Journal of Applied Psychology 1996. Vol. 81. No. 4,400-410 Copyright 1996 by the American Psychological Association, Inc. 0021-9010/96/$3.00 Development and Validation of Work-Family Conflict and Family-Work Conflict Scales Richard G.Netemeyer Louisiana State University James S.Boles Georgia State University Robert McMurrian Louisiana State University Researchers report on a 3-sample study that developed and validated short, self-report scales of work-family conflict (WFC) and family-work conflict (FWC). Using concep- tualizations consistent with the current literature, the researchers offer content domains and definitions of the constructs. Advocated procedures were used to develop the scales and test dimensionality and internal consistency. Estimates of construct validityare pre- sented by relatingthe scales to 16 other on- and off-job constructs. Mean-level difference tests between WFC and FWC also provide evidence of validity. A widely studied topic in organizational behavior is the conflict between work and family. In recent years, several studies have advanced our understanding of how work affects family life and vice versa (Frone, Russell, & Coo- per, 1992 Greenhaus & Beutell, 1985 O'Driscoll, Ilgen, & Hildreth, 1992). Of importance to the study of work- family relations is construct measurement. Although a conceptual distinction between work conflictingwith the family and family conflicting with work has been made, most research has assessed only work-family conflict (WFC). Furthermore, operationalization of WFC has varied widely from study to study, limiting the ability to generalize about the effects of WFC. The purpose of this article is to develop and validate short, self-report mea- sures of WFC and family-work conflict (FWC). Background and Review Two important focal points of adult life are family and work. However, the role expectations of these two do- Richard G. Netemeyerand Robert McMurrian, Department of Marketing, Louisiana State University James S. Boles, De- partment of Marketing, Georgia State University. We thank Art Bedeian for judging items and offering con- structive comments on an earlier version of this article. Wealso thank Cheryl Adkinsand Roger Griffeth for judging items and the Georgia State University Marketing Roundtable for sup- porting this research. Correspondence concerning this article should be addressed to Richard G. Netemeyer,3126BCEBA Building, Department of Marketing,Louisiana State University,Baton Rouge,Louisi- ana 70803. Electronic mail may be sent via Internet to mknete@lsuvm.sncc.lsu.edu. mains are not always compatible, creating conflicts be- tween work and family life. These conflictsare related to outcomes such as job dissatisfaction, job burnout, and turnover (Burke, 1988 Frone et al., 1992 Greenhaus, 1988 Pleck, Staines, & Lang, 1980), as well as to out- comes related to psychological distress (e.g., depression), and life and marital dissatisfaction (Greenhaus & Beu- tell, 1985 Gutek, Searle, & Klepa, 1991 Voydanoff, 1988). Given the increase in dual-earner families, single- parent families, and families with elder-care duties, these outcomes are likely to be even more pronounced in the future. Furthermore, there is mounting evidence that WFC and FWC are related to work productivity and fi- nancial costs incurred by an organization (Cascio, 1991). Although many studies have examined the relation- ships between WFC and other variables, their measures have varied widely. For example, WFC has been opera- tionalized with single-item measures that may lack reli- ability (Rice, Frone, & McFarlin, 1992 Voydanoff, 1988) and with lengthy measures that are possibly cum- bersome to respondents (Burke, 1988 Burke, Weir, & Duwors, 1979). Several studies havecombined WFC and FWC scales into a single measure, ignoring the concep- tual distinction between the two constructs (Cooke & Rousseau, 1984 Kopelman, Greenhaus, & Connolly, 1983 Thomas & Ganster, 1995 Wiley, 1987). Other studies have offered separate WFC and FWC measures, but have used items that reflect potential outcomes of the constructs rather than their content domain, that is, items that assess somatic, physical, and mental symp- toms that are due to WFC, FWC, or both (Bedeian, Burke, & Moffett, 1988 O'Driscoll et al. 1992 Parasur- aman, Greenhaus, Rabinowitz, Bedeian, & Mossholder, 400

WORK AND FAMILY CONFLICT 401 1989). Still other studies have used WFC measures that simply havenot been subjected to rigorous scale-develop- ment procedures (Frone et al., 1992 Gutek et al., 1991 Judge, Boudreau, & Bretz, 1994). Similar observations hold for FWC. To better understand the interplay be- tween these two constructs and their antecedents and out- comes, researchers require sound measures. WFC and FWC Denned The conceptual approach taken in the present research is based on the premise that WFC and FWC are distinct but related forms of interrole conflict (Greenhaus & Beutell, 1985 Kahn, 1981 Kahn, Wolfe, Quinn, Snoek, & Rosenthal, 1964 Pleck et al., 1980). Interrole conflict has been viewedas a form of conflict in which "role pres- sures associated with membership in one organization are in conflict with pressures stemming from member- ship in other groups" (Kahn et al., 1964, p. 20). From work-family and family-work perspectives, this type of conflict reflects the degree to which role responsibilities from the work and family domains are incompatible, that is, "participation in the work (family) role is made more difficult by virtue of participation in the family (work) role" (Greenhaus & Beutell, 1985, p. 77). As such, the demands of one role make performance of the other role more difficult (Katz & Kahn, 1978). Although several sources of WFC and FWC havebeen identified, most researchers agree that the general de- mands of a role, the time devoted to a given role, and the strain produced by a given role are domain elements of WFC and FWC (Bachrach, Bamberger, & Conley, 1991 Cooke & Rousseau, 1984 Greenhaus, 1988 Greenhaus & Beutell, 1985 Gutek et al., 1991 Kahn & Byosiere, 1992 Kahn et al., 1964 Pleck et al., 1980 Voydanoff, 1988). The general demands of a role refer to the respon- sibilities, requirements, expectations, duties, and com- mitments associated with a given role. (These terms have been used interchangeably throughout the literature.) Time-based conflict occurs when the amount of time devoted to the work (family) role interferes with per- forming family- (work-)related responsibilities. Specifi- cally, excessive work (family) time conflicts may make it difficult to comply with family (work) responsibilities. Strain-based conflict occurs when strain created by the work (family) role interferes with performing family (work) responsibilities. For example, irritabilityand anx- iety created by workinterfere with performing family du- ties and vice versa. As such, we used the following definitionsto guide our scale development. WFC is a form of interrole conflict in which the general demands of, time devoted to, and strain created by the job interfere with performingfam- ily-related responsibilities. FWC is a form of interrole conflict in which the general demands of, time devoted to, and strain created by the family interfere with per- forming work-related responsibilities. Relations With Other Variables and Mean-Level Difference Investigating the construct validity of the WFC and FWC scales developed in this study called for a number of predictions to be advanced. These predictions pertain to the relationships between WFC and FWC and other on- and off-job constructs, as well as mean-level differ- ences between WFC and FWC scores. The following sec- tions summarizeour predictions and the rationale for our predictions. On-Job Constructs Research suggests there is an inverse relationship be- tween organizational commitment and WFC and FWC (O'Driscoll et al., 1992 Greenhaus & Beutell, 1985) and between job satisfaction and WFC and FWC (Frone et al., 1992 Rice et al., 1992). Thus, negative correlations between the WFC and FWC scales and organizational commitment andjob satisfaction are predicted. It has been suggested that WFC and FWC should be positively associated with job burnout, job tension, job role conflict, and job role ambiguity (Bedeian et al., 1988 Frone et al., 1992 Maslach & Jackson, 1981). Thus, positive correlations between these variables and WFC and FWC are predicted. Prior research also indi- cates that WFC is more strongly related to job burnout and job tension than FWC (O'Driscoll et al., 1992 Frone et al., 1992 Judge et al., 1994 Maslach & Jackson, 1981). Given these findings, we predict that WFC corre- lates more stronglywithjob burnout andjob tension than with FWC. Research shows that intention-to-leave-an-organiza- tion and search-for-another-jobare positively related to WFC and FWC (Burke, 1988). Therefore, we predict positive correlations between intention-to-leave-an-orga- nization and search-for-another-job and the WFC and FWC measures. Finally, the numbers of hours worked per week should be more highly correlated with WFC than with FWC (Gutek et al., 1991). Because salespeople werethe respondents for the third sample in this study, the following predictions were ad- vanced. Sales self-efficacy influences expectations about one's ability to perform a job successfully (Bandura, 1986). Thus, it is predicted that FWC should have a neg- ative correlation with self-efficacy. The sales literature also suggests an inverse relation between FWC and sales performance (Behrman & Perreault, 1984). Thus, it is predicted that FWC should have a negative correlation with sales performance. To our knowledge, no empirical research has assessed the differences in relationships be-