See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/227060714

Masculinity, Femininity, and the Bem Sex Role Inventory in Turkey

Article  in  Sex Roles · January 2005

DOI: 10.1007/s11199-005-1197-4

CITATIONS READS

119 3,893

2 authors:

Türker Özkan Timo Lajunen

Middle East Technical University Middle East Technical University
156 PUBLICATIONS   3,952 CITATIONS    132 PUBLICATIONS   7,463 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Country-wide seat belt observation View project

Global Road Safety Project - Turkey View project

All content following this page was uploaded by Türker Özkan on 28 March 2014.

The user has requested enhancement of the downloaded file.

Sex Roles, Vol. 52, Nos. 1/2, January 2005 (

C 2005)

DOI: 10.1007/s11199-005-1197-4

Masculinity, Femininity, and the Bem Sex Role

Inventory in Turkey

Türker Özkan1,2,3 and Timo Lajunen1

The aim of this study was to examine the masculinity and femininity scales of Bem Sex
Role Inventory (BSRI) among Turkish university students. Five hundred thirty-six students
(280 men and 256 women) volunteered to complete the short-form of the BSRI and an-
swer demographic questions. In factor analyses, the original factor structure (Bem, 1981)
was found both in the men’s and women’s data. Comparisons of the factor structures with
target rotation (Procrustes rotation) and comparison indexes showed no difference between
the factor structures found among men and women. The internal consistency of the masculin-
ity and femininity scales was acceptable, and t-tests showed that women scored higher on the
femininity scale, and men scored higher on the masculinity scale. There were significant dif-
ferences between men and women only on two masculinity items, but significant differences
were found in 8 (of 10) femininity items.

KEY WORDS: BSRI; factor structure; Turkey.

Gender stereotypes refer to “the beliefs peo- by society. An individual’s gender role was defined
ple hold about members of the categories man or as a function of the expression of masculine and fem-
woman” (Archer & Lloyd, 2002, p. 19). Many social inine traits rather than biological sex. Hence, traits
psychological studies have shown that these gender were called “masculine” if they were evaluated to be
stereotypes vary among different cultures and ethnic more suitable for men than women in society. Sim-
groups (Harris, 1994). The first aim of the present ilarly, “feminine” traits were those that were evalu-
study was to examine those stereotypes in Turkish ated to be more suitable for women than men. In ad-
cultural context. dition to masculinity and femininity scores, the BSRI
One of the most frequently used instruments for can be used to calculate scores that indicate “an-
measuring gender stereotypes is the Bem Sex Role drogyny” and “undifferentiated” classifications. An-
Inventory (BSRI) (Bem, 1974). In the BSRI, self- drogynous people are those who score similarly high
descriptions are used to measure the extent to which on the masculine and feminine scales (Bem, 1974),
men and women describe themselves in terms of per- whereas a person who shows low levels of both mas-
sonality traits that make up the stereotypes for their culine and feminine traits is called “undifferentiated”
own and the other sex (Archer & Lloyd, 2002). Thus, (Spence, Helmreich, & Stapp, 1975).
gender stereotypic traits of men and women were de- A multitude of studies have been conducted to
fined according to their social desirability determined investigate similarities and differences among coun-
tries in gender stereotypic traits and to assess the
1 Department of Psychology, Middle East Technical University, cross-cultural validity and the structure of the BSRI
Ankara, Turkey. (e.g., Ballard-Reisch & Elton, 1992; Blanchard-
2 Traffic Research Unit, Department of Psychology, University of
Fields, Suhrer-Roussel, & Hertzog, 1994; Lara-Cantu
Helsinki, Finland.
3 To whom correspondence should be addressed at Department & Suzan-Reed, 1988; Martin & Ramanaiah, 1988;
of Psychology, Middle East Technical University, 06531 Ankara, Schmitt & Millard, 1988; Waters, Waters, & Pincus,
Turkey; e-mail: ozturker@metu.edu.tr. 1977; Wong, McCreary, & Duffy, 1990). However,

103 0360-0025/05/0100-0103/0

C 2005 Springer Science+Business Media, Inc.
104 Özkan and Lajunen

the findings have been mixed. Although some au- pendence” (a combination of task-related indepen-
thors have suggested that gender stereotypic traits dence and relatedness) or “balanced differentiation
are universal (Basow, 1984; Pitariu, 1981), others and integration” (İmamoğlu, 1987, 1991). The special
have not demonstrated the universality (Kaschak & role of Turkish culture as a synthesis of “modern”
Sharratt, 1983; Ward & Sethi, 1986). This incon- Western values and the “traditional” values of the
sistency might be due to several factors. Different East should be manifested also in gender stereotypes.
research methods, samples, and problems with trans- Kağıtçıbaşı and Sunar (1992) pointed out that
lation, and adaptation of the instruments to non- the socialization of gender roles begins in the Turkish
English speaking cultures might lead to conflicting family even before the child is born. In Kağıtçıbaşı’s
results. In addition to methodological differences, (1982a) study, Turkish parents preferred a son (84%)
inconsistency of results might be related to the char- to a daughter (16%) in a forced choice question. Pref-
acteristics of gender stereotypes themselves. For ex- erence for a son especially in the rural traditional
ample, gender stereotypes do not remain unchange- context seems to be related to parents’ wish that a
able even within one culture, but change with time male child would carry the family name to next gen-
together with general cultural values (Twenge, 1997). eration, contribute to the family’s welfare through fi-
The second aim of the present study was to ob- nancial and practical help, and take care of the ag-
tain current data on gender stereotypes in Turkish ing parents. However, a daughter is perceived as “the
society. property of strangers” (Kağıtçıbaşı, 1982a, 1982b,
Bem (1979) emphasized the role of culture by 1982c; Kağıtçıbaşı & Sunar, 1992). These expecta-
defining the purpose of the BSRI to “assess the tions are likely to be the driving forces to make a
extent to which the culture’s definitions of desir- child to fit his or her gender stereotype. For instance,
able female and male attributes are reflected in an Turkish parents let their sons behave more indepen-
individual’s self-description” (p. 1048). By this defi- dently and aggressively, whereas more dependence
nition, it is reasonable to expect that the definition of and obedience is expected from their daughters; this
gender stereotypes will vary among cultures and eth- difference increases with the child’s age (Başaran,
nic groups (Harris, 1994; Landrine, 1985). Williams 1974).
and Best (1990) found that the gap between men and Sex segregation continues because the fam-
women on the variance of the gender stereotypes ily’s morality and honor (namus) depends on the
was small in highly developed countries, whereas it chastity of women. It results in supportive same
was larger among countries in which the difference sex-kinship and friendship networks (Kağıtçıbaşı,
between men and women in educational achieve- 1982a; Kağıtçıbaşı & Sunar, 1992; Kandiyoti, 1982),
ments was large. In addition, they suggested that gen- which provide women with emotional support and
der roles are closely associated with socioeconomic strength. Gender role differentiation can be seen
development, the importance of religion, urbaniza- in the division of labor between man and woman.
tion, and high latitudes. In this way, gender stereo- For example, men are responsible for farm-related
types would be results of many different cultural and tasks, physically heavy jobs, and external rela-
environmental factors. tions. Women are responsible for household tasks,
Traditionally, Turkey has been seen as a ge- gardening, care of domestic animals, and child-
ographical and cultural bridge between East and care. It is considered as a shame if men do
West. This mediator role of Turkey is seen in so- “women’s work” (Kağıtçıbaşı, 1982c; Kağıtçıbaşı &
cial values too. In studies of collectivism and indi- Sunar, 1992; see Ortaylı, 2002, for a comprehensive
vidualism, Turkish culture has repeatedly been de- monograph for family relations and role of family
scribed as a “culture of relatedness” (Kağıtçıbaşı, members).
1996). As a result, the modern urban family is defined In addition to different work roles, men and
as an emotionally (but not economically) interdepen- women are considered to have different person-
dent unit with “a combination, or coexistence, of in- ality traits. In Sunar’s (1982) study, Turkish men
dividual and group (family) loyalties” (Kağıtçıbaşı, evaluated Turkish women as more childish, more
1996, p. 89). Kağıtçıbaşı (1996) also suggested that dependent, less intelligent, more emotional, more ir-
child socialization is characterized by a trend toward rational, more submissive, less straightforward, more
“autonomous-relational” rather than an independent passive, more ignorant, more honest, more industri-
or interdependent self. Other scholars have called ous, and weaker than men. Gürbüz (1985) found that
this Turkish family characteristic “agentic interde- the BSRI items “affectionate,” “cheerful,” “gentle,”
Bem Sex Role Inventory in Turkey 105

“sympathetic,” “soft-spoken,” “eager to soothe hurt upgrade the knowledge of gender stereotypes in
feelings,” “sensitive to the needs of others,” and Turkey.
“loyal” were equally descriptive for both sexes. Also,
“independent,” “aggressive,” and “individualistic,”
which are instrumental characteristics, were undesir- METHOD
able for both sexes, whereas “dependency” was de-
sirable for both sexes in Turkey. These findings sup- Participants
port the notion of Turkey as having a “culture of
relatedness.” Five hundred thirty-six student volunteers par-
It seems that gender stereotypes in Turkish soci- ticipated in this study. The mean age of the male
ety differ from those of Western countries. Accord- university students was 21.94 years (range = 19–
ing to some studies, the content of Turkish gender 33, SD = 1.85), and the mean age of the female
stereotypes can be mostly accounted for by instru- university students was 21.56 years (range = 18–
mental and expressive dimensions (Gürbüz, 1985; 36, SD = 2.42). Only two students did not report
Kağıtçıbaşı & Sunar, 1992; Sunar & Fişek, in press). their age.
Kavuncu (1987) conducted a validity and reliability
study of the BSRI in Turkey. Even though test–retest Measures
reliabilities of masculinity and femininity scales were
found as high as 0.89 and 0.75, respectively, the The BSRI was developed to measure mascu-
BSRI was not considered a valid instrument for line, feminine, and androgynous personality styles
men. Small sample size and a problematic crite- among men and women. The original BSRI includes
rion variable (i.e., the masculine and feminine sub- 60 items (20 masculine, 20 feminine, and 20 neu-
scales of the MMPI) were given as reasons for the tral). The scale reliability coefficients reported in the
lack of validity. By taking into account these con- BSRI manual range from 0.75 to 0.90. In the present
cerns, Dökmen (1991) showed in her study that the study, gender stereotypes were measured with the
BSRI is a valid instrument for measuring masculin- short-form of the BSRI (Bem, 1981). The masculine
ity and femininity among both sexes. It was found scale (10 items) includes characteristics that are per-
that men scored higher on the BSRI masculinity scale ceived as men’s characteristics (e.g., assertive, strong
than women did, but no significant differences be- personality, and dominant). The feminine scale
tween the sexes on the BSRI femininity scale were (10 items) includes characteristics that are perceived
found. as women’s characteristics (e.g., emotional, sympa-
In addition to the historical role of Turkey as thetic, and understanding). The rest of the inventory
a melting pot of Western and Islamic values, re- (10 items) is composed of neutral items, which are
cent rapid social transition in Turkey makes it an es- perceived neither as men’s nor women’s characteris-
pecially interesting country in which to study gen- tics (e.g., conscientious, unpredictable, and reliable).
der issues. In the last decade, Turkey has gone Participants assessed how well each of the 30 person-
through a period of fast urbanization, industrializa- ality characteristics describes themselves by using a
tion, and Westernization supported by large-scale 7-point scale (1 = almost never true, 7 = almost al-
exposure to European and North American cul- ways true). The short-form of BSRI was translated
ture through the mass media. The current candi- to Turkish by using the translation-back translation
date status of Turkey for the European Union has method.
seemed to accelerate this movement. In addition,
international and regional migration, increasing ed-
ucational opportunities, the emphasis on secular- Procedure
ism, the newly acted civil code of equal property
division, the recognition of the value of house- The data were collected in psychology courses
wives’ unpaid labor, and the increased protection at the Middle East Technical University in Ankara.
of the rights of working women might also have The participants were assured of anonymity and con-
facilitated changes in gender roles and stereotypes. fidentiality. They were also offered an extra course
Therefore, the findings of the previous studies about credit point for their participation. The participants
gender stereotypes in Turkey may not hold any filled out the short-form of BSRI (Bem, 1981) and
more. It is necessary to replicate these studies and items related to demographic variables.
106 Özkan and Lajunen

RESULTS ficients (Cronbach’s alpha) for the masculinity and

the femininity subscales were 0.80 and 0.73, respec-
Factor Structure of the Short-Form of BSRI tively. In the women’s data, reliability coefficients for
Among Turkish Men and Women masculinity and femininity subscales were 0.80 and
0.66, respectively. Although the alpha reliability co-
Factor analyses (principal axis factor analyses efficients for both the men’s and women’s data were
with promax rotation) were conducted on the 10 lower than in original studies (Bem, 1981), they were
masculine and 10 feminine items separately for the acceptable.
men’s and women’s data. The neutral items from the
BSRI were not included in factor analyses. We de-
cided that there is no reason to expect that the per- Target Rotation and Agreement Coefficients
ception of the neutral adjectives would have changed
over time. The criteria used to determine the number Visual comparison of the men’s and women’s
of factors were Cattell’s scree plot and parallel anal- data indicate that the BSRI has the same factor struc-
ysis. The results of the factor analysis for the men’s tures among male and female students. In addition
data can be found in Table I and for women’s data in to visual inspection, target rotations of the men’s
Table II. and women’s factor matrices were carried out to test
Both the scree plot and the parallel analysis sup- the similarity of the factor structures. Proportionality
ported the two-factor model in both women’s and (Tucker’s phi) coefficients were calculated to assess
men’s data. Oblique promax rotation was used be- the similarity of the BSRI factor matrices found in
cause correlations between two factors were found, the men’s and women’s data sets. Proportionality co-
r = 0.27 in women’s data and r = 0.48 in the men’s efficient values above 0.90 indicate sufficient similar-
data. In both data sets, the masculinity items loaded ity between the factors. The values for Tucker’s phi
on the first factor and the femininity items loaded were 0.97 for the masculinity factor and 0.95 for the
on the second. In the men’s data, reliability coef- femininity factor. Hence, the factor structures found

Table I. Factor Structure of the BSRI Among Men

Factor loadings
Item-total Factor 1 Factor 2
correlation Masculinity Femininity Items
.71 .84 Dominant (22)
.58 .76 Assertive (7)
.60 .63 Has leader abilities (16)
.48 .63 Willing to take risks (19)
.40 .50 Independent (4)
.53 .47 Self-sufficient (13)
.52 .43 Strong personality (10)
.30 .42 Willing to take a stand (25)
.41 .38 Defends own beliefs (1)
.36 .34 Eager to soothe hurt feelings (17)
.66 .79 Compassionate (14)
.64 .72 Affectionate (23)
.50 .63 Gentle (29)
.46 .60 Understanding (11)
.42 .54 Tender (2)
.43 .49 Loves children (26)
.28 .46 Take into account other people’s feelings (8)
−.25 −.46 Aggressive (28)
.58 .34 .43 Warm (20)
.48 .33 Sympathetic (5)
Eigenvalues 5.57 2.48
Variance accounted 27.86 12.39
for (%)

Note. Factor loadings below .30 were omitted for the sake of clarity.
Bem Sex Role Inventory in Turkey 107

Table II. Factor Structure of the BSRI Among Women

Factor loadings
Item-total Factor 1 Factor 2
correlation Masculinity Femininity Items
.70 .81 Has leadership abilities (16)
.59 .73 Dominant (22)
.62 .69 Assertive (7)
.55 .66 Willing to take risks (19)
.46 .54 Independent (4)
.50 .52 Strong personality (10)
.44 .44 Defends own beliefs (1)
.40 .42 Self-sufficient (13)
.24 .34 −.31 Willing to take a stand (25)
.32 .32 Eager to soothe hurt feelings (17)
.60 .67 Compassionate (14)
.59 .65 Affectionate (23)
.45 .60 Gentle (29)
.46 .57 Understanding (11)
.50 .53 Sympathetic (5)
.47 .51 Warm (20)
−.36 −.51 Aggressive (28)
.30 .45 Sensitive to needs of others (8)
.32 .42 Tender (2)
.31 .36 Loves children (26)
Eigenvalues 4.79 3.12
Variance accounted 23.98 15.63
for (%)

Note. Factor loadings below .30 were omitted for the sake of clarity.

among male and female students were virtually iden- femininity items (items 2, 5, 8, 11, 14, 20, 23, and 26).
tical and allowed comparisons of the scores. There were also significant differences between men
and women on six “neutral” items of the BSRI (items
3, 9, 15, 18, 24, and 27) (Table III).
Sex Differences in Item and Scale Scores
DISCUSSION
The masculinity and femininity scores were
compared both within (pairwise t-test) and between In some previous studies, the validity of the fac-
men and women. Men scored lower on masculin- tor structure of the short-form of the BSRI has been
ity (M = 48.10, SD = 7.97) than on femininity (M = questioned (Wilcox & Francis, 1998). In this study,
53.62, SD = 7.84), t(279) = −10.46, p < .001. Sim- the applicability of the BSRI factor structure and
ilarly, women scored higher on femininity (M = gender stereotypes were investigated in a Turkish
56.67, SD = 6.62) than on masculinity (M = 46.97, sample. The results of the present study conducted
SD = 7.81), t(255) = −16.81, p < .001. Comparisons among Turkish male and female students support
between men and women showed no statistically sig- the original BSRI masculinity–femininity structure.
nificant difference on the masculinity scale of the Moreover, target rotation together with agreement
BSRI, t(534) = −1.67, p = .096, whereas there was indexes showed the factor structures found among
significant difference on the femininity scale of the men and women to be virtually identical. Reliabil-
BSRI, t(534) = 4.84, p < .001. ity analyses showed that the internal consistencies of
Table III lists item means and SD for men and the femininity and masculinity scales were acceptable
women and the corresponding t-test values. Table III both among men and women, although they were
shows that significant differences were found only on lower than those reported in the original studies con-
two items of the masculinity scale of the BSRI (items ducted in the USA.
25 and 28), whereas a statistically significant differ- It is interesting that the originally feminine
ence between men and women was found on eight item “eager to soothe hurt feelings” loaded on the
108 Özkan and Lajunen

Table III. Means and SD of the BSRI Items Among Turkish Male and Female
University Students
Men Women t
Items Mean (SD) Mean (SD) df = 534
Masculinity items 48.10 (7.97) 46.97 (7.81) 0.48
1. Defends own beliefs 5.74 (1.23) 5.91 (1.03) 1.78
4. Independent 5.02 (1.37) 5.00 (1.27) −0.19
7. Assertive 4.75 (1.39) 4.83 (1.33) 0.66
10. Strong personality 5.75 (1.21) 5.87 (0.99) 1.30
13. Self-sufficient 5.54 (1.08) 5.46 (1.06) −0.78
16. Has leader abilities 4.99 (1.56) 4.87 (1.50) −0.89
19. Willing to take risks 5.00 (1.55) 4.75 (1.53) −1.84
22. Dominant 4.60 (1.43) 4.39 (1.50) −1.60
25. Willing to take a stand 4.31 (1.64) 3.78 (1.70) −3.66c
28. Aggressive 2.43 (1.43) 2.10 (1.37) −2.68c
Femininity items 53.62 (7.84) 56.67 (6.62) 4.07c
2. Tender 5.23 (1.39) 5.87 (1.02) 6.03c
5. Sympathetic 5.03 (1.28) 5.41 (1.11) 3.66c
8. Sensitive to needs of others 5.49 (1.35) 5.97 (1.08) 4.44c
11. Understanding 5.76 (1.05) 5.96 (0.99) 2.23a
14. Compassionate 5.55 (1.20) 5.89 (0.99) 3.46c
17. Eager to soothe hurt feelings 5.15 (1.41) 5.15 (1.47) −0.03
20. Warm 5.40 (1.30) 5.63 (1.21) 2.17a
23. Affectionate 5.28 (1.21) 5.69 (1.12) 4.08c
26. Loves children 5.38 (1.61) 5.66 (1.60) 1.98a
29. Gentle 5.35 (1.13) 5.46 (1.12) 1.14
“Neutral” items
3. Conscientious 5.86 (1.23) 6.33 (0.73) 5.66c
6. Unpredictable 3.84 (1.67) 3.73 (1.68) −0.75
9. Reliable 6.15 (0.97) 6.32 (0.76) 2.33a
12. Jealous 4.29 (1.76) 4.45 (1.77) 1.06
15. Sincere 5.84 (1.03) 6.21 (0.85) 4.51c
18. Secretive 5.92 (1.27) 6.15 (1.06) 2.24a
21. Adaptable 5.64 (1.22) 5.79 (1.14) 1.47
24. Conceited 3.33 (1.70) 2.92 (1.54) −2.89b
27. Tactful 2.32 (1.30) 2.03 (1.26) −2.59b
30. Conventional 4.49 (1.65) 4.34 (1.68) −1.10
ap < .05.
bp < .01.
c p < .001.

masculine factor both in the men’s and women’s They found that the most important characteristics
data sets. The masculine item “aggressive” had a related to being a “good person” in Turkey were
strong negative loading in the femininity factor in “self-sacrificing” and “non-egoistic.” Although Türk-
both data sets but was not related to masculinity. Smith et al. (2000) did not investigate the character-
These findings might indicate that, among Turkish istics of an “ideal person” separately for men and
university students, an ideal man is expected to stay women, it seems that Turkish men are expected to
calm in troublesome situations without showing his be self-sacrificing and able to control their feelings
aggressive urges. According to a study conducted in in a difficult situation. “Aggressiveness” is still un-
Turkey by Fişek (1994), women are allowed to ex- desirable for both sexes (Gürbüz, 1985), and open
press their negative feelings more easily than men, displays of anger, either toward the parents or other
whereas men are expected to “be strong” and pro- authority figures, such as teachers, are not tolerated
vide emotional support when needed. Türk-Smith, (Kağıtçıbaşı & Sunar, 1992; Sever, 1985; Sunar, 2002;
Tevrüz, Artan, Smith, and Christopher (2000) sum- Sunar & Fişek, in press). On the other hand, some
marized several studies about Turkish students’ def- characteristics, which were earlier reported desirable
inition of an “ideal person” or a “good person.” for both sexes (i.e., affectionate, sympathetic, and
Bem Sex Role Inventory in Turkey 109

sensitive to needs of others) were feminine charac- gender roles. In addition to these social changes, the
teristics in the present study. masculinity of Turkish culture may require successful
One of the most striking findings of the present and work-oriented women to show masculine traits
study is that some instrumental characteristics (i.e., (e.g., assertiveness) in addition to traditional femi-
“independent,” “assertive,” “strong personality,” nine traits. Hence, modern Turkish women are ex-
“has leadership abilities,” “willing to take risks,” pected to endorse both masculine and feminine char-
“dominant,” “self-sufficient,” “defends own beliefs”) acteristics, i.e., androgyny. Modern Turkish women
are now desirable for both sexes. This result might are expected to be more flexible, more adaptable,
reflect the socialization process and change in val- and more free to be themselves in response to var-
ues. In the past, a woman’s social status derived ious environmental situations.
from her husband, number of children, and old age. In addition, six of the neutral items that Bem se-
Nowadays, success outside home seems to be an im- lected for the short-form of the BSRI showed signifi-
portant source of social status (Ortaylı, 2002). Sunar cant differences between men and women. This find-
(2002) found that although there are important ar- ing is in line with earlier results that gender-neutral
eas of continuity between traditional and modern BSRI items actually are not neutral in every culture
families, there are important differences as well. All (Eller & Dodder, 1989; Lara-Cantu & Suzan-Reed,
three generations showed in her study a trend of 1988). As the primary focus of the present study was
increasing encouragement of emotional expression masculinity and femininity, the question about the
and independence for their children. As urban fam- “neutral” items of the BSRI remains open. More re-
ilies endorse more “psychological” values (e.g., a search is needed to clarify the role of these filler items
loving relationship between parents and children) on BSRI.
and the instrumental value of the children as labor The present study has some methodological lim-
force or as a carrier of family’s name to new gener- itations that should be taken into account. The sam-
ations decrease, the importance of individual success ple of the study included only students at the Mid-
and achievement increases (Kağıtçıbaşı, 1996; Sunar, dle East Technical University (METU), which limits
2002; Sunar & Fişek, in press). It seems that these the generalizability of the results. METU is one of
new emphases will help to preserve the nature of the most respected universities in Turkey, and only
Turkish culture as a “culture of relatedness” even as approximately 1% of all applicants are accepted to
the society Westernizes in other ways. our undergraduate programs. Also, the language of
Both men and women scored higher on feminin- instruction in METU is English, and it is possible that
ity than on masculinity. Comparisons between men Western values are more dominant in METU than in
and women showed, however, that women scored other universities in Turkey. Hence, it is very likely
higher on femininity than men, whereas no differ- that the general public has a more traditional view
ences between the sexes were found on masculinity of gender roles than do our university students. In
scores. In previous studies conducted among Turkish the future, more research and different samples are
university students about 10 years ago, men scored needed before the BSRI can be applied with less-
higher on the BSRI masculinity scale than women educated and rural populations in Turkey.
(Dökmen, 1991). Hence, it seems that Turkish fe-
male students have adopted a more masculine gen- ACKNOWLEDGMENTS
der role within the last 10 years. This change in femi-
nine and masculine traits seems to be consistent with This work was supported by the Turkish
Twenge’s (1997) meta-analysis results, which indi- Academy of Sciences, in the framework of the Young
cated that women’s self-ratings on masculinity have Scientist Award Program (TL/TÜBA-GEBIP/2001-
been increasing and gender differences on masculin- 2-14), the Scientific and Technical Council of Turkey
ity have been decreasing over time. Cultural change (TÜBTAK project no: SBB-3023), and the H. J.
in Turkey seems to be the most possible explanation Eysenck Memorial Fund, and the Graduate School
for the change in a woman’s gender roles. Changes in of Psychology in Finland.
legal rights for women (e.g., women’s marital rights
and rights in divorce), expanded educational oppor- REFERENCES
tunities and the rapidly increasing number of female
students at universities, urbanization, and new values Archer, J., & Lloyd, B. (2002). Sex and gender. Cambridge:
might have influenced the traditional structure of the Cambridge University Press.
110 Özkan and Lajunen

Başaran, F. A. (1974). Psiko-sosyal gelişim (psycho-social develop- Kavuncu, A. N. (1987). Bem cinsiyet rolü envanterinin Türk
ment). Ankara, Turkey: Ankara University Press. toplumuna uyarlama çalışmaları (Turkish norms for BSRI).
Ballard-Reisch, B., & Elton, M. (1992). Gender orientation and Unpublished master’s thesis, Hacettepe Üniversitesi Eğitim
the Bem Sex Role Inventory: A psychological construct revis- Fakültesi, Ankara, Turkey.
ited. Sex Roles, 27, 291–306. Landrine, H. (1985). Race X class stereotypes of women. Sex
Basow, S. A. (1984). Cultural variations in sex typing. Sex Roles, Roles, 13, 65–75.
10, 577–585. Lara-Cantu, M., & Suzan-Reed, M. (1988). How valid is the social
Bem, S. L. (1974). The measurement of psychological androg- desirability scale of Bem’s Sex Role Inventory? Psychological
yny. Journal of Consulting and Clinical Psychology, 42, 155– Reports, 62, 553–554
162. Martin, H., & Ramanaiah, N. (1988). Confirmatory factor analysis
Bem, S. L. (1979). Theory and measurement of androgyny: A reply of the Bem Sex Role Inventory. Psychological Reports, 62,
to the Pedhazur-Tetenbaum and Locksley-Colten critiques. 343–350.
Journal of Personality and Social Psychology, 37, 1047–1054. Ortaylı, İ. (2002). Osmanlı Toplumunda Aile (Family in Ottoman
Bem, S. L. (1981). Bem Sex Role Inventory: Professional manual. Society). Istanbul: Pan Yayıncılık.
Palo Alto, CA: Consulting Psychologists Press. Pitariu, H. (1981). Validation of the CPI femininity scale in
Blanchard-Fields, F., Suhrer-Roussel, L., & Hertzog, C. (1994). A Romania. Journal of Cross-Cultural Psychology, 12, 111–
confirmatory factor analysis of the Bem Sex Role Inventory: 117.
Old questions, new answers. Sex Roles, 30, 423–457. Schmitt, B., & Millard, R. (1988). Construct validity of the Bem
Dökmen, Z. (1991). Bem cinsiyet rolü envanterinin geçerlik ve Sex Role Inventory (BSRI): Does the BSRI distinguish be-
güvenirlik çalışması (Validity and reliability study of BSRI). tween gender-schematic and gender-aschematic individuals?
Dil ve Tarih Coğrafya Fakültesi Dergisi, 35, 81–89. Sex Roles, 19, 581–588.
Eller, J. H., & Dodder, R. A. (1989). Relationship between gender Spence, J. T., Helmreich, R., & Stapp, J. (1975). Ratings of self and
and gender-neutral characteristics in three societies. Journal peers on sex role attributes and their relation to self-esteem
of Social Psychology, 129, 561–563. and conceptions of masculinity and femininity. Journal of Per-
Fişek, G. O. (1994). Paradoxes of intimacy: Analysis in terms sonality and Social Psychology, 32, 29–39.
of gender and culture. Boğaziçi Journal: Review of Social, Sever, L. (1985). Change in women’s perceptions of parental child
Economic and Administrative Studies, 8, 177–186. rearing practices, attitudes and beliefs in the context of social
Gürbüz, E. (1985). A measurement of sex-trait stereotypes. Unpub- change in Turkey: A three generation comparison. Unpub-
lished master’s thesis, Boğaziçi University, İstanbul, Turkey. lished master’s thesis, Boğaziçi University, Istanbul, Turkey.
Harris, A. C. (1994). Ethnicity as a determinant of sex role iden- Sunar, D. (1982). Female stereotypes in the U.S. and Turkey:
tity: A replication study of item selection for the Bem Sex An application of functional theory to perceptions in power
Role Inventory. Sex Roles, 31, 241–273. relations. Journal of Cross-Cultural Psychology, 13, 445–
İmamoğlu, E. O. (1987). An interdependence model of human de- 460.
velopment. In Ç. Kağıtçıbaşı (Ed.), Growth and progress in Sunar, D. (2002). Change and continuity in the Turkish middle
cross-cultural psychology (pp. 138–145). Lisse, Holland: Swets class family. In E. Özdalga & R. Liljestrom (Eds.), Autonomy
& Zeitlinger. and dependence in family: Turkey and Sweden in critical per-
İmamoğlu, E. O. (1991). Individualism and collectivism in a model spective (pp. 217–238). Istanbul: Swedish Research Institute.
and scale of balanced differentiation and integration. Journal Sunar, D., & Fişek, G. O. (in press). Contemporary Turkish Fam-
of Psychology, 132, 95–105. ilies. In U. Gielen & J. Roopnarine (Eds.), Families in global
Kağıtçıbaşı Ç. (1982a). The changing value of children in perspective. Allyn & Bacon.
Turkey (No. 60-E). Honolulu: East West Population Institute Türk-Smith, S., Tevrüz, S., Artan, I. E., Smith, K. D., &
Publication. Christopher, J. C. (2000, July). Turkish prototypes of the good
Kağıtçıbaşı, Ç. (1982b). Old-age security value of children: Cross- person. Paper presented at the 15th Congress of the Inter-
national socio-economic evidence. Journal of Cross-Cultural national Association for Cross-Cultural Psychology, Pultusk,
Psychology, 13, 133–142. Poland.
Kağıtçıbaşı, Ç. (1982c). Sex roles, value of children and fertility. Twenge, J. (1997). Changes in masculine and feminine traits over
In Ç. Kağıtçıbaşı (Ed.), Sex roles, family, and community in time: A meta-analysis. Sex Roles, 36, 305–325.
Turkey (pp. 151–180). Bloomington, IN: Indiana University Ward, C., & Sethi, R. R. (1986). Cross-cultural validation of the
Press. Bem Sex Role Inventory: Malaysian and South African re-
Kağıtçıbaşı, Ç. (1996). Family and human development across cul- search. Journal of Cross-Cultural Psychology, 17, 171–188.
tures: A view from the other side. Mahwah, NJ: Erlbaum. Waters, C., Waters, L., & Pincus, S. (1977). Factor analysis of mas-
Kağıtçıbaşı, Ç., & Sunar, D. (1992). Family and socialization in culine and feminine sex-typed items from the Bem Sex Role
Turkey. In J. L. Roopnarine & D. B. Carter (Eds.), Annual Inventory. Psychological Reports, 40, 567–570.
advances in applied developmental psychology: Vol. 5. Parent- Williams, J. E., & Best, D. L. (1990). Measuring sex stereotypes: A
child socialization in diverse cultures (pp. 75–88). Norwood, multinational study. Newbury Park, CA: Sage.
NJ: Ablex. Wong, F., McCreary, D., & Duffy, K. (1990). A further validation
Kandiyoti, D. (1982). Urban change and women’s roles in Turkey: of the Bem Sex Role Inventory: A multitrait-multimethod
An overview and evaluation. In Ç. Kağıtçıbaşı (Ed.), Sex study. Sex Roles, 22, 249–259.
roles, family, and community in Turkey (pp. 151–180). Bloom- Wilcox, C., & Francis, L. J. (1998). Beyond gender stereotyping:
ington, IN: Indiana University Press. Examining the validity of the Bem Sex-Role Inventory among
Kaschak, E., & Sharratt, S. (1983). A Latin American Sex Role 16- to 19- year old females in England. Personality and Indi-
Inventory. Cross-Cultural Psychology Bulletin, 18, 3–6. vidual Differences, 23, 9–13.
View publication stats