Content uploaded by Luo Lu
Author content
All content in this area was uploaded by Luo Lu on Nov 19, 2019
Content may be subject to copyright.
Testing an integrated model of the work–family interface in
Chinese employees: A longitudinal study
Luo Lu,1Ting-Ting Chang,2Shu-Fang Kao3and Cary L. Cooper4
1National Taiwan University,Taipei City, 2Lunghwa University of Science and Technology,New Taipei City, 3Hsuan
Chuang University,Hsin-chu City, Taiwan, and 4Lancaster University,Lancaster, UK
The purpose of this study was to examine an integrated model of the work–family interface (WFI) linking
work–family demands (workload and family conflict), resources (supervisory support and family support) and
role satisfaction in a Chinese context. The four-factor structure of WFI comprises direction of influence (work
to family vs family to work) and types of effect (work–family conflict vs work–family enrichment). A longitu-
dinal design was used to collect data from 409 Chinese employees at three time points, separating measures of
antecedents (T1), WFI (T2) and outcomes (T3) in time. The results based on structural equation modelling
(SEM) reveal that: (1) the direction and types of effect were two underlying dimensions of the WFI, supporting
the four-factor structure; (2) demands were more strongly related to conflict, while resources were more strongly
related to enrichment; (3) work–family conflict and enrichment were related to role satisfaction, regardless of the
direction of influence.
Key words: Chinese employees, work and family demands, work and family interface, work and family
resources, work and family satisfaction.
Research on the work and family interface (WFI) has
focused on a four-factor structure that comprises direction
of influence and types of effect (e.g. Frone, 2003;
Greenhaus & Powell, 2006). Specifically, work may inter-
fere with family (work-to-family conflict, WFC) and vice
versa (family-to-work conflict, FWC); and work may
enrich family (work-to-family enrichment, WFE) and vice
versa (family-to-work enrichment, FWE). There has been
supporting evidence for this four-fold taxonomy from US
and European samples (Innstrand, Langballe, Espnes,
Falkum & Gjerl, 2008; Proost, De Witte, De Witte &
Schreurs, 2010). However, as most of the work is been
carried out in the developed West, we know relatively
little about the WFI experiences of those living in non-
Western societies. For example, the only Chinese study
supporting the four-fold taxonomy used a small sample
(N=189) of working parents in Mainland China (Lu, Siu,
Spector & Shi, 2009a). While almost all of the existing
studies have focused on either the antecedents–WFI
linkage (e.g. Aryee, Srinivas & Tan, 2005; Proost et al.,
2010) or the WFI–consequences linkage (e.g. Innstrand
et al., 2008), very few have tested the integrated model of
the four-fold WFI including both its antecedents and
consequences (Cowlishaw, Birch, McLennan & Hayes,
2014; Witt & Carlson, 2006). However, Cowlishaw et al.
(2014) tested a structural model of WFI among Australian
volunteer firefighters, who are arguably different from
paid employees. Also, their model did not include any
variables from the family domain. Witt and Carlson
(2006) did not test a structural model; instead they
focused on the moderating effects of personality (consci-
entiousness) and work environment (perceived organiza-
tional support) on the relationships between various
aspects of the WFI and job performance, in a sample of
American employees. It is thus important to obtain
empirical support for an integral model of WFI, which
will inform us of the overall relationships among different
aspects of the WFI and their pertinent antecedents and
outcomes in both the work and family domains. Testing
such a comprehensive model in the Chinese context,
which has not yet been done, will be of added value in
generalizing knowledge gained in the Western cultures to
a very different societal background.
WFI: a demand–resource dual approach
Conservation of resources theory (COR, Hobfoll, 1989)
proposes that people seek to acquire and maintain
resources. On the one hand, work–family conflict occurs
because resources are lost in the process of juggling both
work and home life (Grandey & Cropanzano, 1999). If
people cannot cope with work and family demands, or if
Correspondence: Luo Lu, Department of Business Administra-
tion, National Taiwan University, No.1, Sec. 4, Roosevelt Road,
Taipei 106, Taiwan. Email: luolu@ntu.edu.tw
Received 10 November 2013; revision 26 March 2014; accepted
29 March 2014.
bs_bs_bannerAsian Journal of Social Psychology
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
Asian Journal of Social Psychology (2014) DOI: 10.1111/ajsp.12081
people deplete all resources, stress or conflict will develop.
On the other hand, adequately managing multiple roles may
create energy and enhance the availability of resources.
When resources generate new resources, work–family
enrichment occurs. Such a dual approach is also in line with
the Job Demands–Resources (JD–R) model, which deline-
ates the health impairment process and the motivational
process, and examines the impact of work demands and
resources on work–family conflict and enrichment (ten
Brummelhuis & Bakker, 2012). These integrated frame-
works highlighting the loss and gain processes of resources
mobilization have already received some support from lon-
gitudinal studies in the Chinese context (Lu, 2011; Lu &
Kao, 2013).
Applying the JD–R model, physical and psychic exer-
tions associated with work overload and family responsi-
bilities may trigger the health impairment process, causing
negative WFI (i.e. positively associated with work and
family conflict). Meta-analytic reviews summarizing find-
ings from the West (Byron, 2005; Michel, Kotrba,
Mitchelson, Clark & Baltes, 2011) have confirmed that both
work and family demands were consistently related to WFC
and FWC respectively. This is because physical and
psychic exertions associated with both work and family role
overload are related to the time and strain bases of the
work–family conflict (Carlson, Kacmar & Williams, 2000).
A recent meta-analysis (Michel et al., 2011) confirmed that
among the significant relations found in the work domain,
work–role overload had the largest relationship with WFC
(ρ=0.55). Michel et al. define work-role overload as ‘the
perception of having too many work-role tasks and not
enough time to do them’ (p. 693), which is nicely captured
by the Quantitative Workload Inventory (QWI) developed
by Spector and Jex (1998). Thus, we used workload meas-
ured by QWI as an indicator of work demands in the
present study.
In the family domain, an earlier meta-analysis (Byron,
2005) found that family conflict was strongly related to
FWC (ρ=0.32). Family conflict has been reported as a
major source of family pressure and highly demanding for
Taiwanese couples (Kao & Lu, 2006). In Byron’s (2005)
meta-analysis, family conflict encompassed marital con-
flict, parental conflict, marital tension, relationship agree-
ment and marital anger, which substantially overlaps with
our conceptualization and measurement of family conflict
as disagreement among family members (not restricted to
married couples) due to lack of congruence on important
aspects of Chinese family life (e.g. filial piety; Kao & Lu,
2006). We thus attempted to capture the psychological
reality of family pressure by assessing the level of family
conflict, which has not often been done in WFI research
(Michel et al., 2011).
Corroborating the above-cited Western findings, heavy
workload has been consistently identified as a precursor of
WFC (Lu, Hwang & Kao, 2005; Lu, Kao, Chang, Wu &
Cooper, 2008) while family conflict has been related to
FWC (Lu & Kao, 2013) for Chinese employees. Thus, we
focused on workload and family conflict as indicators of
work and family demands.
In contrast, resources available in the work and family
domains may energize the motivational process, as pre-
dicted by the JD–R model, which then facilitates better
adjustment (i.e. less work and family conflict). A most
recent meta-analysis (Michel et al., 2011) established that
work social support is an antecedent of WFC while family
social support is an antecedent of FWC.
Applying the JD–R model to the positive aspect of WFI,
demands in work and family domains may trigger the
health impairment process, preventing enrichment
between the two domains. This is because demands
deplete resources, which may hamper both the amount and
the effectiveness of resource transfer between the two
domains (i.e. less work and family enrichment). In con-
trast, resources in work and family domains may trigger
the motivational process, facilitating enrichment between
the two domains. Consistent with the COR theory and the
JD–R model, Hobfoll (1989) regards social support as a
critical form of resources, which generate feelings of love,
care and value. Such a positive state of mind then enables
individuals to transfer more effectively resources gained in
the family domain to the work domain, and vice versa (ten
Brummelhuis & Bakker, 2012). The beneficial effects of
supervisory support (one form of work resources) for
reducing work-to-family conflict are even stronger for Tai-
wanese employees working in organizations with greater
power distance compared to their British counterparts
(Lu et al., 2009b). Siu et al. (2010) also found a direct
beneficial effect of supervisory support for WFE in a
sample of Mainland Chinese workers.
In the family domain, family support has been identi-
fied as an important form of social support promoting per-
sonal adjustment in the West (e.g. Cohen & Syme, 1985)
and in Taiwan (Lu, 2006). Indeed, family support has
been found as an antecedent of both WFE and FWE in the
West (Grzywacz & Mark, 2000; Greenhaus & Powell,
2006).
Corroborating Western findings, supervisor support at
work not only helped Chinese employees to mitigate work
and family conflict (Chang & Lu, 2009), but also enhanced
work and family enrichment (Lu, 2011; Siu et al., 2010). As
found in a recent meta-analysis (Kossek, Pichler, Bodner &
Hammer, 2011), work–family-specific support had a
greater impact on various aspects of the WFI. We thus
focused on work–family supervisor support and family
support as indicators of resources from the work and family
domains.
COR and the JD–R model can also explain the effects of
WFI on work and family outcomes. Applying the central
2Luo Lu et al.
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
idea of resource mobilization, the work–family conflict as a
stressor may deplete individual resources and worsen both
work and family satisfaction (Grandey & Cropanzano,
1999), while work–family enrichment as the process of
obtaining new resources may facilitate both work and
family satisfaction (ten Brummelhuis & Bakker, 2012). In
this study we focused on work and family role satisfaction,
as they are most consistently related to all aspects of the
WFI in meta-analyses in the West (Allen, Herst, Bruck &
Sutton, 2000; McNall, Nicklin & Masuda, 2010). The few
studies with Chinese samples also produced the same
results (e.g. Aryee, Fields & Luk, 1999; Lu, 2011). Moreo-
ver, work-originated WFI was proven to be more strongly
associated with work-related outcomes than family-related
outcomes, and vice versa (Ford, Heinen & Langkamer,
2007). Based on the resource theories and existing empiri-
cal evidence, we thus hypothesized:
H1: WFI will mediate the relationships between work and
family demands and role satisfaction. Specifically,
workload will be positively related to WFC which in
turn will be negatively related to job satisfaction (H1a);
family conflict will be positively related to FWC which
in turn will be negatively related to family satisfaction
(H1b); workload will be negatively related to WFE
which in turn will be positively related to job satisfac-
tion (H1c); family conflict will be negatively related to
FWE which in turn will be positively related to family
satisfaction (H1d).
H2: WFI will mediate the relationships between work and
family resources and role satisfaction. Specifically,
supervisory support will be negatively related to WFC
which in turn will be negatively related to job satisfac-
tion (H2a); family support will be negatively related to
FWC which in turn will be negatively related to family
satisfaction (H2b); supervisory support will be posi-
tively related to WFE which in turn will be positively
related to job satisfaction (H2c); family support will be
positively related to FWE which in turn will be posi-
tively related to family satisfaction (H2d).
Work and family permeability:
a Chinese case
The Chinese culture has traditionally considered being
‘hardworking’ as a virtue, and working long hours is still
the norm in contemporary Chinese societies (Lu, 2011).
However, compared to individualists, collectivists such as
Chinese are more flexible in how they view work and
family issues, and the demarcation between work and
family is far from rigid in daily life (Yang, Chen, Choi &
Zou, 2000). Arecent qualitative study revealed that Taiwan-
ese employees often view work as a means of maintaining
and improving the living standards of their families, or as a
way of fulfilling their duties and commitments to glorifying
the family name (Lu, Chang & Chang, 2012). Afew studies
indeed found that Chinese employees took a more inte-
grated rather than segmented approach to work and family
issues, and were more tolerant of the spillover between the
two domains (Olson-Buchanan & Boswell, 2006; Yang
et al., 2000). These cultural expectations and flexible role
boundaries may explain the lack of association between
WFC and strain found in some Asian samples (e.g. Aryee
et al., 1999). Previous studies conducted in Chinese socie-
ties have also noted that the cross-domain effects of WFI on
work and family outcomes are rather prevalent. For
instance, work interfering with family was associated with
family satisfaction (Lu et al., 2005), while family interfer-
ing with work was associated with work-related outcomes
(Lu et al., 2008). One recent study conducted in Mainland
China provided direct support for the cross-domain effects
of family-to-work conflict on work outcomes (Li, Lu &
Zhang, 2013). Similarly, we would expect that the strong
linkage between work and family will also amplify the
positive mutual influences across the two domains, such
that when one’s family experiences benefit work perfor-
mance, one will have greater satisfaction for both work and
family (FWE →job satisfaction, FWE →family satisfac-
tion). Based on the above analysis of the distinct features of
the Chinese cultural context and existing empirical evi-
dence, we thus hypothesized:
H3: WFI will have cross-domain effects on role outcomes.
Specifically, WFC will be negatively related to family
satisfaction (H3a); FWC will be negatively related to
job satisfaction (H3b); WFE will be positively related
to family satisfaction (H3c); FWE will be positively
related to job satisfaction (H3d).
The proposed research model is presented in Figure 1.
Although WFI as a mediator has been tested in past studies
(e.g. Frone, Yardley & Markel, 1997; Jansen, Kant,
Kristensen & Nijhuis, 2003; Noor, 2003; Wayne,
Grzywacz, Carlson & Kacmar, 2007), these researchers did
not combine both work–family conflict and enrichment/
facilitation simultaneously in their models. Thus, the first
purpose of the present study was to extend Western theories
of the four-factor structure of WFI to a Chinese sample,
testing a comprehensive model including work and family
antecedents and consequences along with all four aspects of
the WFI. Second, we set out to investigate the mediating
roles of all four aspects of the WFI on the paths from work
and family demands and resources to work and family
satisfaction. Testing both conflict and enrichment, in both
directions, will allow us to explore cross-domain effects
and how conflict and enrichment relate to one another.
Testing an integrated WFI model in a diverse Chinese
sample will offer a significant contribution to the validation
and generalization of the results gained in the Western
studies. Finally, we adopted a three-wave study design to
An integrated model of the WFI 3
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
model the processes through which WFI arises and affects
employees’ role satisfaction. Such an approach can remedy
the methodological limitation caused by the scarcity of
longitudinal data in the existing work and family literature,
which should provide a stronger basis for causality infer-
ences than cross-sectional data. A longitudinal dataset can
also mitigate the risk of potential common method biases
(Podsakoff, MacKenzie, Lee & Podsakoff, 2003).
Method
Procedure and participants
We adopted a three-wave longitudinal design to test the
hypotheses. The participants were recruited from executive
education programs in several universities in Taiwan. All of
them, whether they were married or single, were full-time
employees working in different organizations in diverse
industries. Furthermore, all were living with their families.
The surveys were administered three times to the respond-
ents with a six-month interval between each pair of time
points. The span of 6 months was designed to provide
ample separation between our measures while not spacing
surveys so far apart as to unnecessarily increase participant
attrition. The first-wave survey included scales for all the
antecedent variables, namely, workload, family conflict,
supervisory support and family support. The second-wave
survey included scales for all aspects of the WFI, namely,
WFC, FWC, WFE and FWE. The third-wave survey
included scales for outcome variables, namely, job satisfac-
tion and family satisfaction. Questionnaires of the three
surveys were matched through self-identifiable coding.
At the beginning of the study, each participant received
an introductory letter informing them of the purpose of the
study, the commitment required, and assuring them of ano-
nymity. The response rate was 100% in the first-wave data
collection, 91% in the second-wave, and 90% in the third-
wave. At the end of the study period, 409 participants
contributed data for all three times. The high response rate
is probably attributable to the fact that the surveys were
conducted in class sessions and participants returned com-
pleted questionnaires on site.
As a precaution, we systematically examined differences
between employees in the panel sample and the dropouts
with regard to demographic characteristics as well as the
mean scores on the study variables. Analyses revealed no
significant difference. We thus concluded that no serious
selection problems owing to panel loss had occurred. The
sample was 48.7% male (N=199) and 51.3% female
(N=210), with a mean age of 36.14 (SD =10.16,
range =19–60), and mean job tenure of 8.50 years
(SD =8.58). About half of the sample (50.6%, N=207)
was married. Mean years of formal education was 15.99
(SD =2.08). About a quarter of the respondents (24.2%)
were managers at various levels. More people worked in
manufacturing (20.8%), the service industry (11.5%), and
culture and education (17.1%), than other occupations (e.g.
civil service, bio-medical industry).
It is worth noting why we included single respondents in
the sample. We believe that role experiences should not be
exclusively those of care providers, that is, spouses or
parents; the experiences of living in extended families with
parents and other family members can interfere or enhance
the quality of work life, and vice versa. This is especially
salient for Chinese employees, as most young adults live
with parents before marriage in Taiwan. Indeed, recent
Taiwanese studies have found that the experiences of work–
family conflict (Chang & Lu, 2009) and enrichment (Lu,
2011) did not vary between married and single Chinese
workers.
Measures
The survey was administered in Chinese, and all the
Chinese-version scales have been used in previous studies
with satisfactory reliability and validity. These references
will be given along with the original English version when
every scale is introduced below. In this paper, all scales
used five-point rating scales, and higher scores represented
higher levels of the designated constructs.
Work and family demands. Quantitative workload was
used to indicate work demands. Five statements from the
Quantitative Workload Inventory (QWI, Spector & Jex,
1998) were listed describing quantitative aspects of
work demands (e.g. ‘How often is there a great deal to be
Workload
Supervisory
support
Family conflict
Family support
WFC
FWC
FWE
WFE
Job satisfaction
Family satisfaction
0.03 -0.19*
-0.10*
-0.07
-0.05
0.10*
-0.12**
-0.03
0.31
***
0.20***
-0.16***
0.25***
0.08
0.09*
0.36***
0.21
*
0.69***
0.74***
Figure 1 The hypothesized research model and the
summary of standardized path coefficients for the modi-
fied model (N= 409).
Notes: All the one-way arrows represent the hypoth-
esized research model.
Solid lines represent significant coefficients, dotted
lines represent non-significant coefficients. The double
arrows represent the correlations between residual
terms.
4Luo Lu et al.
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
done?’). The Chinese version of QWI was used in Lu and
Kao’s study (2013) with good psychometric property. The
internal consistency of the QWI was 0.82 in the present
sample.
Family conflict was used to indicate family demands.
The Family Conflict Scale (FCS, Lu & Kao, 2013) meas-
ures the extent of conflict caused by disagreement in major
aspects of Chinese family life (e.g. ‘How often is there a
disagreement regarding caring for older parents in your
family?’). The four items we used pertained to filial piety,
household chores, communication/expression and friends/
social life. Previous research in Taiwan has demonstrated
that both married and non-married people could relate to
these aspects of family life, but they may use different
reference points to draw on their own ‘family’ experiences
(Chang & Lu, 2009). This scale was originally developed in
Chinese. The internal consistency of the FCS was 0.84 in
the present sample.
Work and family resources. Supervisory support was
assessed according to a three-item scale developed by Clark
(2001), tapping perceived understanding and psycho-
emotional support provided by direct supervisors regarding
workers’ family role obligations (e.g. ‘My supervisor
listens when I talk about my family’). The Chinese version
of the supervisory support scale was used in Chang and
Lu’s study (2011) with good psychometric properties.
Family support, in which four statements (e.g. ‘Sympa-
thetic understanding and concern’) were listed describing
informational, emotional, feedback and practical support
received from family members (O’Driscoll, Brough &
Kalliath, 2004). The Chinese version of the family support
scale was recently used in Lu’s study (2011) with accept-
able validity. The internal consistency of the supervisory
support scale was 0.90 and that of the family support scale
was 0.87 in the present sample.
WFC and FWC. The Work–Family Conflict Scale (WFCS,
Netemeyer, Boles & McMurrian, 1996) was used to assess
WFC and FWC separately. Sample items included: ‘The
amount of time my job takes up makes it difficult to fulfil
family responsibilities’ (WFC) and ‘I have to put off doing
things at work because of demands on my time at home’
(FWC). The Chinese version of WFCS was recently used in
Lu and Kao’s study (2013) with acceptable validity. The
internal consistency of the WFC scale was 0.94 and that of
the FWC scale was 0.91 in the present sample.
WFE and FEW. The Work–Family Enrichment Scale
(Carlson, Kacmar, Wayne & Grzywacz, 2006) was used to
assess WFE and FWE respectively. Sample items included:
‘My involvement in my work provides me with a sense of
accomplishment and this helps me be a better family
member’ (WFE) and ‘My involvement in my family helps
me expand my knowledge of new things and this helps me
be a better worker’ (FWE). The Chinese version of WFES
was recently used in Lu’s study (2011) with acceptable
validity. The internal consistency of the WFE scale was
0.86 and that of the FWE scale was 0.81 in the present
sample.
Work and family satisfaction. Three items from the Michi-
gan Organizational Assessment Questionnaire (Cammann,
Fichman, Jenkins & Klesh, 1979) were used to assess job
satisfaction (e.g. ‘In general, I like working here’). The
Family Satisfaction Scale (Edwards & Rothbard, 1999) was
used to assess family satisfaction (e.g.,‘My family life is
very enjoyable’). The Chinese versions of these two scales
were used for the Chinese sample with acceptable validity
(Lu & Kao, 2013). The internal consistency of the job
satisfaction scale was 0.80 and that of the family satisfac-
tion scale was 0.80 in the present sample.
In addition, information on sex (coded male =0,
female =1), age, marital status (coded married =1,unmar-
ried =0), education years, job tenure and job position
(coded managers =1,employees =0) were recorded. These
were intended as control variables.
Results
Descriptive statistics
Table 1 displays means, standard deviations, coefficient
alphas and correlations among all the variables. None of the
demographic variables, except marital status, was system-
atically related to our research variables. However, the cor-
relation pattern was similar for married and single
respondents. As a precaution, we also checked that the
results were generally the same when only married
respondents were used for model testing, with a few non-
significant paths probably due to the smaller sample size
(see below for model specification). Thus to facilitate
model estimation, the demographics were excluded from
all further analyses.
Model testing
As suggested by Anderson and Gerbing (1988), a two-step
approach to SEM analysis was employed in the current
study. Measurement models were first tested to examine the
distinctiveness of the measures, then the nested structural
model test was employed to test the research hypotheses.
We employed confirmatory factor analysis to compare
separate measurement models. Specifically, we compared
the hypothesized 10-factor model (measurement model,
workload, supervisory support, family conflict, family
support, WFC, FWC, WFE, FWE, job satisfaction and
family satisfaction) with four alternative models: the
An integrated model of the WFI 5
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
eight-factor model (combining WFC and FWC, WFE and
FWE), the seven-factor model (combining WFC, FWC,
WFE and FWE), the six-factor model (combining work-
load, supervisory support, family conflict, and family
support, job satisfaction and family satisfaction), and the
one-factor model. Table 2 displays fit indices for each of the
nested models listed in order, with chi-squared change sta-
tistics. The results clearly suggest that the measurement
model (10-factor) fits the data best.
The hypothesized Research Model (Model A, the struc-
tural model shown in Fig. 1) was compared to a Base
Model (Model B) in which all path coefficients to and from
work–family conflict and enrichment were constrained to
zero. Some SEM experts recommend against the use of the
GFI, AGFI, χ2/df ratio and NFI, as these fit indices can be
substantially affected by factors extrinsic to model
misspecification (e.g. sample size and number of indicators
per factor) and do not generalize well across samples;
instead they advocate the use of the CFI and RMSEA
(e.g. Anderson & Gerbing, 1988). We thus followed these
recommendations.
We found that Model A (χ2=1830.45, df =686,
p<0.001, CFI =0.89, RMSEA =0.07) fitted the data sig-
nificantly better than Model B (χ2=2017.19, df =697,
p<0.001, CFI =0.87, RMSEA =0.07; Δdf =11, Δ
χ2=186.74, p<0.001). However, the results of Model A
are less than desirable (CFI <0.90, RMSEA >0.06). We
thus modified the research model by allowing the error
terms of WFC and FWC, WFE and FWE to correlate
(shown in Fig. 1). This is because theoretically WFC and
FWC may be indicators of a higher-order factor, namely
work–family conflict; similarly, WFE and FWE may be
indicators of a higher-order factor, namely work–family
enrichment (see Table 1). The modified research model
(A1) produced a significantly better fit to the data
(χ2=1457.79, df =684, p<0.001, CFI =0.93, RMSEA =
0.05) compared to Model A (Δdf =2, Δχ2=372.66,
p<0.001) and Model B (Δdf =13, Δχ2=559.40,
p<0.001). The value of CFI for Model A1 was above 0.90
and the value of RMSEA was below 0.06. Therefore, we
concluded that Model A1 provided the most parsimonious
fit to the data.
The paths and parameter estimates for Model A1 are
shown in Figure 1. Among relations between antecedents
and WFI, all estimated paths were significant except for
paths between workload and WFE, and family support and
FWC. Among relations between WFI and role satisfaction,
all estimated paths were significant except for paths
between WFC and family satisfaction, FWC and family
satisfaction, and WFE and job satisfaction. Taken together,
Hypotheses 1a, 1d, 2a and 2d were supported, but Hypoth-
eses 1b, 1c, 2b and 2c were not supported. In addition, the
cross-domain paths between FWC and job satisfaction,
WFE and family satisfaction, and FWE and job satisfaction
Table 1 Means, standard deviations (SD ) and correlations for the research variables (N= 409)
MSD 12345678910111213141516
1. Gender – – –
2. Age 36.14 10.16 0.03 –
3. Marital status – – −0.02 0.68*** –
4. Years of
education
15.99 2.08 −0.10 −0.13* −0.02 –
5. Job tenure 8.50 8.58 0.07 0.77*** 0.52*** −0.23** –
6. Job position – – −0.09 0.25*** 0.21*** 0.07 0.14* –
7. Workload 18.17 2.92 0.09 0.07 0.13 0.19*** 0.05 0.11* (0.82)
8. Supervisory
support
8.79 2.55 −0.02 0.12 0.14* −0.13 0.18** 0.09 −0.04 (0.87)
9. Family conflict 10.02 3.12 −0.06 −0.09 −0.15* 0.07 −0.13 0.01 0.12 −0.08 (0.83)
10. Family support 16.24 4.64 0.13 0.08 0.15* 0.16** 0.03 −0.04 0.00 0.16** −0.26*** (0.90)
11. WFC 12.56 3.57 −0.04 0.04 0.10 0.11* 0.06 0.10 0.36*** −0.08 0.21*** −0.07 (0.93)
12. FWC 10.70 2.90 −0.11 0.10 0.15* 0.04 0.13 0.08 0.13 0.00 0.20*** −0.09 0.65*** (0.91)
13. WFE 9.41 2.01 0.01 0.12 0.12 0.00 0.12 0.09 0.03 0.32*** −0.16** 0.22*** −0.10 −0.05 (0.86)
14. FWE 10.15 1.84 0.04 0.11 0.08 −0.02 0.10 0.12* 0.04 0.20*** −0.23*** 0.31*** −0.03 −0.15** 0.66*** (0.81)
15. Job satisfaction 9.67 1.10 0.05 0.05 0.13 0.11* 0.05 0.06 0.07 0.11 −0.15** 0.19*** 0.00 −0.01 0.09 0.04 (0.73)
16. Family
satisfaction
12.13 1.95 0.07 0.05 0.09 0.05 0.14* −0.03 −0.08 0.13** −0.35*** 0.36*** −0.14** −0.17*** 0.24*** 0.34*** 0.31*** (0.96)
Notes:Sex:0=male, 1 =female; Marital status: 1 =married, 0 =unmarried; Job position: 1 =manager, 0 =non-manager; FWC, family-to-work conflict; FWE, family-to-work enrichment; WFC, work-to-
family conflict; WFE, work-to-family enrichment. Cronbach’s alpha reliabilities are in parentheses on the diagonal. *p<0.05, **p<0.01, ***p<0.001.
6Luo Lu et al.
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
were also significant. Thus, Hypotheses 3b, 3c and 3d were
supported, but Hypothesis 3a was not supported.
Discussion
One aim of the present three-wave study was to empirically
test an integrated model of work–family interface, high-
lighting the mediating roles of all four aspects of the WFI
(i.e. WFC, FWC, WFE, FWE). Existing studies have
assessed effects of antecedents on WFI (e.g. Aryee et al.,
2005; Proost et al., 2010), or those of WFI on outcomes
(e.g. McNall et al., 2010). To the best of our knowledge, the
present study is the first to link the antecedents, outcomes
and all four WFI mediators in a comprehensive structural
model. Extending the previous studies mentioned above,
we systematically included demands, resources and role
satisfaction in both work and family domains. We also
examined processes involving the dual direction and dual
nature of the WFI. Another feature of the present study is
the use of Chinese employees for validating and extending
Western-based theories and inferences on work and family
issues. Instead of using a convenient but restricted sample
from one or two organizations, we recruited a large sample
from diverse organizational and industry backgrounds, thus
enhancing the generalizability of our findings (c.f. Lu et al.,
2009a). Last but not least, we enhanced the methodological
strength of the work–family research through the inclusion
of a longitudinal dataset, which is uncommon in the litera-
ture. Longitudinal data not only provide a stronger basis for
causality inferences than cross-sectional data, but they also
mitigate the risk of potential common method biases
(Podsakoff et al., 2003).
The most important findings that emerged from our study
are as follows: (1) the direction and nature of influence
were two underlying dimensions of the WFI; (2) demands
were mostly related to conflict, while resources were
mostly related to enrichment; and (3) both conflict and
enrichment were related to role satisfaction, regardless of
the nature of the work or family origin.
Regarding the first finding, our confirmatory factor
analysis supported the validity of the fourfold taxonomy of
work and family interface in the Chinese context. Specifi-
cally, combining WFC and FWC into one factor, WFE and
FWE into one factor, or all four aspects into one factor,
resulted in a worse fit than examining the four as distinct
constructs. Further, the four aspects of the WFI acted as
separate mediators linking different antecedents with dif-
ferent outcomes in the structural model (Fig. 1). Our results
corroborate Western findings that WFC, FWC, WFE and
FWE are four discernable constructs of the WFI and each
deserves research attention (Frone, 2003; Grzywacz &
Mark, 2000). It seems, therefore, that the dynamism of
balancing work and family is essentially the same for
employees working and living in diverse cultural and eco-
nomic systems, lending support for the universality of the
current work and family theories.
Regarding the second finding, we found empirical
support for half of the hypothesized mediating paths
(Hypotheses 1 and 2). If we focused on the segment of
antecedents →WFI, six out of eight paths were significant
(Fig. 1), generally corroborating the ‘domain-specific’ pat-
terns of antecedents and WFI established by meta-analyses
studies in the West (Allen et al., 2000; Michel et al., 2011).
Our results thus lend support to the demand–resource dual
approach advocated by the COR and JD–R theories, in a
Chinese context. Our findings also corroborate the results
of some existing research. For example, Proost et al. (2010)
found that job demands and job control (a job resource)
were positively negatively related to work-family conflict,
respectively. In addition, we noticed that demands more
than resources seemed to be associated with work–family
Table 2 Model fit summary and measure models comparison (N= 409)
Model χ2df χ2/df p GFI CFI NFI RMSEA
10-factor model (M1) 1224.69 657 1.86 <0.000 0.87 0.95 0.89 0.05
Eight-factor model (M2) 2102.84 674 3.12 <0.000 0.75 0.86 0.81 0.07
Seven-factor model (M3) 3288.95 681 4.83 <0.000 0.63 0.75 0.71 0.10
Six-factor model (M4) 3675.90 687 5.35 <0.000 0.64 0.71 0.67 0.10
One-factor model (M5) 8160.89 702 11.63 <0.000 0.39 0.28 0.27 0.16
Notes:χ2, chi-square; df, degree of freedom; GFI, goodness of fit index; CFI, comparative fit index; NFI, normed fit index; RMSEA, root
mean square error of approximation. The 10-factor model (M1) assumes that workload, supervisory support, family conflict, family
support, WFC, FWC, WFE, FWE, job satisfaction and family satisfaction are 10 distinct factors. The eight-factor model (M2) is the same
as M1 except that WFC and FWC were combined, and WFE and FWE were also combined. The seven-factor model (M3) is the same as
M1 except that WFC, FWC, WFE and FWE were combined. The six-factor model (M4) is the same as M1 except that workload,
supervisory support, family conflict and family support were combined, and job satisfaction and family satisfaction were also combined.
The one-factor model (M5) assumes that all items for workload, supervisory support, family conflict, family support, WFC, FWC, WFE,
FWE, job satisfaction and family satisfaction loaded on the same factor.
An integrated model of the WFI 7
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
conflict (2 vs 1), whereas resources more than demands
seemed to be associated with work–family enrichment (2 vs
1). Since this is the first test of an integrated model of WFI,
we will need to conduct more studies to conclude whether
demands are indeed more salient for conflict and resources
are more salient for enrichment.
In contrast to the ‘domain-specific’ relations regarding
the antecedents, we hypothesized additional cross-domain
effects of WFI on role satisfaction (Hypothesis 3). We
found support for five out of eight hypothesized paths
linking up WFI and outcomes (Fig. 1). That is, work-
originated WFI is associated with both work and family
outcomes while family-originated WFI is also associated
with both work and family outcomes. Our results provide
some support for the ‘cross-domain’ patterns of WFC and
outcomes established by meta-analysis in the West (Ford
et al., 2007). Our findings also corroborate the results of
some recent research. For example, Boyar and Mosley
(2007) found in the USA that both work interfering with
family and family interfering with work negatively pre-
dicted work and family satisfaction. Similar cross-domain
effects have been reported with non-Western samples too.
For example, Chang and Lu (2009) reported that for
Chinese employees in Taiwan, WFC was negatively related
to family satisfaction, while FWC was negatively related to
job satisfaction. Li et al. (2013) also reported cross-domain
effects of FWC on work outcomes for Mainland Chinese
workers. These existing findings and the results of the
present study seem to suggest that role satisfaction is jointly
enhanced by work–family enrichment and depressed by
work–family conflict, regardless of the initiating domain.
Finally our findings advance research of the JD–R model
by adding family demands and family support in parallel to
work demands and work resources. Providing evidence
linking the JD–R model, the work–family conflict model,
and the work–family enrichment model, we have shown
that the work–family theories developed in a Western
context are largely generalizable to Chinese workers.
However, the fact that some of the hypothesized
mediational paths (e.g. H1b, H1c) and cross-domain paths
(e.g. H3a) were not supported in the current dataset, should
remind us that the quest for understanding the complex
processes and mechanisms underlying work–family expe-
riences is far from over. A comprehensive model depicting
exactly what antecedents go through which mediator to
impact on what outcomes can only be mapped out with
more systematic research. Also, it may be fruitful to work
with a partial mediation model that encompasses both
direct and indirect effects as shown in this study.
Limitations
Some limitations of this study need to be acknowledged.
First, we used self-reports to collect data, which may
increase the possibility of contamination of the reported
relationships through common method variance (CMV).
However, the results of confirmatory factor analyses show
that all of our study constructs could be empirically distin-
guished, and the one-factor CMV model fitted the data
poorly (Table 2). In addition, our three-wave design sepa-
rating assessments of antecedents, WFI and outcomes in
time lowers the likelihood of finding correlations due to
consistency in responses. Nonetheless, it would be benefi-
cial to include additional sources of data such as
co-workers, supervisors and family members.
Another limitation of the present study is that we only
surveyed Chinese workers in Taiwan; thus caution needs to
be exercised in generalizing our conclusions to other
Chinese societies, such as Mainland China, which has its
own political, economic and social characteristics. Never-
theless, a recent comparative study found that the experi-
ences of WFC and FWC were not different between
Taiwanese and Mainland Chinese employees (Kao, Lu &
Lu, 2008).
Finally, the correlation pattern (Table 1) shows high
correlations between WFC and FWC (0.65) and between
WFE and FWE (0.66), implying that they may be meas-
uring similar constructs. In fact, the high correlation
between WFC and FWC is a consistent finding in the
work–family research in the West (cf. Byron, 2005;
Michel et al., 2011) and in Chinese societies (e.g. Lu
et al., 2008). A similar pattern has been observed for the
relationship between WFE and FWE (e.g. Carlson,
Grzywacz & Zivnuska, 2009; Lu, 2011). Nonetheless,
through comparison of nested models, researchers have
argued that the four-factor model (WFC, FWC, WFE,
FWE) fits the data better than the one-factor or two-factor
models (distinguishing only conflict and enrichment, or
distinguishing only work-to-family and family-to-work)
(Carlson et al., 2009; Lu, 2013). It thus seems that the
four aspects of WFI can be discerned both theoretically
and empirically, though more rigorous measurements are
still desired.
Practical implications
As work and family demands are related to work–family
conflict, organizations should attempt to reduce employees’
workload and provide counselling to help them resolve
family discord. Such actions will alleviate work and family
interferences in both directions (e.g. workload →WFC;
family conflict →FWC). Furthermore, our findings should
remind managers of the importance of employees’ family
lives because family life is an important source of both
demand and support, which through the work–family inter-
face, contributes to employees’ work satisfaction (e.g.
family conflict →FWC →job satisfaction; family support
→FWE →job satisfaction).
8Luo Lu et al.
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
For employees striving to balance work and family roles,
getting support from direct supervisors may be crucial, and
gaining support from family members may be equally
important. Family support can help transfer gains in the
family realm to work, which in turn fosters more positive
experiences of both work and family life (e.g. family
support →FWE →job satisfaction/family satisfaction).
Family resources have largely been overlooked in the extant
work and family research and practice, which is an oppor-
tunity for future work. The challenge now for organizations
and individuals is to break the destructive flow of demands
→conflict →dissatisfaction, and to nurture the construc-
tive flow of resources →enrichment →satisfaction. This
way, work and family can be allies.
Acknowledgement
This research was supported by a grant from the National
Science Council, Taiwan, NSC100-2410-H-002–081-MY3.
References
Allen, T. D., Herst, D. E., Bruck, C. S. &
Sutton, M. (2000). Consequences associated
with work-to-family conflict: A review and
agenda for future research. Journal of Occu-
pational Health Psychology, 5, 278–308.
Anderson, J. C. & Gerbing, D. W. (1988).
Structural equation modeling in practice: A
review and recommended two-step
approach. Psychological Bulletin, 103,411–
423.
Aryee, S., Fields, D. & Luk, V. (1999). A cross-
cultural test of a model of the work–family
interface. Journal of Management, 25, 491–
511.
Aryee, S., Srinivas, E. S. & Tan, H. H. (2005).
Rhythms of life: Antecedents and outcomes
of work–family balance in employed
parents. Journal of Applied Psychology, 90,
132–146.
Boyar, S. L. & Mosley, D. C. (2007). The
relationship between core self-evaluations
and work and family satisfaction: The medi-
ating role of work–family conflict and facili-
tation. Journal of Vocational Behavior, 71,
265–281.
Byron, K. (2005). A meta-analytic review of
work–family conflict and its antecedents.
Journal of Vocational Behavior, 67, 169–
198.
Cammann, C., Fichman, M., Jenkins, D. &
Klesh, J. (1979). The Michigan
Organizational Assessment Questionnaire.
Ann Arbor, MI: University of Michigan.
Carlson, D. S., Grzywacz, J. G. & Zivnuska, S.
(2009). Is work–family balance more than
conflict and enrichment? Human Relations,
62, 1459–1486.
Carlson, D. S., Kacmar, K. M., Wayne, J. H. &
Grzywacz, J. G. (2006). Measuring the posi-
tive side of the work–family interface:
Development and validation of a work–
family enrichment scale. Journal of Voca-
tional Behavior, 68, 131–164.
Carlson, D. S., Kacmar, K. M. & Williams, L.
J. (2000). Construction and initial validation
of a multidimensional measure of work–
family conflict. Journal of Vocational
Behavior, 56, 249–276.
Chang, T. T. & Lu, L. (2009). The effects of
resource loss and resource gain on work–
family conflict: The perspective of resource
conservation. NTU Management Review, 20,
69–98.
Chang, Y. Y. & Lu, L. (2011). Sex differences
in the relationships between work and
family demands, resources, and work–
family conflict. International Journal of
Commerce and Strategy, 3, 25–38.
Clark, S. C. (2001). Work cultures and work/
family balance. Journal of Vocational
Behavior, 58, 348–365.
Cohen, S. & Syme, S. L., eds. (1985). Social
Support and Health. Orlando, FL: Academic
Press.
Cowlishaw, S., Birch, A., McLennan, J. &
Hayes, P. (2014). Antecedents and outcomes
of volunteer work–family conflict and facili-
tation in Australia. Applied Psychology: An
International Review, 63, 168–189.
Edwards, J. R. & Rothbard, N. P. (1999). Work
and family stress and well-being: An exami-
nation of person–environment fit in the work
and family domains. Organizational
Behavior and Human Decision Processes,
77, 85–129.
Ford, M. T., Heinen, B.A. & Langkamer, K. L.
(2007). Work and family satisfaction and
conflict: A meta-analysis of cross-domain
relations. Journal of Applied Psychology,
92, 57–80.
Frone, M. R. (2003). Work-family balance. In
J. C. Quick & L. E. Tetrick (Eds.), Hand-
book of occupational health psychology
(pp. 143–162). Washington, DC: American
Psychological Association.
Frone, M. R., Yardley, J. K. & Markel, K. S.
(1997). Developing and testing an integra-
tive model of the work–family interface.
Journal of Vocational Behavior, 50, 145–
167.
Grandey, A. A. & Cropanzano, R. (1999). The
conservation of resources model applied to
work–family conflict and strain. Journal of
Vocational Behavior, 54, 350–370.
Greenhaus, J. H. & Powell, G. N. (2006).
When work and family are allies: Atheory of
work–family enrichment. The Academy of
Management Review, 31, 72–92.
Grzywacz, J. G. & Mark, N. F. (2000).
Reconceptualizing the work–family inter-
face: An ecological perspective on the cor-
relates of positive and negative spillover
between work and family. Journal of Occu-
pational Health Psychology, 5, 111–126.
Hobfoll, S. E. (1989). Conservation of
resources: A new attempt at conceptualizing
stress. The American Psychologist, 44, 513–
524.
Innstrand, S. T., Langballe, E. M., Espnes,
G. A., Falkum, E. & Gjerl, O. (2008). Posi-
tive and negative work–family interaction
and burnout: A longitudinal study of recip-
rocal relations. Work & Stress, 22, 1–15.
Jansen, N. W. H., Kant, I., Kristensen, T. S. &
Nijhuis, F. J. N. (2003). Antecedents and
consequences of work–family conflict: A
prospective cohort study. Journal of Occu-
pational and Environmental Medicine, 45,
479–491.
Kao, S. F. & Lu, L. (2006). The relationship of
conjugal congruence in psychological
traditionality/modernity to marital adjust-
ment. Indigenous Psychological Research in
Chinese Societies, 25, 47–100.
Kao, S. F., Lu, L. & Lu, C. Q. (2008). Demands
and resources in work/family domains,
work/family conflict and their conse-
quences: A regional comparison between
Taiwanese and Mainland Chinese. Chinese
Business Review-USA, 8, 39–47.
Kossek, E. E., Pichler, S., Bodner, T. &
Hammer, L. B. (2011). Workplace social
support and work–family conflict: A
An integrated model of the WFI 9
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association
meta-analysis clarifying the influence of
general and work–family-specific supervisor
and organizational support. Personnel
Psychology, 64, 289–313.
Li, C., Lu, J. & Zhang, Y. (2013). Cross-
domain effects of work–family conflict on
organizational commitment and perfor-
mance. Social Behavior and Personality, 41,
1641–1654.
Lu, J. F., Siu, O. L., Spector, P. E. & Shi, K.
(2009a). Antecedents and outcomes of a
fourfold taxonomy of work–family balance
in Chinese employed parents. Journal of
Occupational Health Psychology, 14, 182–
192.
Lu, L. (2006). The transition to parenthood:
Stress, resources and gender differences in a
Chinese society. Journal of Community Psy-
chology, 34, 471–488.
Lu, L. (2011). A Chinese longitudinal study on
work/family enrichment. Career Develop-
ment International, 16, 385–400.
Lu, L. (2013). The dynamism of balancing
work and family in a developing society:
Evidence from Taiwan. Research in Applied
Psychology, 59, 49–79.
Lu, L., Chang, T. T. & Chang, Y. Y. (2012). The
meanings of work and family and their roles
in coping with work and family conflict:
Practicing the Chinese bicultural self. Indig-
enous Psychological Research in Chinese
Societies, 37, 141–189.
Lu, L., Hwang, M. T. & Kao, S. F. (2005). The
bi-directional conflict of work and family:
Antecedents, consequences and moderators.
Research in Applied Psychology, 27, 133–
166.
Lu, L. & Kao, S. F. (2013). The reciprocal
relations of pressure, work/family interfer-
ence, and role satisfaction: Evidence from a
longitudinal study in Taiwan. Human
Resource Management, 52, 353–373.
Lu, L., Kao, S. F., Chang, T. T., Wu, H. P.
& Cooper, C. L. (2008). Work/family
demands, work flexibility, work/family con-
flict, and their consequences at work: A
national probability sample in Taiwan. Inter-
national Journal of Stress Management, 15,
1–21.
Lu, L., Kao, S. F., Cooper, C., Allen, T. D.,
Lapierre, L. M., O’Driscoll, M., et al.
(2009b). Work resources, work-to-family
conflict, and its consequences: ATaiwanese–
British cross-cultural comparison. Interna-
tional Journal of Stress Management, 16,
25–44.
McNall, L. A., Nicklin, J. M. & Masuda, A. D.
(2010). A meta-analytic review of the con-
sequences associated with work–family
enrichment. Journal of Business and Psy-
chology, 25, 381–396.
Michel, J. S., Kotrba, L. M., Mitchelson, J. K.,
Clark, M. A. & Baltes, B. B. (2011). Ante-
cedents of work–family conflict: A meta-
analytic review. Journal of Organizational
Behavior, 32, 689–725.
Netemeyer, R. G., Boles, J. S. & McMurrian,
R. (1996). Development and validation of
work–family conflict and family–work con-
flict scales. Journal of Applied Psychology,
81, 400–410.
Noor, N. M. (2003). Work- and family-related
variables, work–family conflict and
women’s well-being: Some observations.
Community, Work & Family, 6, 297–319.
O’Driscoll, M. P., Brough, P. & Kalliath, T. J.
(2004). Work/family conflict, psychological
well-being, satisfaction and social support:
A longitudinal study in New Zealand. Equal
Opportunities International, 23, 36–56.
Olson-Buchanan, J. B. & Boswell, W. R.
(2006). Blurring boundaries: Correlates
of integration and segmentation between
work and nonwork. Journal of Vocational
Behavior, 68, 432–445.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y.
& Podsakoff, N. P. (2003). Common method
biases in behavioral research: A critical
review of the literature and recommended
remedies. Journal of Applied Psychology,
88, 879–903.
Proost, K., De Witte, H., De Witte, K. &
Schreurs, B. (2010). Work–family conflict
and facilitation: The combined influence of
the job demand–control model and achieve-
ment striving. European Journal of Work
and Organizational Psychology, 19, 615–
628.
Siu, O. L., Lu, J. F., Brough, P., Lu, C., Bakker,
A. B., Kalliath, T., et al. (2010). Role
resources and work–family enrichment:
The role of work engagement. Journal of
Vocational Behavior, 77, 470–480.
Spector, P. E. & Jex, S. M. (1998). Develop-
ment of four self-report measures of job
stressors and strain: Interpersonal conflict at
work scale, organizational constraints scale,
quantitative workload inventory, and physi-
cal symptoms inventory. Journal of Occupa-
tional Health Psychology, 3, 356–367.
ten Brummelhuis, L. L. & Bakker, A. B.
(2012). A resource perspective on the work–
home interface: The work–home resources
model. American Psychologist, 67, 545–
556.
Wayne, J. H., Grzywacz, J. G., Carlson, D. S.
& Kacmar, K. M. (2007). Work–family
facilitation: A theoretical explanation and
model of primary antecedents and conse-
quences. Human Resource Management
Review, 17, 63–76.
Witt, L. A. & Carlson, D. S. (2006). The work–
family interface and job performance: Mod-
erating effects of conscientiousness and
perceived organizational support. Journal of
Occupational Health Psychology, 11, 343–
357.
Yang, N., Chen, C. C., Choi, J. & Zou, Y.
(2000). Sources of work–family conflict:
A Sino–U.S. comparison of the effects of
work and family demands. Academy of
Management Journal, 43, 113–123.
10 Luo Lu et al.
© 2014 Wiley Publishing Asia Pty Ltd with the Asian Association of Social Psychology and the Japanese Group Dynamics Association