• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

The choices between our work and nonwork roles: Applying behavioral reasoning theory and the life-role-value concept

ProQuest Dissertations and Theses, 2009
Dissertation
Author: Rebecca Renee Zusman
Abstract:
The ways in which employees' work and nonwork roles conflict with each other have been studied in numerous contexts. However, research has not focused on the specific behavioral decisions that may influence work-family conflict. Therefore, the general purpose of this dissertation was to examine work-family conflict as an outcome of the decision to work beyond normal working hours, through the application of behavioral reasoning theory (Westaby, 2005). Furthermore, life-role-values were introduced as potential moderators of the decision-making process. Regarding specific hypotheses, this study first predicted that the decision to work beyond normal working hours would be related to work-family conflict. This hypothesis is in line with expectancy violation theories, which posit that expectations guide behavior and that negative arousal results when these expectations are broken. Next, behavioral reasoning theory provided a framework to explore the factors that contributed to the decision to work beyond normal working hours. Based on this theory, it was proposed that behavioral reasons (for and against the behavior) and global motives (i.e., attitude toward the behavior, subjective norm, and/or perceived behavioral control) would predict intention which in turn would predict behavior. Finally, the moderating effect of life-role-values on the relationship between attitude and behavioral intention was examined. It was hypothesized that a stronger link between these variables would emerge when work-role-values were strong than when they were weak. A cross-sectional survey design was utilized, involving 336 participants from 11 organizations. Hypotheses were tested using hierarchical regression analyses. Overall, it was found that: (1) the behavior of working beyond normal work hours predicted work-family conflict; (2) intention predicted behavior; (3) attitude predicted intention; (4) behavioral reasons predicted global motives; and (5) behavioral reasons independently predicted intention, over and above that explained by constructs in the theory of planned behavior. Against predictions, life-role-values did not moderate the attitude-intention relationship. Supplemental analyses also found that behavioral reasons for and against working beyond normal working hours directly predicted work-family conflict, suggesting that work-family conflict can be explained by the internal conflict associated with the decision to work beyond normal working hours. Implications for organizations, managers, and employees are discussed.

TABLE OF CONTENTS LIST OF TABLES iv LIST OF FIGURES vii ACKNOWLEDGEMENTS viii CHAPTER I: INTRODUCTION 1 Overview 1 Cultural Context 3 Conceptual Definitions 5 CHAPTER II: LITERATURE REVIEW 9 Work-Nonwork Theories ..9 Segmentation 9 Compensation 11 Rational View 13 Negative Spillover 14 Facilitation 15 Work-Nonwork Decisions 16 Decision Process Theory of Work and Family 16 Work-Family Decision Trees..... 18 Work-Nonwork Research Limitations 20 Overview of Model 22 Working Beyond Normal Working Hours 23 Behavioral Reasoning Theory 26 Intentions 27 Global Motives 28 Reasons 30 Life-Role-Values 33 Life-Role-Value Definitions 33 Life-Role-Values and Decisions 35 Summary 39 CHAPTER III: METHODOLOGY 41 Research Participants 41 Measures 46 Reasons For/Against Behavior 47 Global Motives 49 Behavioral Intentions 50 Behavior 50 Life-Role-Values 51 i

TABLE OF CONTENTS (continued) Work-Nonwork Conflict 56 Job/Life Satisfaction 58 Demographics 58 Procedure 59 Data Analysis 60 Hypothesis Testing 60 CHAPTER IV: RESULTS 62 Statistical Analyses for Study Hypotheses ...62 Descriptive Statistics 62 Reliability Analyses 62 Common Method Bias 63 Demographic Control Variables 63 Supplemental Control Variables 64 Tests of Hypotheses 70 Predicting Work-Family Conflict from Behavior 70 Predicting Behavior from Intention 72 Predicting Intention from Global Motives 73 Predicting Global Motives from Reason Variables 74 Predicting Intention from Reason Variables 78 Moderating Effects of Life-Role-Values 81 Mediation Analyses 85 Global Motives -> Intention -> Behavior 86 Reasons -> Global Motives -> Intention 88 Intention -> Behavior -> Conflict 95 Post-hoc Analyses 98 Alternate Predictors of Work-Family/Family- Work Conflict 98 Alternate Predictors of Behavior and Behavioral Intention 100 Alternate Predictors of Global Motives 103 CHAPTER V: DISCUSSION 108 Overview of Findings 109 Work-Family Conflict 110 Behavioral Reasoning Theory 111 Behaviors, Intentions, and Global Motives 112 Behavioral Reasons 117 Life-Role-Values 119 Other Supplemental Findings 120 Limitations and Suggestions for Future Research 123 Practical Implications 127 Summary 130 ii

TABLE OF CONTENTS (continued) REFERENCES 131 APPENDICES 148 Appendix A: Definition of Terms 148 Appendix B: Proposal to Organizations 149 Appendix C: Measures 151 Appendix D: Request for Participation 156 Appendix E: Recommended Text for Organizations to Send to Colleagues 157 Appendix F: Cover Letter to Participants 158 Appendix G: Detailed Research Description and Informed Consent 159 Appendix H: Concluding Message 161 Appendix I: Raffle Details 162 Appendix J: Online Survey 163 Appendix K: Documentation of Institutional Review Board Approval 174 Appendix L: Zero-Order Correlations with List-Wise Deletion 175 Appendix M: Demographic Control Variables 179 Appendix N: Supplemental Control Variables 180 Appendix O: Descriptive Results of "Reasons For" Items 181 Appendix P: Descriptive Results of "Reasons Against" Items 182 iii

LIST OF TABLES Table 1 Demographic Characteristics of All Respondents 43 Table 2 Response Rate by Organization 46 Table 3 Factor loadings, communalities (h2), eigenvalues, percents of variance, and cumulative percents for exploratory factor analysis and oblique rotation on life-role-value items 55 Table 4 Component correlation matrix 55 Table 5 Descriptive Statistics and Zero-Order Correlations with Pair-wise Deletion for All Demographic and Study Variables 66 Table 6 Hypothesis la: Regression testing the effects of Behavior on Work- Family Conflict 71 Table 7 Hypothesis lb: Regression testing the effects of Behavior on Family- Work Conflict 71 Table 8 Hypothesis 2: Regression testing the effects of intention on behavior 73 Table 9 Hypothesis 3: Regression testing the effects of Global Motives on Intention 74 Table 10 Hypothesis 4: Regression testing the effects of Behavioral Reasons on Attitude 76 Table 11 Hypothesis 4: Regression testing the effects of Behavioral Reasons on Perceived Control 77 Table 12 Hypothesis 4: Regression testing the effects of Behavioral Reasons on Subjective Norm 78 Table 13 Hypothesis 5: Regression testing the effects of behavioral reasons on intention (Model 1) 80 Table 14 Hypothesis 5: Regression testing the effects of behavioral reasons on intention (Model 2) 81 Table 15 Hypothesis 6: Regression testing the moderating effects of attitude and work-role-values on intention 83 Table 16 Hypothesis 6: Regression testing the moderating effects of attitude and nonwork-role-values on intention 84 Table 17 Hypothesis 6: Regression testing the three-way interaction of attitude, work-role-values, and nonwork-role-values on intention 85 IV

LIST OF TABLES (continued) Table 18 Direct Effect Analysis: Regression testing the effect of attitude on behavior 87 Table 19 Direct Effect Analysis: Regression testing the effect of perceived control on behavior 88 Table 20 Direct Effect Analysis: Regression testing the effect of reasons for behavior on intention 90 Table 21 Mediation Analysis: Regression testing the effects of reasons for the behavior and attitude on intention 91 Table 22 Mediation Analysis: Regression testing the effects of reasons for the behavior and perceived control on intention 92 Table 23 Direct Effect Analysis: Regression testing the effect of reasons against behavior on intention 93 Table 24 Mediation Analysis: Regression testing the effects of reasons against the behavior and attitude on intention 94 Table 25 Mediation Analysis: Regression testing the effects of reasons against the behavior and perceived control on intention 95 Table 26 Direct Effect Analysis: Regression testing the effect of intention on work-family conflict 96 Table 27 Mediation Analysis: Regression testing the effects of intention and behavior on work-family conflict 97 Table 28 Direct Effect Analysis: Regression testing the effect of intention on family-work conflict 97 Table 29 Exploratory Analysis: Regression testing the effects of behavioral reasoning theory variables and supplemental variables on work-family conflict 99 Table 30 Exploratory Analysis: Regression testing the effects of behavioral reasoning theory variables and supplemental variables on family-work conflict 100 Table 31 Exploratory Analysis: Regression testing the effects of behavioral reasoning theory variables and supplemental variables on behavior 102 v

LIST OF TABLES (continued) Table 32 Exploratory Analysis: Regression testing the effects of behavioral reasoning theory variables and supplemental variables on intention 103 Table 33 Exploratory Analysis: Regression testing effects of behavioral reasoning theory variables and supplemental variables on attitude 105 Table 34 Exploratory Analysis: Regression testing the effects of behavioral reasoning theory variables and supplemental variables on subjective norm 106 Table 35 Exploratory Analysis: Regression testing the effects of behavioral reasoning theory variables and supplemental variables on perceived control 107 vi

LIST OF FIGURES Figure 1 A Decision-Making Model of Working beyond Normal Hours and Work-Family Conflict... 23 Figure 2 Hypothesized Attitude/Intention Relationship by Life-Role-Value 37 Figure 3 Summary of Theoretical Findings 110 vii

ACKNOWLEDGEMENTS There are a number of individuals I would like to acknowledge, for without their help and support this dissertation would not have been possible. First and foremost, I am deeply indebted to Dr. James (Jim) Westaby, my sponsor, advisor, and mentor. It was an honor and privilege to conduct research with the author of the theory on which my study was based. Jim provided me with the tireless guidance, encouragement, and discipline I needed to persevere through the dissertation process. Perhaps more importantly, he pushed me to develop my thought processes - and by extension, myself- more profoundly than I ever knew I could. Since my entry into this program, Jim has made my doctoral education the challenging and rewarding experience I had always hoped it would be. Thank you, Jim. I would also like to thank the other members of my committee, Drs. Patricia Raskin, Stephen Peverly, Jane Monroe, and Steven Schinke. Each shared with me their unique perspectives and feedback, which greatly influenced my thinking and strengthened my final product. The support I have received from my family and friends over the years is no small part of my accomplishment. To start, I am grateful to my parents, Barry and Margo Zusman, for instilling in me a deep value for education from the very beginning. That, along with their unconditional love and pride, helped mold me into who I am today and bring me to this point. My "extended" family has also provided great support to me over the years: Michelle Greenberg Baum and Erica Sorg have stood by me through all the ups and downs, helping me celebrate my intermediate successes, survive my setbacks, and even check my references. Many other friends and colleagues also contributed to this endeavor. I am particularly grateful to Drs. Deanna Siegel Senior, Orla NicDomhnaill, and Matthew Kleinman for blazing the way and encouraging me to meet them on the other side of the finish line. I was also fortunate to have the support of Marina Field and Amy Beacom, who both offered not only their friendship but also to review my writing in a very short timeframe. Last but not least, I owe a special thank you to Dr. Adam Barsky. Adam gave me the inspiration to begin my doctoral studies and the confidence I needed to complete them. viii

1 CHAPTER I: INTRODUCTION Overview The primary purpose of this dissertation was to test a model of life roles that views work-nonwork conflict as the outcome of important individual decisions. As companies continue to develop flexible workplaces and encourage boundariless careers, a decision-making focus may become increasingly important (Poelmans, 2005). The proposed model builds on existing decision-making theories of work and family (e.g., Poelmans, 2005; Powell & Greenhaus, 2006) by applying behavioral reasoning theory (Westaby, 2005). A key component to the predictive power of behavioral reasoning theory is the significance of reasons for and against performing a particular behavior in predicting behavioral intentions. Since it is not realistic to address all possible work/non- work decisions, this study examined a specific critical behavior that is implicated in many issues of work-family conflict: the decision to work beyond normal working hours. By viewing work-family issues through this particular behavioral decision, a better understanding of the process underlying work-family conflict may be gleaned. A secondary purpose of this study was to extend existing literature by exploring the impact of life-role-values on the decision-making process, as it relates to work-family conflict. Carlson and Kacmar (2000) have noted the importance of life-role-values in work-family literature because they may guide and motivate action. Though tradeoffs between work and family are inevitable, they may be mollified in part by viewing the situations as opportunities to become clearer about individuals' values (Friedman & Greenhaus, 2000). In turn, if individuals perceive a fit or congruency between their

2 values and their behaviors, their work and personal lives may complement and enrich each other. If fit is not perceived, the domains will likely compete and conflict will arise (Kossek & Lautsch, 2008). Thus, through further awareness and potential change of one's values, one may be able to strategically make decisions and engage in actions that limit his/her perceived levels of work-family conflict. A final contribution of this study is its inclusion of a more demographically diverse group of participants than previous studies. In their methodological review of work-family research, Casper and colleagues (Casper, Eby, Bordeaux, Lockwood, & Lambert, 2007) discussed that samples are generally homogeneous and fail to include singles, single-parents, extended families, and other non-traditional familial configurations. As a result, it is difficult to evaluate the generalizablity of the existing research to workers who are diverse in terms of these demographic characteristics. To address this, the current study used a diverse sample in which gender, marital status, parental status, and familial arrangements were varied. To build the case for these assertions, the cultural context surrounding work and family in the United States is first described. Definitions1 and theoretical grounding of work-family conflict, behavioral reasoning theory, and life-role-values are then discussed, along with supporting evidence. Finally, existing models of work and family decision-making are reviewed and the proposed model of work and family decisions is presented. 1 For a complete list of terms and definitions used in this paper, please refer to Appendix A.

3 Cultural Context The relationships between work and family have changed over the past several decades, in terms of (1) rates of maternal employment, single-parent households, and dual-income families (Boris & Lewis, 2006); (2) the ways and places work is performed (Perry-Jenkins, Repetti, & Crouter, 2000); and (3) advances in social policies (Frye & Breaugh, 2004). Women, who were previously expected to be home caring for their children in accordance with the male breadwinner/female caregiver model (a.k.a., functionalist theory; Parsons, 1949), are now justified in the workforce with the rationalization that they can simply engage in higher quality, more intensive mothering at the end of the work-day (Boris & Lewis, 2006). As of 2001, 7.5 million of women in the workforce were single mothers, marking a 25 percent rise in this population from 1990 (Kantrowitz, Wingert, Scelfo, Springen, Figueroa, Brant, & Abrams, 2001). Comparative statistics between 1977 and 2002 demonstrate the shift from the male breadwinner/female caregiver model to a dual-earner model: in 1977, less than half of female spouses with children under age 18 living at home were employed, and 74 percent of men felt that they should be out earning the money while the women stayed home to mind the house and children. In 2002, these numbers shifted to 67 percent and 42 percent, respectively (Bond, Thompson, Galinsky, & Prottas, 2002). Technological advances - such as email, cell phones, Blackberries, and telecommuting - have added to this shift in gender responsibilities and to the blurring of work and family roles by creating new ways and places for work to be performed (Desrochers, Hilton, & Larwood, 2005; Friedman, 2005; Milliken & Dunn-Jensen, 2005; Perry-Jenkins et al., 2000). Work and family policies have also seen advances, with

increased family-friendly options such as on-site childcare and flexible work schedules (Frye & Breaugh, 2004), though "the United States continues to have.. .a half-hearted response to the needs of working families..." (Perry-Jenkins et al., 2000, p. 992). At present, in comparison to other industrialized nations, the United States still offers far less public assistance to families with wage-earning and care-giving responsibilities (Boris & Lewis, 2006). For example, there is no provision under current U.S. laws for any paid maternity/paternity leave, though all the countries of the European Union offer at least 14 weeks (Kelly, 2006). In response to these cultural changes, work-family relationships have gained academic attention across of a variety of disciplines, including business management (e.g., Rapoport, Bailyn, Fletcher, & Pruitt, 2002), economics (e.g., Albelda & Tilly, 1997; Schor, 1992), marketing (e.g., Netemeyer, Maxham, & Pullig, 2005), social psychology (e.g., Barnett & Rivers, 1996; Deutsch, 2006), sociology (e.g., Gerstel & Sarkisian, 2006; Heymann, 2000), and women's studies (e.g., Smithson & Stokoe, 2005; Williams, 2000). Research has been conducted qualitatively (e.g., Hochschild, 1997) and quantitatively (e.g., Eby, Casper, Lockwood, Bordeaux, & Brinley, 2005) utilizing a broad range of professions, including engineers, accountants, teachers, executives, salespeople, police officers, prison guards, nurses, social workers, and entrepreneurs (Allen, Herst, Bruck, & Sutton, 2000; Kossek & Ozeki, 1998). These studies have covered a variety of antecedents, outcomes, mediators, and moderators. Examples of antecedents explored include family responsibility, martial status, and job involvement (Eby et al., 2005). The outcomes examined have ranged across the work domain (e.g., job performance, productivity, tardiness, absenteeism, poor morale, organizational commitment, and

turnover) and the nonwork domain (e.g., physical and mental health risks, psychological strain, parenting, and marital/life satisfaction; Allen et al., 2000, Greenhaus & Beutell, 1985; Grzywacz & Marks, 2000; Kelly & Voydanoff, 1985). Mediators include stress, organization fit, and organization commitment (Bacharach, Bamberger, & Conley, 1991; Barnett, Gareis, & Brennan, 1999; Tompson & Werner, 1997). Finally, moderators studied include age, gender, marital status, and parental status (Martins, Eddleston, & Veiga, 2002). Conceptual Definitions Borrowing from Edwards and Rothbard (2000), work is defined as "instrumental activity intended to provide goods and services to support life" (p. 179). Consistent with Eby et al. (2005), the definition of work variables in this dissertation will be associated with gainful employment, but not restricted to full-time employment. Family is defined as membership in and contribution to a social structure comprised of "persons related by biological ties, marriage, social custom, or adoption" (Edwards & Rothbard, 2000, p. 179). In fact, the Latin wordfamilia has its roots in an Indo-European word meaning "house," thus implying that all household residents, blood-related or otherwise, were considered family (Riche, 2006; Rothausen, 1999). In practice, however, only biologically and legally-related social organizations are typically considered families (Rothausen, 1999). Notably, this practice excludes cohabitating couples and singles. The term family may be therefore limiting, if one is trying to distinguish between activities related to work-life versus personal-life. A further potential limitation of family 2 According to U.S. Census data, there was a 72% increase in cohabitation from 1990 to 2000 (Kantrowitz etal.,2001).

6 as a descriptor of one's personal life domain is that it encompasses only the roles associated with family (e.g., spouse, parent, sibling, child, etc.) and not other personal roles, such as student, citizen, or "leisurite" (Super, 1984, p. 76). Cross-national research has demonstrated that there are in fact five major life domains - work, homemaking (i.e., home and family), study, leisure, and community service (i.e., citizenship; Kulenovic & Super, 1995). Thus, in an effort to use terminology that (1) includes the oft-neglected populations of cohabitating couples and singles, (2) includes non-work life activities that are not related to family, and (3) maintains a degree of consistency with concepts used in prior work-family research, the broader term nonwork (Champoux, 1978; Kirchmeyer, 1992; Shamir, 1983; van Steenbergen, Ellemers, Mooijaart, 2007) is used interchangeably with family in this dissertation to describe life activities outside of work. Regardless of the practical scope of "family," the work and family roles interface with each other and can lead to work-family conflict. Work-family conflict is defined as a type of interrole conflict in which the separate "demands of participation in one domain are incompatible with demands of participation in the other domain" (Adams, King, & King, 1996, p. 411). This incompatibility can lead to deleterious consequences, such as (1) job, life, and marital dissatisfaction; (2) job burnout; (3) turnover; (4) depression; (5) unpleasant moods; and (6) health complaints (Adams & Jex, 1999; Allen et al., 2000; Perrewe, Hochwarter, & Kiewitz, 1999). The relationships between work and family are bidirectional. That is, the work role can conflict with the family role or vice versa. When an imbalance in the work- family relationship arises from variables in the family domain, the conflict is classified as family interference with work (FIW; Gutek, Searle, & Klepa, 1991; Kelloway, Gottlieb &

7 Barham, 1999). Likewise, when a variable from the work domain leads to work-family conflict, the conflict is referred to as work interference with family (WIF; Gutek et al., 1991; Kelloway et al., 1999). A meta-analytic review (Mesmer-Magnus & Viswesvaran, 2005) provides support for this discriminant validity. While FIW and WIF are strongly correlated (Frone, Russell, & Cooper, 1992a), WIF is more prevalent than FIW (Gutek et al., 1991; Kinnunen & Mauno, 1998). This emphasis on WIF may be because demands from the work domain are easier to test quantitatively (Kelloway et al., 1999) or because work boundaries are less permeable than family boundaries (Frone, Russell, & Cooper, 1992b; Kinnunen & Mauno, 1998). The present study examined FIW and WIF concurrently. Cross-cultural differences, such as individualism versus collectivism (Hofstede & Bond, 1984; Triandis, 1995), may also account for the prominence of WIF research. Most work-family conflict studies have been conducted in Western cultures, not Eastern (Hill, Yang, Hawkins, & Ferris, 2004). Western cultures tend to fall towards the individualism side of the continuum, as characterized by their emphasis on autonomous, individually-oriented achievement, whereas Eastern cultures focus on meeting expectations of others, particularly the family (Markus & Kitayama, 1991). This distinction may be an important variable in work-family research, because individuals' experience of and values around the work and family domains may be very different across cultures. For example, Larson et al. (2001) found that traditional Indian family roles shaped men's lives so strongly that hardly any relationship between work and home 3 Notable exceptions include a comparison of Chinese and American work-family conflict (Yang, Chen, Choi, & Zou, 2000); a study of work and family lives for men in India (Larson, Verma, & Dworkin, 2001); and Hill and colleagues' (2004) cross-cultural test of work-family relations across 48 countries, 11 of which were Eastern.

8 experiences existed at all. As such, the WIF versus FIW emphasis discussed above may represent the fact that Western society may have a greater focus on work-related commitments whereas Eastern cultures may have a greater focus on family-related commitments. Globalization has led to a distinct increase in multinational corporations (Friedman, 2005; Hill et al., 2004), so a cross-cultural sample might offer valuable contributions to the literature. However, that endeavor is beyond the scope of the current research.

9 CHAPTER II: LITERATURE REVIEW Work-Nonwork Theories Many theories have been developed to help explain conflict between the work and family domains. Prevalent work-family conflict theories are: segmentation (Ashforth, Kreiner, & Fugate, 2000; Bulger, Matthews, & Hoffman, 2007; Lambert, 1990; Zedeck & Mosier, 1990), compensation (Lambert, 1990; Zedeck & Mosier, 1990), rational view (Duxbury, Higgins, & Lee, 1994), and negative spillover (Bartolome & Evans, 1980; Lambert, 1990; Maume & Houston, 2001; Zedeck & Mosier, 1990). An overview of each of these theories is presented below, along with the limitations. Additionally, a brief review of work-family facilitation is provided. Segmentation The segmentation concept asserts that individuals can and do compartmentalize the different aspects of their lives, so that each domain is bounded and does not impact the other. Each domain is considered independently, and satisfaction can be derived from one, the other, or both (Burke, 1986). This perspective is typically taken by socio- biologists and evolutionary psychologists, with the view that successful men are those who are aggressive and competitive by nature and do not invest heavily in their offspring while successful women, contrastingly, are those who are nurturing and caring by nature, and who do invest heavily in their offspring (Barnett & Hyde, 2001). Deviations from this model are then believed to lead to mental health consequences (Barnett & Hyde, 2001).

10 Regardless of the gender implications, segmentation is an active, intentional effort to keep work and family completely separate so that they do not affect each other and so that work-family conflict is avoided (Lambert, 1990). To prevent negative effects from occurring, work and family roles are separated - theoretically by individuals leaving their work at work and their personal life at home. While this strategy may be effective in the short-term, long-term utilization is likely to be impractical, if not impossible (Burke, 1986; Lobel, 1991). Rather, attempted segmentation may lead to increased conflict that will only be exacerbated as time goes by. Segmentation is conceptually addressed by work-family border theory (Clark, 2000) and by boundary theory (Ashforth et al., 2000). These theories aver that employees develop boundaries or borders around their work and nonwork domains that vary in strength. Strength is determined by boundary permeability (i.e., the extent to which elements from one domain are found in the other domain) and by boundary flexibility (i.e., the extent to which a boundary from one domain can be relaxed to meet the demands of the other domain; Bulger et al., 2007). It is the strength of the boundaries that determines the outcomes of the interactions between the two domains, and determines whether or not an individual experiences work-family conflict. In support of this theory, Clark (2002) found that border permeability was positively correlated with work-family conflict and that border flexibility was negatively correlated with work- family conflict.4 Additional research has also shown that low boundary flexibility and high boundary permeation lead to higher interference between work and nonwork domains (Bulger et al., 2007). 4 Correlations with work-family conflict were significant at the 0.05 level for family flexibility (r = -0.27), work flexibility (r = -0.16), and family permeability (r = 0.19). Though the correlation for work permeability was also positive (r = 0.07), it was not significant (Clark, 2002).

11 Boundary theory research focuses on the ways people develop mental barriers around their life roles, specifically by examining the temporal and spatial boundaries around the domains and how they are performed. Though Hall and Richter (1988) concluded that "people have a preference for psychological separation from work and home parallel to the physical separation" (p. 217), this area has not been closely examined (Rothbard, Phillips, & Dumas, 2005). This omission may be due to the implicit assumption that individuals' desires are aligned with their boundary management strategies for the segmentation or integration of their life roles. However, this assumption does not address individuals' underlying values. Additionally, it does not examine whether or not boundary management strategies are developed with the intention of effectively supporting individuals' values. Compensation Unlike segmentation theory, compensation theory assumes a reciprocal relationship between the work and family domains (Champoux, 1978; Staines, 1980). The idea is that individuals try to compensate for lack of satisfaction in one domain by seeking more satisfaction from the other (Burke, 1986; Lambert, 1990). For example, an individual may rest at home to compensate for fatigue caused at work (Zedeck & Mosier, 1990). Thus, an inverse relationship between work and nonwork is theorized, with individuals making differential investments to the various domains; work-family conflict occurs when the balance between the two domains is not optimal. Though this utilitarian view has been employed mostly to understand individuals' reactions to unfulfilling or unsatisfying jobs (Lambert, 1990), it also helps explain why a worker might become

12 more involved in occupational work if he/she is experiencing problems at home. The individual would be compensating for the lack of satisfaction from home life (e.g., a rocky marriage) by seeking more satisfaction at work. If this compensatory strategy is successful, work-family conflict would be avoided (Lambert, 1990). Empirical work by Champoux (1978) demonstrated that individuals who use a compensatory approach for dealing with their work and nonwork domains can be classified into two groups: compensatory-work-oriented or compensatory-nonwork- oriented. The former group consisted of individuals who received enough variety, creativity, and challenge from their work role that they didn't need to seek additional gratification form their nonwork activities. The latter group was characterized by individuals who experienced deprivations at work and sought to compensate for that deficit by finding additional gratification in their nonwork roles.5 Similarly, Zedeck and Mosier (1990) identified two components of compensation: supplemental compensation and reactive compensation. Supplemental compensation is described as the sought-after experiences, psychological states, and behaviors that are absent in work activities and fulfilled by nonwork activities. Reactive compensation is described as the deprivations caused by the work domain that can be made up for in nonwork activities. A limitation of compensation theory is that it assumes an inverse relationship between work and nonwork satisfaction. The possibility of experiencing either satisfaction or dissatisfaction in both domains concurrently is not explained. 5 Subsequent research comparing compensation theory and spillover theory found more support for spillover theory (Staines, 1980).

13 Rational View The rational view of work-family conflict hypothesizes that the amount of perceived conflict an individual experiences increases proportionally with the amount of time he/she spends in either the work or family domain (Duxbury et al., 1994; Gutek et al., 1991). Thus, the more time one spends on the roles associated with each of these domains, the more conflict this individual will perceive. For example, people will likely experience more work-family conflict if they spend 80 hours per week fulfilling obligations of their work role than if they spend only 40 hours per week. The foundation of the rational view is similar to the "scarcity hypothesis", which states that human energy is of a fixed and limited quantity (Goode, 1960). Stated differently, conflict arises between work and family because there are not enough resources (e.g., time, attention, energy, affect, etc.) to fulfill all obligations in each domain (Duxbury et al., 1994; Greenhaus, Bedeian, & Mossholder, 1987). Further, the total time spent performing these roles is positively associated with role overload - when multiple roles cannot be performed adequately because the time and energy required for the associated activities is too great (Duxbury et al., 1994). Thus, extensive commitments in one domain will necessarily detract from the resources dedicated to the other domain. The rational view also posits that men will devote more hours to work, while women will spend more hours on family activities, which can lead to asymmetry in the way this conflict is experienced. This is because women continue be more active caregivers than men (e.g., women dedicate many more hours to household activities than men; Gutek et al., 1991), and men continue to focus primarily on work matters (McElwain, Korabik, & Rosin, 2005; Vaananen, Kumpulainen, Kevin, Ala-Mursula,

Full document contains 195 pages
Abstract: The ways in which employees' work and nonwork roles conflict with each other have been studied in numerous contexts. However, research has not focused on the specific behavioral decisions that may influence work-family conflict. Therefore, the general purpose of this dissertation was to examine work-family conflict as an outcome of the decision to work beyond normal working hours, through the application of behavioral reasoning theory (Westaby, 2005). Furthermore, life-role-values were introduced as potential moderators of the decision-making process. Regarding specific hypotheses, this study first predicted that the decision to work beyond normal working hours would be related to work-family conflict. This hypothesis is in line with expectancy violation theories, which posit that expectations guide behavior and that negative arousal results when these expectations are broken. Next, behavioral reasoning theory provided a framework to explore the factors that contributed to the decision to work beyond normal working hours. Based on this theory, it was proposed that behavioral reasons (for and against the behavior) and global motives (i.e., attitude toward the behavior, subjective norm, and/or perceived behavioral control) would predict intention which in turn would predict behavior. Finally, the moderating effect of life-role-values on the relationship between attitude and behavioral intention was examined. It was hypothesized that a stronger link between these variables would emerge when work-role-values were strong than when they were weak. A cross-sectional survey design was utilized, involving 336 participants from 11 organizations. Hypotheses were tested using hierarchical regression analyses. Overall, it was found that: (1) the behavior of working beyond normal work hours predicted work-family conflict; (2) intention predicted behavior; (3) attitude predicted intention; (4) behavioral reasons predicted global motives; and (5) behavioral reasons independently predicted intention, over and above that explained by constructs in the theory of planned behavior. Against predictions, life-role-values did not moderate the attitude-intention relationship. Supplemental analyses also found that behavioral reasons for and against working beyond normal working hours directly predicted work-family conflict, suggesting that work-family conflict can be explained by the internal conflict associated with the decision to work beyond normal working hours. Implications for organizations, managers, and employees are discussed.