34 */52.!

,
of
#/,,%'% 3#)%.#% 4%!#().'
7
he issue oI women`s under-
representation in science,
mathematics, andengineering
(SM&E) felds in the United
States, in both education andthe work-
Iorce (Moore 2001; NSB 2002; NSE
2000, 2003, 2004a, 2004b), continues
to be a theme in current research. In
spite oI substantial gains, gaps still ex-
ist between women and men in terms
oI their college-level enrollment in the
SM&E majors. Women are still less
likely than men to choose a career that
involves SM&E, and are more likely
than men to earn bachelor`s degrees
in nonscience and nonengineering
felds. Among those who do choose a
major in SM&E, the majority is still
concentrated in certain felds such as
biology, psychology, and the social
sciences (NSE 2000, 2004a, 2004b).
Among the reIorm eIIorts in the
United States, national initiatives oI-
ten target K12 education to encour-
age all students, but especially girls
(and other disadvantaged groups),
to participate more actively in sci-
ence education (NRC 1996, 2000;
NSE 2003). Although the rationale
underlying these eIIorts has been to
enhance students` interest in the sci-
ences, these eIIorts are not as visible
at the college level. According to
Matyas (1992), Iew programs in the
United States directly target Iemale
undergraduate students.
There is a small amount oI re-
search into the eIIects oI undergradu-
ate research and mentoring programs
(Campbell and Skoog 2004), and
course-based eIIorts (Atkin, Green,
and McIaughlin 2002), suggesting
"Y !JDA +AHVECI 3HERRY ! 3OUTHERLAND AND 0ENNY * 'ILMER
!JDA +AHVECI (a|dakahvecimarmara.
edu.tr) IS AN INSTRUCTOR OF SCIENCE EDUCATION
AT -ARMARA 5NIVERSITY IN )STANBUL 4URKEY
3HERRY! 3OUTHERLANDIS ANASSOCIATEPROFESSOR
OF SCIENCE EDUCATION AND 0ENNY * 'ILMER IS
PROFESSOR OF CHEMISTRY AND BIOCHEMISTRY AT
&LORIDA3TATE5NIVERSITYIN4ALLAHASSEE &LORIDA
In this research our purpose was to examine the effectiveness of a program in
retaining women in science, mathematics, and engineering mafors. We compared
undergraduate women participating in a support and mentoring program with
nonprogram students in terms of their retention in these mafors. 1he pretest and
posttest results over one academic year revealed that the programwas successful,
adding to the call for support programs or orientation courses directed toward
women. 1he hndings had implications for many young men, as well.
Retaining Undergraduate Women in
Science, Mathematics, and ngineering
07)3%-4HE 0ROGRAM FOR 7OMEN IN
3CIENCE %NGINEERING AND -ATHEMATICS
35 NOVM8R/DCM8R 2006
that such programs are successIul in
increasingtherepresentationoI women
in science. However, the bulk oI these
studies are descriptive in nature, and
there is scant comparative inIormation
about the relative impact oI these pro-
grams. Eor instance, Malcom`s (1984)
evaluation Iocuses on programs Ior
enhancing K12 Iemales` access to
and achievement in mathematics and
science, and includes identiIying
criteria to assess the exemplariness oI
those programs. By contrast, in this
research our Iocus is on comparing
the retention oI undergraduate women
participating in an SM&Esupport and
mentoring programwith other, closely
matched undergraduate students who
are not in the support and mentoring
program (nonprogram students). By
comparing program and nonprogram
students, our purpose is to examine
the success and eIIectiveness these
programs have.
The Program for women
in Science, Lngineering,
and Mathematics
Our research investigated the infu-
ence oI a living-learning community
established to encourage Iemale col-
lege students to participate in the
SM&E majors. The Program Ior
Women in Science, Engineering, and
Mathematics (PWISEM), Iounded
in 2001 at a research university in
the American southeast, targets un-
dergraduates (particularly frst-year
students) to Ioster the participation
oI women in these felds, and enhance
their retention.
Students enter the program in
their frst year and live together in
an on-campus residence hall. While
all Iirst-year program students are
required to reside in the hall, aIter
their frst year, women can remain in
the program but reside elsewhere. In
addition to a shared residence, which
provides students with built-in study
partners, program students have the
opportunity to participate in a variety
oI PWISEM activities. The central
Iormal activity oI the program is a
one-credit course entitled Women in
Science Colloquium, in which women
7AL 1
Number of PWISMand HGC students in their intended SM& majors.
Intended majors PWISM HGC men HGC women
Athletic training/sports medicine l 0 0
8iochemistry 2 2 2
8iology l0 8 ll
8iomedical engineering l 0 0
Chemical engineering l l 2
Chemistry l 2 l
Civil engineering 0 l 0
Computer engineering 0 l 0
Computer science l 0 0
Llectrical engineering l l 0
Lngineering (undecided) 3 3 l
Lnvironmental engineering l 0 l
Lxercise science 3 l 0
Mechanical engineering 0 3 0
Meteorology 3 0 0
Nursing 4 0 2
Physics 2 2 0
Science (undecided) l 0 0
Undecided 0 7 3
Other 0 2 6
7O7AL 35 34 29
7AL 2
Students' interest in SM& in the beginning and at the end of the year.
Pretest Posttest
Mean
Standard
deviation
Mean
Standard
deviation
Pw|SLM students 4.40 0.70 4.36 0.96
HGC women students 4.45 l.06 4.l7 l.30
HGC men students 4.65 0.70 3.85 l.50
Signicance` 0.42l N/A 0.336 N/A
` Condence interval: 95%.
scientists Irom around the university
come to the hall and describe their re-
search. The program also encourages
SM&E success and retention through
lectures, panel discussions, mentor-
ing, advising assistance, research
internships, tutoring, and feld trips.
Thus, the program places a premium
on student interactions with Iaculty
and with other students.
36 */52.!,
of
#/,,%'% 3#)%.#% 4%!#().'
Pesearch design
Inthis researchweemployedthe'static-
grouppretest-posttest design(Eraenkel
andWallen 2003, p. 273), in which two
already existing groups are used. We
compared the PWISEM students with
the Honors General Chemistry (HGC)
students over the period oI one aca-
demicyear. HGCstudents wereselected
as acomparisongroupduetotheir simi-
larities with PWISEM students; both
groups were diIIerent Irom the general
university Iirst-year student popula-
tion in that they were high-achieving
students (as determined by their SAT
scores andhighschool GPAs), andboth
groups intended to pursue SM&E ma-
jors. Their central diIIerences included
that the HGC group consisted oI both
males and Iemales and that PWISEM
students participated in activities.
Participants
There were 35 frst-year PWISEM
students involved in this research.
Their ages ranged Irom18 to 19, typi-
cal Ior frst-year students. There were
63 HGC Iemale and male students
who participated in the study, and they
had an age range oI 17 through 21.
In Table 1, we summarize students`
intended majors as selI-reported in the
beginning oI their Ireshman year.
Data collection and analysis
In this research we compared program
and nonprogram students in terms
oI their
1. interest,
2. confdence,
3. determination in pursuing one oI
the SM&E majors (CIRP 2000),
4. views on science and scientists
(NORC 2004),
5. interest in and
6. understanding oI science and tech-
nology (NORC 2004),
7. GPAs, and
8. intended/declared majors.
Because the HGC student group in-
volvedmalestudents, it was possibleIor
us to make cross-comparisons among
the groups with respect to gender.
In order to measure these con-
structs, we developed a questionnaire
with insight Iromtwo diIIerent instru-
ments (CIRP 2000; NORC 2004).
Each oI these two instruments have
established validity and reliability
measures, which reIer to the ability oI
'drawing correct conclusions based on
the data obtained Iroman assessment
(Eraenkel and Wallen 2003, p. 158),
and 'the consistency oI the scores
obtained (p. 165). The questionnaire
included 23 items, which reIlected
the constructs described above, to be
rated on Iikert scales. In addition, we
collected student demographic data in
order to ensure that the same students
took the pre- and posttests. Among
these data were students` intended or
declared majors. Data Ior which we
did not have a pre/post pairing were
discarded. We administered a pretest at
the outset oI the Iall semester to both
groups and the posttest was adminis-
tered at the end oI the academic year.
In our analysis, we recognized that
there were three distinct groups within
PWISEMand HGC: (1) the PWISEM
students, (2) the HGCwomenstudents,
and (3) the HGC men students. To
compare the similarities or diIIerences
across the three groups, we perIormed
one-way ANOVA (analysis oI vari-
ance) Ior mean diIIerences across the
groups in the pretest, and then per-
Iormed it again Ior the posttest. Here
we used inIerential statistics 'to make
judgments oI the probability that an
observed diIIerence between groups
is a dependable one or one that might
7AL 4
CrosstabuIation of students' major choices in the beginning of the academic year.
SM& NonSM& Undecided 7O7AL
Pw|SLM students 93.5% (29) 0% (0) 6.5% (2) l00.0% (3l)
HGC women students 76.9% (20) ll.5% (3) ll.5% (3) l00.0% (26)
HGC men students 86.2% (25) 0% (0) l3.8% (4) l00.0% (29)
7AL 5
CrosstabuIation of students' major choices at the end of the academic year.
SM& NonSM& Undecided 7O7AL
Pw|SLM students 90.3% (28) 3.2% (l) 6.5% (2) l00.0% (3l)
HGC women students 69.2% (l8) 30.8% (8) 0% (0) l00.0% (26)
HGC men students 82.8% (24) l3.8% (4) 3.4% (l) l00.0% (29)
7AL 3
Students' condence in SM& in the beginning and at the end of the year.
Pretest Posttest
Mean
Standard
deviation
Mean
Standard
deviation
Pw|SLM students 3.94 0.80 3.94 0.9l
HGC women students 4.03 0.68 4.ll 0.83
HGC men students 4.24 0.74 4.00 0.65
Signicance` 0.258 N/A 0.776 N/A
` Condence interval: 95%.
` Condence interval: 95%, nonsignicant relationship
` Condence interval: 95%, signicant relationship (P 0.047)
37 NOVM8R/DCM8R 2006
have happened by chance (Trochim
2000, p. 381).
Pindings
The response rate oI PWISEM stu-
dents Ior the pretest was 100 (all oI
the program students completed the
survey). HGCstudents had a response
rate oI 64, meaning that 63 out oI
99 students completed the survey.
Eor the posttest, the response rate Ior
PWISEMstudents was 94.3(33 out
oI 35 students completed the survey).
HGC students had a response rate oI
60.3 (38 out oI the 63 students who
had completed the pretest completed
the posttest). To increase the response
rate oI HGC students, we used an
electronic version oI the survey Ior
those who did not return the hard-
copy version or were absent in class
during the distribution.
According to the results oI the
pre- and posttests, the three groups
oI undergraduate studentsprogram
women, nonprogram women, and
nonprogram menembodied no
signifcant diIIerences in terms oI
interest, confdence, and determina-
tion to pursue one oI the SM&E ma-
jors. They had signifcantly diIIerent
scores on some oI the items about
views oI science and scientists, as
well as on select items that measured
their interest in and understanding
oI science and technology. The most
prominent diIIerence among the
groups was in their responses to the
item asking Ior intended/declared
majors in the posttest.
Similarityininterest, condence,
views of scientists
In terms oI their interest and conf-
dence in pursuing one oI the SM&E
majors, there were no signifcant diI-
Ierences either in the pretest or the
posttest (Tables 2 and 3). Students
were asked to rate their 'current
level oI interest and 'current level
oI confdence in pursuing a major in
science, mathematics, or engineering
on a 5-point scale, 1 being 'low and
5 being 'high. All students across
the three groups expressed almost the
same level4 out oI 5oI interest
and confdence both in the beginning
and the end oI the academic year.
Although the diIIerence in terms oI
interest among the groups was not
signifcant, as seen in Table 2, there
had been a decline in the interest mean
scores oI all three groups Irom the
pretest to the posttest. The decline oI
interest was consistent Ior all groups
and might indicate important implica-
tions, which we discuss later.
We obtained students` cumula-
tive GPA records Ior the academic
year during which this research took
place Irom the OIfce oI Admissions
and Records at the university. There
was no signifcant diIIerence in GPA
across the three groups. All oI the
groups had GPA means above 3.30
(on a 4-point scale), an indication oI
their academic success.
In terms oI their views oI science
and scientists, the three groups oI
students again were similar. However,
there were diIIerences regarding sev-
eral items. Eor example, in the begin-
ning oI the academic year, PWISEM
students were signifcantly more likely
(confdence interval: 95) to disagree
that 'scientists are apt to be odd and
peculiar people, whereas HGC men
students were more likely to agree
|F(2, 95) 5.53, p 0.005|. Regarding
this item on the posttest, students had
similar responses, and tended to dis-
agree. Also, PWISEM students were
signifcantly more likely to disagree
that 'scientists are not likely to be very
religious people than the HGC men
students in the beginning |F(2, 95)
4.05, p 0.021|; however, there was
no diIIerence at the end.
Regarding interest in and under-
standing oI science and technology,
in general, all students were alike.
The only exception was in their in-
terest in 'issues about the use oI new
inventions and technologies. HGC
men students were signifcantly more
likely to be 'very interested in these
issues than the HGCwomen group in
the beginning oI the academic year
|F(2, 95) 4.55, p 0.013|. At the
end oI the academic year, all students
expressed similar interest in the use oI
new inventions and technologies.
Differences in ma|or choice
Although students were similar in
terms oI their GPA scores and interest
in, confdence, and views oI science
and scientists, they were diIIerent in
terms oI major choice. According to
the results oI crosstabulation and a
Chi-square test perIormedIor students`
intended/declared majors in the begin-
ning oI their Ireshman year, there was
no statistically signifcant relationship
among the groups (Table 4).
In the beginning oI the academic
year, the majority oI students Irom
all three groups intended to pursue an
SM&Emajor (93.5oI the PWISEM
students, 76.9 oI the HGC women
students, and 86.2 oI the HGC men
students). Noticeably, at the beginning
oI the year there were no students
Iromeither the PWISEMor the HGC
men student groups who wanted a
nonSM&E major, while 11.5 oI
the HGC women students were in
this category.
However, at the end oI the same
academic year, there were more stu-
dents deciding Ior a nonSM&E major,
and this was particularly true Ior the
nonprogram women group (Table 5).
Moreover, according to the results oI
the crosstabulationandChi-square test,
the relationship among the groups and
their declared majors was signifcant,
|x
2
(4, N 86) 9.618, p 0.047|. At
the end oI the academic period, a con-
siderable number oI HGC women stu-
dents (30.8) and HGC men students
(13.8) declared a nonSM&E major.
Also, compared with the begin-
ning oI the academic year, there
were Iewer students in the undecided
category Irom both the HGC women
and HGC men groups. Namely, more
students Irom these groups decided
on the major they wanted to pursue,
but most oI these students decided
not to have an SM&E major. In other
words, aIter one year, signifcantly
more PWISEM students remained
in the SM&E felds than students in
either oI the HGC groups.
Conclusions
The women support and retention
program at the center oI this study
38 */52.!,
of
#/,,%'% 3#)%.#% 4%!#().'
(PWISEM) was successIul in retain-
ing Iemale students in their intended
SM&E majors. When compared with
nonprogram students, program stu-
dents were more likely to choose an
SM&E major. This diIIerence is in-
teresting given that there was no sig-
nifcant diIIerence among the groups
in terms oI their interest, confdence,
or academic potential (as measured
by SAT and GPAs). In other words,
all oI the undergraduates involved
in this study were interested and
confdent in pursuing SM&E, they
had similar views about science and
scientists, and similar interests in
and understandings oI science and
technology, but more program stu-
dents decided to remain in SM&E
aIter their frst year at college. Thus,
this research adds to the call Ior sup-
port programs or orientation courses
directed toward women.
There is clear indication that
Iactors other than interest or conf-
dence play a major role in students`
decisions to be involved in SM&E. As
other research has reported, supportive
environments, close student-Iaculty/
scientist relationships (Meyer 2002),
opportunities Ior research experiences,
mentoring, and academic networking
(Atkin, Green, and McIaughlin 2002;
Campbell and Skoog 2004; Down-
ing, Crosby, and Blake-Beard 2005),
and teacher attitudes (Cutler 2004;
Eriedman 1999), may be Iundamental
Iactors in PWISEM`s success. These
Iactors are also among the three suc-
cess criteriastrongacademic empha-
sis, multiple strategies, and a systems
approachdeveloped by Stage et al.
aIter their evaluation oI such programs
(Stage et al.`s study, as cited in Eraser
andWalberg 1995). Eurther researchin
examining and assessing programs de-
signed Ior undergraduate women may
suggest other criteria and strategies Ior
development and improvement.
The profle oI students involved
in this study suggests that there is
need Ior concern about the retention
oI both men and women in SM&E
majors. The consistent decline in the
interest oI all students (Table 2) is an-
other aspect that brings into question
the SM&E culture itselI. As Seymour
suggests, 'some aspects oI the learn-
ing environments in which many
women Ieel most comIortablepar-
ticularly those which are interac-
tive, cooperative, experiential, and
learner-Iocusedare also congenial
to many young men (1995, p. 470).
Seymour also suggests that chang-
ing the pedagogy Irom a teacher-
centered to a learner-centered ap-
proach promises to reduce the loss oI
both women and men students Irom
the SM&Emajors. II we are to reduce
the loss oI potential talent in SM&E,
we need to establish programs such
as PWISEMas frst-aid interventions
as we work toward transIorming sci-
ence teaching itselI. O
References
Atkin, A.M., R. Green, and I.
McIaughlin. 2002. Patching the
leaky pipeline. Journal of College
Science 1eaching 32 (2): 10208.
Campbell, A., and G. Skoog. 2004.
Preparing undergraduate women Ior
science careers. Journal of College
Science 1eaching 33 (5): 2426.
Cooperative Institutional Research
Program (CIRP). 2000. Survey of
incoming hrst-year freshmen and
transfers. Elorida State University:
OIfce oI the Vice President Ior
Student AIIairs.
Cutler, A. 2004. Mommy love and
the sciences. Journal of College
Science 1eaching 34 (3): 89.
Downing, R.A., E.J. Crosby, and S.
Blake-Beard. 2005. The perceived
importance oI developmental rela-
tionships on women undergradu-
ates`pursuit oI science. Psychology
of Women Quarterly 29: 41926.
Eraenkel, J.R., and N.EWallen. 2003.
How to design and evaluate re-
search in education. 5th ed. New
York: McGraw-Hill.
Eraser, B.J., and H.J. Walberg, eds.
1995. Improving science educa-
tion. Chicago: The National Soci-
ety Ior the Study oI Education.
Eriedman, D.I. 1999. Science, yes!
A program to excite the scientist
in the teacher. Journal of College
Science 1eaching 28 (4): 23944.
Malcom, S.M. 1984. Equity and excel-
lence. Compatible goals. Washing-
ton, DC: American Association Ior
the Advancement oI Science.
Matyas, M.I. 1992. Promoting under-
graduate studies inscience andengi-
neering. In Science and engineering
programs. Ontarget for women?eds.
M.I. Matyas andI.S. Dix. Washing-
ton, DC: National Academy Press.
Meyer, G.M. 2002. Encouraging
Iemale undergraduate students:
What can a science department
do? Journal of College Science
1eaching 32 (2): 98101.
Moore, R. 2001. The 'pretty redhead
who changed science education.
Journal of College Science 1each-
ing 31 (3): 194-96.
National Opinion Research Center
(NORC). 2004. General social
survey. University oI Chicago.
www.norc.uchicago.edu/profects/
gensoc.asp.
National Research Council (NRC).
1996. National science education
standards. Washington, DC: Na-
tional Academy Press.
NRC. 2000. Inquiryandthenational sci-
ence education standards. Washing-
ton, DC: National Academy Press.
National Science Board (NSB). 2002.
Science & engineering indica-
tors2002. Arlington, VA. www.
nsI.gov/sbe/srs/seind02/start.htm.
National Science Eoundation (NSE).
2000. Women, minorities, and per-
sons with disabilities in science and
engineering. 2000. Arlington, VA.
NSE. 2003. New formulas for Amer-
icas workforce. Girls in science
and engineering. Arlington, VA.
NSE. 2004a. Science and engineering
degrees. 19662001. Arlington, VA.
NSE. 2004b. Women, minorities, andper-
sons with disabilities in science and
engineering. 2004. Arlington, VA.
Seymour, E. 1995. The loss oI women
Irom science, mathematics, and
engineering undergraduate majors:
An explanatory account. Science
Education 79 (4), 43773.
Trochim, W.M. 2000. 1he research
methods knowledge base. http.//
trochim.human.cornell.edu/kb/
contents.htm.