-
and family life.One of the most inuential papers in the area of
work and family introduced a work-family conict model wherein
work-
family conict mediates the effect of work (family) role conict
and involvement on family (work) satisfaction (Frone, Rus-sell,
& Cooper 1992). While this model has received empirical support
in a few primary studies (e.g., Aryee, Fields, &
0001-8791/$ - see front matter Published by Elsevier Inc.
* Corresponding author. Fax: +1 305 348 3879.E-mail address:
[email protected] (J.S. Michel).
Journal of Vocational Behavior 74 (2009) 199218
Contents lists available at ScienceDirect
Journal of Vocational
Behaviordoi:10.1016/j.jvb.2008.12.005Work-family integration
Published by Elsevier Inc.
1. Introduction
Of the numerous life domains individuals participate in, few, if
any, are as comprehensive and prevalent as the work andfamily
domains. For example, recent US Census gures indicate that 75.9
percent of US adults aged 2064 are in the laborforce (81.9 percent
of men, 70.0 percent of women; Clark &Weismantle, 2003), and
68.1 percent of US households are familyhouseholds (Simmons &
ONeill, 2001). Further, work and family roles have recently been
confounded by the increase ofdual-earner households, single-parent
households, and other nontraditional gender roles; which, as a
whole, have greatlyincreased work and family demands for many
individuals (Bond, Galinsky, & Swanberg, 1998; Duxbury &
Higgins, 1991;Greenhaus, Callanan, & Godshalk, 2000). As a
result, scholars have produced a considerable body of theoretical
(e.g., Edwards& Rothbard, 2000; Greenhaus & Powell, 2006),
empirical (e.g., Ilies, Schwind, Wagner, Johnson, DeRue, &
Ilgen, 2007), andreview literature (e.g., Eby, Casper, Lockwood,
Bordeaux, & Brinely, 2005; Kossek & Ozeki, 1998) on the
intersection of worka r t i c l e i n f o
Article history:Received 14 November 2008Available online 25
December 2008
Keywords:Linking mechanismsMeta-analysisSegmentationWork and
familyWork-family conictWork-family linkagesa b s t r a c t
This paper is a comprehensive meta-analysis of over 20 years of
work-family conictresearch. A series of path analyses were
conducted to compare and contrast existingwork-family conict
models, as well as a new model we developed which integrates
andsynthesizes current work-family theory and research. This new
model accounted for 40%of the variance in job satisfaction, 38% of
the variance in family satisfaction, and 35% ofthe variance in life
satisfaction. In a critical examination of work-family linkages, a
seriesof analyses excluding work-family conict constructs and
pathways resulted in a well-t-ting and more parsimonious model that
still accounted for 39% of the variance in job sat-isfaction, 37%
of the variance in family satisfaction, and 33% of the variance in
lifesatisfaction. Results indicate that direct effects drive
work-family conict models whileindirect effects provide little
incremental explanation in regards to satisfaction outcomes.A
comparative test of work-family conict models and
criticalexamination of work-family linkages
Jesse S. Michel a,*, Jacqueline K. Mitchelson b, Lindsey M.
Kotrba c,James M. LeBreton d, Boris B. Baltes e
a Florida International University, Department of Psychology,
11200 S.W. 8th Street, Miami, FL 33199, USAbAuburn University,
Department of Psychology, Auburn, AL 36849, USAcDenison Consulting,
Ann Arbor, MI 48104, USAd Purdue University, Department of
Psychological Sciences, West Lafayette, ID 47907, USAeWayne State
University, Department of Psychology, Detroit, MI 48202, USA
journal homepage: www.elsevier .com/ locate / jvb
-
thoroughly explore the intersection of work and family life. We
include several related models of work-family conict which,
200 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 199218similar to Frone et al. (1992), hypothesize that
work-family conict plays an intermediary role linking various
antecedents(e.g., work social support) to various consequences
(e.g., life satisfaction). The rst alternative model was proposed
by Carl-son and Kacmar (2000) which extends Frone and colleagues
work through the inclusion of additional work and family
stress-ors, specically role ambiguity and time demands. The second
model, proposed by Carlson and Perrew (1999), views workand family
stressors as a partial mediator between work and family social
support and involvement and work-family con-ict. Finally, we offer
a new model which integrates and synthesizes existing work-family
conict ndings. As such, thisstudy will test and evaluate a series
of work-family conict models thus answering the call of Eby et al.
(2005) for additionaltesting of theoretical work-family modelsboth
current models and the development of new models.
A nal goal of this paper is to critically examine and evaluate
the state of the literature by dissecting the most saturated
orintegrative model presented in this review in regards to the
explanatory power of specic work-family linkages. Extendingthe work
of Michel and Hargis (2008), which found that direct effect or
segmentation based linkages (e.g., work antecedentsto work
outcomes) accounted for far more variance in outcome variables than
indirect effect or conict based linkages (e.g.,work antecedents to
family outcomes via WIF), we seek to more completely examine the
linkage mechanisms within ourmodel by examining nested work-family
linkages within a complete integrative model (i.e., a single
work-family modelof indirect and direct effects) versus separate
quasi competing models (i.e., separate models of indirect effects
and direct ef-fects). Accordingly, in this paper we propose to (1)
review and update primary structural models within the
work-familyconict literature, (2) provide the rst meta-analytic
test of each of these models, and (3) critically examine and
evaluatethe explanatory power of model linkages (cf. Edwards &
Rothbard, 2000).
In the following sections we provide a brief background of the
work-family conict construct, review three primary work-family
conict models, propose a new model which integrates the core
features from previous models and empirical liter-ature, and
describe the results of a meta-analytic and path analytic
comparative test of extant models, our new integrativetheoretical
model, and critical examination of work-family linkages.
2. The work-family conict construct
Research on work and family has sought to explain work-family
conict from multiple theoretical approaches such asboundary theory,
compensation theory, ecological systems theory, social identity
theory, and spillover theory, to name afew. However, researchers
generally state that role theory (Kahn, Wolfe, Quinn, Snoek, &
Rosenthal, 1964; Katz & Kahn,1978) has provided the broad
theoretical umbrella for much of the work-family conict literature.
Role theory implies thatwork and family roles result from the
expectations of others, and what is believed to be appropriate
behavior for a particularposition (e.g., subordinate, coworker,
spouse, father; Kahn et al., 1964). Role theory indicates that both
work and family do-mains entail multiple roles where numerous
demands are placed on the individual, often resulting in conict
(e.g., interroleconict; Kahn et al., 1964). Rooted in role theory,
and derived from a scarcity hypothesis (xed amount of resources,
such astime and energy), conict theory posits that the work and
family domains can be incompatible resulting from differentnorms
and requirements (Burke, 1986; Evans & Bartolome, 1984; Zedeck
& Mosier, 1990); thus, increased role performancein one domain
(such as work) results in decreased role performance in the other
domain (such as family). Consequently,work-family conict is
popularly dened as a form of interrole conict in which the role
pressures from the work and familydomains are mutually incompatible
in some respect (Greenhaus & Beutell, 1985, p. 77).
Recent research has conceptualized work-family conict as a
multidimensional construct with aspects of WIF (alsotermed
work-to-family) and FIW (also termed family-to-work; e.g., Gutek,
Searle, & Klepa, 1991; Kelloway, Gottlieb, &Barham, 1999;
ODriscoll, Ilgen, & Hildreth, 1992; Williams & Alliger,
1994). Research has also shown that WIF and FIW haveLuk, 1999), it
has not been examined in a comprehensive way at a meta-analytic
level. For example, Ford, Heinen, and Langk-amer (2007) conducted a
quantitative review of work-family conict, but they only examined
components of the Frone et al.(1992) model. Specically, the effects
of work and family were examined as disparate models by separating
the effect ofwork on family satisfaction and the effect of family
on job satisfaction, thus providing (1) the extent to which factors
inthe work domain were related to satisfaction with family
lifemediated by work interference with family conict (WIF),and (2)
the extent to which factors in the family domain were related to
satisfaction with work lifemediated by familyinterference with work
conict (FIW).
One purpose of the current paper is to offer a comprehensive
examination of the Frone et al. (1992) model in its entirety.In
doing so, we may better understand the complexity between work and
family beyond the review of Ford and colleagues inseveral ways.
First, we examine work-family relationships previously unexamined.
These include relationships between (a)work and family antecedents,
(b) WIF and FIW, and (c) satisfaction outcomes (cf. Frone et al.,
1992). Second, we examine thedirect effects of (a) work antecedents
to job satisfaction, and (b) family antecedents to family
satisfaction (cf. Frone et al.,1992). Third, we include the nal
outcome of life satisfaction (cf. Frone et al., 1992). And forth,
we provide a concurrentexamination of the Frone et al. model,
versus multiple examinations of portions of the model, thus
controlling for variableinterrelationships. As such, this
quantitative review will extend the work of primary studies by
examining the stability in ourresults via meta-analytic path
analysis (Viswesvaran & Ones, 1995), along with extending the
work of Ford and colleagues inseveral important ways.
An additional goal of this paper is to provide a series of path
analyses on alternative work-family conict models to more
-
like exerts a direct effect or exerts an indirect effect, and we
use these terms in our integrative model, we use the original
J.S. Michel et al. / Journal of Vocational Behavior 74 (2009)
199218 201authors terminology in reference to their models.Before
reviewing the work-family conict models, a brief review of the
variables contained in these models is appropriate.
Pertaining to the selection of model variables, we use the
previous work of full-range work-family conict models to guideour
study (Carlson & Kacmar, 2000; Carlson & Perrew, 1999;
Frone et al., 1992). Due to space considerations, we do
notelaborate on the theoretical underpinnings of each variable, as
these specications were articulated in the articles that
orig-inally introduced the models being tested in this study. For a
review, readers are directed to Carlson and Kacmar (2000), Carl-son
and Perrew (1999), and Frone et al. (1992).
3.1. Conceptual denitions of model variables
In addition to the mediating WIF and FIW variables, each of the
models incorporate some or all of the following anteced-ent
variables: work social support (family social support) refers to
instrumental aid, emotional concern, informational, andappraisal
functions of others in the work (family) domain that serve to
heighten ones feelings of self-importance (Carlson &Perrew,
1999; House, 1981; Matsui, Ohsawa, & Onglatco, 1995); work
involvement (family involvement) refers to the levelof
psychological and cognitive preoccupation with, engagement in, and
immersion in ones work (family) role (Diefendorff,Brown, Kamin,
& Lord, 2002; Frone, 2003; Greenhaus & Parasuraman, 1999;
Kanungo, 1982; Yogev & Brett, 1985); work roleconict (family
role conict) refers to the extent to which an individual
experiences incompatible role pressures within thework (family)
domain (Bacharach, Bamberger, & Conley, 1990; Beehr, 1995; Kahn
et al., 1964; Katz & Kahn, 1978; Kopelman,Greenhaus, &
Connolly, 1983); work time demands (family time demands) refers to
time devoted to the work (family) role(e.g., Carlson & Frone,
2003; Frone, Yardley, & Markel, 1997; Judge, Boudreau, &
Bretz, 1994; Kirchmeyer, 1992; Major, Klein,& Ehrhart, 2002),
often confused with work role overload (family role overload; e.g.,
Carlson & Kacmar, 2000; Carlson & Per-rew, 1999), which is
the perception of having too many work role (family role) tasks and
not enough time to do them (Bach-arach et al., 1990; Caplan, Cobb,
& French, 1975; French & Caplan, 1973; Kahn, 1980); and
work role ambiguity (family roleambiguity) refers to the lack of
necessary information (specicity and predictability) about duties,
objectives, and responsi-bilities needed for a particular work role
(family role) or the lack of work role (family role) clarity (Beehr
& Glazer, 2005;Cooper, Cooper, & Eaker, 1988; Elloy &
Smith, 2003; Gupta & Jenkins, 1985; Kahn et al., 1964; Peterson
et al., 1995; Schuler,1980; Usita, Hall, & Davis, 2004).
In addition to these antecedents, each of the identied models
incorporate some or all of the following outcome variables:job
satisfaction (family satisfaction) refers to the degree to which an
individual is satised (positive feelings, emotional expe-rience)
with the work (family) aspects of their life (Cranny, Smith, &
Stone, 1992; Hopkins, 1983; Locke, 1976; Smith, Ken-dall, &
Hulin, 1969); and life satisfaction refers to the degree to which
an individual is satised (positive feelings, emotionalexperience)
with their general quality of life (Diener, Emmons, Larsen, &
Grifn, 1985; Rice, McFarlin, Hunt, & Near, 1985).
3.2. Model 1: A parsimonious stressor and involvement model
(Frone et al., 1992)
The Frone et al. (1992) model was the rst highly cited
structural model theorizing separate WIF and FIW constructs as
amediating component between work and family domain antecedents,
and work domain, family domain, and life outcomes.More specically,
this model conceptualizes WIF and FIW as mediating components
between job stressors (measured aswork pressure, lack of autonomy,
and work role ambiguity), job involvement, family stressors
(measured as parental work-load and child misbehavior), and family
involvement, and the outcomes of job distress, family distress, and
depression (mea-sured as reverse scored job satisfaction, family
satisfaction, and life satisfaction). In a study to test the
cross-culturalgeneralizability of this model, Aryee et al. (1999)
utilized a slightly adapted version which replaced the antecedents
ofjob and family stressors (work pressure, lack of autonomy, and
work-role ambiguity; parental workload and child misbehav-ior) with
job and family conicts (work role conict and family role
conict).
While Frone et al. (1997) provide and test a version of this
model, the Frone et al. (1997) model was not included in
thefollowing review because it does not meet our criteria for a
full-range model. For example, in this 1997 model, outcomesunique
relationships with work and family antecedents and outcomes (e.g.,
Anderson, Coffey, & Byerly, 2002; Byron, 2005;Carlson &
Kacmar, 2000; Frone et al., 1992; Grandey & Cropanzano, 1999;
Kelloway et al., 1999; Kossek & Ozeki, 1998;ODriscoll et al.,
1992; Parasuraman, Purohit, Godshalk, & Beutell, 1996), and
provide incremental variance over one another(see Mesmer-Magnus
& Viswesvaran, 2005).
3. Work-family conict as a mediator
As work-family conict has been a dominant construct within the
work-family literature, a number of structural modelshave been
advanced. However, most theoretical models contain three core
components: (1) a set of work and family domainantecedents; (2) a
combination of work domain, family domain, and life outcomes; and
(3) a mediating work-family conictconstruct. It is this conguration
that we refer to as full-range models. We should note that we are
using the terms medi-ation and partial mediation in reference to
the original authors conceptions. However, these terms imply causal
relation-ships and the conditions for causal inference are quite
strong (James, Muliak, & Brett, 1982). While we would prefer
phrases
-
have a recursive effect on work-family conict (e.g., WIF to
Family Dissatisfaction to FIW). Further, we chose to examine
theslightly modied Aryee et al. (1999) version of the Frone et al.
(1992) model as it provides greater variable overlap with re-cent
conict models at the operational level, thus allowing for greater
condence in our inferences of model adequacy acrossthe structural
models being compared in this multi model test.
3.3. Model 2: An expanded stressor model (Carlson & Kacmar,
2000)
TheCarlson andKacmar (2000)model conceptualizesWIF and
FIWasmediating components betweenwork role ambiguity,work role
conict, work time demands (measured as work role overload), job
involvement, family involvement, family timedemands (measured as
family role overload), family role conict, and family role
ambiguity, and the outcomes of job satisfac-tion, family
satisfaction, and life satisfaction. Unique to thismodel is the
inclusionof domain role ambiguity and timedemands.
3.4. Model 3: A partial mediating stressor model (Carlson &
Perrew, 1999)
In the original Carlson and Perrew (1999) model, WIF and FIW
were combined to make a unidimensional work-familyconict construct.
In our adapted version, and consistent with previous research
(e.g., Carlson, Kacmar, & Williams, 2000),work-family conict is
conceptualized as two separate WIF (leading to family outcomes) and
FIW (leading to work out-comes) constructs. Thus, in our adapted
model, WIF and FIW are mediating components between work social
support, workinvolvement, work role conict, work time demands
(measured as work role overload), work role ambiguity, family
socialsupport, family involvement, family role conict, family time
demands (measured as family role overload), and family
roleambiguity, and the outcomes of job satisfaction and family
satisfaction. Further, this model posits the partial mediation
of
202 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 199218work role conict, work time demands, and work role
ambiguity between work social support and work involvement, andWIF
and job satisfaction on the work domain side; along with the
partial mediation of family role conict, family time de-mands, and
family role ambiguity between family social support and family
involvement, and FIW and family satisfaction onthe family domain
side. Unique to this model is the inclusion of domain social
support, the partial mediation of domain roleconict, time demands,
and role ambiguity, along with the exclusion of life
satisfaction.
3.5. Model 4: Development of an integrative work-family conict
model
Based on current work-family research, we provide a model in
Fig. 1 that integrates the literature. This integrative modelis
founded on what we refer to as three forms or types of quasi
linking mechanisms (cf. Edwards & Rothbard, 2000). We referto
these as quasi linking mechanisms and classify them by effect
(e.g., indirect effect, direct effect) as each revolves aroundthe
interplay of work and family, yet could be explained or partially
explained by a number of theoretical approaches, suchas work-family
conict or segmentation (see Ashforth, Kreiner, & Fugate, 2000;
Edwards & Rothbard, 2000).
3.5.1. Indirect effect linkagesWork-family conict posits that
work and family antecedents lead to WIF and FIW, which then leads
to family and work
consequences. These linkages are represented by pathways 12 for
WIF (work antecedents to WIF to family outcomes) and34 for FIW
(family antecedents to FIW to work outcomes). These indirect
effects are the core feature of work-family conict
Work Antecedents-Work Social Support-Work Involvement-Work Role
Conflict-Work Time Demands-Work Role Ambiguity
Family Antecedents-Family Social Support-Family
Involvement-Family Role Conflict-Family Time Demands-Family Role
Ambiguity
Work Outcomes
-Job Satisfaction
Family Outcomes
-Family Satisfaction
Life Outcomes
-Life Satisfaction
WIF
FIW
1
2
3
4
11
12
17
16
10
913 15
5
68
7
14
Fig. 1. Integrative work-family conict model (Model 4).
-
satisfaction is an outcome of WIF (r = .24, k = 8; Allen, Herst,
Bruck, & Sutton, 2000).
J.S. Michel et al. / Journal of Vocational Behavior 74 (2009)
199218 203A nal set of work-family conict linkages involve the
prediction of life satisfaction by WIF and FIW. These linkages
arerepresented by pathways 9 (fromWIF) and 10 (from FIW). These
pathways are based on the previous work of Frone et al. andhave
received modest support.
3.5.2. Direct effect linkagesIn addition to indirect effect
linkages (e.g., work antecedents to family outcomes via WIF), our
integrative model also in-
cludes direct effect linkages (e.g., work antecedents to work
outcomes), which are typically included in conict models
(e.g.,Carlson & Kacmar, 2000; Carlson & Perrew, 1999; Frone
et al., 1992). These are represented by pathways 11 (work
anteced-ents to work outcomes) and 12 (family antecedents to family
outcomes). As much of the literature on organizational behav-ior
focuses on direct effect linkages (e.g., Hackman & Oldham,
1976), they have receive extensive support (e.g., Michel
&Hargis, 2008).
3.5.3. Construct level cross-domain effectsThe nal set of
linkages in our integrative model represents the relationships
between work and family that fall outside of
indirect and direct effect linkages. These consist of
relationships between (1) work and family antecedentslinkage 13,
(2)WIF and FIWlinkage 14, and (3) satisfaction outcomeslinkages
1517. In essence, these linkages account for a similaritybetween
the work and family roles and are based on the previous work-family
models previously reviewed (e.g., Carlson &Kacmar, 2000; Frone
et al., 1992).
4. Study purpose
The purpose of this study is to provide a series of
meta-analytic path analyses on the previously reviewed
work-familyconict models to better understand the complex interplay
between work and family life. Specically, we propose to testand
evaluate the parsimonious stressor and involvement model presented
by Frone et al. (1992; Model 1), the expandedstressor model
presented by Carlson and Kacmar (2000; Model 2), and the partial
mediating stressor model presented byCarlson and Perrew (1999;
Model 3), in addition to a new integrative model developed for this
study (Model 4). ThoughModels 13 each found support in primary
studies, meta-analytic path analysis enables us to examine how well
these struc-tural models actually generalize across a comprehensive
meta-analytic data set of observed validities. Likewise, this
studywill provide the rst examination of our new integrative model.
Though each of these models shares the characteristicsof a
full-range model, they have subtle yet signicant differences. As
such, we hope to better understand these similaritiesand
differences with this research, as model adequacy will be evaluated
on model t, parameter estimates, and variance ex-plained in the
outcome variables.
In addition, we propose to critically examine and evaluate the
state of the literature by scrutinizing our integrative modelas it
is (1) based on current work-family theory and empirical ndings,
and (2) incorporates all of the linkages held by theprevious models
reviewed. This examination extends previous literature on
work-family linking mechanisms (see Michel &Hargis, 2008),
where direct effect models (e.g., work antecedents to work
outcomes) accounted for far more variance in out-come variables
than indirect effect models (e.g., work antecedents to family
outcomes via WIF), by examining linkages with-in a single
theoretical model, therefore allowing us to determine the
explanatory power of model linkages by determiningwhere the
variance explained in outcomes occurs. In doing so, we hope to
better determine the theoretical and empiricalforces underlying
models of work and family.
5. Methods
5.1. Literature search
The search for studies was conducted in two stages. In the rst
stage, computer-based literature searches were conductedon the
databases of ABI/INFORM and PsychINFO (including Dissertation
Abstracts). Keyword searches were conducted on thefollowing terms:
work family conict in conjunction with job satisfaction, family
satisfaction, life satisfaction, work social
sup-port/organizational support, family social support/social
support, work involvement/job involvement, family involvement,
workrole conict, family role conict, work time demands, family time
demands, work role ambiguity, and family role ambiguity. Allmodels
(e.g., Carlson & Kacmar, 2000; Carlson & Perrew, 1999;
Frone et al., 1992), and are the sole linkages tested by theFord et
al. (2007) meta-analytic path analysis. Empirical support for these
relationships has been modest. For example, in aseries of
meta-analytic multiple regressions, Ford et al. found that work
antecedents and WIF accounted for approximately7% of the variance
in family satisfaction, while family antecedents and FIW accounted
for approximately 7% of the variance injob satisfaction.
A second series of work-family conict linkages revolve around
empirical ndings that have yet to be included in modelsof
work-family conict. These linkages are represented by pathways 5
(work antecedents to FIW), 6 (family antecedents toWIF), 7 (WIF to
job satisfaction), and 8 (FIW to family satisfaction). Empirical
support for these relationships has been con-sistent. For example,
based on meta-analytic data, job stress is an antecedent of FIW (rc
= .29, k = 19; Byron, 2005) and job
-
electronic searches were limited to post-1985 material since
directional WIF and FIW were rst introduced by Greenhausand Beutell
(1985). Manual cross-referencing was also used to ensure relevant
studies missed in the electronic searches wereconsidered for
inclusion. This second stage consisted of a manual
cross-referencing of studies included in relevant qualitativeand
quantitative reviews (e.g., Allen et al., 2000; Byron, 2005; Eby et
al., 2005; Kossek & Ozeki, 1998), and related journalspecial
issues (e.g., Journal of Vocational Behavior: Special Issue on Work
and Family Balance, 1997; Journal of OccupationalHealth Psychology:
Special Issue on Relationship BetweenWork and Family Life, 1999;
International Journal of Stress Management:Special Issue on Work
and Personal Life Integration, 2004).
5.2. Inclusion criteria
Studies were included if they included a measure of WIF and/or
FIW and included two or more variables theorized in thestructural
models of interest. When relevant articles published within the
past seven years (the required duration to retaindata per APA
guidelines) were located, and data were not provided, the
corresponding author was contacted to obtain miss-ing data. When
identical datasets were reported in multiple studies, the most
recent study was included. When portions ofdata from the same
sample were reported in more than one publication or study, unique
bivariate relationships and bivariaterelationships with the larger
sample size were included. This process yielded a total of 211
studies with 263 samples and2060 effect sizes that were included in
the current meta-analysis.
5.3. Coding of studies
Of the studies that met the inclusion criteria, 60 articles were
randomly chosen for three subject matter experts (SME) tocode. The
average percent agreement between all three SME was 93.30%. After
coding procedures, the SME resolved all issuesand discrepancies
until absolute agreement was reached for the 15 constructs included
in the identied structural models.As outlined in Table 1, a nal
coding scheme was developed and the remaining samples were coded
based on these criteria.Whenever additional judgment calls were
required (i.e., any issue outside the information provided in Table
1), the studies inquestion were placed aside and discussed by the
three SME until absolute agreement was reached and all coding
issues wereresolved.
204 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 1992185.4. Computation of effect sizes
Since one of our research goals were to critically examine
work-family linkages and determine the utility of the work-family
conict construct, we chose to use observed versus psychometrically
corrected validity coefcients in our analyses.Accordingly, sample
size weighted correlation estimates were conducted via the
meta-analytic techniques of Hedges and
Table 1Summary of Included Measures.
Construct Measures included
WIF All forms of WIF (e.g., self developed; Frone et al., 1992;
Gutek et al., 1991; Kopelman et al., 1983)FIW All forms of FIW
(e.g., self developed; Burley, 1990; Frone et al., 1992; Gutek et
al., 1991)Work social support Coworker social support,
emotional/informational/instrumental support from work, group
cohesiveness, lack of personal support
at work/nonsupport (rev.), leader/managerial support, level of
work group support, mentor supportiveness, perceivedorganizational
support, psychosocial support, supervisor support (e.g., self
developed; Beehr, King, & King, 1990; Eisenberger,Huntington,
Hutchison, & Sowa, 1986; Shinn, Wong, Simko, &
Ortiz-Torres, 1989)
Work involvement Job involvement, work identity, work
involvement (e.g., self developed; Kanungo, 1982; Lodahl &
Kejnar, 1965; Quinn & Staines,1979)
Work role conict Work role conict (e.g., self developed; Caplan,
Cobb, French, Harrison, & Pinneau, 1980; Kahn et al., 1964;
Kopelman et al., 1983;Rizzo, House, & Lirtzman, 1970)
Work time demands Number of hours spent on paid work activities,
time commitment to work (e.g., self developed)Work role ambiguity
Work role ambiguity (e.g., self developed; Caplan et al., 1980;
Kahn et al., 1964; Rizzo et al., 1970)Family social
supportDomestic support, emotional/informational/instrumental
support from family, family cohesion, family and friends support,
lack ofspousal support (rev.), spousal support (e.g., self
developed; Caplan, Cobb, French, Harrison, & Pinneau, 1975;
King, Mattimore,King, & Adams, 1995; Procidano & Heller,
1983)
Family involvement Family identity, family involvement (e.g.,
self developed; adapted job involvement measures; Amatea, Cross,
Clark, & Bobby, 1986;Yogev & Brett, 1985)
Family role conict Family role conict (e.g., self developed;
adapted work-role conict measures; Kopelman et al., 1983; Nye &
MacDougall, 1959)Family time
demandsAge of youngest child, number of children living at home,
number of hours spend on family activities, parental (time)
demands,time commitment to family (e.g., self developed; Bedeian,
Burke, & Moffett, 1988)
Family roleambiguity
Family role ambiguity (e.g., adapted work role ambiguity
measures)
Job satisfaction Business dissatisfaction (rev.), job
satisfaction, quality of work life, work satisfaction (e.g., self
developed;Hackman & Oldham,1975; Quinn & Staines, 1979;
Smith et al., 1969)
Family satisfaction Family satisfaction, home satisfaction,
marital satisfaction/adjustment, parenting satisfaction, quality of
family life, relationshipsatisfaction/agreement (e.g., self
developed; Spanier, 1976; Staines & Pleck, 1984)
Life satisfaction Life satisfaction, quality of life (e.g., self
developed; Diener et al., 1985; Quinn & Shepard, 1974; Quinn
& Staines, 1979)
-
1983; Hu & Bentler, 1995; Hu & Bentler, 1999; Jreskog
& Srbom, 1993; Marsh, Hau, & Wen, 2004). Specically, the
Good-
J.S. Michel et al. / Journal of Vocational Behavior 74 (2009)
199218 205ness-of-t Index (GFI), Normed Fit Index (NFI),
Comparative Fit Index (CFI), and Root Mean Square Error of
Approximation(RMSEA) were used to assess model t or model mist. In
addition to these indices, the Normal Theory Weighted LeastSquares
Chi-Square (v2) statistic and the Critical N (CN) were evaluated.
Finally, to compare the series of non-nested modelsin the current
study, the Akaike Information Criterion (AIC) was also
evaluated.
6. Results
Our meta-analytic correlation matrix is presented in Table 2.
Ninety-ve percent condence intervals indicate that85.44% of the
meta-analytic correlations were signicantly signicant. However,
given the study purpose, we focus on modelt, parameter estimates,
and variance explained in outcomes versus bivariate
relationships.
6.1. Model t and parameter estimates of work-family conict
models
Model 1 (parsimonious stressor and involvement model) was
supported with overall t statistics suggesting very goodmodel t:
v2(13) = 179.57, CN = 323.94, GFI = .98, NFI = .96, CFI = .96,
RMSEA = .08, and AIC = 243.57 (see Table 3). In additionto model t,
it is appropriate to examine pathway magnitudes to better interpret
a model, thus we provide parameter esti-mates in Fig. 2. Work role
conict was a signicant predictor of WIF (pathway = .33) and job
satisfaction (pathway = .40),and family role conict was a signicant
predictor of FIW (pathway = .21) and family satisfaction (pathway =
.41). Workinvolvement and family involvement had small effects on
WIF (pathway = .08) and FIW (pathway = .00), but moderate ef-fects
on job satisfaction (pathway = .30) and family satisfaction
(pathway = .21). These results suggest that work and familymay be
best conceptualized as predictors of same domain satisfaction
versus work-family conict. WIF and FIW seem tohave a small to
moderate reciprocal relationship (pathways of .20 and .17), but
have relatively low prediction on satisfactionoutcomes (ranging
from .01 to .13). Finally, job satisfaction (pathway = .34) and
family satisfaction (pathway = .37) eachhad a moderately high
relationship with life satisfaction.
Model 2 (expanded stressor model) was supported with overall t
statistics suggesting very good model t:v2(28) = 383.08, CN =
266.03, GFI = .97, NFI = .95, CFI = .95, RMSEA = .08, and AIC =
509.08 (see Table 3). Parameter estimatesare displayed in Fig. 3.
As more variables and pathways were added to this model, work role
conict remained a signicantpredictor of WIF (pathway = .26) and job
satisfaction (pathway = .26), and family role conict remained a
signicant pre-dictor of FIW (pathway = .16) and family satisfaction
(pathway = .38). Similarly, work involvement and family
involvementcontinued to have small effects on WIF (pathway = .03)
and FIW (pathway = .02), but moderate effects on job
satisfaction(pathway = .27) and family satisfaction (pathway =
.20). WIF and FIW continued to have a small to moderate reciprocal
rela-tionship, yet magnitudes changed signicantly (.17.26 for FIW
to WIF, .20.09 for WIF to FIW). Likewise, relationships fromOlkin
(1985). Johnsons (1993) DSTAT was used to convert the study
statistics into effect sizes adjusted for sample size. Sincethe
product-moment correlation (r) possesses some undesirable
statistical properties (Alexander, Scozzaro, & Borodkin,1989;
Rosenthal, 1994), and the unbiased estimator of effect size (d)
possesses many desirable properties (Hedges, 1982;Hedges &
Olkin, 1985), the unbiased estimator of effect size (d) was used to
provide meta-analytic mean weighted effectsizes. Though each sample
correlation was converted to an unbiased estimator of effect size
(d) for meta-analytic analysis,nal estimates were converted back to
the product-moment correlations (r) for path analyses.
5.5. Outlier analysis
Outliers can signicantly distort meta-analytic results (Hedges
& Olkin, 1985), and this problem can be compounded
withmeta-analytic path analysis. Accordingly, two approaches to
identify and address potential outliers were employed. First,within
each cell of the meta-analytic correlation matrix, effect sizes
were converted to z-scores and extreme values(2.58) were deemed
outliers. This procedure yielded 26 of the 2060 effect sizes as
statistical outliers at the p = .01 level.Second, we employed a
visual inspection of the data (Hedges & Olkin, 1985). This
process yielded three effect sizes that weredeemed outliers. Based
on these two approaches, 29 of the 2060 effect sizes or 1.41% of
the data were deemed potential out-liers. All analyses were
conducted with and without the identied outliers. Results were near
identical for model t indicesand parameter estimates. Accordingly,
due to space considerations, only analyses with outliers excluded
are reported. Thefull correlation matrix with the outliers included
may be obtained from the corresponding author.
5.6. Path analyses and model evaluation
Path analysis of the meta-analytic correlation matrix was
conducted in LISREL 8.80 (Jreskog & Srbom, 1993). All
latentconstructs were treated as single item indicators. As with
previous meta-analytic path analyses (e.g., Brown &
Peterson,1993; Chen, Casper, & Cortina, 2001; Ford et al. 2007;
Hom, Caranikas-Walker, Prussia, & Griffeth, 1992; Michel &
Hargis,2008), all models were tested utilizing the maximum
likelihood estimation method and the harmonic mean. To provide
acomprehensive examination of model t, a variety of t indices were
chosen based on type of t index (e.g., absolute versusincremental)
and recommendations within the structural equation modeling
literature (e.g., Bentler & Bonett, 1980; Hoelter,
-
206 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 199218Table 2Meta-analytic correlation matrix of model
variables.
Variable WIF FIW WSS WI WRC WTD WRA FSS FI FRC FTD FRA JS FS
LS
WIF k n
FIW .39* k 121 n 47,305
WSS .23* .09* work-family conict and satisfaction outcomes
remained constant (ranging from .05 to .09), as did job
satisfaction (path-way = .37) and family satisfaction (pathway =
.39) to life satisfaction. Unique to this model was the inclusion
of ambiguityand time demands variables. Work role ambiguity had a
moderate relationship to job satisfaction (pathway = .30) butlow
relationship to WIF (pathway = .05). Inverse to this was the role
of work time demands, which was moderately relatedtoWIF (pathway =
.20) but not job satisfaction (pathway = .04). On the family side,
much different results were found as fam-ily role ambiguity was
moderately related to both FIW (pathway = .21) and family
satisfaction (pathway = .19), while fam-ily time demands had small
effects on FIW (pathway = .08) and family satisfaction (pathway =
.05).
Model 3 (partial mediating stressor model) was not supported
with overall t statistics suggesting inadequate model t:v2(49) =
920.03, CN = 160.26, GFI = .94, NFI = .87, CFI = .88, RMSEA = .09,
and AIC = 1,032.03 (see Table 3). Accordingly, exam-ination of
parameter estimates could provide misleading results and are not
included.
k 55 37 n 21,255 13,424
WI .08* .01 .15* k 40 28 22 n 10,944 8434 5621
WRC .36* .20* .36* .00 k 24 11 6 11 n 5933 3628 1413 2703
WTD .26* .02* .04* .24* .15* k 64 51 25 21 5 n 28,507 18,549
11,098 5953 1572
WRA .24* .19* .24* .07* .49* .11* k 19 13 9 10 18 8 n 5585 4731
3183 3257 4613 2915
FSS .13* .17* .17* .02* .05* .03* .06* k 38 29 39 21 5 20 6 n
14,171 7732 12,314 5611 1550 9871 1999
FI .01 .01 .05* .12* .04* .02 .00 .14* k 24 22 12 34 7 12 5 16 N
7708 6970 3573 9876 2220 3839 2325 3905
FRC .24* .26* .03 .07* .29* .01 .11* .35* .04* k 9 4 2 5 12 2 4
2 5 n 2817 2056 575 1502 2771 1248 1294 575 1502
FTD .03* .08* .01 .06* .05* .01* .01 .02 .05* .25* k 68 56 16 19
4 48 7 21 13 2 n 26,196 20,607 5891 5238 1411 17,034 2763 4386 3581
1248
FRA .17* .25* .06* .07* .16* .18 .33* .29* .13* .28* .21 k 3 3 2
2 4 1 4 2 2 4 1 n 1069 1069 575 808 1294 261 1294 575 808 1294
261
JS .25* .15* .43* .30* .41* .03* .45* .18* .02 .14* .02* .05* k
85 60 39 33 24 30 14 28 18 11 29 3 n 29,587 21,523 15,397 8592 5766
12,186 4089 10,648 5417 2524 9453 1069
FS .21* .23* .13* .07* .12* .03* .13* .48* .24* .45* .00 .35*
.19* k 45 30 13 15 14 13 5 16 17 11 16 3 44 n 14,504 10,808 3854
4465 3429 5050 1616 4102 4765 2524 5851 1069 13,741
LS .30* .20* .32* .02 .31* .04* .26* .19* .07* .29* .00 .24*
.44* .46* k 35 24 4 10 11 9 4 7 11 9 11 2 37 26 n 9162 6657 963
2320 2375 2696 1392 1472 2700 1809 3240 808 9476 6832
Note:WIF, work interference with family conict; FIW, family
interference with work conict; WSS, work social support; WI, work
involvement; WRC, workrole conict; WTD, work time demands; WRA,
work role ambiguity; FSS, family social support; FI, family
involvement; FRC, family role conict; FTD, familytime demands; FRA,
family role ambiguity; JS, job satisfaction; FS, family
satisfaction; LS, life satisfaction; k, number of samples; n,
number of participants.* p < .05
-
J.S. Michel et al. / Journal of Vocational Behavior 74 (2009)
199218 207Table 3Model t indices from primary path analyses.
Model df v2 CN GFI NFI CFI RMSEA AIC
Model 1Parsimonious stressor and involvement model 13 179.57
323.94 .98 .96 .96 .08 243.57
Model 2Expanded stressor model 28 383.08 266.03 .97 .95 .95 .08
509.08
Model 3Partial mediating stressor model 49 920.03 160.26 .94 .87
.88 .09 1032.03
Model 4aModel 4 (integrative work-family conict model) was
supported with overall t statistics suggesting excellent model
t:v2(20) = 272.42, CN = 287.81, GFI = .98, NFI = .97, CFI = .97,
RMSEA = .08, and AIC = 472.42 (see Table 3). Due to the complex-ity
of this model, parameter estimates are provided in Table 4 versus a
gure. As more variables and pathways were added tothis model, work
role conict remained a signicant predictor of WIF (pathway = .20)
and job satisfaction (pathway = .16),and family role conict
remained a signicant predictor of FIW (pathway = .13) and family
satisfaction (pathway = .28).Work involvement and family
involvement continue to have small effects on WIF (pathway = .05)
and FIW (pathway = .02),but moderate effects on job satisfaction
(pathway = .24) and family satisfaction (pathway = .16). Work role
ambiguity con-tinued to have a moderate relationship to job
satisfaction (pathway = .28) but low relationship to WIF (pathway =
.06),while work time demands was moderately related to WIF (pathway
= .20) but not job satisfaction (pathway = .06). On thefamily side,
family role ambiguity was moderately related to both FIW (pathway =
.19) and family satisfaction (path-
Work RoleConflict
LifeSatisfaction
WIF
Family RoleConflict
FamilySatisfaction
JobSatisfaction
-.40 (.02)
.34 (.02)
FIW.37 (.02) -.10 (.02)
-.06 (.02)
.20 (.05).17 (.05)
WorkInvolvement
FamilyInvolvement
.33 (.02)
.08 (.02)
.21 (.02)
-.00 (.02)
-.13 (.02)
-.01 (.02)
.11 (.02)
.21 (.02)
-.41 (.02)
.30 (.02)
Intercorrelations Among Exogenous Variables CRFIFIWCRW
-CRW-)20.(00.IW
-)20.(21.-)20.(40.IF-)20.(40.-)20.(70.)20.(92.CRF
Note: WRC = Work role conflict, WI = Work involveme nt, FI =
Family involvement, FRC = Family role conflict.
Fig. 2. Parsimonious stressor and involvement model with path
coefcients (Model 1). Note: values in parentheses are standard
errors for the pathcoefcient. Variance explained in outcome
variables: job satisfaction = 26%, family satisfaction = 25%, life
satisfaction = 34%. N = 2119.
Integrative work-family conict model 20 272.42 287.81 .98 .97
.97 .08 472.42
Model 4bDirect effects only model 20 312.83 250.39 .98 .96 .96
.08 454.83
Note: df, degrees of freedom; v2, normal theory weighted least
squares chi-square; CN, critical N; GFI, goodness-of-t index; NFI,
normed t index; CFI,comparative t index; RMSEA, root mean square
error of approximation; AIC, Akaikes information criterion. N =
2119.
-
208 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 199218
Work RoleAmbiguity
LifeSatisfaction
WIF
FamilyInvolvement
JobSatisfaction
FIW
Work RoleConflict
WorkInvolvement
Family TimeDemands
Work TimeDemands
.27 (.02) .04 (.02)
-.26 (.02)
-.30 (.02)
.05 (.02)
.08 (.02)
.02 (.02)
.26 (.02)
.20 (.02)
.03 (.02)
-.05 (.02) .37 (.02)
.26 (.04) .09 (.05) way = .14), while family time demands had
small effects on FIW (pathway = .09) and family satisfaction
(pathway = .04).The linage between WIF and FIW was high (.27), but
again the pathways to satisfaction were low (ranging from .01
to.13). Finally, relationships from job satisfaction (pathway =
.37) and family satisfaction (pathway = .39) to life
satisfactionremained constant. Unique to this model was the
inclusion of social support as a predictor of work-family conict
and sat-isfaction. Work social support had a small relationship to
WIF (pathway = .13) and a moderate relationship to job
satisfac-tion (pathway = .25); similarly, family social support had
a small relationship to FIW (pathway = .06) and a
moderaterelationship to family satisfaction (pathway = .29). Most
of the additional indirect effect pathways in this model were
signif-icant yet small. The greatest support for these linkages,
besides model t, were the results for work and family role
conict;specically, the added pathways from work role conict to FIW
(pathway = .10) and family role conict to WIF (path-way = .14).
Finally, there was moderate evidence for construct level
cross-domain effects as many of the work domainand family domain
constructs covaried (e.g., .33 for role ambiguity).
6.2. Variance explained in satisfaction outcomes by work-family
conict models
Each of the well tting models accounted for a large portion of
variance in outcome variables. Model 1 (parsimoniousstressor and
involvement model) explained 26% of the variance in job
satisfaction, 25% of the variance in family satisfaction,and 34% of
the variance in life satisfaction. As more predictor variables were
added to Model 2 (expanded stressor model),variance explained in
job satisfaction increased to 33%, variance explained in family
satisfaction increased to 30%, but var-iance explained in life
satisfaction decreased to 32% due to the elimination of WIF and FIW
direct effects. Model 4 (integra-
FamilySatisfaction
Family RoleConflict
Family RoleAmbiguity
.20 (.02) .05 (.02)
-.38 (.02)
-.19 (.02)
.21 (.02)
.16 (.02) -.09 (.02) .39 (.02)
Intercorrelations Among Exogenous Variables WRA WRC WTD WI FI
FTD FRC FRA
WRA - WRC .49 (.02) - WTD .11 (.02) .15 (.02) - WI -.07 (.02)
.00 (.02) .24 (.02) - FI .00 (.02) .04 (.02) -.02 (.02) -.12 (.02)
- FTD -.01 (.02) -.05 (.02) -.01 (.02) -.06 (.02) .05 (.02) - FRC
.11 (.02) .29 (.02) .01 (.02) .07 (.02) -.04 (.02) .25 (.02) - FRA
.33 (.02) .16 (.02) .18 (.02) .07 (.02) -.13 (.02) -.21 (.02) .28
(.02) - Note: WRA = Work role ambiguity, WRC = Work role conflict,
WTD = Work time demands, WI = Work involvement, FI = Family
involvement, FTD = Family time demands, FRC = Family role conflict,
FRA = Family role ambiguity.
Fig. 3. Expanded stressor model with path coefcients (Model 2).
Note: values in parentheses are standard errors for the path
coefcient. Varianceexplained in outcome variables: job satisfaction
= 33%, family satisfaction = 30%, life satisfaction = 32%. N =
2119.
-
J.S. Michel et al. / Journal of Vocational Behavior 74 (2009)
199218 209Table 4Parameter estimates for the integrative
work-family conict model (Model 4).
Parameter estimates
Intercorrelations among exogenous variables Work-family
spillover
WSS WI WRC WTD WRA FSS FI FRC FTD FRA Satisfaction spillover
(correlationrelationship)
WSS JS FSWI .15
(.02) JS
WRC .36(.02)
.00(.02)
FS .06(.01)
WTD .04(.02)
.24(.02)
.15(.02)
Conict spillover (correlationrelationship)
WRA .24(.02)
.07(.02)
.49(.02)
.11(.02)
WIF FIW
FSS .17(.02)
.02(.02)
.05(.02)
.03(.02)
.06(.02)
WIF
FI .05(.02)
.12(.02)
.04(.02)
.02(.02)
.00(.02)
.14(.02)
FIW .27(.02)
FRC .03(.02)
.07(.02)
.29(.02)
.01(.02)
.11(.02)
.35(.02)
.04(.02)
FTD .01(.02)
.06(.02)
.05(.02)
.01(.02)
.01(.02)
.02(.02)
.05(.02)
.25(.02)
FRA .06(.02)
.07(.02)
.16(.02)
.18(.02)
.33(.02)
.29(.02)
.13(.02)
.28(.02)
.21(.02)
tive work-family conict model) explained 40% of the variance in
job satisfaction, 38% of the variance in family satisfaction,and
35% of the variance in life satisfaction. Collectively, these
results suggest that Models 1, 2, and 4, provide
incrementalexplanation in the prediction of satisfaction
outcomes.
6.3. Explanatory power of model linkages
To provide a more critical examination of work-family conict
models, we systematically analyzed nested models or link-ages
within our integrative framework. In isolation, the indirect
effects of linkage 2 (WIF) explained 4% of the variance infamily
satisfaction while linkage 4 (FIW) explained 2% of the variance in
job satisfaction. Though not traditionally modeledin work-family
conict models, our additional linkage 7 (WIF) explained 6% of the
variance in job satisfaction while linkage 8(FIW) explained 5% of
the variance in family satisfaction. The nal WIF and FIW linkages 9
and 10 were both modest pre-dictors of life satisfaction accounting
for 9% and 4%. Turning to direct effect linkages, we found that
linkage 11 (work ante-cedents) explained 39% of the variance in job
satisfaction while linkage 12 (family antecedents) explained 38% of
thevariance in family satisfaction.
To determine the utility of the work-family conict construct, in
regards to structural models predicting satisfaction out-comes, we
examined our integrative model with and without WIF, FIW, and
pathways 110, and 14, thus removing all work-family conict
variables and pathways. Model t without these variables and
pathways was nearly identical to the full work-family conict model:
v2(20) = 312.83, CN = 250.39, GFI = .98, NFI = .96, CFI = .96,
RMSEA = .08, and AIC = 454.83. In addition,the remaining pathway
estimates remained highly stable with no more than .02 differences
in magnitudes between modelswith and without work-family conict
constructs and pathways. Further, variance explained in job
satisfaction reduced from40% to 39%, variance explained in family
satisfaction reduced from 38% to 37%, and variance explained in
life satisfactionreduced from 35% to 33%. Collectively, these
results suggest that work-family conict and corresponding indirect
effect link-
Path estimates (direction of pathway is represented by column to
row)
WSS WI WRC WTD WRA FSS FI FRC FTD FRA WIF FIW JS FSWIF .13
(.02).05(.02)
.20(.03)
.20(.02)
.06(.02)
.04(.02)
.03(.02)
.14(.02)
.01(.02)
.03(.02)
FIW .02(.02)
.01(.02)
.10(.03)
.04(.02)
.06(.03)
.06(.02)
.02(.02)
.13(.03)
.09(.02)
.19(.02)
JS .25(.02)
.24(.02)
.16(.02)
.06(.02)
.28(.02)
.10(.02)
.01(.02)
FS .29(.02)
.16(.02)
.28(.02)
.04(.02)
.14(.02)
.07(.02)
.05(.02)
LS .13(.02)
.01(.02)
.34(.02)
.37(.02)
Note:WIF, work interference with family conict; FIW, family
interference with work conict; WSS, work social support; WI, work
involvement; WRC, workrole conict; WTD, work time demands; WRA,
work role ambiguity; FSS, family social support; FI, family
involvement; FRC, family role conict; FTD, familytime demands; FRA,
family role ambiguity; JS, job satisfaction; FS, family
satisfaction; LS, life satisfaction. Values in parentheses are
standard errors for thepath coefcient. Variance explained in
outcome variables: job satisfaction = 40%, family satisfaction =
38%, life satisfaction = 35%. N = 2119.
-
effects accounted for 7% of the variance in job satisfaction and
7% of the variance in family satisfaction independently, by
210 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 199218accounting for 40% of the variance in job
satisfaction, 38% of the variance in family satisfaction, and 35%
of the variancein life satisfaction concurrently (see integrative
work-family conict model). It is important to note that this
variance ac-ages (see Edwards & Rothbard, 2000; Michel &
Hargis, 2008) possess small incremental explication in the
prediction of sat-isfaction outcomes.
7. Discussion
This study examined the complex interplay between work and
family through a conict theory perspective. Our resultssuggest that
Model 1 (the parsimonious stressor and involvement model), Model 2
(the expanded stressor model), and Mod-el 4 (the integrative
work-family conict model) are generalizable to our meta-analytic
correlation matrix, while Model 3(the partial mediating stressor
model) was not generalizable. Consistent path analytic ndings
across models suggest thefollowing: (1) work role conict and work
time demands are primary predictors of WIF, while work social
support, workinvolvement, work role conict, and work role ambiguity
are primary predictors of job satisfaction; (2) family role
conictand family role ambiguity are primary predictors of FIW,
while family social support, family involvement, family role
con-ict, and family role ambiguity are primary predictors of family
satisfaction; (3) work antecedents are related to FIW andfamily
antecedents are related to WIF, but these relations are small; (4)
there tends to be cross-domain spillover among con-structs (e.g.,
role ambiguity, work-family conict); (5) WIF and FIW have low
relationships with satisfaction outcomes; and(6) job satisfaction
and family satisfaction are strong predictors of life
satisfaction.
We also critically examined our integrative work-family conict
model (Model 4) in regards to explanatory power of spe-cic
work-family linkages (cf. Edwards & Rothbard, 2000; Michel
& Hargis, 2008). At the core of work-family conict modelsare
what we have referred to as indirect effects. In isolation, work
and family antecedents had small indirect effects on familyand work
satisfaction outcomes via WIF and FIW (24% explained variance).
This is surprising considering the signicant,though modest,
bivariate results supporting these linkages (e.g., Kossek &
Ozeki, 1998). In addition, we found that WIFand FIW were actually
better predictors of same domain satisfaction than cross-domain
satisfaction (56% explained vari-ance). This was highly unexpected
as WIF and FIW explain greater variance in the satisfaction
outcomes that they shouldnot explain, theoretically, then the
satisfaction outcomes that they should explain (e.g., WIF is a
better predictor of job sat-isfaction at 6% explained variance than
family satisfaction at 2% explained variance). In regards to direct
effects, results indi-cate that work and family domain antecedents
are strong predictors of same domain satisfaction (3839%
explainedvariance). Further, when examined as a complete interface,
indirect effects had an incremental predictive explanation of amere
12% explained variance in satisfaction outcomes. These ndings
suggest that when taking into account direct effectlinkages,
indirect effect linkages of work-family conict possess little
incremental explanatory power in regards to satisfac-tion
outcomes.
7.1. Contributions to theory and practice
There are several implications from this study based on specic
parameter estimates. For example, family time demands(hours worked,
number of children, and age of youngest child) had a low impact on
FIW (pathways of .08 for the stressor/involvement based Model 2 and
.04 for integrative Model 4). This is an important implication as
employees with family de-mands, particularly working mothers, often
experience inequality within organizations (e.g., Bagilhole, 2006;
Bo, 2006). Fur-ther, researchers have suggested that these family
demands are directly responsible for feelings of FIW and
genderdifferences in the workforce (e.g., Keene & Reynolds,
2005). Our results, however, indicate that family time demands
havevery little impact on work rolesas indicated by FIW. This nding
has potential impact on the quest for equality in the workplace in
regards to family and gender roles, as traditional perceptions of
family time demands being a job cost may be inquestion. Another
interesting nding is the relationship between social support and
role conict; specically, work socialsupport and work role conict
(.36 in integrative Model 4), and family social support and family
role conict (-.35 in inte-grative Model 4). Should the causal
mechanisms of these relationships be determined, organizations may
benet by poten-tially increasing work social support in an effort
to decrease work role conict. This logic may extend to the family
domain aswell. According to the pathways presented here, from work
role conict to job satisfaction (pathway of .16 in integrativeModel
4) and from family role conict to family satisfaction (pathway of
.28 in integrative Model 4), role conict hasimportant implications
to the outcomes of satisfaction considered in this study.
Though the results of do not strongly support the work-family
conict construct in regards to satisfaction outcomes, thisstudy
does provided clarication on the relationships between work and
family, particularly in regards to the additive valueof different
work-family variables and model linkages. In doing so, this study
answered the call of Eby et al. (2005) for addi-tional model
testing and development by documenting nontrivial differences
between existing models and through the inte-gration of current
theory and empirical ndings in the culmination and examination of
our integrative work-family conictmodel (Model 4). In particular,
we found that the model structure and variable inclusion of Model 1
(parsimonious stressorand involvement model) and Model 2 (expanded
stressor model) t the data well and were predictive, and each model
pro-vided incremental explanation. Likewise, our integrative
work-family conict model (Model 4) sheds further light on
thecomplex interrelationships between work and family. Our ndings
also extend the work of Ford et al. (2007), where indirect
-
J.S. Michel et al. / Journal of Vocational Behavior 74 (2009)
199218 211counted for is not necessarily due to an increase in the
number of predictors, but instead the modeling of direct effects
tra-ditionally found in work-family conict models.
However, an unexpected implication involves our results
regarding the indirect effects of WIF and FIW on
satisfactionoutcomes. Across the well-tting models, WIF and FIW had
low magnitudes to satisfaction outcomes; in fact, these ndingsnever
exceed pathways of .13. Though these ndings differ from bivariate
meta-analyses, i.e., our magnitudes are muchsmaller, these models
still possess signicant though small relationships from WIF and FIW
to satisfaction outcomes whilecontrolling for various work and
family domain antecedents, accounting for 12% of unique variance.
As our results indicatethe presence of potential moderators, and
this question is beyond the scope of the current study, it seems
appropriate forfuture research to better determine under what
conditions that these linkages are strong (e.g., contextual
factors; social cog-nitions). In addition, research could determine
why WIF and FIW were more related to same domain satisfaction
(e.g., WIFand job satisfaction) than cross-domain satisfaction
(e.g., WIF and family satisfaction) as WIF and FIWmeasures were
explic-itly developed to tap into cross-domain effects. This nding
suggests that when expectations of interference from one do-main
are higher than expected, the satisfaction levels in that same
domain are negatively impacted. Further research mayalso consider
that affect at home and affect at work (cf. Ilies, Schwind,
&Wagner, in press; Ilies et al., 2007) may be an under-lying
and relatively unassessed component of these and other work-family
linkages.
7.2. Potential limitations
Like most meta-analytic studies, methodological limitations are
inevitable. For the current meta-analytic path analysis,two
potential limitations need mentioning. The stability of the
meta-analytic bivariate relationships is variable. Since thisreview
investigated the interrelationships of 15 variables, a
meta-analytic correlation matrix of 105 cells was generated.These
cells contain discrepancies in the stability of their estimation of
the population parameters. For example, the WIFand FIW relationship
(k = 121, n = 47,305) should be a much more robust population
approximation than the family roleambiguity and work time demands
(k = 1, n = 261) relationship. Though variability in parameter
estimate stability is alwaysa weakness, this limitation should not
be a major problem for the current study. For example, a recent
meta-analytic pathanalysis described as truly exemplary (see Bobko
& Roth, 2003, p. 82) contained empty cells in 21% of the
meta-analyticmatrix, while another 21% of the cells contained only
one study. Considering the current work contains estimates for all
105cells, while only two cells (2%) had one sample, the variability
in population parameter estimate stability in the
currentmeta-analytic matrix is not inconsistent with that found in
other, similar studies.
In addition, we used structural equation modeling with
cross-sectional data to t a series of structural models. Several
ofthe models showed good t, however, this just means that these
models are plausible explanations for the observed patternsof
covariance (James, Muliak, & Brett, 1982). Accordingly, other
equally plausible models may exist that provide similar orbetter t.
Accordingly, future research should seek to establish if the
general structure implied by these models are consis-tent with a
truly causal model (this would entail collecting and structuring
studies so as to best satisfy the requisite condi-tions for causal
inference; cf. James et al., 1982).
8. Conclusions
We have reviewed, developed, tested, and critiqued multiple
full-range work-family conict models and model linkages.Each model
was evaluated and compared on model t, parameter estimates, and
variance explained in outcomes. Theembedded linkages within our
integrative model were then critically examined through a series of
path analyses. Resultsindicated that direct effects drive these
models and indirect effects provide only small amounts of
incremental explanationin satisfaction outcomes. Research should
further explore the potential mechanisms linking work and family
domains, bothindirect and direct, concurrently and in
isolation.
References
Alexander, R. A., Scozzaro, M. J., & Borodkin, L. J. (1989).
Statistical and empirical examination of the chi-square test for
homogeneity of correlations in meta-analysis. Psychological
Bulletin, 106, 329331.
Allen, T. D., Herst, D. E. L., Bruck, C. S., & Sutton, M.
(2000). Consequences associated with work-to-family conict: A
review and agenda for future research.Journal of Occupational
Health Psychology, 5, 278308.
Amatea, E. S., Cross, E. G., Clark, J. E., & Bobby, C. L.
(1986). Assessing the work and family role expectations of career
oriented men and women: The life rolesalience scales. Journal of
Marriage and the Family, 48, 831838.
Anderson, S. E., Coffey, B. S., & Byerly, R. T. (2002).
Formal organizational initiatives and informal workplace practices:
Links to work-family conict and jobrelated outcomes. Journal of
Management, 28, 787810.
Aryee, S., Fields, D., & Luk, V. (1999). A cross-cultural
test of a model of the work-family interface. Journal of
Management, 25, 491511.Ashforth, B. E., Kreiner, G. E., &
Fugate, M. (2000). All is a days work: Boundaries and micro role
transitions. Academy of Management Review, 25, 472491.Bacharach, S.
B., Bamberger, P., & Conley, S. (1990). Work processes, role
conict, and role overload: The case of nurses and engineers in the
public sector.
Work and Occupations, 17, 199228.Bagilhole, B. (2006).
Family-friendly policies and equal opportunities: A contradiction
in terms? British Journal of Guidance & Counseling, 34,
327343.Bedeian, A. G., Burke, B. G., & Moffett, R. G. (1988).
Outcomes of work-family conict among married male and female
professionals. Journal of Management,
14, 475491.Beehr, T. A. (1995). Psychological stress in the
workplace. London: Routledge.Beehr, T. A., & Glazer, A. (2005).
Organizational role stress. In J. Barling, E. K. Kelloway, & M.
R. Frone (Eds.), Handbook of work stress. Thousand Oaks, CA:
Sage.Beehr, T. A., King, L. A., & King, D. W. (1990). Social
support and occupational stress: Talking to supervisors. Journal of
Vocational Behavior, 36, 6181.
-
212 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 199218Bentler, P. M., & Bonett, D. G. (1980). Signicance
tests and goodness of t in the analysis of covariance structures.
Psychological Bulletin, 88, 588606.Bo, I. (2006). Working life and
family life: Ambiguous communication at work. Community, Work &
Family, 9, 123141.Bobko, P., & Roth, P. L. (2003).
Meta-analysis and validity generalization as research tools: Issues
of sample bias and degrees of mis-specication. In K. R.
Murphy (Ed.), Validity generalization: A critical review.
Mahwah, NJ: Lawrence Erlbaum Associates, Inc.Bond, J. T., Galinsky,
E., Swanberg, J. E. (1998). The 1997 national study of the changing
workplace. New York: Families and Work Institute.Brown, S. P.,
& Peterson, R. A. (1993). Antecedents and consequences of
salesperson job satisfaction: Meta-analysis and assessment of
causal effects. Journal
of Marketing Research, 30, 6377.Burke, R. J. (1986).
Occupational and life stress and the family: Conceptual frameworks
and research ndings. International Review of Applied Psychology,
35,
347369.Burley, K. A. (1990). Work-family conict and marital
adjustment in dual career couples: A comparison of three time
models (Doctoral dissertation,
Claremont Graduate School, 1989). Dissertation Abstracts
International, 50, 4803.Byron, K. (2005). A meta-analytic review of
work-family conict and its antecedents. Journal of Vocational
Behavior, 67, 169198.Caplan, R. D., Cobb, S., & French, J. R.
P. Jr., (1975). Relationships of cessation of smoking with job
stress, personality, and social support. Journal of Applied
Psychology, 60, 211219.Caplan, R. D., Cobb, S., French, J. R.
P., Jr., Harrison, R. V., & Pinneau, S. R. Jr., (1975). Job
demands and worker health: Main effects and occupational
differences.
Washington, DC: U.S. Government Printing Ofce.Caplan, R. D.,
Cobb, S., French, J. R. P., Harrison, R. V., & Pinneau, S. R.
Jr., (1980). Job demands and worker health: Main effects and
occupational differences.
Research Report Series. Ann Arbor, MI: University of Michigan,
Institute for Social Research.Carlson, D. S., & Frone, M. R.
(2003). Relation of behavioral and psychological involvement to a
new four-factor conceptualization of work-family
interference. Journal of Business and Psychology, 17,
515535.Carlson, D. S., & Kacmar, K. M. (2000). Work-family
conict in the organization: Do life role values make a difference?
Journal of Management, 26, 10311054.Carlson, D. S., Kacmar, K. M.,
& Williams, L. J. (2000). Construction and initial validation
of a multidimensional measure of work-family conict. Journal of
Vocational Behavior, 56, 249276.Carlson, D. S., & Perrew, P.
L. (1999). The role of social support in the stressor-strain
relationship: An examination of work-family conict. Journal of
Management, 25, 513540.Chen, G., Casper, W. J., & Cortina,
J. M. (2001). The roles of self-efcacy and task complexity in the
relationships among cognitive ability, conscientiousness,
and work-related performance: A meta-analytic examination. Human
Performance, 14, 209230.Clark, S. L., & Weismantle, M. (2003).
Employment Status: 2000. Census 2000 Brief, Series C2KBR-18.Cooper,
C. L., Cooper, R. D., & Eaker, L. H. (1988). Living with
stress. London: Penguin.Cranny, C. J., Smith, P. C., & Stone,
E. F. (Eds.). (1992). Job Satisfaction: How people feel about their
jobs and how it affects their performance. New York: Lexington
Books.Diefendorff, J. M., Brown, D. J., Kamin, A. M., &
Lord, R. G. (2002). Examining the roles of job involvement and work
centrality in predicting organizational
citizenship behaviors and job performance. Journal of
Organizational Behavior, 23, 93108.Diener, E., Emmons, R. A.,
Larsen, R. J., & Grifn, S. (1985). The satisfaction with life
scale. Journal of Personality Assessment, 49, 7175.Duxbury, L. E.,
& Higgins, C. A. (1991). Gender differences in work-family
conict. Journal of Applied Psychology, 76, 6073.Eby, L. T., Casper,
W. J., Lockwood, A., Bordeaux, C., & Brinely, A. (2005). Work
and family research in IO/OB: Content analysis and review of the
literature
(19802002). Journal of Vocational Behavior, 66, 124197.Edwards,
J. R., & Rothbard, N. P. (2000). Mechanisms linking work and
family: Clarifying the relationship between work and family
constructs. Academy of
Management Review, 25, 178199.Eisenberger, R., Huntington, R.,
Hutchison, S., & Sowa, D. (1986). Perceived organizational
support. Journal of Applied Psychology, 71, 500507.Elloy, D. F.,
& Smith, C. R. (2003). Patterns of stress, work-family conict,
role conict, role ambiguity and overload among dual-career and
single-career
couples: An Australian study. Cross Cultural Management, 10,
5566.Evans, P., & Bartolome, F. (1984). The changing pictures
of the relationship between career and family. Journal of
Occupational Behavior, 5, 921.Ford, M. T., Heinen, B. A., &
Langkamer, K. L. (2007). Work and family satisfaction and conict: A
meta-analysis of cross-domain relations. Journal of Applied
Psychology, 92, 5780.French, J. R. P., & Caplan, R. D.
(1973). Organizational stress and individual strain. In A. J.
Marrow (Ed.), The failure of success. New York: AMACOM.Frone, M. R.
(2003). Work family balance. In J. C. Quick & L. E. Tetrick
(Eds.), Handbook of occupational health psychology. Washington
D.C.: American
Psychological Association.Frone, M. R., Russell, M., &
Cooper, M. L. (1992). Antecedents and outcomes of work-family
conict: Testing a model of the work-family interface. Journal
of
Applied Psychology, 77, 6578.Frone, M. R., Yardley, J. K., &
Markel, K. S. (1997). Developing and testing an integrative model
of the work-family interface. Journal of Vocational Behavior,
50,
145167.Grandey, A. A., & Cropanzano, R. (1999). The
conservation of resources model applied to work-family conict and
strain. Journal of Vocational Behavior, 54,
350370.Greenhaus, J. H., & Beutell, N. J. (1985). Sources of
conict between work and family roles. Academy of Management Review,
10, 7688.Greenhaus, J. H., Callanan, G. A., & Godshalk, V. M.
(2000). Career management. Fort Worth: The Dryden Press.Greenhaus,
J. H., & Parasuraman, S. (1999). Research on work, family, and
gender: Current status and future directions. In G. N. Powell
(Ed.), Handbook of
gender and work. Thousand Oaks: Sage.Greenhaus, J. H., &
Powell, G. N. (2006). When work and family are allies: A theory of
work-family enrichment. Academy of Management Review,
31, 7292.Gupta, N., & Jenkins, G. D. (1985). Dual-career
couples: Stress, stressors, strains and strategies. In T. A. Beehr
& R. S. Bhagat (Eds.), Human stress and cognition
in organizations: An integrated perspective. New York:
Wiley.Gutek, B. A., Searle, S., & Klepa, L. (1991). Rational
versus gender role explanations for work-family conict. Journal of
Applied Psychology, 76, 560568.Hackman, J. R., & Oldham, G. R.
(1975). Development of the Job Diagnostic Survey. Journal of
Applied Psychology, 60, 159170.Hackman, J. R., & Oldham, G. R.
(1976). Motivation through the design of work: Test of a theory.
Organizational Behavior and Human Performance, 16,
250279.Hedges, L. V. (1982). Estimation of effect size from a
series of independent experiments. Psychological Bulletin, 92,
490499.Hedges, L. V., & Olkin, I. (1985). Statistical methods
for meta-analysis. Orlando, FL: Academic Press.Hoelter, J. W.
(1983). The analysis of covariance structures: Goodness-of-t
indices. Sociological Methods & Research, 11, 325344.Hopkins,
A. H. (1983). Work and job satisfaction in the public sectors.
Totowa, NJ: Rowman and Allonheld.Hom, P. W., Caranikas-Walker, F.,
Prussia, G. E., & Griffeth, R. W. (1992). A meta-analytical
structural equations analysis of a model of employee turnover.
Journal of Applied Psychology, 77, 890909.House, G. S. (1981).
Work stress and social support. Reading, MA: Addison-Wesley.Hu, L.,
& Bentler, P. M. (1995). Evaluating model t. In R. H. Hoyle
(Ed.). Structural equation modeling: Concepts, issues, and
applications. Thousand Oaks, CA:
Sage.Hu, L., & Bentler, P. M. (1999). Cutoff criteria for t
indexes in covariance structure analysis: Conventional criteria
versus new alternative. Structural Equation
Modeling, 6, 155.Ilies, R., Schwind, K. M., & Wagner, D. T.
(in press). The spillover of daily job satisfaction onto employees
family lives: The facilitating role of work-family
integration. Academy of Management Journal.
-
J.S. Michel et al. / Journal of Vocational Behavior 74 (2009)
199218 213Ilies, R., Schwind, K. M., Wagner, D. T., Johnson, M. D.,
DeRue, D. S., & Ilgen, D. R. (2007). When can employees have a
family life? The effects of dailyworkload and affect on work-family
conict and social behaviors at home. Journal of Applied Psychology,
92, 13681379.
James, L. R., Muliak, S. S., & Brett, J. M. (1982).
Mediators, moderators, and tests for mediation. Journal of Applied
Psychology, 69, 307321.Johnson, B. T. (1993). DSTAT 1.10: Software
for the meta-analytic review of research literatures. Hillsdale,
NJ: Erlbaum.Jreskog, K. G., & Srbom, D. (1993). LISREL 8:
Analysis of linear structural relationships by maximum likelihood,
instrumental variables and least squares methods
(8th ed.). Mooresville, IN: Scientic Software.Judge, T. A.,
Boudreau, J. W., & Bretz, R. D. (1994). Job and life attitudes
of male executives. Journal of Applied Psychology, 79, 767782.Kahn,
R. (1980). Conict, ambiguity, and overload: Three elements in job
stress. In D. Katz, R. Kahn, & J. Adams (Eds.), The study of
organizations
(pp. 418428). San Francisco: Jossey-Bass.Kahn, R. L., Wolfe, D.
M., Quinn, R. P., Snoek, J. D., & Rosenthal, R. A. (1964).
Organizational stress: studies in role conict and ambiguity. New
York: Wiley.Kanungo, R. N. (1982). Measurement of job and work
involvement. Journal of Applied Psychology, 67, 341349.Katz, D.,
& Kahn, R. (1978). The social psychology of organizations. New
York: Wiley.Keene, J. R., & Reynolds, J. R. (2005). The job
costs of family demands: Gender differences in negative
family-to-work spillover. Journal of Family Issues, 26,
275299.Kelloway, R. W., Gottlieb, B. H., & Barham, L.
(1999). The source, nature, and direction of work and family
conict: A longitudinal investigation. Journal of
Occupational Health Psychology, 4, 337346.King, L. A.,
Mattimore, L. K., King, D. W., & Adams, G. A. (1995). Family
support inventory for workers: A new measure of perceived social
support for family
members. Journal of Organizational Behavior, 16,
235258.Kirchmeyer, C. (1992). Perceptions of nonwork-to-work
spillover: Challenging the common view of conict-ridden domain
relationships. Basic and Applied
Social Psychology, 13, 231249.Kopelman, R., Greenhaus, J. H.,
& Connolly, T. F. (1983). A model of work, family, and
interrole conict: A construct validation study. Organizational
Behavior
and Human Performance, 32, 198215.Kossek, E. E., & Ozeki, C.
(1998). Work-family conict, policies, and the job-life satisfaction
relationship: A review and directions for organizational
behavior-
human resources research. Journal of Applied Psychology, 83,
139149.Locke, E. A. (1976). The nature and causes of job
satisfaction. In M. D. Dunnette (Ed.), Handbook of industrial and
organizational psychology (pp. 12971349).
Chicago: Rand-McNally.Lodahl, T. M., & Kejnar, M. (1965).
The denition and measurement of job involvement. Journal of Applied
Psychology, 49, 2433.Major, V. S., Klein, K. J., & Ehrhart, M.
G. (2002). Work time, work interference with family, and
psychological distress. Journal of Applied Psychology, 87,
427436.Marsh, H. W., Hau, K., & Wen, Z. (2004). In search of
golden rules: Comment on hypothesis-testing approaches to setting
cutoff values for t indexes and
dangers in overgeneralizing Hu and Bentlers (1999) ndings.
Structural Equation Modeling, 11, 320341.Matsui, T., Ohsawa, T.,
& Onglatco, M. I. (1995). Work-family conict and the
stress-buffering effects of husband support and coping behavior
among
Japanese married working women. Journal of Vocational Behavior,
47, 178192.Mesmer-Magnus, J. R., & Viswesvaran, C. (2005).
Convergence between measures of work-to-family and family-to-work
conict: A meta-analytic
examination. Journal of Vocational Behavior, 67, 215232.Michel,
J. S., & Hargis, M. B. (2008). Linking mechanisms of
work-family conict and segmentation. Journal of Vocational
Behavior, 73, 509522.Nye, F., & MacDougall, E. (1959). The
dependant variable in marital research. Pacic Sociological Review,
2, 6770.ODriscoll, M. P., Ilgen, D. R., & Hildreth, K. (1992).
Time devoted to job and off-job activities, interrole conict, and
affective experiences. Journal of Applied
Psychology, 77, 272279.Parasuraman, S., Purohit, Y. S.,
Godshalk, V. M., & Beutell, N. J. (1996). Work and family
variables, entrepreneurial career success and psychological
well-
being. Journal of Vocational Behavior, 48, 275300.Peterson, M.
F., Smith, P. B., Akande, A., Ayestaran, S., Bochner, S., Callan,
V., et al (1995). Role conict, ambiguity, and overload: A 21-nation
study. Academy
of Management Journal, 38, 429452.Procidano, M. E., &
Heller, K. (1983). Measures of perceived social support from
friends and from family: Three validation studies. American Journal
of
Community Psychology, 11, 124.Quinn, R. P., & Shepard, L. J.
(1974). The 197273 quality of employment survey: Descriptive
statistics, with comparison data from the 196970 survey of
working
conditions. Ann Arbor, MI: University of Michigan, Institute for
Social Research.Quinn, R. P., & Staines, G. L. (1979). The 1977
quality of employment survey. Ann Arbor, MI: University of
Michigan, Institute for Social Research.Rice, R. W., McFarlin, D.
B., Hunt, R. G., & Near, J. P. (1985). Organizational work and
the perceived quality of life. Academy of Management Review,
10,
296310.Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970).
Role conict and ambiguity in complex organizations. Administrative
Science Quarterly, 15, 150163.Rosenthal, R. (1994). Statistically
describing and combining studies. In H. Cooper & L. V. Hedges
(Eds.), The handbook of research synthesis (pp. 231244). New
York: Russell Sage.Schuler, R. S. (1980). Denition and
conceptualization of stress in organizations. Organizational
Behavior and Human Performance, 25, 184215.Shinn, M., Wong, N. W.,
Simko, P. A., & Ortiz-Torres, B. (1989). Promoting the
well-being of working parents: Coping, social support, and exible
job
schedules. American Journal of Community Psychology, 17,
3155.Simmons, T., & ONeill, G. (2001). Households and Families:
2000. Census 2000 Brief, Series C2KBR/01-8.Smith, P. C., Kendall,
L. M., & Hulin, C. L. (1969). The measurement of satisfaction
in work and retirement: A strategy for the study of attitudes.
Chicago: Rand
McNally.Spanier, G. B. (1976). Measuring dyadic adjustment: New
scales for assessing the quality of marriage and similar dyads.
Journal of Marriage and the Family,
38, 1528.Staines, G. L., & Pleck, J. H. (1984). Nonstandard
work schedules and family life. Journal of Applied Psychology, 69,
515523.Usita, P. M., Hall, S. S., & Davis, J. C. (2004). Role
ambiguity in family caregiving. Journal of Applied Gerontology, 23,
2039.Viswesvaran, C., & Ones, D. S. (1995). Theory testing:
Combining psychometric meta-analysis and structural equations
modeling. Personnel Psychology, 48,
865885.Williams, K., & Alliger, G. M. (1994). Role
stressors, mood spillover, and perceptions of work-family conict in
employed parents. Academy of Management
Journal, 37, 837868.Yogev, S., & Brett, J. (1985). Patterns
of work and family involvement among single- and dual-earner
couples. Journal of Applied Psychology, 70, 754768.Zedeck, S.,
& Mosier, K. L. (1990). Work in the family and employing
organization. American Psychologist, 45, 240251.
References marked with an asterisk indicate studies include in
the meta-analysis.
*Adams, G. A., & Jex, S. M. (1999). Relationships between
time management, control, work-family conict, and strain. Journal
of Occupational HealthPsychology, 4, 7277.
*Adams, G. A., King, L. A., & King, D. W. (1996).
Relationship of job and family involvement, family social support,
and work-family conict with job and lifesatisfaction. Journal of
Applied Psychology, 81, 411420.
*Allen, T. D. (2001). Family-supportive work environments: The
role of organizational perceptions. Journal of Vocational Behavior,
58, 414435.*Anderson, S. E., Coffey, B. S., & Byerly, R. T.
(2002). Formal organizational initiatives and informal workplace
practices: Links to work-family conict and job-
related outcomes. Journal of Management, 28, 787810.
-
*214 J.S. Michel et al. / Journal of Vocational Behavior 74
(2009) 199218Anderson-Kulman, R. E., & Paludi, M. A. (1986).
Working mothers and the family context: Predicting positive coping.
Journal of Vocational Behavior, 28,241253.
*Aryee, S. (1992). Antecedents and outcomes of work-family
conict among married professional women: Evidence from Singapore.
Human Relations, 45,813837.
*Aryee, S., Fields, D., & Luk, V. (1999). A cross-cultural
test of a model of the work-family interface. Journal of
Management, 25, 491511.*Aryee, S., & Luk, V. (1996). Work and
nonwork inuences on the career satisfaction of dual-earner couples.
Journal of Vocational Behavior, 49, 3852.*Aryee, S., Luk, V.,
Leung, A., & Lo, S. (1999). Role stressors, interrole conict,
and well-being: The moderating inuence of spousal support and
coping
behaviors among employed parents in Hong Kong. Journal of
Vocational Behavior, 54, 259278.*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, 132146.*Bacharach, S. B., Bamberger, P., &
Conley, S. (1991). Work-home conict among nurses and engineers:
Mediating the impact of role stress on burnout and
satisfaction at work. Journal of Organizational Behavior, 12,
3953.*Baltes, B. B., & Heydens-Gahir, H. (2003). Reduction of
work-family conict through the use of Selection, Optimization, and
Compensation behaviors.
Journal of Applied Psychology, 88, 10051018.*Barich, A. W.
(1995). The moderating roles of coping and control on the
experience of work-family conict (Doctoral dissertation, Southern
Illinois
University, 1994). Dissertation Abstracts International, 56,
1691.*Barrah, J. L., Shultz, K. S., Baltes, B., & Stolz, H. E.
(2004). Mens and womens eldercare-based work-family conict:
Antecedents and work-related outcomes.
Fathering, 2, 305330.*Batt, R., & Valcour, P. M. (2003).
Human resources practices as predictors of work-family outcomes and
employee turnover. Industrial Relations, 42,
189220.*Beatty, C. A. (1996). The stress of managerial and
professional women: Is the price too high? Journal of
Organizational Behavior, 17, 233251.*Bedeian, A. G., Burke, B. G.,
& Moffett, R. G. (1988). Outcomes of work-family conict among
married male and female professionals. Journal of Management,
14, 475491.*Behson, S. J. (2002a). Coping with family-to-work
conict: The role of informal work accommodations to family. Journal
of Occupational Health Psychology, 7,
324341.*Behson, S. J. (2002b). Which dominates? The relative
importance of work-family organizational support and general
organizational context on employee
outcomes. Journal of Vocational Behavior, 61, 5372.*Bernas, K.
H., & Major, D. A. (2000). Contributors to stress resistance:
Testing a model of womens work-family conict. Psychology of Women
Quarterly, 24,
170178.*Beutell, N. J., & Wittig-Berman, U. (1999).
Predictors of work-family conict and satisfaction with family, job,
career, and life. Psychological Reports, 85,
893903.*Bhuian, S. N., Menguc, B., & Borsboom, R. (2005).
Stressors and job outcomes in sales: A triphasic model versus a
linear-quadratic-interactive model. Journal
of Business Research, 58, 141150.*Blau, G. (1995). Inuence of
group lateness on individual lateness: A cross-level examination.
Academy of Management Journal, 38, 14831496.*Block, B. L. (1995).
An exploration of the relationships among work-family conict,
relevant personality variables, and well-being (Doctoral
dissertation,
The City University of New York, 1995). Dissertation Abstracts
International, 56, 1729.*Bohle, P., & Tilley, A. J. (1998).
Early experience of shiftwork: Inuences on attitudes. Journal of
Occupational and Organizational Psychology, 71, 6179.*Boles, J. S.,
& Babin, B. J. (1996). On the front lines: Stress, conict, and
the customer service provider. Journal of Business Research, 37,
4150.*Boles, J. S., Howard, W. G., & Donofrio, H. H. (2001). An
investigation into the inter-relationships of work-family conict,
family-work conict and work
satisfaction. Journal of Managerial Issues, 13, 376390.*Boles,
J. S., Wood, J. A., & Johnson, J. (2003). Interrelationships of
role conict, role ambiguity, and work-family conict with different
facets of job
satisfaction and the moderating effects of gender. The Journal
of Personal Selling and Sales Management, 23, 99113.*Bolino, M. C.,
& Turnley, W. H. (2005). The personal costs of citizenship
behavior: The relationship between individual initiative and role
overload, job
stress, and work-family conict. Journal of Applied Psychology,
90, 740748.*Borovsky, D. M. (1999). Models and operationalizations
of work-family conict: A comparative analysis and extension to
organizational outcomes in a
sample of unionized workers. (Doctoral dissertation, Wayne State
University, 1999). Dissertation Abstracts International, 59,
6507.*Boyar, S. L., Maertz, C. P., Jr., Pearson, A. W., &
Keough, S. (2003). Work-family conict: A model of linkages between
work and family domain variables and
turnover intentions. Journal of Managerial Issues, 15,
175190.*Brough, P., & Kelling, A. (2002). Women, work &
well-being family-work conict. New Zealand Journal of Psychology,
31, 2938.*Bruck, C. S. (2003). Theoretical antecedents to
work-family conict: A comprehensive approach. (Masters thesis,
University of South Florida, 2000)..
Dissertation Abstracts International, 64, 449.*Bruck, C. S.,
& Allen, T. D. (2003). The relationship between big ve
personality traits, negative affectivity, type A behavior, and
work-family conict. Journal
of Vocational Behavior, 63, 457472.*Bruck, C. S., Allen, T. D.,
& Spector, P. E. (2002). The relation between work-family
conict and job satisfaction: A ner-grained analysis. Journal of
Vocational
Behavior, 60, 336353.*Buffardi, L. C., & Erdwins, C. J.
(1997). Child-care satisfaction: Linkages to work attitudes,
interrole conict, and maternal separation anxiety. Journal of
Occup