barriers for undergraduate
women particularly may include teaching environments that portray
science and engineering as highly competitive, masculine domains
(Margolis and Fisher 2002); isolate women from social concerns (Rosser
1993); or take a “weed out” orientation for students in their progression
through the curriculum (Seymour and Hewitt 1997).
Previous studies of programs for undergraduate women in science and
engineering have concentrated largely on single-institution programs that
focus on mentoring, tutoring, and student support (Brainard 1993; Brainard,
McIntyre, and Carlin 1995); living-learning communities (Allen 1999;
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 593
Fisher, Young, and Hein 2000; Hathaway, Sharp, and Davis 2001;
Kahveci, Southerland, and Gilmer 2008); and research internships that
include discussions of gender and engineering (Han, Sax, and Kim 2007).
Studies of multiple programs are exceptional and include Brainard,
Kelly, and Wahl’s (1993) evaluations of women in engineering programs’
objectives, target populations, years in existence, and budgetary levels;
Knight and Cunningham’s (2004) description of the views of directors of
programs for women in engineering on “advice for those who may be
interested” in such programs; and the Final Report of the Women’s Experiences
in College Engineering (WECE) Project (Goodman Research Group
2002) about “effective ways to retain female engineering students.” A
recent analysis of programs in the United Kingdom focused on initiatives
to support women in science, engineering, and technology in that country.
It explored both the aims and frameworks—that is, personal/individual
choices of women and structural/cultural factors governing women’s
participation and success—of the initiatives (Phipps 2008).
In a previous article (Fox, Sonnert, and Nikiforova 2009) that was based
on data from site visits and interviews, we investigated the characteristics
of two extreme subsets of programs in the United States: the five programs
with the most successful and the five with the least successful outcomes in
terms of improving the proportion of undergraduate degrees awarded to
women in science and engineering over time. We found that the most
successful programs had a more structural, as opposed to individual,
perspective on, and orientation to, women in scientific fields. The most
successful programs adapted the institutional environment in a mutually
dependent relationship between the program and its environment, whereas
the least successful programs adopted the institutional environment and its
goals and priorities in a one-way relationship. Successful programs found
ways to integrate with the core academic context through activities such
as hands-on research opportunities for students. Programs with the least
successful outcomes focused on activities, such as peer-mentoring among
students, which were more distant from the academic and structural core
of the university.
In the present article, we extend this inquiry, taking a comprehensive
approach to the full range of programs and analyzing quantitative
indicators of the programs’ characteristics. The specific research questions
are these. First, what are the programs’ operational definitions of the
problems facing women undergraduate majors in science and engineering
in terms of the individual versus structural dimensions? What are the
leading obstacles reported, and what do they indicate about an individual,
as opposed to structural, orientation of the programs? Second, what are
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

594 GENDER & SOCIETY / October 2011
the major and minor activities undertaken? How do these activities
represent individual, as opposed to structural, solutions for the issue/problem
of undergraduate women in science and engineering? Do the solutions
parallel or diverge from definitions of the problem, and what, in turn, does
this reveal about initiatives to improve the condition of women in science
and engineering? Third, what are the programmatic goals? Are these goals
predominantly individual or structural? To what extent do impacts fall short
of goals, and why? Fourth, what are the organizational characteristics of
programs, and how do they relate to structural-individual orientations of the
programs? Overall, what are implications of these patterns for understanding
gender, science, and higher education—and for organizational initiatives
to improve the condition of undergraduate women in these critical fields?
In addressing these questions, we present the first conceptually oriented
analyses of the universe of programs for undergraduate women in science
and engineering in the United States. Thus, we aim to meet the challenge
(Dietz, Anderson, and Katzenmeyer 2002, 400) for a study of U.S. programs
for women in science and engineering that situates the study within a
framework that links theory, research questions, and analyses—and advances
understanding of gender disparity in organizations of higher education.
METHOD
Data
We collected data through a mail survey of the directors of the universe
of undergraduate-level programs (N = 49) for women in engineering,
women in science, and women in science and engineering that existed at
the time of the survey (2002) in institutions of higher education. The
Women in Engineering Program Advocates Network (WEPAN) provided
a roster for all programs. Four of the undergraduate-level programs on the
roster existed outside of academic institutions and we eliminated them;
three programs no longer existed. The resulting 49 programs were located
at 45 institutions, with four of the institutions having two programs each.
One of the (49) programs was in the process of becoming defunct at the
time of the survey. Thus, the eligible number of programs became 48. The
response rate to the survey was 79 percent (38/48).
The directors of these programs are academic professionals, among
whom 45 percent have doctoral degrees. However, only two (5 percent) are
tenured faculty members, and another nine (24 percent) hold a nontenured
faculty position; the remaining 71 percent of program directors do not
hold faculty appointments.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 595
Variables
In accord with the four dimensions of study identified in the Introduction,
the variables include (1) the views/orientation of the program regarding
obstacles for undergraduate women as majors in science and engineering;
(2) the prevalence of certain activities of the program; (3) the importance
of particular goals for the program, and the program’s reported impact for
each goal; and (4) the organization of the program and relationship to the
institutional setting, including the program’s reporting line, funding, and
seniority or founding date within the institution, and level of faculty
participation in the program. We created the specific variables on the basis
of pilot site visits to three programs, through which we ascertained the
relevant and typical responses of the program directors.
We assessed the programs’ orientations to (or views of) issues of
undergraduate women in science and engineering through questions about
the level of importance of obstacles (“not an obstacle,” “slightly important,”
“moderately important,” “very important”) for undergraduate women as
majors in science and/or engineering. The reported obstacles divide into
two groups: views of deficiencies of undergraduate women and views of
deficiencies within the institutional structure of the university.
We captured the views of deficiencies in personal characteristics of
undergraduate women through the directors’ ratings of the level of
importance of each of the following as obstacles for undergraduate women
as majors in science and/or engineering: (low levels of) (a) self-confidence
(sense of assurance about self), (b) independence, (c) academic ability,
(d) motivation to succeed in science and/or engineering, (e) commitment
to academic work, (f) information about careers in science and/or engineering,
and (g) commitment to careers in science and/or engineering.
We measured the views of deficiencies in the institutional environment
through ratings of the importance of each of the following as obstacles:
(the lack of) (a) supportive peer relationships for undergraduate women,
(b) faculty advisors for undergraduate women, (c) a supportive classroom
climate for undergraduate women, (d) faculty commitment to the success
of undergraduate women, (e) administrative commitment to the success
of undergraduate women, and (f) a campus climate of gender equity for
undergraduate women.
The survey also contained a separate question listing the combined
individual and structural obstacles, and asked directors which they
considered to be the number one, “single most important,” obstacle in the
program’s orientation. In addition, we constructed a variable that, for each
program, subtracts the mean rating of the importance of the individual
obstacles from the mean rating of the importance of structural obstacles.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

596 GENDER & SOCIETY / October 2011
Positive values on this variable thus represent structural orientations, and
negative values, individual orientations of programs. We refer to this
variable as structural-individual orientation.
Regarding the prevalence of activities, the questionnaire recorded the
directors’ ratings (“not an activity,” “minor,” or “major” activity) of (a)
undergraduate peer-to-peer mentoring, (b) graduate students mentoring
undergraduates, (c) faculty mentoring undergraduates, (d) study and/or
social lounges or spaces provided, (e) academic tutoring, (f) career
seminars or talks, (g) residential, living-learning halls, (h) social activities
such as dinners and group outings, (i) scholarships offered, (j) curricular
component (offering of courses), (k) research projects on women in
science and/or engineering, (l) linkages with campus programs for other
minorities (racial/ethnic), and (m) training or development for faculty
(e.g., diversity training).
We constructed another variable subtracting the mean of the individual
activities (represented by activities a, b, e, f, h, and i, above) from the mean
of structural activities (including faculty activities and represented by
activities c, d, g, j, k, l, and m, above). Positive values on this variable
represent structural leanings, and negative values, individual leanings, in
activities. We refer to this variable as structural-individual activities.
We assessed the programs’ particular goals through program directors’
rating of importance (“not a focus,” “slightly important,” “moderately
important,” or “very important”) of (a) recruiting undergraduate women,
(b) developing undergraduate women’s positive attitudes (e.g., confidence,
independence), (c) strengthening undergraduate women’s motivation and
commitment to careers in science and/or engineering, (d) improving
undergraduate women’s academic abilities in science and/or engineering,
(e) creating a programmatic “home” or “community” for undergraduate
women, (f) improving classroom climate for undergraduate women,
(g) changing faculty attitudes and/or behavior toward undergraduate
women, (h) creating faculty development programs focused on diversity,
(i) advancing a “feminist” science and/or engineering, (j) providing
research for understanding women in science and/or engineering, and (k)
providing a visible signal that the institution is committed to advancing
women in science and/or engineering, We also assessed the reported
impact of the program regarding these goals through the directors’ ratings
(“no impact,” “slight,” “moderate,” or “strong”).
We constructed an additional variable representing the average of the
structural goals (represented by goals a, b, c, and d, above) minus the
average of the individual goals (represented by goals e, f, g, h, i, j, and k,
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 597
above). Positive values on this variable indicate structural, and negative
values represent individual, leanings in goals. We refer to this variable as
structural-individual goals. Goals encompassing faculty are within the
structural category because the specified goals with faculty (goals f, g, h,
and j) involve core institutional objectives.
We assessed elements of the programs’ organization and connections
to the larger institution. Information about the reporting line was gathered
through a question about the person to whom the program reports; we
categorized the responses into persons within the Office of the Provost, a
Dean, Associate Dean, and other administrators lower than the position of
Associate Dean. We measured level of funding with a question about the
total amount of funding (on a current annual basis) from all sources. A
question about the director’s employment fraction yielded the two
categories of either a “part-time” or “full-time” appointment. We
ascertained program age through a question about the year in which the
program was founded. Finally, we measured levels of faculty participation
with a question about the overall level of faculty participation (attendance
at events, participation in activities) in the program, with response categories
of “almost none,” “low,” “moderate,” and “high.”
Methods of Analysis
For statistical analyses, we treated the rating scales as continuous
variables so that we could calculate mean responses and compare them
through t-tests. The small number of cases precluded more sophisticated
statistical techniques such as multiple regression and factor analysis.
1
Because we surveyed the entire population of programs for undergraduate
women in science and engineering in the United States, and received
responses from almost 80 percent of the programs, we applied a finite
population correction factor,
()/( )Nn N−−1
, that diminishes standard errors
and tightens confidence intervals around estimates (Kokoska and
Zwillinger 2000, 146). The standard errors of means, calculated in this
way, appear in Figures 1 to 3 as error bars. We also calculated paired
sample t-tests of differences
2
between mean levels of the importance of
goals and the impacts with respect to the goals (Figure 4) with standard
errors that are adjusted by the finite population correction factors.
We further used the finite population corrections for t-tests of differences
of the mean structural-individual orientations by the five variables of
programs’ organizational characteristics/connections (Table 1). For these
tests, we recoded the five organizational variables as dichotomous
variables as follows: (1) reporting level to Office of Provost, Dean, or
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

598 GENDER & SOCIETY / October 2011
Figure 1: Mean Levels of Importance of “Individual Obstacles”
Associate Dean (high), and to other administrators (low); (2) funding
levels above $50,000 (high), and levels of $50,000 or less (low); (3)
founding date in 1992 or later (new), and before 1992 (old); (4) director’s
appointment, full-time (high), and part-time (low); and (5) level of faculty
participation that is high, medium, or low (high), and that is none (low).
FINDINGS AND DISCUSSION
Definitions of the Problem
The mean ratings of importance for each individual obstacle, ordered
from highest to lowest, appear in Figure 1; and likewise, for each
structural obstacle, in Figure 2. The overall mean rating of importance for
structural obstacles is near “moderately important” (2.8) (Figure 2), while
the mean of individual obstacles is close to “slightly important” (2.3)
(Figure 1). These overall means suggest that programs lean somewhat
more to a structural rather than to an individual definition of the problems
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 599
Figure 2: Mean Levels of Importance of “Institutional Obstacles”
facing undergraduate women in science and engineering. Correspondingly,
the summary variable, average structural minus individual orientation, is
a positive value, 0.52.
Within each of the two categories of deficiencies, interesting variations
appear. Figure 1, which represents the individual obstacles, shows that
program directors report two individual deficiencies to be, on average, at
least moderately important obstacles: women’s “self-confidence” (3.2)
and their knowledge/information about careers in science (3.0). Two
obstacles are moderately to slightly important: “career commitment” (2.7)
and “motivation to succeed” (2.4). “Commitment to academic work” (1.7)
and “independence” of students (1.7) are less than slightly important.
Notably, “academic ability” ranks last in importance of the individual
characteristics and is close to not at all important (1.3). Thus, the mean
importance ratings of individual characteristics are fairly wide-ranging,
from not at all important to more than moderately important.
Among the structural deficiencies (Figure 2), “classroom climate” leads
as moderately important (3.1), followed by the availability of “peer relationships”
(2.9), “faculty advisors” (2.9), and “faculty commitment to the success of
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

600 GENDER & SOCIETY / October 2011
undergraduate women” (2.8). The lack of a “campus climate of gender
equity” is between a moderately and slightly important (2.7) obstacle.
Interestingly, “administrative commitment to the success of undergraduate
women” is the least important structural deficiency—only somewhat more
than slightly important (2.4). The range of the mean values for structural
obstacles is narrower than that for individual obstacles: It extends only from
more than slightly important to more than moderately important.
On the question about the number one, “single most important,” obstacle,
the leading obstacle is “women’s self-confidence” (with 27 percent of the
directors reporting this as the number one obstacle), followed by “supportive
classroom climate” (21 percent). These two obstacles represent an interesting
coupling of reported barriers, with one being a clearly individual factor, and
the other, structural. These obstacles may operate together—with classroom
climate potentially influencing women’s self-confidence.
Activities—Programmatic “Solutions”
The activities of the programs appear in Figure 3. Leading in prevalence
are career development talks and seminars and peer mentoring among
Figure 3: Mean Prevalence of Activities
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 601
undergraduates, followed by social activities such as dinners and group
outings. Each of these three types of activities received an average rating
that lies in the middle between a “major” and “minor” activity (2.4 to 2.5
on the scale) (Figure 3). The next most prevalent activities involve linkages
with other programs, academic tutoring, scholarships, and alternate
environments created through either residential, living-learning halls for
undergraduate women or through study and/or social lounges provided by
the program. On the average, these activities are at least minor activities
(close to, or higher than, 2.0). Faculty-centered activities—including
diversity training for faculty, faculty mentoring of undergraduates, and
curricular course components—are all less than minor activities, on
average (less than 2.0). The least common activity is graduate students
mentoring undergraduates.
It is notable that the reported program activities largely leave untouched
key structural obstacles reported in the previous section: namely,
classroom climate, as well as faculty advisors and faculty commitment to
undergraduate education. In these ways, the typical activities of programs
do not necessary align with their typical definitions of the problem of
women in science and engineering.
Correspondingly, the average of the structural-individual activities
variable is negative (–0.31). This indicates that, on average, programs lean
toward individual activities. Furthermore, a virtually null correlation
Figure 4: Mean Levels of Importance of Goals and of Impacts with Goals
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

602
TABLE 1. Programs’ Structural-Individual Orientation by Institutional Characteristics
t-Test for Difference of Means
Institutional Characteristic
Percentage
(%)
Mean
Structural-
Individual
Orientation
Dichotomous
Groupings
Mean Structural-
Individual
Orientation P-value
Reporting level
Office of provost 18.4 0.56
}
Dean 26.3 0.47 High 0.60
Associate dean 36.8 0.72
Other administrator 18.4 0.17 Low 0.17 0.004**
Funding level
>$200k/year 43.2 0.53
}
High 0.61
>$50k/year but ≤$200k/year 32.4 0.73
≤$50k/year 24.3 0.16 Low 0.16 0.002**
Founding date
1969-1991 36.8 0.34 Low 0.34
1992-1999 63.2 0.63 High 0.63 0.017*
Director’s appointment
Full-time 71.1 0.56 High 0.56
Part-time 28.9 0.42 Low 0.42 0.272
Level of faculty participation
Almost none 5.3 –0.17 Low –0.17
Low 36.8 0.54
}
Moderate 47.4 0.51 High 0.56 0.006**
High 10.5 0.84
*p ≤ .05. **p ≤ .01.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 603
(–0.03) between the structural-individual activities and the structural-
individual orientations of programs indicates also that the activities
programs undertake do not align with their definitions of the problem of
women in scientific fields. This suggests that activities are driven by
different considerations, pragmatism, or necessity.
Goals—and Impacts
Goals represent intents of the program, giving direction and guidelines
toward programmatic decision making. They can facilitate as well as limit
programmatic actions. The goals and the reported impacts with respect to
each goal appear in Figure 4.
3
The five leading goals (reported as moderately to very important for
programs) comprise creating homes/communities for women students
(3.7), strengthening women’s motivation/commitment to careers in science/
engineering (3.7), developing women’s positive attitudes (3.6), providing
a visible signal that the institution is committed to advancing women in
science/engineering (3.5), and recruiting women students (3.5). Two of
these leading goals, focusing on career commitments and on attitudes,
reflect an individual perspective. Recruiting women students is an
individual goal to the extent that it focuses on “increasing numbers of
women students” without necessarily addressing the features of the
settings into which they are recruited. The remaining two leading goals,
creating homes/communities and providing a visible signal, are at least
partly structural in nature. However, creating homes/communities for
students focuses primarily on providing alternative—or niche—environments,
rather than changing a prevailing institutional environment. Providing a
signal about the institution’s commitment to advancing women in science/
engineering may be a catalyst for real institutional change, but may also
be merely “window-dressing.”
The next set of goals, reported to be at least moderately important, are
improving classroom climate (3.3) and changing faculty attitudes toward
undergraduate women (3.1). These goals are deeply structural because
they go to core elements of the institutional environments in which women
(and men) students are educated, and because achieving such goals would
constitute potentially fundamental changes in the academic environment.
The fact that programs at least aspire to such goals reflects a certain level
of understanding about the centrality of faculty in improving the condition
of women as undergraduate majors in science/engineering.
Goals that are between moderately and slightly important are providing
research on women in science/engineering (2.8), improving academic
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

604 GENDER & SOCIETY / October 2011
abilities in science/engineering (2.8), and creating faculty development
programs, such as diversity training (2.6). Advancing feminist science is
an only slightly important (2.0) goal.
4
In Figure 4, the programs’ reported impacts with respect to each goal
appear along with the reported importance of that goal. For each of the
11 goals, the mean reported impact is significantly lower than the mean
reported importance of the goal (p <.001 for eight pairs of goals/impacts,
and p <.01 for three pairs). Among all goals specified, the disparity
between mean importance of the goal and the mean impact is greatest for
three goals that pertain to faculty: changing classroom climate, changing
faculty attitudes and behavior toward undergraduate women students, and
creating faculty development programs focused on diversity. These disparities
between reported levels of goals and impacts suggest that fundamental
academic modifications (or even transformations) that involve faculty are
especially challenging.
5
Organizational Features and Institutional Contexts
The programs are located, disproportionately, within research universities
where research is a priority, federal funding of research is high, and
commitment to graduate education prevails. Eighty-four percent of the
responding programs are located in research universities with high or very
high research activity (using the Carnegie Classification of Institutions of
Higher Education for 2003/2004, the period close to the time of the survey
of the programs). The location of programs within research settings is
partially a by-product of the location of colleges (or units) of engineering
within research universities. “Women in Engineering” and “Women in
Science and Engineering” programs, together, represent 89 percent of the
programs in this study; only 11 percent of the programs are for “Women
in Science.”
The level of the unit/person to which programs report is important
because reporting to higher levels within the university gives a program
higher level connections and potentially greater institutional visibility. The
largest group (36.8 percent) of the programs report to an Associate Dean,
and 18.4 percent report to administrators below this level (such as Assistant
Dean, Director of Minority Engineering, or Center Director) (Table 1).
Less than a fifth (18.4 percent) report to units or persons within the Office
of the Provost, which is the most central and broadest level within the
university, spanning all academic units of the institution. The remainder
(26 percent) of the programs report to a Dean (within Engineering,
Sciences, or Undergraduate Studies). The predominant reporting lines are
thus to a mid-to-lower level within universities.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 605
Programs’ funding levels (from all sources of funding) cluster at annual
budgets of over $200,000 (43.2 percent of programs). Nearly a third (32.2
percent) of programs have budgets that exceed $50,000 per year but less
than $200,000. About a quarter (24.3 percent) of programs have annual
budgets that are $50,000 or less (Table 1). Programs’ funding sources vary.
Nearly all (97.4 percent) of the programs receive some funding from
university and/or college sources. Most programs receive some funding
through other sources as well: 88 percent receive funding through
corporations, 81 percent through donors/alumnae, and 79 percent through
grants. Of the 16 programs that receive more than $200,000 in funding,
three of these programs are receiving more than $200,000 from their
university/college (along with other sources of funding).
The establishment of programs for women in science and engineering
escalated during the 1990s, and most of the reporting programs were
founded in that period of expansion. Almost two-thirds (63.2 percent) were
established in 1992 or thereafter (Table 1). Furthermore, of the (24)
programs established in 1992 or later, seven were founded after 1996.
Hence, programs for undergraduate women in science and engineering are
relatively new units within university settings. It is likely that universities
created the programs partly in reaction to political pressures for equity in
higher education, and to provide a “visible sign” or signal of institutional
responsiveness. Despite being relatively new units, the (71 percent) majority
of directors of programs have full-time appointments (Table 1). This marks
programs, administratively, as more than add-on undertakings that exist
along with a person’s other appointment.
Because research universities are strongly influenced by the professional
and expert authority of faculty (Birnbaum 1992), faculty participation in
programs is potentially important. However, as already indicated, it is
challenging to achieve. Faculty participation in programs is reported to be
“low” by 36.8 percent of the programs and “moderate” by 47.4 percent.
“High” faculty participation is unusual (10.5 percent). “Almost no faculty
participation” characterizes 5.3 percent of the programs (Table 1).
Thus, in sum, the programs’ organizational characteristics and relationship
to their institutional context indicate that typical programs (1) report to
persons/units at mid-to-low level in the administrative hierarchy of the
university, (2) have relatively high funding from variable sources, (3)
fairly recent founding dates, (4) full-time directorships, and (5) moderate
to low participation of the faculty.
Given our attention to programs’ more or less structural (as opposed to
individual) orientations, it is interesting to examine how orientations
relate to the five organizational indicators. First, programs that report to
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

606 GENDER & SOCIETY / October 2011
the Office of the Provost, Deans, or Associate Deans have mean values of
0.56, 0.47, and 0.72, respectively, on the structural-individual orientation
variable, pointing to structural leanings (Table 1). It is the programs that
report to units/persons below the level of Associate Deans that have the
least structural orientation, with a mean structural-individual level of
0.17 (p < .01). (Note that this difference—and the following significant
differences—remain significant also after a Bonferroni correction for post
hoc comparisons.)
Second, programs with higher levels of funding tend toward the
structural side. The structural-individual orientation variable has a mean of
0.73 for those with funding between $50,001 and $200,000 annually, and
a mean of 0.53 for those with annual funding higher than $200,000. The
notable difference in orientation/leanings (p < .01) is between programs
with these two higher levels of funding and programs with the lowest
level of funding ($50,000 or less). Those programs with funding $50,000
or less annually have a low mean level of structural-individual orientation
(0.16).
Third, programs founded more recently have higher structural leanings.
Those programs founded in 1992 or later have a mean structural-individual
orientation of 0.63, compared with 0.34 for programs founded prior to
1992 (p < .05). The trend of increasing structural orientation among more
recent programs is confirmed among the post-1996 programs, which have
a yet higher mean of 0.85 on the structural-individual orientation.
Fourth, programs with full-time directors have somewhat higher
structural leanings (0.56) compared with those with part-time directors
(0.42), but the difference is not significant (p < .27).
Fifth, as one might expect, it is programs with “almost no” faculty
participation that have a clearly individual orientation (with a mean of
–0.17); and programs with “high” faculty participation have the strongest
structural orientation (0.84). Programs with “moderate” or “low” faculty
participation have structural leanings that are similar, with means, respectively,
of 0.54 and 0.51. The difference in orientations is significant (p < .01)
between programs with almost none and those with other levels of faculty
participation.
Thus, the weakest structural orientation to issues of women in science
accompanies programs with a marginal to low institutional connectivity
in (1) reporting to units/persons at the lowest levels in the institutional
hierarchy, (2) having the lowest budget levels, and (3) the lowest levels of
faculty participation. It is not clear, however, if the position on such
institutional margins is cause, effect, or both cause and effect of a weak
structural orientation to issues facing women as undergraduate majors in
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 607
scientific fields, or if an underlying cause exists that brings about
marginality as well as weak structural orientation.
SUMMARY AND CONCLUSIONS
In their orientations to the issue of women in science and engineering,
programs lean toward structural, as opposed to individual, definitions of
the problem. They also point to “classroom climate”—a core structural
feature of higher education—as among the most important obstacles.
However, the most prevalent activities or solutions undertaken by
programs are those that may be carried out relatively independently by the
program directors with staff and participating students. Activities that
reach into the realm of faculty—diversity training for faculty, faculty
mentoring of undergraduates, and curricular course components—are all
less than minor activities. Furthermore, the negative mean (–0.31) of the
structural-individual activities variable indicates individual leanings of
the programs’ activities, on the average.
Thus, the typical activities programs undertake do not necessarily
align with their typical definitions of the problem, particularly insofar as
they involve faculty. Circumstances appear to compel programs into
doing what is possible, even if, by the director’s own definition, the crux
of the issue of women in science and engineering lies somewhere else.
This misalignment may compromise the effectiveness of programs.
The discrepancy (or misalignment) between orientations and implementations
appears again in the patterns of the reported importance of goals compared with
impacts achieved. The disparity between the importance of, and the impact
achieved with respect to, the goals is greatest for two goals that involve faculty:
changing classroom climate and changing faculty attitudes and behavior
toward undergraduate women. On the one hand, at least in their intents, the
programs have goals that involve faculty (given that these two goals
involving faculty are reported to be at least moderately important). The fact
that programs at least aspire to such goals pertaining to faculty reflects a level
of understanding that these core structural elements of the university are
crucial determinants of the success of women in science and engineering.
On the other hand, in their reported impacts, programs fall short of goals
pertaining to faculty by a particularly wide margin.
It can be especially challenging to achieve impacts with faculty within
science and engineering fields. The research activity and external funding
of science and engineering departments—which derive from the faculty—
have been critical to the status and national rankings of universities
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

608 GENDER & SOCIETY / October 2011
(see Benezet 1977; Long and Fox 1995; Salancik and Pfeffer 1974). This,
in turn, reduces the motivation of university administrators to take steps
to alter the decentralized character of universities where departments
operate relatively autonomously and where individual Principal Investigators
have a high amount of control over what happens in their laboratories and/
or classrooms. Thus, the norms and practices concerning faculty and
students exist within relatively sovereign departmental or laboratory units
that are fairly free from administrative interference or interventions (Fox
2000). Programs for women in science and engineering have to maneuver
within such conditions of comparatively high faculty control and low
institutional surveillance of departments.
Further, and fundamentally, these institutions of higher education can
be characterized as gendered institutions. Assumptions about gender
underlie the divisions of labor and authority in these institutions, and the
ways in which status and rewards accrue (Acker 1990; Britton 2000). For
programs, this means that activities and impacts that contest the latitude and
authority of science and engineering faculty represent potential challenges
to the prevailing gender hierarchies in the institutions. For
example, interventions in classroom climate and faculty attitudes toward
undergraduate women can pose a challenge to institutions in which men
hold the higher ranks and exercise relatively unfettered authority over
setting curricula and pedagogy, and evaluating students as well as their
peers (Bird 2010; Fox 2000). Thus, these gendered organizational
environments can temper programs’ typical activities and their impacts
with goals, particularly those that involve core structural elements of
classroom climate and faculty’s attitudes toward women students.
In addition, features of the programs’ relationships to their organizational
context are associated with the structural-individual orientation of the
programs. The strongest individual orientations to issues of women in
science and engineering exist among programs that are the most institutionally
marginal in their reporting lines, budgets, and faculty participation. Such
individual orientations pose little or no challenge to existing institutional
arrangements and prevailing gender hierarchies. Low budgets and a mid-to-low
reporting structure among programs can create dependencies among
programs that potentially constrain their activities and impacts for
women in science and engineering. This may result in compliance with
institutional demands for conformity with the institutional status quo and,
specifically, with the assumption that the environment is operating
neutrally (Acker 1990) and that women are the actors who need to change
in order to succeed in science and engineering (Bird 2010). Such an
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 609
individual orientation may be among the underlying factors that explain
why many organizational efforts to recruit and retain women fail to result
in substantial gains for women (Ely and Meyerson 2000; Fox, Sonnert,
and Nikiforova 2009).
What are the implications of these patterns for program directors who
aim to improve the condition of undergraduate women in these critical
scientific fields? First, a structural definition of the problem of women in
science and engineering is an appropriate way to frame the issue. Therefore,
programs may do well to look beyond confining themselves to individually
oriented activities, such as peer mentoring and social events, that are
pragmatically “easy” and do not challenge any existing institutional
arrangements.
Second, organizational factors, specifically a high level of material
and social resources, accompany a structural, as opposed to individual,
orientation to obstacles for women in science and engineering. Although,
as indicated, orientations by themselves do not parallel activities or
solutions, shoring up the level of budget through industrial sources,
foundations, or agencies has the potential to lessen the dependencies of
programs on financial support of their own institutions. Financial
dependence on home institutions may come with the cost of compelling
(or, at least, constraining) a program so that it reflects the institutional
status quo with respect to the gender disparity in science and engineering.
In addition, gaining reporting lines that reach higher in the organizational
chart represent alignments that can make programs more active and
influential in their environments, increasing the likelihood of positive
outcomes for undergraduate women. Furthermore, alliances with science
and engineering faculty through partnerships in externally funded projects
and course offerings are also elements associated with structural orientations
that can enable programs to be more active and influential in shaping
environments and outcomes for women in science and engineering.
Just as creating a structure does not guarantee change (Edelman 1992),
establishing a program does not guarantee a positive outcome for women
students. In designing programs, it is critical to consider key organizational
concerns, namely, the relationship between programs and the institutional
settings in which they are situated, and the ways this relationship affects
opportunities to address and redress structures of gender disparity within
organizations. At the level of women faculty rather than students, the
National Science Foundation’s ADVANCE Institutional Transformation
(IT) program is a key example of an initiative that contains strong
organizational elements—funding levels, reporting lines, and connections
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

610 GENDER & SOCIETY / October 2011
to science and engineering faculty— as well as a focus on structural change
(Fox 2008).
The results reported here identify ways in which structural orientations
to the issues, problems, and solutions of reducing gender disparity will
both encounter challenges within organizations and pose the potential
for redressing gender inequity. Because gender disparity characterizes
numerous other organizational contexts in society, the results may also
serve to inform strategic initiatives to improve the condition of women
in other and broader organizational contexts marked by gender hierarchies
(Britton 2000), especially since women’s integration appears to have
stalled in many fields (England 2010).
NOTES
1. Green (1991) analyzes numerous rules about the minimum number of subjects
needed for multiple regressions, and obtains some support for the rule that N ≥ 50
+ 8m, where m is the number of predictors. Our N of 38 programs responding to
the survey (out of 48 total programs) falls clearly short of that rule. Because
factor analysis relies on correlation matrices and correlations that are unstable at
low sample sizes, the N of 38 is insufficient also for this technique (Tabachnick
and Fidell 2001, 588).
2. The rationale for using the paired t-tests is this. First, the two questions
on (1) the importance of goals and (2) the impact with respect to goals appear
side-by-side in the questionnaire. The question on impacts refers to the goals with
the wording: “To what extent do you think your program has an impact with the
same broad goals?” Second, the scales (1 through 4) for the responses for the
importance of goals (“not a focus,” “slightly important,” “moderately important,”
“very important”) and for impacts with the same goals (“none,” slight,” “moderate,”
“strong”) are designed in parallel to each other.
3. Pairwise t-tests involve cases that are complete (without missing data) for
goals and matching impacts.
4. The summary measure, structural-individual orientation, indicates that,
on average, programs tend to have structural orientations to the problem. The
summary measure for structural-individual goals, however, has a negative mean
value, –0.46, indicating individual leanings in programs’ goals, on average. The
correlation between these two summary structural-individual measures
(orientations, goals) is positive (0.36) and statistically significant (p = .03),
suggesting that structural leanings in orientations and structural leanings in goals
tend to go together. The scatterplot for the two measures, however, shows three
outlying programs in the upper right quadrant (positive-value orientations, positive-
value goals) that may be affecting the association, so that the significance of this
correlation may be questionable. Furthermore, the correlation is positive (0.25)
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 611
between the structural-individual summary variables for activities and goals, but
it is not significant. As reported in the text, the structural-individual summary
variables of orientation and of activities are not at all correlated.
5. Because for all goals the levels of importance are consistently higher than
the levels of impact, and because the meaning of the scales for importance of a
goal and for impact of a goal might not be identical, it is advisable to consider the
relationship of importance and impact, net of their respective mean levels. This
complementary approach focuses attention on the gaps between the relative
evaluations of the goals’ importance and impact. In taking this approach, we
standardized the variables of importance of goals and the variables of impact of
goals by subtracting the average value of all importance variables from each
importance variable and, analogously, subtracting the average value of all impact
variables from each impact variable (and by giving all variables a standard
deviation of 1.0). In this analysis (with the finite population correlation factors),
the gap between importance and impact was significant for the following goals:
strengthening undergraduate women’s motivation and commitment to career,
improving classroom climate for undergraduate women, changing faculty attitudes
and behavior toward undergraduate women, and advancing feminist science.
REFERENCES
Acker, Joan. 1990. Hierarchies, jobs, bodies: A theory of gendered organizations.
Gender & Society 4 (2): 139-58.
Allen, Caitlyn. 1999. Wiser women: Fostering undergraduate success in science
and engineering with a residential academic program. Journal of Women and
Minorities in Science and Engineering 5 (3): 265-77.
Astin, Helen S., and Linda J. Sax. 1996. Developing scientific talent in undergraduate
women. In The equity equation: Fostering the advancement of women in the
sciences, mathematics, and engineering, edited by C. S. Davis, A. Ginorio,
C. Hollenshead, B. Lazarus, and P. Rayman. San Francisco: Jossey-Bass.
Benezet, Louis. 1977. Uses and abuses of departments. In Academic departments,
edited by D. E. McHenry and Associates. San Francisco: Jossey-Bass.
Bird, Sharon. 2010. Unsettling universities’ incongruous, gendered bureaucratic
structures: A case-study approach. Gender, Work, and Organization 18 (2).
[on-line version]
Birnbaum, Robert. 1992. How colleges work: The cybernetics of academic
organization and leadership. San Francisco: Jossey-Bass.
Brainard, Suzanne. 1993. Student ownership: The key to successful intervention
programs. Initiatives 55 (3): 23-30.
Brainard, Suzanne, Suzanne Laurich McIntyre, and Linda Carlin. 1995. Retaining
female undergraduate students in engineering and science. Journal of Women
and Minorities in Science and Engineering 2 (4): 255-67.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

612 GENDER & SOCIETY / October 2011
Brainard, Suzanne, James Kelly, and Patricia Wahl. 1993. National evaluation of
existing women in engineering programs. Women in Engineering Program
Advocates Network, Working paper 93–92.
Britton, Dana. 2000. The epistemology of the gendered organization. Gender &
Society 14 (3): 418-34.
Charles, Maria, and Karen Bradley. 2002. Equal but separate? A cross-national
study of sex segregation in higher education. American Sociological Review
67: 573-99.
Clewell, Beatrix Chu, Bernice Taylor Anderson, and Margaret Thorpe. 1992.
Breaking the barriers: Helping female and minority students to succeed in
mathematics and science. San Francisco: Jossey-Bass.
Clewell, Beatrix, and Myra Ficklen. 1986. Improving minority retention in higher
education: A search for effective institutional practices. Princeton, NJ:
Educational Testing Service.
Commission on Professionals in Science and Technology (CPST). 2008.
Professional women and minorities: A total human resources data compendium.
Washington, DC: CPST.
Cozzens, Susan, and Edward Woodhouse. 1995. Science, government, and the
politics of knowledge. In Handbook of science and technology studies, edited
by S. Jasanoff, G. Markle, J. Petersen, and T. Pinch, 533-53. Thousand Oaks,
CA: Sage.
Cronin, Catherine, and Angela Roger. 1999. Theorizing progress: Women in
science, engineering and technology in higher education. Journal of Research
in Science Teaching 36 (6): 637-61.
Derber, Charles, William Schwartz, and Yale Magrass. 1990. Power in the highest
degree: Professionals and the rise of a new mandarin order. New York:
Oxford University Press.
Dietz, James S., Bernice Anderson, and Conrad Katzenmeyer. 2002. Women and
the crossroads of science: Thoughts on policy, research, and evaluation. Journal
of Women and Minorities in Science and Engineering 8 (3&4): 395-408.
Edelman, Lauren B. 1992. Legal ambiguity and symbolic structures. American
Journal of Sociology 97:1531-76.
Ely, Robin, and Debra Meyerson. 2000. Theories of gender in organizations: A
new approach to organizational analysis and change. Research in Organizational
Behavior 22:103-51.
England, Paula. 2010. The gender revolution: Uneven and stalled. Gender &
Society 24:149-66.
Fisher, Jennifer, John Young, and Jean Hein. 2000. Retaining women in the
sciences: Evidence from Douglass College’s Project SUPER. Journal of Women
and Minorities in Science and Engineering 6 (4): 349-72.
Fox, Mary Frank. 1998. Women in science and engineering: Theory, practice,
and policy in programs. Signs: Journal of Women in Culture and Society 24
(Autumn): 201-23.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 613
Fox, Mary Frank. 1999. Gender, hierarchy, and science. In Handbook of the sociology
of gender, edited by J. S. Chafetz, 441-57. New York: Kluwer Academic/Plenum.
Fox, Mary Frank. 2000. Organizational environments and doctoral degrees
awarded to women in science and engineering departments. Women’s Studies
Quarterly 28 (Spring/summer): 47-61.
Fox, Mary Frank. 2001. Women, science, and academia: Graduate education and
careers. Gender & Society 15 (October): 654-66.
Fox, Mary Frank. 2008. Institutional transformations and the advancement of
women faculty: The case of academic science and engineering. In Higher
education: Handbook of theory and research, edited by J. C. Smart, vol. 23,
73-103. Dordrecht, Netherlands: Kluwer Academic.
Fox, Mary Frank, Gerhard Sonnert, and Irina Nikiforova. 2009. Successful
programs for undergraduate women in science and engineering: Adapting
versus adopting the institutional environment. Review of Higher Education 50
(4): 333-53.
Fox, Mary Frank, and Paula Stephan. 2001. Careers of young scientists: Preferences,
prospects, and realities by gender and field. Social Studies of Science 31
(February): 109-22.
Goodman Research Group. 2002. Final Report of the Women’s Experiences in
College Engineering (WECE) Project. Cambridge, MA.
Green, Samuel B. 1991. How many subjects does it take to do a regression
analysis. Multivariate Behavioral Research 26 (3): 499-510.
Han, June C., Linda Sax, and Karen Kim. 2007. Having the talk: Engaging
engineering students in discussions on gender and inequity. Journal of Women
and Minorities in Science and Engineering 13:145-63.
Hathaway, Russel, Sally Sharp, and Cinda-Sue Davis. 2001. Programmatic
efforts affect retention of women in science and engineering. Journal of
Women and Minorities in Science and Engineering 7 (2): 107-24.
Kahveci, Ajda, Sherry Southerland, and Penny Gilmer. 2008. From marginality
to legitimate peripherality: Understanding the essential functions of a women’s
program. Science Education 92 (1): 33-64.
Knight, Meredith Thompson, and Catherine Cunningham. 2004. Building a
structure of support: An inside look at the structure of women in engineering
programs. Journal of Women and Minorities in Science and Engineering 10
(1): 1-20.
Kokoska, Stephen, and Daniel Zwillinger. 2000. CRC Standard probability and
statistics tables and formulae. Boca Raton, FL: Chapman & Hall/CRC.
Long, J. Scott. 2001. From scarcity to visibility: Gender differences in the careers
of doctoral scientists and engineers. Washington, DC: National Academy Press.
Long, J. Scott, and Mary Frank Fox. 1995. Scientific careers: Universalism and
particularism. Annual Review of Sociology 21:45-71.
Margolis, Jane, and Alan Fisher. 2002. Unlocking the clubhouse: Women in
computing. Cambridge, MA: MIT Press.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

614 GENDER & SOCIETY / October 2011
Matyas, Marsha Lakes, and Linda Skidmore Dix, eds. 1992. Science and engineering
programs: On target for women? Washington, DC: National Academy Press.
Merton, Robert. 1973. Multiple discoveries as a strategic research site. In The
sociology of science, edited by N. W. Storer, 371-82. Chicago: University of
Chicago Press.
Montgomery, Scott L. 1994. Minds for the making: The role of science in
American education. New York: Guilford.
National Academy of Sciences. 2003. Pan-organizational summit on the U.S.
science and engineering workforce. Washington, DC: National Academy Press.
Phipps, Alison. 2008. Women in science, engineering, and technology: Three
decades of UK initiatives. Stoke on Trent, UK: Trentham Books.
Reskin, Barbara, and Denise Bielby. 2005. A sociological perspective on gender
and career outcomes. Journal of Economic Perspectives 19 (1): 71-86.
Robinson, J. Gregg, and Judith McIlwee. 1989. Women in engineering: A promise
unfulfilled? Social Problems 36 (December): 455-72.
Rosser, Sue. 1993. Female friendly science: Including women in curricular content
and pedagogy. Journal of General Education 42 (3): 191-220.
Salancik, Gerald, and Jeffrey Pfeffer. 1974. The bases and uses of power in
organizational decision-making: The case of a university. Administrative
Science Quarterly 19 (4): 453-73.
Seymour, Elaine, and Nancy M. Hewitt. 1997. Talking about leaving: Why
undergraduates leave the sciences. Boulder, CO: Westview.
Smith-Doerr, Laurel. 2004. Women’s work: Gender equality vs. hierarchy in the
life sciences. Boulder, CO: Lynne Rienner.
Sonnert, Gerhard, and Gerald Holton. 1995a. Gender differences in science
careers: The Project Access Study. New Brunswick, NJ: Rutgers University
Press.
Sonnert, Gerhard, and Gerald Holton. 1995b. Who succeeds in science? The
gender dimension. New Brunswick, NJ: Rutgers University Press.
Tabachnick, Barbara G., and Linda S. Fidell. 2001. Using multivariate statistics.
4th ed. Boston, MA: Allyn & Bacon.
Wajcman, Judy. 1991. Feminism confronts technology. University Park, PA:
Pennsylvania State University Press.
Xie, Yu, and Kimberlee Shauman. 2003. Women in science: Career processes and
outcomes. Cambridge, MA: Harvard University Press.
Mary Frank Fox is ADVANCE Professor in the School of Public Policy, and
codirector of the Center for the Study of Women, Science, and Technology,
at Georgia Institute of Technology. Her research focuses on women and men
in academic and scientific organizations and occupations, and is published
in more than 50 different journals, books, and collections.
Gerhard Sonnert is a research associate at the Harvard-Smithsonian Center
for Astrophysics. He received master’s and doctorate degrees in sociology
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from

Fox et al. / WOMEN IN SCIENCE AND ENGINEERING 615
from the University of Erlangen, Germany, and a master’s in public
administration from Harvard University. He is a sociologist of science
interested in science policy and in gender in science.
Irina Nikiforova is a doctoral student in sociology of science in the School
of History, Technology, and Society at Georgia Institute of Technology. Her
research interests include scientific careers, gender, and recognition in
science, computing, and engineering. Her dissertation focuses on the Turing
Award winners in computing.
at GEORGIA TECH LIBRARY on September 16, 2011gas.sagepub.comDownloaded from