Sex Roles, Vol. 53, Nos.

7/8, October 2005 (

C 2005)

DOI: 10.1007/s11199-005-7134-8

Relative Contributions of Childcare, Spousal Support,

and Organizational Support in Reducing Work–Family
Conflict for Men and Women: The Case of Turkey1

Zeynep Aycan2,4 and Mehmet Eskin3

The overarching purpose of the study was to investigate the role of three types of social
support (i.e., spousal, childcare, and organizational support) in relation to work–family con-
flict (WFC) in dual-earner families with children ages 0–6 years. The relationship of WFC
to psychological well-being and well-being in the domain of family was explored. Finally, the
relationship of spousal support to psychological well-being and marital satisfaction was exam-
ined. A total of 434 participants (237 mothers, 197 fathers) in dual-earner families in Turkey
with at least one preschool child participated in the study. The relative impact of childcare,
spousal support, and organizational support on WFC and outcome variables was tested si-
multaneously for both men and women through structural equation modeling. Spousal sup-
port was related to WFC for women, whereas both spousal and organizational support were
related to WFC for men. WFC had a negative relationship with psychological well-being,
marital satisfaction, and parental role performance for both men and women. Implications of
the results in relation to changing gender roles in a cultural context that is characterized by
high collectivism and low gender egalitarianism are discussed.

KEY WORDS: work–family conflict; social support; men and women; Turkey.

Women’s participation in the workforce is roles, which is generally referred to as “work–family

increasing all around the world (cf., Adler & Izraeli, conflict” (WFC). WFC is a type of inter-role conflict
1994; Davidson & Burke, 2004). This trend is re- that occurs as a result of incompatible role pressures
sulting in the transformation of traditional gender from the work and family domains (Greenhaus
roles and raising concerns about the psychological & Beutell, 1985; Kahn, Wolfe, Quinn, Snoek, &
well-being of women and men who are experiencing Rosenthal, 1964). The direction of the conflict
role overload and work–family conflict (Elloy & between work and family is inherently bidirectional
Smith, 2003; Staines, Pleck, Shepard, & O’Connor, (Gutek, Searle, & Klepa, 1991). That is, work may
1978). During the last three decades, numerous interfere with the family domain (work-to-family
studies have been conducted on the predictors and conflict; W-to-FC) and family may interfere with the
outcomes of the conflict between work and family work domain (family-to-work conflict; F-to-WC). In
1 The
the W-to-FC, the demands of work interfere with
paper received a Caroline Dexter Best Paper Award Nomi-
the performance of family responsibilities, whereas
nation, Academy of Management Annual Meeting, New Orleans,
August 2004. in F-to-WC, familial demands interfere with the
2 Department of Psychology, Koç University, Sariyer, Istanbul, performance of work-related responsibilities.
Turkey. The purpose of the present study was to exam-
3 Department of Psychiatry, School of Medicine, Adnan Menderes
ine the role of social support in WFC for men and
University, Aydin, Turkey. women in dual-earner families in Turkey with chil-
4 To whom correspondence should be addressed at Department

of Psychology, Koç University, Sariyer, Istanbul 34450, Turkey;

dren ages 0–6 years. We had three specific aims to
e-mail: [email protected]. examine: (1) the role of three types of social support

453 0360-0025/05/1000-0453/0

C 2005 Springer Science+Business Media, Inc.
454 Aycan and Eskin

(spousal, childcare, and organizational support) in do (e.g., Beutell & Greenhaus, 1980; Pleck, Staines,
relation to WFC (W-to-FC and F-to-WC); (2) the & Lang, 1980). Research in recent years points to
relation of WFC (W-to-FC and F-to-WC) to psy- the fact that WFC is not an issue that only concerns
chological well-being as well as well-being in the women (e.g., Duxbury, Higgins, & Lee, 1994; Eagle,
domain of family; and (3) the relation of spousal Miles, & Icenogle, 1997; Loerch, Russell, & Rush,
support to psychological well-being and marital 1989). As women with children are increasingly em-
satisfaction. ployed, new roles and role changes emerge for both
The majority of research on WFC has been men and women in such a way that women are more
conducted in Western industrialized societies, but involved in work outside the home and men are
there is an increasing recognition of the role of cul- more involved in work within the home (Barnett &
ture in WFC (cf., Aryee, Luk, Leung, & Lo, 1999). Baruch, 1987; Eagle et al., 1997; Frone & Rice, 1987).
Work and family issues are related to cultural be- The role of social support in reducing WFC for men
liefs, norms, and values, especially with respect to and women has been examined extensively in the
gender roles (Aryee, 1992; Lobel, 1991; Rosenbaum literature with specific emphasis on spousal and or-
& Cohen, 1999; Skitmore & Ahmad, 2003; Treas & ganizational support (e.g., Carlson & Perrewé, 1999;
Widmer, 2000; Williams & Best, 1990). Turkey is a Thomas & Ganster, 1995). However, the relative im-
country that bridges east and west geographically, as portance of different sources of support in relation to
well as culturally. The country is in a state of eco- WFC for men and women has not been explored sep-
nomic and cultural transition. Although 98% of the arately in a single study. The present study was de-
Turkish population is Muslim, Turkey has officially signed to fill this void by examining the role of three
been a secular state since the early 1920s with the sources of social support simultaneously: childcare
adoption of the Parliamentary Democratic Govern- support, spousal support, and organizational support.
ment System. After the end of the Islamic Ottoman
Empire, the Turkish Republic started the era of mod- Social Support and WFC: Theoretical Background
ernization in Turkey with a strong emphasis on liber-
alization and the emancipation of women. Entry into Social support refers to interpersonal relation-
the professions by women was very important in the ships and social interactions that help to protect in-
modernization of the Republic. dividuals from the effects of stress (Kessler, Price, &
According to recent data (Aycan, 2004), Turkish Wortman, 1985; Nielson, Carlson, & Lankau, 2001).
women make up 32% of professionals in scientific Both the availability and the quality of social support
and technical jobs, 35% of managerial personnel, are important social assets for individual adjustment
and 11% of entrepreneurs, directors, and top man- and well-being. Social support has consistently been
agement positions. Turkish women also represent a related to increased health and well-being (Cohen,
substantial proportion of professions such as phar- 1988; House, Landis, & Umberson, 1988). Not only
macists (60%), physicians (19%), dentists (30%), the actual availability of social support but also per-
lawyers (34%), and professors (23%) (Gürüz, 2001). ceived likelihood of receiving social support have
With the increasing involvement of women in the been found to be related to greater job satisfaction,
workforce, cultural values and norms with respect to lower job-related tension, and lower probability of
gender roles have been undergoing a rapid transi- terminating employment (Bedeian, Mossholder, &
tion. Women and men in professional jobs are try- Touliatos, 1986).
ing to adjust to the “modern” norms of gender roles The role of social support has been an issue
while keeping the traditional values of familialism of debate in the literature. Social support has been
and collectivism intact. This transition makes Turkey postulated both as a moderator between stressors
a unique cultural context in which to study WFC. and psychological well-being and as a main effect
According to Lopata’s (1966) model of family that influences psychological well-being (e.g., Cohen
stages, demands on the time and energy of parents & Wills, 1985). Carlson and Perrewé (1999) stud-
are the strongest in families with infants or children ied family and work support as an independent vari-
at the preschool age. There is strong evidence that able and as a moderator in the role stressors and
mothers of children aged 0–6 years are at great risk WFC, and they concluded that social support could
of psychological distress (Barnett, 1993), and they ex- be best conceptualized as a variable that directly in-
perience more conflict than mothers of older children fluences perceived stressors. More recently, Grandey
Social Support and WFC 455

and Cropanzano’s (1999) conservation of resources included in our study: supervisory support, organi-
model has been applied to WFC (e.g., Allen, 2001). zational policies and practices to reduce WFC, and
The model suggests that individuals are motivated time demands and inflexibility in employees’ jobs.
to acquire and maintain resources (i.e., social sup- Our conceptualization of organizational support re-
port networks) while coping with stress, which im- sembles the construct of “work–family culture” (e.g.,
plies that social support has a direct effect on WFC. Thompson, Beauvais, & Lyness, 1999, p. 392) that
We took a similar approach in our research, as we comprises managerial support for work–family bal-
proposed that three main sources of support (i.e., ance, career consequences associated with utilizing
spousal, organizational, and childcare support) were work–family benefits, and organizational time expec-
directly related to WFC. tations that may interfere with family responsibilities.
Spousal support is the help, advice, understand- First, there is the support provided by the supervisor
ing, and the like that spouses provide for one an- (e.g., Eby et al., 2002; Thomas & Ganster, 1995). As
other. Two forms of spousal support (emotional and in the case of spousal support, supervisory support
instrumental) have been conceptualized and empir- can be conceptualized as having two components: in-
ically tested (Adams, King, & King, 1996). Emo- strumental and emotional support (Frone, Yardley,
tional support includes emphatic understanding and & Markel, 1997).
listening, affirmation of affection, advice, and gen- Instrumental supervisory support refers to the
uine concern for the welfare of the partner. In- provision of direct assistance and advice with the in-
strumental support is tangible help from the part- tent of helping an employee to meet his or her family
ner in household chores and childcare. Increased responsibilities (Frone et al., 1997). Supportive su-
spousal support is shown to be associated with pervisors are instrumental in making and interpret-
lower levels of WFC (Aryee et al., 1999; Burke ing organizations’ work–family policies (Eby et al.,
& Greenglass, 1999; Erdwins, Buffardi, & Casper, 2002). On the other hand, emotional supervisory sup-
2001; Rosenbaum & Cohen, 1999), higher levels port refers to emphatic understanding and listening,
of psychological well-being (e.g., Ganster, Fusilier, sensitivity toward the WFC issues, and genuine con-
& Mayes, 1986), greater life satisfaction (LaRocco, cern for the well-being of the employee and his or her
House, & French, 1980), and better marital adjust- family (Frone et al., 1997). The importance of super-
ment (Frone, Russell, & Cooper, 1992; Rapoport & visory support is emphasized to such an extent that
Rapoport, 1971). Other researchers concurred that some researchers asserted that the well-being of fam-
spousal support was particularly influential in reduc- ilies lies largely in the hands of first-line supervisors
ing F-to-WC (Adams et al., 1996; Thomas & Ganster, (Rodgers & Rodgers, 1989).
1995). The second component of organizational sup-
Instrumental spousal support eases the bur- port is family-supportive organizational policies that
den of family demands and enables individuals are defined as services and allowances such as flex-
to devote more time to work, whereas emotional time, job sharing, and childcare facilities that are
spousal support enhances feelings of self-efficacy designed to help employees to make arrangements
both at home and at work (Parasuraman, Purohit, to balance their work and family responsibilities
& Godshalk, 1996). In light of the literature, we ex- (Thomas & Ganster, 1995). The final component is
pected spousal support to have a direct and posi- the organizational time demands that refer to the du-
tive relationship with psychological well-being and ration and flexibility of work hours (Behson, 2002).
marital satisfaction. Work overload that is coupled with inflexible and
long work hours increases WFC (Aryee et al., 1999;
H1: Spousal support would be positively associ- Clark, 2001; Duxbury et al., 1994).
ated with psychological well-being and marital The final source of support included in the
satisfaction. present study is childcare support. Although child-
care support has not been extensively studied by
Organizational support has been recognized as WFC researchers, it is an important support for dual-
playing a key role in employees’ ability to bal- earner families with young children. Failure to find
ance work and family responsibilities (Allen, 2001; satisfactory childcare arrangements is a primary rea-
Behson, 2002). The literature discusses three com- son why women leave jobs (Rosin & Korabik, 1990),
ponents of organizational support, which we have and it is a source of strain-based conflict (Wallace,
456 Aycan and Eskin

1999). Both women and men are reported to spend port of Frone et al.’s model, well-being in the family
almost one-half of their time at work unproductively domain (e.g., quality of family life) is associated neg-
due to childcare problems (Fernandez, 1986). Satis- atively with WFC (Duxbury & Higgins, 1991; Mauno
faction with childcare support leads to lower levels & Kinnunen, 1999). In our study, quality of family life
of WFC (Ahmad, 2002; Bedeian, Burke, & Moffett, has two components: satisfaction with parental role
1988), as well as better concentration at work and en- performance and marital satisfaction. Satisfaction
hanced job performance and satisfaction (Aryee & with parental role performance has three different
Luk, 1996). indicators: satisfaction with parenthood, satisfaction
In light of the literature, we proposed that three with time spent with children, and employment-
main sources of support were associated with two related guilt. Numerous studies show that the ex-
types of WFC differentially as stated in the following perience of WFC is related to lower marital satis-
hypotheses. faction and adjustment (e.g., Burley, 1995; Duxbury
& Higgins, 1991; Frone et al., 1992; Kinnunen &
H2: Spousal support would be more instrumental in Mauno, 1998).
reducing F-to-WC than W-to-FC. WFC also decreases satisfaction with role per-
H3: Childcare support would be more instrumental formance both at work and at home (cf., Bedeian
in reducing F-to-WC than W-to-FC. et al., 1988; Cooke & Rousseau, 1984; Williams &
H4: Organizational support would be more instru- Anderson, 1991). Individuals who are experiencing
mental in reducing W-to-FC than F-to-WC. WFC are less likely to be satisfied with their roles
as parents. Parental guilt has not been studied ex-
Psychological Well-Being tensively by WFC researchers, but it emerged as a
and Quality of Family Life significant theme in qualitative studies (e.g., Aycan
& Eskin, 2000; Napholz, 2000; Simon, 1995). In the
The outcome variables in the present study are WFC context, guilt arouses anxiety that occurs as a
parents’ psychological well-being (life satisfaction result of perceived failure to fulfill prescribed gen-
and depression) and the quality of family life (sat- der roles (cf., Chapman, 1987; Duxbury & Higgins,
isfaction with parental role performance and marital 1991). According to Staines’ (1980) fixed-sum-of-
satisfaction). Allen, Herst, Bruck, and Sutton (2000) scarce-resources theory, women’s involvement in the
conducted a comprehensive review of the outcomes work role may result in guilt regarding their perfor-
associated with WFC. The findings of their meta- mance as parents.
analysis revealed that several of the outcome vari- In accordance with Frone et al.’s theory, we pre-
ables in the present study had a strong relationship dicted that interference of work-to-family would be
with WFC: the weighted mean correlation of WFC associated with lower quality of family life. In accor-
with life satisfaction was −.28, depression was .32, dance of the theory, we also expected that both types
and marital satisfaction/functioning was −.23. of conflict would be associated with lower psycholog-
One of the most authoritative theoretical frame- ical well-being.
works in the WFC literature was proposed by Frone
and his colleagues (Frone et al., 1992, 1997). Accord- H5: Compared to F-to-WC, W-to-FC would be more
ing to their theory, work stressors lead to higher in- strongly associated with lower quality of family
terference of work-to-family, and, in turn, increase life (i.e., satisfaction with parental role perfor-
family distress or dissatisfaction. Similarly, family mance and marital satisfaction).
stressors lead to higher interference of family-to- H6: Both types of WFC would be negatively associ-
work and cause work distress or dissatisfaction. Dis- ated with psychological well-being.
tress in both family and work domains is associated
with depression. Gender Differences
In support of Frone et al.’s model, studies have
consistently showed that WFC decreases life satisfac- Gender differences in WFC remain inconclu-
tion (e.g., Kossek & Ozeki, 1998; Netemeyer, Boles, sive. Several researchers (e.g., Duxbury et al., 1994;
& McMurrian, 1996; Thomas & Ganster, 1995) and Frone et al., 1992; Gutek et al., 1991) found that
increases mental health problems such as depression, women experience interference from work-to-family
anxiety, burnout, and substance abuse (e.g., Frone, more than men do, whereas others (e.g., Duxbury &
2000; Vinokur, Pierce, & Buck, 1999). Again, in sup- Higgins, 1991; Kinnunen & Mauno, 1998) found no
Social Support and WFC 457

gender differences. One of the most influential theo- fluenced by the support they receive from the do-
ries in explaining gender differences is Pleck’s (1977) main that they feel primarily responsible for: family
asymmetric boundary permeability theory. The the- for women and work for men (cf., Cinamon & Rich,
ory first predicts that work and family boundaries are 2002). Due to the transitional nature of Turkey’s eco-
asymmetrically permeable; that is, W-to-FC is more nomic and social development, urban middle-class
prevalent than F-to-WC. Indeed, studies have shown men and women in Turkey are experiencing a transi-
that individuals typically report more W-to-FC than tion in gender roles. Now more than before, women
F-to-WC (e.g., Greenhaus & Beutell, 1985; Gutek are more involved in work, and men are more in-
et al., 1991). Pleck contends that men allow greater volved in family. This involvement is expected to pro-
interference from work-to-family than do women. duce stress because of the belief that work involve-
However, in today’s competitive work environment ment threatens a woman’s role as mother and wife
women also allow work to interfere with family, and that family involvement threatens a man’s role
which results in higher conflict for women than for as a committed employee (cf., Gutek et al., 1991).
men because gender role expectations in society do Support from the primary domain therefore, is crit-
not allow it (Parasuraman et al., 1996). According to ical in reducing the perceived threat to prescribed
Karasek (1979), women have less control over their gender roles and in increasing the sense of being
ability to satisfy work and family expectations, which, “approved of” and “accepted” by those in the pri-
in turn, leads to higher WFC (Duxbury & Higgins, mary social domain.
1991; Duxbury et al., 1994). Working mothers are
H7: For women, spousal support would be a more
more strongly affected by parental demands than are
important source of support in reducing F-to-WC
working fathers because they have to fulfill multiple
than childcare and organizational support.
roles simultaneously (e.g., parent, self, worker, and
H8: For men, organizational support would be a
spouse), rather than sequentially as men do (Hall,
more important source of support in reducing
1972). Not only do women have more family de-
W-to-FC than childcare and spousal support.
mands than men do, but they also experience more
sanctions than men do for noncompliance with fam- The hypotheses are combined and presented in
ily demands (Greenhaus & Beutell, 1985). the conceptual model to be tested in the present
Rosenbaum and Cohen (1999) found that study (see Fig. 1). According to the proposed medi-
spousal support is extremely important for women ated model, the three types of social support are re-
in cultures where there is low gender egalitarian- lated to lower F-to-WC and W-to-FC, which, in turn,
ism. In such cultures, women, more than men, tend are associated with better psychological well-being,
to internalize traditional gender roles and believe higher satisfaction with parental role performance,
that marital and parental relationships suffer as a and higher marital satisfaction. In addition to the me-
consequence of women’s work outside the home diated effect of spousal support, the model proposes
(Emmons, Biernat, Tiedje, Lang, & Wortman, 1990). that it is also directly related to psychological well-
Turkey was found to be low in gender egalitarian- being and marital satisfaction.
ism (Fikret-Paşa, Kabasakal, & Bodur, 2001). Pro-
fessional women in Turkey more strongly endorse METHOD
the viewpoint that “women’s place is at home and
near her husband” than do men (Aycan, 2004). It Participants and Procedure
was, therefore, expected that spousal support would
emerge as the most important source of support for The sample comprised 434 participants: 237
the women in Turkey. mothers (54.6%) and 197 fathers (45.4%) in dual-
Research shows that men’s well-being is also as- earner families with at least one child between the
sociated chiefly with spousal support (e.g., Barnett ages of 0 and 6 years. The sample was drawn from
& Hyde, 2001; Greenberger & O’Neil, 1993). How- four large banks. We purposefully chose the banking
ever, when spousal support was juxtaposed with or- sector for this study because it is one of the largest
ganizational support, we expected the latter to play and fastest growing sectors in Turkey, and because
a more important role in men’s WFC. This is in it has high rates of women employees at every
line with Etzioni’s (1984) contention that stress is level. We sent out letters to general managers of
buffered by work sources of support for men but by 18 banks, and 4 of them granted us permission to
family sources of support for women. People are in- collect data from their employees. Human resource
458 Aycan and Eskin

Fig. 1. Proposed conceptual model of the study.

management (HRM) departments of the banks

provided us with a list of male and female employees Table I. Demographic Characteristics of the Sample
who met our sampling criteria: being married, having Women Men
a spouse who is also working, having at least one N = 237 N = 197
child between the ages of 0 and 6, working full-time Age
on permanent basis, working in the organization M 31.8 35.4
for at least 1 year, and holding a white-collar job SD 3.9 4.6
position. We received names of 2,300 employees Education (%)
Less than high school 4.9 3.1
(45% women; 55% men). We randomly drew the High school 17.9 23.0
names of 400 male and 400 female employees, University 72.8 64.0
and then distributed our questionnaires to them Master’s 2.5 3.7
with the assistance of the HRM departments. Each Doctorate 1.9 6.2
questionnaire also included a self-stamped enve- Job position (%)
Manager 44 58
lope with the return address of the first author. Nonmanager 56 42
At the end of the second week, a reminder note Tenure (months)
was sent out to everyone to increase the response M 87 89
rate. We received a total of 443 questionnaires, SD 51.05 61.81
and discarded 9 of them due to excessive missing Number of children (%)
One 71.5 59.5
data. The response rate was 59.25% for women and Two 27 35.4
49.25% for men; this is above the typical response Primary childcare support (%)
rate (i.e., around 30%) for questionnaire-based Daycare 15.2 18.6
field studies (e.g., Rosenfeld, Edwards, & Thomas, Nanny 23.4 19.4
1993). Maternal grandmother 38.2 32.2
Paternal grandmother 22 29.8
Table I presents the demographic character- Sister of one of the parents 1.2 —
istics of both women and men. The only gender
Social Support and WFC 459

differences in the demographic characteristics of the and/or home-based care (1 = not satisfied at all; 5 =
sample were age, job position, and number of chil- very satisfied).
dren. As can be seen from the table, the mean
age of the fathers was slightly higher than that of Spousal Support
mothers, t(459) = −9.59, p < .001. The distribu-
tion of educational attainment of mothers and fa- The 44-item Family Support Inventory devel-
thers was similar. Two-thirds of the sample held a oped by King, Mattimore, King, and Adams (1995)
university degree that is representative of the pop- was used to assess perceived support received from
ulation of professionals or white-collar employees the spouse. This measure has two subscales: a 29-item
in corporate business life in Turkey. The percent- emotional sustenance subscale and a 15-item instru-
age of men with a managerial position was slightly mental assistance subscale. Emotional sustenance
higher than that of women, χ2 (1) = 8.26, p < .05. focuses on spousal behaviors or attitudes geared to-
Compared to male respondents, a higher percent- ward providing encouragement, understanding, at-
age of female respondents had one child between tention, positive regard, and guidance with problem
the ages of 0 and 6 years, χ2 (1) = 15.26, p < .01. solving. Included in this scale are behaviors such as
The mean age of children (including all children willingness to listen to, talk to, give advice, and show
in the family) was 59.51 months (SD 35.5 months). care and concern for the well-being of the spouse.
The data came from branches of banks in 38 dif- Instrumental assistance, on the other hand, includes
ferent cities in Turkey, including the three economi- behaviors and attitudes aimed at facilitating day-to-
cally most developed cities: Istanbul (55%), Ankara day family/household operations by sharing house-
(10%), and Izmir (10%). Although Turkey is a rela- hold tasks, actively relieving the partner of undue
tively homogenous country in terms of its ethnic, re- responsibility for family obligations and duties, and
ligious, and cultural makeup, there tends to be some structuring family life so as to accommodate the part-
regional differences in values (cf., Güvenç, 1994). ner’s work schedule or job requirements. The extent
Therefore, obtaining data from a variety of different of agreement with each item is rated on a 5-point
cities increased the representativeness of the sam- Likert scale that ranges from 1 (strongly disagree) to
ple. To further ensure the representativeness of the 5 (strongly agree). One-half of the items is reverse
sample, we compared the demographic character- coded to minimize the response bias. Higher scores
istics of the respondents (e.g., age, educational at- obtained from both subscales indicate higher social
tainment, number, and ages of children) with the support received from the spouse. The internal con-
rest of the employees who met the sampling criteria, sistency (Cronbach’s α) of the emotional sustenance
and we detected no significant differences between scale was .94 and of the instrumental assistance was
groups. .93 for the present sample.

Measurement Organizational Support

The research instrument was a questionnaire Organizational support was measured with three
that comprised 14 parts selected to measure vari- indicators: supervisory support, WFC policies and
ables under three categories: social support, WFC, practices, and time demands and inflexibility.
and outcome variables. Supervisory support was measured by 10 items
that were developed on the basis of a study by
Childcare Support Galinsky, Bond, and Friedman (1996). Items as-
sessed the extent to which managers provided sup-
Because our sample comprised working parents, port to employees with young children. This support
all of them had daycare or a home-based childcare was in form of both showing sympathy and under-
facility available for their children (see Table I). standing (e.g., “My supervisor gives advice on how
This part, therefore, was concerned with the satis- to handle my work and family responsibilities”) and
faction with (rather than availability of) childcare helping the employee to arrange work schedules to
arrangements set up for children under age 7. It was accommodate family demands (e.g., “My supervisor
assessed by one question asked about the extent to allows for flexibility in my working arrangements to
which the respondent was satisfied with the daycare enable me to handle my family responsibilities”).
460 Aycan and Eskin

Each item was rated on a 5-point Likert scale (1 = Emmons, Larsen, and Griffin (1985) was used in this
almost never; 5 = almost always), and high scores in- study as one of the indices of psychological well-
dicate high support. This was a reliable measure of being. Sample items include “In most ways my life
the construct (α = .85) in the present sample. is close to ideal” and “If I could live my life over,
Organizational WFC practices were assessed by I would change almost nothing.” A 5-point Likert
seven items developed to measure the existence of scale was used to respond to each item (5 = strongly
services that the organization provides to assist em- agree; 1 = strongly disagree). The scale was reliable
ployees with childcare. Respondents were asked to for this sample (α = .86). Higher scores obtained
indicate whether or not their organization provided from this scale indicate higher life satisfaction.
services such as daycare at work, daycare subcon- To assess psychological well-being, the 21-item
tracted by the organization, contribution to the ex- Beck Depression Inventory (Beck, Rush, Shaw, &
penses of daycare that the employee arranged him- Emery, 1979) was used to measure the frequency
self or herself, help with finding a caregiver, or train- and intensity of depressive symptoms. This measure
ing on childcare and education. Each item was coded was translated into Turkish and validated for Turk-
dichotomously (0 = no; 1 = yes). An index that ish populations by Hisli (1988). For each question,
ranged from 0 to 7 was created by adding up all there are four options from which to choose. Partic-
the responses. The higher the score on this index, ipants are asked to respond to each question based
the higher the organizational support and services to on their feelings during the last week, including the
assist employees’ families. present day. Responses to each question are coded
Time demands and inflexibility was measured from 0 (the first option) to 3 (the fourth option). The
by an 11-item scale (Duxbury & Higgins, 1991) that higher the score, the higher the depressive tenden-
was designed to measure the extent to which par- cies. Scores on the overall scale can range from 0 to
ticipants experience overload, flexibility, and control 63; 17 is the cutoff score for clinical depression among
in their jobs (e.g., “I have a little say over when I Turkish samples. This measure was found to be reli-
start and stop work,” “In my job, I have too much able for the sample of the present study (α = .87).
to do.”). Items are rated on a 5-point Likert scale
that ranges from 1 (strongly disagree) to 5 (strongly
agree). Higher scores indicate more demand and less Satisfaction with Parental Role Performance
flexibility. The internal consistency of this scale was
α = .84 in the present sample. This construct was captured by three variables:
satisfaction with time spent with children, satisfaction
Work–Family Conflict with parenthood, and employment-related guilt. One
question was designed to measure the level of satis-
The 10-item scale developed by Netemeyer et al. faction with the time spend with children. The ques-
(1996) was used to measure WFC. One-half of the tion read as follows: “To what extent are you satis-
items measures F-to-WC, and the other half mea- fied with the time you spend with your children?”
sures W-to-FC. Sample items include “The demands There were five options from which to choose. The
of my work interfere with my home and family life” responses ranged from 1 (I spend much less time with
(W-to-FC) and “Family-related strain interferes with my children than I wish) to 5 (I spend much more time
my ability to perform job-related duties” (F-to-WC). with my children than I wish).
The response scale is a 5-point Likert scale, where Satisfaction with parenthood emerged as an im-
5 indicates strong agreement and 1 indicates strong portant outcome of WFC in our previous qualita-
disagreement. Higher scores indicate higher conflict. tive studies in Turkey (e.g., Aycan, 2004). There
The internal consistency of both scales was high in are no instruments in the literature specifically de-
the present study (α = .89 for family-to-work, α = .90 signed to measure satisfaction with the role as a par-
for work-to-family). ent. Based on the general concept of “family role
satisfaction” (Kopelman, Greenhaus, & Connolly,
Psychological Well-being 1983), we generated 14 items to measure the extent
to which respondents were satisfied with their per-
There were two indicators of psychological well- formance as parents. Sample items include “I be-
being: life satisfaction and depression. The six- lieve that I am a very good mother/father,” “I be-
item Life Satisfaction scale developed by Diener, lieve that I meet all the needs of my child(ren),”
Social Support and WFC 461

and “Overall, I am very satisfied with my parent- in our questionnaire. The internal consistency was
hood.” The response scale was a 5-point Likert α = .83 for the present study.
scale (1 = strongly disagree; 5 = strongly agree). In the last part of the questionnaire, participants
Higher scores indicate higher satisfaction with par- were asked to provide information about themselves
enthood. The internal consistency of this scale was in such categories as age, gender, education, city of
α = .87. residency, tenure in the current organization, cur-
A nine-item scale was developed to measure re- rent job position, number of children, and ages of
spondents’ experiences of guilt resulting from work- children.
ing and not being able to spend enough time with The scales in English were translated into
their family. We generated items for this scale on Turkish by the researchers and were back-translated
the basis of our focus group discussions and in-depth by another researcher who is bilingual and was
interviews with more than 50 working women in blind to the goals of the present study. This pro-
Turkey (see Aycan, 2004). Sample items include “I cedure is recommended by Brislin (1980) to ensure
feel guilty for going to work and leaving my chil- both linguistic and conceptual equivalence. To test
dren everyday,” “I feel guilty for not being able to the clarity and relevance of the questions, we con-
spend as much time as I wish with my children.” Re- ducted a pilot study with 25 couples (i.e., 25 men
sponses were on a 5-point Likert scale (1 = strongly and 25 women) whose characteristics met our sam-
disagree; 5 = strongly agree). Higher scores indicate pling criteria. We discussed whichever questions ap-
higher guilt in this measure. This scale was found peared not to be clear and relevant to them and
to be a reliable measure of the underlying construct then modified a few questions on the basis of their
(α = .89). suggestions.

Marital Satisfaction RESULTS

To assess the quality of the marital relationship, Descriptive Findings

the Family Assessment Device developed by Gülerce
(1996) was utilized. Gülerce developed and validated Prior to the test of the hypothesized model, we
this measure to capture the transformational nature examined the descriptive findings and gender dif-
of the family structure in Turkish families. The 23- ferences on all study variables. Table II presents
item scale measures three main dimensions of mar- descriptive statistics for all the study variables. A
ital harmony: communication, coherence, and emo- paired sample t-test revealed that both men and
tional bonding between spouses. Examples of the women experienced higher W-to-FC than F-to-WC,
items include the following: “In our family, we can twomen (236) = −14.04, p < .001; tmen (196) = −11.56,
talk about every issue freely and openly,” “There p < .001. As can be seen in Table II, indepen-
are other people (e.g., friends, relatives, parents) to dent sample t-tests revealed a number of interest-
whom I feel closer than my husband/wife” (reverse ing gender differences. Women experienced greater
coded item). The extent of agreement with each item W-to-FC than men did. Compared to women, men
is rated on a 5-point Likert scale that ranges from 1 reported receiving more support from their spouse
(strongly disagree) to 5 (strongly agree). One-third of and supervisor at work. Men also experienced lower
the items is reverse coded to minimize the response employment-related guilt and higher satisfaction
bias. Higher scores obtained from this scale indicate with their parental role performance.
greater satisfaction with marital relationships. The Table III presents Pearson’s correlation coeffi-
reliability of the measure for the present sample was cients among all the study variables for both men
α = .88. and women. For both women and men, spousal sup-
Although the questionnaires were completed port (both instrumental and emotional) was strongly
anonymously, the literature cautions us against and negatively correlated with F-to-WC, whereas
the tendency to respond in a socially desir- the correlation between childcare support and F-
able way in collectivistic cultures such as Turkey to-WC was modest and only for women. Organiza-
(cf., van de Vijver & Leung, 1997). To ensure that tional support (especially the time demand compo-
this bias would not put our findings at risk, we in- nent) was strongly correlated with W-to-FC for both
cluded the seven-item version of Marlowe–Crowne men and women. Overall, both types of WFC as well
Social Desirability Scale (Crowne & Marlowe, 1964) as spousal support were correlated with indices of
462 Aycan and Eskin

Table II. Desciptive Statistics for all Study Variables

Scale Men Women Independent sample
Max. M SD M SD t-test
Family-to-work conflict 5 1.86 .87 1.96 .89 1.19
Work-to-family conflict 5 2.59 .72 2.78 .70 2.81∗∗
Childcare support 5 3.05 .90 3.41 .68 2.05∗
Spousal support (received)-emotional 5 3.86 .53 3.73 .62 −2.23∗
Spousal support (received)-instrumental 5 3.91 .48 3.35 .81 −8.70∗∗∗
Supervision support 5 3.25 .81 3.01 .72 −3.13∗∗
Organizational WFC practices 7 2.92 .70 2.44 .55 .42
Time demand and inflexibility 5 3.44 .68 3.39 .67 .79
Depression 63 7.88 6.77 9.00 7.17 1.71
Life satisfaction 5 3.26 .70 3.41 .70 2.33∗
Satisfaction with parenthood 5 3.56 .52 3.45 .56 −2.18∗
Satisfaction with time spent with children 5 2.71 .99 2.53 1.03 −1.86
Employment-related guilt 5 2.36 .77 2.69 .81 4.28∗∗∗
Marital satisfaction 5 3.71 .50 3.79 .55 .71

Note. ∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001.

psychological well-being and quality of family life. square residual (RMSR) is based on the average
It should also be noted that the social desirability of the squared discrepancies between the observed
score did not correlate significantly with any of the and implied covariance matrices. A value less than
key study variables. or equal to .10 is acceptable. Goodness-of-fit index
(GFI) is based on the sum of the squared discrep-
Test of the Model ancies between the observed and the implied covari-
ance matrices (Jöreskog, 1993). The adjusted GFI
AMOS (Version 3.6; Arbuckle, 1997) was used (AGFI) adjusts the GFI by the ratio of degrees of
as the structural equation-modeling program. The freedom in the model to the degrees of freedom in
major strength of structural equation modeling is the null model. Normed fit index (NFI) provides in-
that through the use of latent variables, it permits formation about how much better the model fits than
estimation of relationships among theoretically in- a baseline model (i.e., the null model). The incremen-
teresting constructs that are free of the effects of tal fit index (IFI) aims at decreasing the dependency
measurement unreliability. Structural equation mod- of NFI on sample size while simultaneously control-
eling allows simultaneous estimation of a measure- ling for the degrees of freedom. The comparative
ment model that relates observed indicators to latent fit index (CFI) uses uncorrelated variables based on
concepts by providing factor loadings and a struc- noncentrality parameters. GFI, AGFI, NFI, IFI, and
tural model that relates latent concepts to one an- CFI values range from 0 to 1; values above .90 indi-
other by providing path coefficients (Hayduk, 1987). cate good fit. The proposed model provided a good
This procedure was chosen in this study, so that fit to data. The χ2 /df ratio was 1.62; RMSR was .08;
it would be possible to test both the validity of GFI was .97; AGFI was .92; CFI was .96; IFI was .97;
measures and the relationships among underlying NFI was .91.
concepts. We used the standard procedure of multigroup
In the present study, spousal support, organi- modeling in order to test for gender differences in
zational support, psychological well-being, and sat- both the measurement and structural parameters. To
isfaction with parental role performance were desig- test gender differences in the measurement model,
nated as latent constructs. Each latent construct com- factor loadings were set equal between genders, and
prised several observed variables as seen in Fig. 1. their equality was tested by a nested likelihood ra-
Several criteria were used to evaluate the model tio chi-square test. All the other parameters were
fit. Ratio of chi-square to the degrees of freedom freely estimated (separate parameters for men and
is an index of model fit and parsimony. For χ2 /df women). In other words, the chi-square test was
ratio, Jöreskog and Sörbom (1982) recommended used to test whether the parameters were equal be-
a value approaching 2 as acceptable. Root mean tween genders. The testing of gender equality of
 Social Support and WFC

Table III. Pearson’s Correlations among Study Variables

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1. Childcare support — .07 .04 .15∗ .05 −.04 .02 −.10 .06 .03 .02 −.00 .04 .04 .10
2. Spousal support—emotional .22∗∗∗ (.94) .71∗∗∗ .19∗∗ −.15∗ −.15∗ −.09 −.29∗∗∗ .36∗∗∗ −.22∗∗ .28∗∗∗ −.04 .01 .48∗∗∗ .04
3. Spousal support—instrumental .32∗∗∗ .52∗∗∗ (.93) .13 −.02 −.07 −.08 −.38∗∗∗ .30∗∗∗ −.24∗∗∗ .22∗∗ −.06 .02 .49∗∗∗ −.05
4. Supervisory support .09 .17∗∗ .09 (.85) −.09 −.24∗∗∗ −.16∗ −.01 .18∗∗ −.19∗∗ .18∗∗ .21∗∗ −.07 .14∗ .07
5. Organizational WFC practices .12 .02 .13∗ .09 — .14∗ −.02 −.05 −.04 .07 −.08 .03 −.14∗ −.15∗ −.08
6. Time demand and inflexibility −.05 −.03 −.09 −.21∗∗∗ .10 (.84) .51∗∗∗ .28∗∗∗ −.25∗∗∗ .30∗∗∗ −.32∗∗∗ −.26∗∗∗ .16∗ −.19∗∗ −.07
7. Work-to-family conflict −.10 −.10 −.07 −.23∗∗∗ −.16∗ .49∗∗∗ (.90) .38∗∗∗ −.12 .19∗∗ −.35∗∗∗ −.36∗∗∗ .25∗∗∗ −.17∗ −.08
8. Family-to-work conflict −.14∗ −.31∗∗∗ −.28∗∗∗ −.11 −.04 .12∗ .25∗∗∗ (.89) −.13 .24∗∗∗ −.18∗∗ .01 .03 −.36∗∗∗ −.07
9. Life satisfaction .28∗∗∗ .37∗∗∗ .38∗∗∗ .10 .06 −.18∗∗ −.20∗∗ −.32∗∗∗ (.86) −.41∗∗∗ .38∗∗∗ .08 −.13 .33∗∗∗ .05
10. Depression −.14∗ −.23∗∗∗ −.26∗∗∗ −.25∗∗∗ −.17∗∗ .20∗∗ .36∗∗∗ .36∗∗∗ −.43∗∗∗ (.87) −.38∗∗∗ −.06 .32∗∗∗ −.37∗∗∗ −.11
11. Satisfaction with parenthood .16∗∗ .22∗∗∗ .18∗∗ .11 .03 −.23∗∗∗ −.33∗∗∗ −.17∗∗ .50∗∗∗ −.37∗∗∗ (.89) .19∗∗∗ −.34∗∗∗ .27∗∗∗ .10
12. Satisfaction with time spent .02 .01 −.00 .17∗∗ −.02 −.18∗∗ −.41∗∗∗ −.00 .21∗∗∗ −.14∗ .40∗∗∗ (.88) −.33∗∗∗ .01 .02
with children
13. Employment-related guilt −.09 −.16∗ −.12 −.21∗∗∗ −.18∗∗ .25∗∗∗ .52∗∗∗ .27∗∗∗ −.15∗ .42∗∗∗ −.51∗∗∗ −.47∗∗∗ (.87) −.07 .00
14. Marital satisfaction .24∗∗∗ .63 ∗∗∗ .57 ∗∗∗ .16∗ .15∗ −.10 −.26 ∗∗∗ −.42 ∗∗∗ .54 ∗∗∗ −.42∗∗∗ .36∗∗∗ .06 −.27∗∗∗ — .07
15. Social desirability .01 −.04 .01 −.08 −.10 −.09 −.07 −.10 .07 −.10 −.09 .01 −.06 .10 —

Note. ∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001.

Correlations above the diagonal are for men; below the diagonal are for women.
Numbers in parentheses on the diagonal are reliability coefficients of the scales.
463
464 Aycan and Eskin

Fig. 2. The structural equation modeling (standardized path coefficients for men are underlined).

the parameters of the structural model is explained Standardized path coefficients in the structural
below for each relevant hypothesis. model are also presented in Fig. 2 and Table IV. H1
predicted that spousal support would be positively
associated with psychological well-being and marital
Measurement Model
satisfaction for both genders. Results provided
support for this hypothesis for both women and men,
We tested gender differences for each path
and this structural parameter did not significantly
in the measurement model (see Fig. 1). The test
differ between genders for psychological well-being,
of equality of the parameters in the measurement
as indicated by the nested likelihood ratio test,
model across two genders was not significant, χ2 (6) =
χ2 (1) = .28, p = .54, or for marital satisfaction,
7.68, p = .26. Loadings of observed variables on their
χ2 (1) = .62, p = .84.
respective latent constructs that were equal for both
H2 stated that spousal support would be nega-
genders are presented for the estimated model in
tively related to F-to-WC rather than W-to-FC. Re-
Fig. 2.
sults confirmed the hypothesis that spousal support
was not significantly related to W-to-FC, whereas
Structural Model it was negatively associated with F-to-WC for both
men and women. The structural parameter did not
After testing all the hypothesized relationships, significantly differ between genders, χ2 (1) = 1.87,
we deleted the parameters that were not statistically p = .17.
significant for both genders. Model trimming resulted H3 predicted that the childcare support would
in a slight improvement in the model fit (The χ2 /df be significantly related to F-to-WC rather than
ratio = 1.60; RMSR = .08; GFI = .97; AGFI = .93; W-to-FC. Results did not support the hypothesis.
CFI = .96; IFI = .97; NFI = .92). The standardized Childcare support was not significantly related to
beta coefficients in the structural model (see Fig. 2) WFC either for men or for women.
also slightly improved compared to those in the orig- H4 stated that organizational support would
inal model (see Table IV). The explanation rates for be negatively associated with W-to-FC rather than
each endogenous variable are also reported in Fig. 2. F-to-WC. Results partially confirmed the hypothesis.
Social Support and WFC 465

Table IV. Path Coefficients of the Relationships among Key Variables: Gender Comparison
Women Men
Unstandardized Standardized Unstandardized Standardized
coefficient SE coefficient coefficient SE coefficient
Childcare support → F-to-WC −.08 .03 −.04 −.08 .03 −.11
Childcare support → W-to-FC −.07 .28 −.05 −.04 .05 −.05
Spousal support → F-to-WC −.60 .12 −.41∗∗∗ −.77 .15 −.36∗∗∗
Spousal support → W-to-FC −.13 .17 −.07 −.10 .18 −.04
Organizational support → F-to-WC −.14 .20 −.08 −.19 .20 −.11
Organizational support → W-to-FC −.12 .12 −.10 −.97 .46 −.55∗∗∗
Spousal support → psychological well-being 4.09 .86 .52∗∗∗ 4.62 1.23 .44∗∗∗
Spousal support → marital satisfaction .75 .08 .54∗∗∗ .60 .09 .57∗∗∗
F-to-WC → psychological well-being −1.39 .43 −.27∗∗ −.29 .57 −.04
F-to-WC → satis. with parental role perfor. −.10 .05 −.14∗ −.04 .05 −.09
F-to-WC → marital satisfaction −.06 .04 −.10 −.07 .04 −.12
W-to-FC → psychological well-being −1.52 .35 −.34∗∗∗ −.47 .40 −.20∗∗∗
W-to-FC → satis. with parental role perfor. −.38 .04 −.55∗∗∗ −.21 .04 −.66∗∗∗
W-to-FC → marital satisfaction −.09 .03 −.15∗ −.24 .03 −.18∗∗∗

Note. ∗ p < .05; ∗∗ p < .01; ∗∗∗ p < .001.

As hypothesized, organizational support was not re- The next hypothesis predicted that for women
lated to F-to-WC. It was significantly and negatively spousal support would be more strongly related to
related to W-to-FC for men but not for women. The lower F-to-WC than would childcare and organiza-
structural parameter showed that the gender differ- tional support. Indeed, results revealed that spousal
ence was significant, χ2 (1) = 7.22, p < .01. support was the only variable associated with lower
H5 predicted that W-to-FC would be more F-to-WC for women (see Fig. 2), which confirmed
strongly related to satisfaction with parental role per- H7. It should also be noted that, although not hy-
formance and marriage than would F-to-WC. As pothesized, spousal support was found to be the only
expected, for both men and women, W-to-FC was source of support that was related to lower F-to-WC
associated with lower satisfaction with parental role for men, as well (see Table IV). H8 predicted that for
performance; the gender difference in the structural men organizational support would be more strongly
parameter was not significant, χ2 (1) = 2.08, p = .17. associated with W-to-FC than would childcare and
Similarly, W-to-FC was associated with lower marital spousal support. The hypothesis was confirmed; or-
satisfaction (test of gender differences in the struc- ganizational support was the only source of support
tural parameter was not significant, χ2 [1] = 2.25, p = that was related to lower W-to-FC for men.
.34). In support of the hypothesis, F-to-WC was not
associated with marital satisfaction either for men
or for women, and it was only marginally related DISCUSSION
to parental role performance for women but not for
men (the structural parameter showed that the gen- The overarching purpose of the present study
der difference was significant, χ2 [1] = 2.16, p = .52). was to examine the role of social support in WFC for
H6 stated that both types of WFC would be re- both men and women; we focused on psychological
lated to lower psychological well-being to the same well-being and quality of family life as outcomes
extent for both women and men. Indeed, for both of WFC. We also investigated the direct effect of
men and women W-to-FC was associated with lower spousal support on psychological well-being and
psychological well-being (a gender difference was marital satisfaction. The sample of the study com-
not found in the structural parameter, χ2 [1] = 1.57, prised Turkish mothers and fathers in dual-earner
p = .21). F-to-WC, on the other hand, was negatively families with preschool children. As such, this study
related to psychological well-being of women but not contributes to the literature by testing theories and
of men. The gender difference in the structural pa- research findings derived from economically de-
rameter was significant, χ2 (1) = 5.32, p < .05, which veloped Western countries in a country that is
lends partial support to the hypothesis. economically and culturally different.
466 Aycan and Eskin

The findings supported Pleck’s (1977) theory of sized, we found that men also benefited significantly
asymmetric boundary permeability for this sample: from spousal support to offset the effects of F-to-WC.
both men and women allowed greater interference In summary, whereas for women a single source
from work-to-family than from family-to-work. The of support (i.e., spousal support) was critical in re-
findings also showed that women experienced more lation to WFC, for men, both spousal and orga-
W-to-FC than did men, which is in line with earlier nizational support were important: spousal support
research (e.g., Duxbury & Higgins, 1991; Greenhaus helped alleviate F-to-WC, whereas organizational
& Beutell, 1985; Hall, 1972; Karasek, 1979). support reduced W-to-FC. This is congruent with
The first set of hypotheses (H1–H4) dealt with the findings of earlier research (e.g., Greenberger &
the relative importance of three sources of support O’Neil, 1993; Higgins & Duxbury, 1992), which sug-
for W-to-FC and F-to-WC. As expected, we found gest that men in dual-career families need more sup-
that spousal support was associated with lower F- port from both work and family domains to cope with
to-WC rather than W-to-FC. Spousal support was WFC than do their counterparts whose spouses do
also strongly related to psychological well-being and not work. It has been suggested that men require a
marital satisfaction, which confirmed our hypothe- substantial amount of support to cope with organi-
sis. Organizational support, on the other hand, had zational and societal pressures (Higgins & Duxbury,
a stronger association with W-to-FC rather than F- 1992). On the one hand, changing gender roles pre-
to-WC but only for men. However, contrary to our scribe men (especially fathers) to be more involved
expectations, childcare support was not related to in their families (cf., Cinamon & Rich, 2002); on
F-to-WC. the other hand, this is not supported or tolerated by
There may be two explanations for why satisfac- most organizations. Furthermore, men do not have
tion with childcare support did not influence F-to- role models in their immediate social milieu to guide
WC. First, participants in our sample were not dis- them in their quest to balance work and family obli-
satisfied with childcare facilities to the extent that gations (Higgins & Duxbury, 1992), and they do not
they let this interfere with their work. The majority of exhibit help-seeking behavior as much as women do
parents relied on their family networks for childcare (cf., Addis & Mahalik, 2003).
(see Table I), and they reported high satisfaction with It is also possible to offer a cultural perspective
it. Quality daycare centers are rare and expensive to interpret the above findings. Turkey has a pater-
in Turkey. Whereas companies with more than 50 nalistic culture, which implies that people in author-
women employees are obliged by law to provide day- ity are expected to nurture those under their subor-
care, the majority does not offer this service and in- dination in a fatherly manner and to guide them in
stead pay the fine; the companies claim high cost, and their personal and professional life (Aycan, in press).
geographical dispersion among branches precludes Paternalistic managers are concerned with and in-
their offering of services. Home-based childcare (i.e., volved in personal and family lives of their subordi-
nannies) is common and affordable, but the majority nates. This is an expected and desired management
of parents prefer to use family networks for childcare approach in collectivistic and high-power distant cul-
because it is almost free and more reliable. Involve- tures such as Turkey (Aycan, in press). It is likely that
ment of extended family in childcare is a unique char- men expect protection, care, and guidance from their
acteristic of collectivistic countries such as Turkey. superiors at work in matters that concern their fam-
Childcare support may emerge as a key predictor ily life, whereas women expect the same from their
of WFC in countries where such networks are not husbands, who are perceived to be in a position of
available. “authority.” Hence, organizational support is a more
Second, it is possible that, when tested simulta- important source of support for men than it is for
neously with spousal support, childcare support lost women.
its predictive power. In fact, this is in line with our The findings supported our hypotheses regard-
hypotheses that, compared to spousal and organiza- ing the outcomes of WFC. In line with Frone et al.’s
tional support, childcare support is less influential theory (1992, 1997) and our fifth hypothesis, W-
on WFC. As expected, spousal support emerged as to-FC was associated with lower satisfaction with
the most important source of support in reducing F- parental role performance and with lower marital
to-WC for women, whereas organizational support satisfaction. Again, in support of the theory and our
emerged as the most important source of support in sixth hypothesis, both W-to-FC and F-to-WC were
reducing W-to-FC for men. Although not hypothe- related to lower psychological well-being, but the
Social Support and WFC 467

latter held true only for women. F-to-WC was related vergence of findings in WFC research conducted in
to lower psychological well-being and lower satisfac- different cultural contexts.
tion with parental role performance for women only. Although they endorse modern values, Turkish
Women feel distressed and dissatisfied with their urban middle-class families also try to keep the tra-
mothering when their work performance is lower ditional values of familialism and collectivism intact.
than what they desired. It is possible that when per- These values prescribe that the family is of premium
formance at work suffers due to family demands, importance in the society. Our findings reflect this
women feel unaccomplished both at home and at position. First of all, strong family ties in Turkey are
work, which lowers their psychological well-being. a major source of support for childcare, which helps
F-to-WC predicted marital satisfaction for neither families to cope with WFC. However, the most im-
men nor women. Overall, W-to-FC predicted study portant support from the family is that received from
outcomes better than did F-to-WC. Future studies the spouse. Spousal support not only helps families
should include work outcomes (e.g., organizational to cope with F-to-WC, but also allows them to ex-
commitment, job satisfaction, withdrawal behavior) perience psychological well-being and marital sat-
to better capitalize the predictive power of F-to-WC isfaction. Although this finding may be considered
(cf., Hammer, Bauer, & Grandey, 2003). universal, it is particularly striking in Turkey’s cul-
Finally, as expected, spousal support was as- tural context, which is low in gender egalitarianism
sociated with better psychological well-being and (Fikret-Pasa, Kabasakal, & Bodur, 2001). In such cul-
higher marital satisfaction for both men and women. tural contexts, egalitarian relationships in the nuclear
Parasuraman, Greenhaus, and Granrose (1992) con- family (i.e., instrumental and emotional support from
tended that spousal support played an important the spouse) play a crucial role for the well-being of
role, especially in the family satisfaction of women. individuals and families.
Because the provision of support for a partner’s ca- Finally, the importance of the family can be seen
reer is traditionally expected of wives, they tend to in the relationships of WFC with outcome variables.
overvalue the support provided by their husbands We found that interference of work role to family
(Parasuraman et al., 1992). role was a much stronger predictor of psychologi-
cal well-being, satisfaction with parental role perfor-
Cultural and Gender-Related mance, and marital satisfaction for both men and
Implications of the Study women. In other words, the possibility of harming
the family because of work responsibilities was more
This study was conducted in a non-Western cul- disturbing to Turkish dual-earner families than was
tural context, which is in a state of transition from the possibility of harming work life due to family
traditional to modern gender role ideologies (Aycan, responsibilities. In conclusion, our findings point to
2004). The transition started 80 years ago with the both universal and cultural-specific patterns of WFC.
establishment of Turkish Republic that marked the What seems to be specific to the Turkish cultural con-
era of modernization. Since then, Turkey’s culture text is the centrality of family both as a source of sup-
has become characterized by the confluence of tra- port and as a source of conflict and concern for work-
ditional Islamic values with modern Western values. ing people.
This is well-reflected in the findings of the present
study. Although some of our findings are in line CONCLUSIONS AND IMPLICATIONS
with those presented in the Western literature, there
are culture-specific patterns that are worth mention- The findings of the study have the potential to
ing. As a natural outcome of Turkey’s modernization make contributions to the extant literature. First, al-
process, cultural changes with respect to gender roles though social support has traditionally been treated
are in the same direction as those in the Western as a moderator between life demands and stress, the
part of the world. As proposed by the Expansionist results of the present study have demonstrated that
Theory (Barnett & Hyde, 2001), gender roles are ex- it could also be conceptualized as a key factor that
panding, especially in the urban, middle-class, profes- is directly related to WFC. Second, for the first time
sional population of Turkey. Women are more active multiple sources of social support were tested simul-
in professional work life, whereas men are more ac- taneously in a single study in relation to WFC. This
tive in family life, in today’s modern Turkey (Aycan, way, it was possible to test the relative importance
2004). The universality of this trend leads to a con- of childcare, spousal, and organizational support in
468 Aycan and Eskin

reducing WFC. Third, the role of social support for The second limitation of this study is related to
both women and men was compared. Finally, this is its measurement. Data were collected from a sin-
the first study in the published literature on WFC on gle source that may increase the common-method
Turkish men and women. This sample captured the bias. Spousal and organizational support could be
cultural transition in gender roles and how it influ- collected from different sources. However, percep-
enced the WFC process. tion of social support may be more directly related
The most important policy implication of the to WFC than is the presence or absence of it as re-
present study is the finding that organizational sup- ported by the original sources of support (i.e., the
port is a critical factor in reducing W-to-FC for men spouse or the organization). Furthermore, the con-
as well as for women. This points to the need to ceptual closeness of spousal support and marital sat-
change the academic and organizational discourse isfaction could have created spurious correlations
from “organizational WFC policies and practices are between them. Future researchers should test al-
for women” to “organizational WFC policies and ternative models where spousal support would be
practices are for all.” Prompted by this assertion, or- treated as an indication of marital satisfaction. An-
ganizations are more likely to pay attention to such other recommendation for future research is to in-
policies and practices and to show a willingness to clude work outcomes such as job satisfaction, or-
invest in them. Another policy implication of our ganizational commitment, absenteeism, intention to
study derives from the finding that spousal support turnover, and job performance.
plays a key role in WFC. As stated by Loscocco Finally, the cross-sectional design of the present
(1997), despite the changes in the workplace, the study does not allow us to make causal inferences.
family is the primary site where gender norms are Future research would benefit from longitudinal de-
reproduced, and therefore it is a site of potential signs in order to establish causal relationships. Due
change: “Changes in gender consciousness at home to the cross-sectional nature of our data, it would
are the ‘final frontier’ in the quest for greater gender be possible to test alternative models that may fit
equality in work-family linkages” (p. 223). Organiza- the data equally well. In such models, social support
tions can take leadership in promoting gender equal- could be treated as a moderator between WFC and
ity in the family by offering training programs to both its outcomes (e.g., psychological well-being and qual-
male and female employees on changing cultural val- ity of family life). Future researchers should investi-
ues and gender roles, communication and job shar- gate the role of social support by testing it as a main
ing at home, and childrearing practices. This is par- effect and as a moderator in a single study.
ticularly important for countries with low gender
egalitarianism such as Turkey (cf., Kulik & Rayyan, ACKNOWLEDGMENT
2003).
The limitations of the study include sample char- The authors wish to thank Nazli Baydar for her
acteristics, measurement, and the causal inferences guidance and suggestions on earlier drafts of the
in the proposed model. The sample comprised ur- article.
ban, middle-class employees in Turkey. Although
this sample represents the white-collar professional
working population in metropolitan cities, it does not REFERENCES
represent the Turkish population in general. The ma-
jority of economically active women work in agricul- Adams, G. A., King, L. A., & King, D. W. (1996). Relationships of
job and family involvement, family social support, and work-
ture (56.8% of women, 25.2% of men); many fewer family conflict with job and life satisfaction. Journal of Ap-
work in industry (14.4% women, 29.5% of men) and plied Psychology, 81, 411–420.
services (28.8% of women, 45.3% of men) (SIS, La- Addis, M. E., & Mahalik, J. R. (2003). Men, masculinity, and the
context of help seeking. American Psychologist, 58, 5–14.
bor Force Survey Results, 2000). The findings of the Adler, N. J., & Izraeli D. N. (Eds.) (1994). Competitive fron-
present study converged to a great extent with those tiers: Women managers in a global economy. Cambridge:
reported in the Western literature. This is mainly be- Blackwell.
Ahmad, A. (2002, April). Conflict between work and family roles of
cause our sample comprised well-educated, urban, employed women in Malaysia. Paper presented at the Meeting
and professional men and women. The influence of of the Society for Industrial and Organizational Psychology,
culture on WFC could be observed better with sam- Toronto.
Allen, T. D. (2001). Family-supportive work environments: The
ples from nonurban, less educated, and lower income role of organizational perceptions. Journal of Vocational Be-
families. havior, 58, 414–435.
Social Support and WFC 469

Allen, T. D., Herst, D. E., Bruck, C. S., & Sutton, M. (2000). Carlson, D. S., & Perrewé, P. L. (1999). The role of social support
Consequences associated with work-to-family conflict: A re- in the stressor-strain relationship: An examination of work-
view and agenda for future research. Journal of Occupational family conflict. Journal of Management, 25, 513–540.
Health Psychology, 5, 278–309. Chapman, F. S. (1987). Executive guilt: Who’s taking care of the
Arbuckle, J. L. (1997). Amos user’s guide, version 3.6. Chicago: children. Fortune, February, 30–37.
SmallWaters Corporation. Cinamon, R. G., & Rich, Y. (2002). Gender differences in the
Aryee, S. (1992). Antecedents and outcomes of work-family importance of work and family roles: Implications for work-
conflict among married professional women: Evidence from family conflict. Sex Roles, 47, 531–541.
Singapore. Human Relations, 45, 813–837. Clark, S. C. (2001). Work cultures and work/family balance. Jour-
Aryee, S., & Luk, V. (1996). Work and nonwork influences on the nal of Vocational Behavior, 58, 348–365.
career satisfaction of dual-earner couples. Journal of Voca- Cohen, S. (1988). Psychological models of the role of the social
tional Behavior, 49, 38–52. support in the etiology of physical disease. Health Psychology,
Aryee, S., Luk, V., Leung, A., & Lo, S. (1999). Role stressors, 7, 269–297.
inter-role conflict, and well-being: The moderating influence Cohen, S., & Wills, T. A. (1985). Stress, social support, and the
of spousal support and coping behaviors among employed buffering hypothesis. Psychological Bulletin, 98, 310–357.
parents in Hong Kong. Journal of Vocational Behavior, 54, Cooke, R. A., & Rousseau, D. M. (1984). Stress and strain from
259–278. family roles and work-role expectations. Journal of Applied
Aycan, Z. (2004). Key success factors for women in management Psychology, 69, 252–260.
in Turkey. Applied Psychology: An International Review, 53, Crowne, D., & Marlowe, D. (1964). The approval motive. New
453–477. York: Wiley.
Aycan, Z. (in press). Paternalism: Toward conceptual refinement Davidson, M. J., & Burke, R. (Eds.) (2004). Women in manage-
and operationalization. In K. S. Yang, K. K. Hwang, & ment worldwide: Progress and prospects. London: Ashgate.
U. Kim (Eds.), Scientific advances in indigenous psycholo- Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The
gies: Empirical, philosophical, and cultural contributions. satisfaction with life scale. Journal of Personality Assessment,
London: Cambridge University Press. 49, 71–75
Aycan, Z., & Eskin, M. (2000, July). A cultural perspective to Duxbury, L. E., & Higgins, C. A. (1991). Gender differences in
work-family conflict in dual-career families with preschool work-family conflict. Journal of Applied Psychology, 76, 60–
children: The case of Turkey. Paper presented at the Meeting 74.
of the International Association for Cross-Cultural Psychol- Duxbury, L., Higgins, C., & Lee, C. (1994). Work-family conflict:
ogy, Pultusk. A comparison by gender, family type, and perceived control.
Barnett, R. C. (1993). Multiple roles, gender, and psychological Journal of Family Issues, 15, 449–466.
distress. In L. Goldberger & S. Breznitz (Eds.), The handbook Eagle, B. W., Miles, E. W., & Icenogle, M. L. (1997). Inter-role
of stress (2nd ed., pp. 427–445). New York: Free Press. conflicts and the permeability of work and family domains:
Barnett, R. C., & Baruch, G. K. (1987). Social roles, gender, and Are there gender differences? Journal of Vocational Behav-
psychological distress. In R. C. Barnett, L. Biener, & G. K. ior, 50, 168–184.
Baruch (Eds.), Gender and stress (pp.122–143). New York: Eby, L. T., Douthitt, S. S., Perrin, T., Noble, C. L., Atchley, K.
Free Press. P., & Ladd, R. T. (2002). Managerial support for dual-career
Barnett, R. C., & Hyde, J. S. (2001). Women, men, work, and fam- relocation dilemmas. Journal of Vocational Behavior, 60, 354–
ily: An expansionist theory. American Psychologist, 56, 781– 373.
796. Elloy, D. F., & Smith, C. R. (2003). Patterns of stress, work-family
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cogni- conflict, role conflict, role ambiguity and overload among
tive therapy of depression. New York: Guilford. dual-career and single-career couples: An Australian study.
Bedeian, A. G., Burke, B. G., & Moffett, R. G. (1988). Outcomes Cross Cultural Management, 10, 55–67.
of work-family conflict among married male and female pro- Emmons, C. A., Biernat, M., Tiedje, L. B., Lang, E. L.,
fessionals. Journal of Management, 14(3), 475–491. & Wortman, C. B. (1990). Stress, support and coping
Bedeian, A. G., Mossholder, K. W., & Touliatos, J. (1986). Individ- among women professionals with preschool children. In
ual propensities for emotional supportiveness within a dual J. Eckenrode & S. Gore (Eds.), Stress Between Work and
career context: Work and non-work reactions. International Family (pp. 61–93). New York: Plenum Press.
Journal of Management, 7, 7–12. Erdwins, C. J., Buffardi, L. C., & Casper, W. J. (2001). The rela-
Behson, S. J. (2002). Which dominates? The relative importance tionship of women’s role strain to social support, role satisfac-
of work-family organizational support and general organiza- tion, and self-efficacy. Family Relations, 50, 230–238.
tional context on employee outcomes. Journal of Vocational Etzioni, D. (1984). Moderation effect of social support on the
Behavior, 61, 53–72. stress-burnout relationship. Journal of Applied Psychology,
Beutell, N. J., & Greenhaus, J. H. (1980). Some sources and con- 69, 615–622.
sequences of inter-role conflict among married women. Pro- Fernandez, J. P. (1986). Childcare and corporate productivity: Re-
ceedings of the Annual Meeting of the Eastern Academy of solving family-work conflicts. Lexington, MA: DC Health.
Management, 17, 2–6. Fikret-Paşa, S., Kabasakal, H., & Bodur, M. (2001). Society, orga-
Brislin, R. W. (1980). Translation and content analysis of oral nizations, and leadership in Turkey. Applied Psychology: An
and written material. In H. C. Triandis & J. W. Berry (Eds.), International Review, 50, 559–589.
Handbook of cross-cultural psychology (Vol. 2, pp. 389–444). Frone, M. R. (2000). Work-family conflict and employee psychi-
Boston: Allyn & Bacon. atric disorders: The national comorbidty survey. Journal of
Burke, R. J., & Greenglass, E. R. (1999). Work-family conflict, Applied Psychology, 85, 888–895.
spouse support, and nursing staff well-being during organiza- Frone, M. R., & Rice, R. W. (1987). Work-family conflict: The ef-
tional restructuring. Journal of Occupational Health Psychol- fect of job and family involvement. Journal of Occupational
ogy, 4, 327–336. Behavior, 8, 45–53.
Burley, K. A. (1995). Family variables mediators of the relation- Frone, M. R., Russell, M., & Cooper, M. L. (1992). Antecedents
ship between work-family conflict and marital adjustment and outcomes of work-family conflict: Testing a model of the
among dual-career men and women. Journal of Social Psy- work-family interface. Journal of Applied Psychology, 77, 65–
chology, 135, 483–498. 78.
470 Aycan and Eskin

Frone, M. R., Yardley, J. K., & Markel, K. S. (1997). Developing Kopelman, R. E., Greenhaus, J. H., & Connolly, T. F. (1983). A
and testing an integrative model of the work-family interface. model of work, family, and inter-role conflict: A construct val-
Journal of Vocational Behavior, 50, 145–167. idation study. Organizational Behavior and Human Perfor-
Galinsky, E., Bond, J. T., & Friedman, D. E. (1996). The role of mance, 32, 198–215.
employers in addressing the needs of employed parents. Jour- Kossek, E. E., & Ozeki, C. (1998). Work-family conflict, policies,
nal of Social Issues, 52(3), 111–136. and the job-life satisfaction relationship: A review and direc-
Ganster, D. C., Fusilier, M. R., & Mayes, B. T. (1986). Role of tions for organizational behavior human resources research.
social support in the experience of stress at work. Journal of Journal of Applied Psychology, 83, 139–149.
Applied Psychology, 71, 102–110. Kulik, L., & Rayyan, F. (2003). Wage-earning patterns, perceived
Grandey, A. A., & Cropanzano, R. (1999). The conservation of division of domestic labor, and social support: A comparative
resources model applied to work-family conflict and strain. analysis of educated Jewish and Arab-Muslim Israelis. Sex
Journal of Vocational Behavior, 54, 350–370. Roles, 48, 53–66.
Greenberger, E., & O’Neil, R. (1993). Spouse, parent, worker: LaRocco, J. M., House, J. S., & French, J. R. (1980). Social sup-
Role commitments and role related experiences in the con- port, occupational stress, and health. Journal of Health and
struction of adults’ well-being. Developmental Psychology, 29, Social Behavior, 21, 202–218.
181–197. Lobel, S. A. (1991). Allocation of investment in work and fam-
Greenhaus, J. H., & Beutell, N. J. (1985). Sources of conflict be- ily roles: Alternative theories and implication for research.
tween work and family roles. Academy of Management Re- Academy of Management Review, 16, 507–521.
view, 10, 76–88. Loerch, K. J., Russell, J. E. A., & Rush, M. C. (1989). The relation-
Gülerce, A. (1996). A family assessment device for Turkey. In ships among family domain variables and work-family conflict
J. Pandey, D. Sinha, & D. P. S. Bhawuk (Eds.), Asian con- for men and women. Journal of Vocational Behavior, 35, 288–
tributions to cross-cultural psychology (pp. 68–89). Delhi: 308.
Sage. Lopata, H. Z. (1966). The life cycle of the social role of housewife.
Gürüz, K. (2001). Dünyada ve Türkiye’de yükseköğretim: Tarihçe Sociology and Social Research, 51, 5–22.
ve bugünkü sevk ve idare sistemleri [Higher education in Loscocco, K. A. (1997). Work-family linkages among self-
Turkey and the world: History and administration systems]. employed women and men. Journal of Vocational Behavior,
Ankara, Turkey: ÖSYM. 50, 204–226.
Gutek, B. A., Searle, S., & Klepa, L. (1991). Rational versus Mauno, S., & Kinnunen, U. (1999). Conflicts between work and
gender role explanations for work-family conflict. Journal of family life. Community, Work and Family, 2, 147–171.
Applied Psychology, 76, 560–568. Napholz, L. (2000). Balancing multiple roles among a group of ur-
Güvenç, B. (1994). Türk kimliği [The Turkish identity]. Ankara, ban midlife American Indian working women. Health Care
Turkey: Ministry of Culture. for Women International, 21, 255–266.
Hall, D. T. (1972). A model of coping with role conflict: The role of Netemeyer, R. G., Boles, J. S., & McMurrian, R. (1996). Devel-
behavior of college-educated women. Administrative Science opment and validation of work-family conflict and family-
Quarterly, 17, 471–489. work conflict scales. Journal of Applied Psychology, 81, 400–
Hammer, L. B., Bauer, T. N., & Grandey, A. A. (2003). Work- 410.
family conflict and work-related withdrawal behaviors. Jour- Nielson, T. R., Carlson, D. S., & Lankau, M. J. (2001). The sup-
nal of Business and Psychology, 17, 419–436. portive mentor as a means of reducing work-family conflict.
Hayduk, L. A. (1987). Structural equation modeling with LISREL. Journal of Vocational Behavior, 59, 364–381.
Baltimore: John Hopkins University Press. Parasuraman, S., Greenhaus, J. H., & Granrose, C. S. (1992). Role
Higgins, C. A., & Duxbury, L. E. (1992). Work-family conflict: A stressors, social support, and well-being among two-career
comparison of dual-career and traditional-career men. Jour- couples. Journal of Organizational Behavior, 13, 339–356.
nal of Organizational Behavior, 13, 389–412. Parasuraman, S., Purohit, Y. S., & Godshalk, V. M. (1996). Work
Hisli, N. (1988). Validity of the beck depression inventory for and family variables, entrepreneurial career success, and psy-
Turkish populations. Psikoloji Dergisi, 7, 3–13. chological well-being. Journal of Vocational Behavior, 48,
House, J. S., Landis, K. R., & Umberson, D. (1988). Social rela- 275–300.
tionships and health. Science, 241, 540–545. Pleck, J. H. (1977). The work-family role system. Social Problems,
Jöreskog, K. (1993). Testing structural equation models. In K. A. 24, 417–427.
Bollen & J. S. Long (Eds.), Testing structural equation models Pleck, J. H., Staines, G. L., & Lang, L. (1980). Conflicts between
(pp. 295–316). Newbury Park, CA: Sage. work and family life. Monthly Labor Review, 103, 29–33.
Jöreskog, K. G., & Sörbom, D. (1982). Recent developments in Rapoport, R., & Rapoport, R. N. (1971). Further considerations
structural equation modeling. Journal of Marketing Research, on the dual career family. Human Relations, 24, 519–544.
19, 404–417. Rodgers, G., & Rodgers, J. (Eds.) (1989). Precarious jobs in labour
Kahn, R. L., Wolfe, D. M., Quinn, R. P., Snoek, J. D., & Rosenthal, market regulation: The growth of atypical employment in West-
R. A. (1964). Organizational stress: Studies in role conflict and ern Europe. Brussels: Free University of Brussels.
ambiguity. New York: Wiley. Rosenbaum, M., & Cohen, E. (1999). Equalitarian marriages,
Karasek, R. (1979). Job demands, job decision latitude, and mental spousal support, resourcefulness, and psychological distress
strain: Implications for job redesign. Administrative Science among Israeli working women. Journal of Vocational Behav-
Quarterly, 24, 285–307. ior, 54, 102–113.
Kessler, R. C., Price, R. H., & Wortman, C. B. (1985). Social fac- Rosenfeld, P., Edwards, J. E., & Thomas, M. D. (1993). Improving
tors in psychopathology: Stress, social support, and coping organizational surveys. Newbury Park, CA: Sage.
processes. Annual Review of Psychology, 36, 531–572. Rosin, H. M., & Korabik, K. (1990). Marital and family correlates
King, L. A., Mattimore, L. K., King, D. W., & Adams, G. A. of women managers’ attrition from organizations. Journal of
(1995). Family support inventory for workers: A new measure Vocational Behavior, 37, 104–120.
of perceived social support from family members. Journal of Simon, R. W. (1995). Gender, multiple roles, role meaning, and
Organizational Behavior, 16, 235–258. mental health. Health and Social Behavior, 36, 182–194.
Kinnunen, U., & Mauno, S. (1998). Antecedents and outcomes of Skitmore, M., & Ahmad, S. (2003). Work-family conflict: A survey
work-family conflict among employed women and men in Fin- of Singaporean workers. Singapore Management Review, 25,
land. Human Relations, 51, 157–177. 35–53.
Social Support and WFC 471

Staines, G. L. (1980). Spillover versus compensation: A review of Treas, J., & Widmer, E. D. (2000). Married women’s employment
the literature on the relationship between work and family. over the life course: Attitudes in cross-cultural perspective.
Human Relations, 33, 111–130. Social Forces, 78, 1409–1437.
Staines, G. L., Pleck, J. H., Shepard, L. J., & O’Connor, P. (1978). Van de Vijver, F., & Leung, K. (1997). Methods and data analysis
Job insecurity and well-being: A longitudinal study among for cross-cultural research. Thousand Oaks, CA: Sage.
male and female employees in Finland. Psychology of Women Vinokur, A. D., Pierce, P. F., & Buck, C. L. (1999). Work-family
Quarterly, 3, 90–120. conflicts of women in the air force: Their influence on mental
State Institute of Statistics (SIS). (2000). Household labor force health and functioning. Journal of Organizational Behavior,
survey results. Ankara, Turkey: SIS. 20, 865–878.
Thomas, L. T., & Ganster, D. C. (1995). Impact of family- Wallace, J. E. (1999). Work-to-nonwork conflict among married
supportive work variables on work-family conflict and strain: male and female lawyers. Journal of Organizational Behavior,
A control perspective. Journal of Applied Psychology, 80, 6– 20, 797–816.
15. Williams, J. E., & Best, D. (1990). Measuring sex stereotypes: A
Thompson, C. A., Beauvais, L. L, & Lyness, K. S. (1999). When multinational study. London: Sage.
work-family benefits are not enough: The influence of work- Williams, L. J., & Anderson, S. E. (1991). Job satisfaction and or-
family culture on benefit utilization, organizational attach- ganizational commitment as predictors of organizational cit-
ment, and work-family conflict. Journal of Vocational Behav- izenship and in-role behaviors. Journal of Management, 17,
ior, 54, 392–415. 601–617.