USF Libraries
USF Digital Collections

You can't always get what you want, but does it matter? the relationship between pre-child preferences and post-child a...

MISSING IMAGE

Material Information

Title:
You can't always get what you want, but does it matter? the relationship between pre-child preferences and post-child actual labor division fit and well-being
Physical Description:
Book
Language:
English
Creator:
Shockley, Kristen
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Division of household labor
Paid labor
Person-environment fit
Dual-earner couples
Dual-career couples
Work-family management
Career prioritization
Dissertations, Academic -- Psychology -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: Significant shifts in social ideology and legislation have brought about considerable changes in work and family dynamics in the Western world, and the male as breadwinner-wife as homemaker model is no longer the norm. However, despite increasingly gender egalitarian ideals, the division of labor among dual-earner couples tends to adopt a "neo traditional" once children are born, where women devote more time to family labor and men spend more time in paid employment Although asymmetrical divisions of labor have clear workplace and societal consequences in terms of women's earnings, organizational advancement, and inequality, the effects on individual well-being are not well understood. The purpose of the present study was to apply the theoretical lens of person-environment fit to examine how misfit between dual-earner couples' pre-child division of labor preferences and post-child actual divisions of labor relate to affective (career, marital, and family satisfaction) and health-related (depression and physical health symptoms) well-being. Additionally, several conditions were posited to temper the strengths of these relationships (domain centrality, gender, voice in division of labor decision making, and satisfaction with the current division of labor). Participants were 126 dual-earner couples with small children, and hypotheses were testing using polynomial regression analyses. The results suggested that congruence between an individual's own pre-child desires for the division of paid labor and the actual post-child division of paid labor relates to his/her own career and marital satisfaction, depression, and physical health symptoms. Congruence in the family domain is also important, as desire-division of family labor fit related to affective sentiments toward family and one's spouse. With the exception of career satisfaction, these relationships were curvilinear, such that deviations in either direction from perfect fit related to poorer well-being. On the other hand, there was little evidence for spousal effects, as dual-earner well-being did not relate the congruence between division of labor abilities and spousal demands. Finally, evidence of moderation was only found in a few cases, and none were consistent with prediction, highlighting the need for future research on the contextual conditions of P-E fit in the dual-earner context.
Thesis:
Dissertation (PHD)--University of South Florida, 2010.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Kristen Shockley.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004543
usfldc handle - e14.4543
System ID:
SFS0027858:00001


This item is only available as the following downloads:


Full Text

PAGE 1

You can’t always get what you want, but does it matter? The rela tionship between prechild preferences and post-child actual labor division f it and well-being by Kristen M. Shockley A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Psychology College of Arts and Sciences University of South Florida Major Professor: Tammy D. Allen, Ph.D. Walter C. Borman, Ph.D. Michael T. Brannick, Ph.D. Judith B. Bryant, Ph.D. Joseph A. Vandello, Ph.D. Date of approval: June 29, 2010 Keywords: division of household labor, paid labor, person-environment fit, dual-earner couples, dual-career couples, work-fam ily management, car eer prioritization Copyright 2010, Kristen M. Shockley

PAGE 2

Acknowledgements Several people have supported me throughout my educational journey. First, I would like to acknowledge Dr. Tammy Allen for her unbelievable career and psychosocial mentoring throughout the past five years. I cannot imagine making it through graduate school without her! Second, thank you to my dissertation committee for your time and feedback. Third, I would like to thank my parents for their continual support and for raising me in a way that gave me confidence to pursue a Ph.D. I would also like to thank my sisters, Lauren and Shannon, for th eir support and help gathering emails for data collection. Fourth, I would like to thank my boyfri end, Andy, for his continual guidance and assurance. Finall y, I am indebted to all of my dissertation participants and friends who helped obtain those participants Without them, this study would not have been possible.

PAGE 3

i TABLE OF CONTENTS List of Tables iv List of Figures vii Abstract vi Chapter One – Introduction 1 Division of Family and Paid Labor 3 Review of Previous Desire-Divis ion of Labor Fit and Well-being Research 6 Gaps and Limitations of Previous Research 7 Theoretical 7 Methodological 11 Current Study 13 Person-Environment Fit Theory 14 Fundamental premise 14 Types of P-E fit 16 Measurement Application of P-E Fit Theory to Desire-Division of Labor Fit and Well-Being 17 Theoretical Framework: Needs-supplies Fit 17 Defining Needs and Supplies 18 Needs-supplies Fit and Well-being 19 Satisfaction 20 Depression and Physical Health Symptoms 24 Moderators 27 Domain Centrality 27 Gender 29 Voice in Division of Labor Decisions 30 Theoretical Framework: Demands-abilities Fit 32 Defining Demands and Abilities 33 Demands-abilities Fit and Well-being 34 Gender 37 Satisfaction With Current Division of Labor 39 Chapter Two – Method 41 Participants and Procedures 41 Measures 44

PAGE 4

ii Pre-child Desires for Division of Labor 45 Post-child Actual Division of Labor 48 Career satisfaction 50 Family satisfaction 50 Marital satisfaction 50 Depression 51 Physical Health Symptoms 51 Domain Centrality 52 Gender 52 Voice in Division of Labor Decision Making 52 Satisfaction With Current Division of Labor 53 Control Variables 54 Demographic Information 55 Inclusion criteria 55 Chapter Three – Results 56 Preliminary Analyses 56 Data Screening 56 Assumption Testing 56 Analysis of Hypotheses 58 Hypothesis Testing 60 Supplementary Analyses 70 Chapter Four – Discussion 75 Summary and Interpretation of Results 75 Needs-supplies Fit and Satisfaction 75 Needs-supplies Fit and Mental and Physical Health 78 Supplementary Analysis of Fit Line in Needs-supplies Fit and Well-being Relationships 80 Moderators of Needs-supp lies Fit and Well-being 83 Demands-abilities Fit and Well-being 89 Moderators of Demands-abi lities Fit and Well-being 92 Theoretical Implications 96 Practical Implications 99 Limitations 100 Future Directions 104 Conclusion 107 References 108 Appendices 199 Appendix A: Pre-child Desires for Division of Labor Scale Items 200 Appendix B: Post-child Actual Division of Labor Scale Items 203 Appendix C: Career Satisfaction Scale Items 204 Appendix D: Family Satisfaction Scale Items 205 Appendix E: Marital Satisfaction Scale Items 206

PAGE 5

iii Appendix F: Depression Scale Items 207 Appendix G: Physical Health Symptoms Scale Items 208 Appendix H: Domain Centrality Scale Items 209 Appendix I: Voice in Division of Labor Decision Making Scale Items 210 Appendix J: Satisfaction With Cu rrent Division of Labor Scale Items 212 Appendix K: Family Responsibility Scale Items 213 Appendix L: Demographics Scale Items 214

PAGE 6

iv List of Tables Table 1 Summary of previous desi re-division of labor fit and wellbeing research 137 Table 2 Recruitment sources and response rate information for initial respondents 139 Table 3 Recruitment sources and response rate information for spouse respondents 139 Table 4 Descriptive statistics of study variables for Partner A 140 Table 5 Descriptive statistics of study variables for Partner B 142 Table 6 Descriptive statistics of study variables by gender 144 Table 7 Source of reports for hypotheses 144 Table 8 Results of confirmatory factor analyses 145 Table 9 Results of exploratory f actor analysis and inter-item correlations for voice in divisi on of labor decision making (Partner A) 146 Table 10 Results of exploratory factor analysis and inter-item correlations for satisfaction w ith current division of labor (Partner B) 146 Table 11 Correlations be tween study variables 147 Table 12 Correlations betw een study variables used in hypothesis testing 156 Table 13 Polynomial regression equa tions regressing Partner A’s wellbeing on Partner A’s paid labor needs and supplies (Hypotheses1, 3, 5, 7) 157 Table 14 Polynomial regression equa tions regressing Partner A’s wellbeing on Partner A’s family labor needs and supplies (Hypotheses 2, 4, 6, 8) 158

PAGE 7

v Table 15 Hierarchical moderated re gression involving the regression of Partner A’s well-being on th e interaction between career centrality and Partner A’s pa id labor fit (Hypothesis9) 159 Table 16 Hierarchical moderated re gression involving the regression of Partner A’s well-being on the interaction between family centrality and Partner A’s fam ily labor fit (Hypothesis10) 160 Table 17 Hierarchical moderated re gression involving the regression of Partner A’s well-being on the interaction betw een gender and Partner A’s paid labo r fit (Hypothesis11) 161 Table 18 Hierarchical moderated re gression involving the regression of Partner A’s well-being on the interaction betw een gender and Partner A’s family labor fit (Hypothesis12) 162 Table 19 Hierarchical moderated re gression involving the regression of Partner A’s well-being on the inte raction between voice in paid labor decision making and Partner A’s paid labor fit (Hypothesis 13) 163 Table 20 Hierarchical moderated re gression involving the regression of Partner A’s well-being on the interaction between voice in family labor decision making and Partner A’s paid labor fit (Hypothesis 14) 164 Table 21 Polynomial regression equa tions regressing Partner B’s wellbeing on Partner B’s paid labor demands and abilities (Hypotheses 15, 17, 19) 165 Table 22 Polynomial regression equa tions regressing Partner B’s wellbeing on Partner B’s family labor demands and abilities (Hypotheses 16, 18, 20) 166 Table 23 Hierarchical moderated re gression involving the regression of Partner B’s well-being on the interaction betw een gender and Partner B’s paid labo r fit (Hypothesis 21) 167 Table 24 Hierarchical moderated re gression involving the regression of Partner B’s well-being on the interaction betw een gender and Partner B’s family la bor fit (Hypothesis 22) 168 Table 25 Hierarchical moderated re gression involving the regression of Partner B’s well-being on the interaction between satisfaction with current division of paid la bor and Partner B’s paid labor 169

PAGE 8

vi fit (Hypothesis 23) Table 26 Hierarchical moderated re gression involving the regression of Partner B’s well-being on the interaction between satisfaction with current division of family labor and Partner B’s family labor fit (Hypothesis 24) 170 Table 27 Summary of hypothesis testing 171 Table 28 T-test results comparing Part ner As that not are not matched to partner As that are matched preand postrandomization. 173 Table 29 Description of shape of fit line for significant moderator relationships 174

PAGE 9

vii List of Figures Figure 1 Model of division of family labor hypotheses 175 Figure 2 Model of division of paid labor hypotheses 176 Figure 3 Predicted pattern of rela tionships in Hypotheses 1, 2, 3, 4, 15, and 16 177 Figure 4 Predicted pattern relation ships in Hypotheses 5, 6, 7, 8, 17, 18, 19, 20 177 Figure 5 Predicted pattern of rela tionships for parts a and b of Hypotheses 9 and 10 178 Figure 6 Predicted patte rn of relationships for parts c and d of Hypotheses 9 and 10 178 Figure 7 Predicted pattern of rela tionships for parts a and b of Hypotheses 11, 12, 13, 14, and part a of Hypotheses 23, and 24 179 Figure 8 Predicted pattern of rela tionships for parts c and d of Hypotheses 11, 12, 13, 14, and parts b and c of Hypotheses 23, and 24 180 Figure 9 Predicted pattern of relati onships for part a Hypotheses 21 and 22 181 Figure 10 Predicted patte rn of relationships for parts c and d of Hypotheses 21 and 22 181 Figure 11 Response surface for paid labor need-supplies fit and career satisfaction for Partner A (Hypothesis1) 182 Figure 12 Response surface for family labor need-supplies fit and family satisfaction for Partner A (Hypothesis2) 182 Figure 13 Response surface for paid labor need-supplies fit and marital satisfaction for Partner A (Hypothesis3) 183 Figure 14 Response surface for family labor need-supplies fit and marital 183

PAGE 10

viii satisfaction for Partner A (Hypothesis4) Figure 15 Response surface for paid la bor need-supplies fit and depression for Partner A (Hypothesis5) 184 Figure 16 Response surface for paid labor need-supplies fit and physical health symptoms for Partner A (Hypothesis7) 184 Figure 17 Response surface for moderating role of career centrality between the relationship of pa id labor need-supplies fit and marital satisfaction for Partner A (Hypothesis9b) 185 Figure 18 Response surface for moderating role of gender between the relationship of paid labor need-s upplies fit and physical health symptoms for Partner A (Hypothesis11d) 186 Figure 19 Response surface for modera ting role of voice in division of paid labor decision making betw een the relationship of paid labor need-supplies fit and marital satisfaction for Partner A (Hypothesis13b) 187 Figure 20 Response surface for family labor demands-abilities fit and physical health symptoms for Partner B (Hypothesis20) 188 Figure 21 Response surface for moderating role of gender between the relationship of paid labor demands-abilities fit and marital satisfaction for Partner B (Hypothesis21a) 189 Figure 22 Response surface for moderating role of gender between the relationship of family labor dema nds-abilities fit and depression for Partner B (Hypothesis22b) 190 Figure 23 Response surface for moderating role of gender between the relationship of family labor de mands-abilities fit and physical health symptoms for Partner B (Hypothesis22c) 191 Figure 24 Response surface for modera ting role of satisfaction with current division of paid labor be tween the relationship of paid labor demands-abilities fit and marital satisfaction for Partner B (Hypothesis23a) 192 Figure 25 Response surface for modera ting role of satisfaction with current division of paid labor be tween the relationship of paid labor demands-abilities fit and depression for Partner B (Hypothesis23b) 193

PAGE 11

ix Figure 26 Response surface for modera ting role of satisfaction with current division of paid labor be tween the relationship of paid labor demands-abilities fit and physical health symptoms for Partner B (Hypothesis23c) 194 Figure 27 Response surface for modera ting role of satisfaction with current division of family labor between the relationship of family labor demands-abilities fit and depression for Partner B (Hypothesis24b) 195 Figure 28 Response surface for paid labor need-supplies fit and career satisfaction for Partner A (Hypothesis1) 196 Figure 29 Response surface for paid labor need-supplies fit and marital satisfaction for Partner A (Hypothesis3) 196 Figure 30 Response surface for paid la bor need-supplies fit and depression for Partner A (Hypothesis5) 197 Figure 31 Response surface for paid labor need-supplies fit and physical health symptoms for Partner A (Hypothesis7) 197 Figure 32 Response surface for family labor need-supplies fit and marital satisfaction for Partner A (Hypothesis4) 198 Figure 33 Response surface for family labor need-supplies fit and physical health symptoms for Partner A 198

PAGE 12

x You can’t always get what you want, but does it matter? The relationship between prechild preferences and post-child actual labor division f it and well-being Kristen M. Shockley Abstract Significant shifts in social ideology and legislation have brought about considerable changes in work and family dynamics in the Western world, and the male as breadwinner–wife as homemaker model is no longer the norm. However, despite increasingly gender egalitarian ideals, the division of la bor among dual-earner couples tends to adopt a “neo traditional” once children are born, where women devote more time to family labor and men spend more time in paid employment Although asymmetrical divisi ons of labor have clear workplace and societal consequences in terms of women’s ear nings, organizational advancement, and inequality, the effects on indi vidual well-being are not well understood. The purpose of the present study was to apply the theoretical lens of pers on-environment fit to examine how misfit between dual-earner couples’ prechild division of la bor preferences and post-child actual divisions of labor relate to affective (career, marital, and family satisfaction) and health-related (depression and physical health symptoms) well-being. Additionally, several conditi ons were posited to temper the strengths of these relationships (domain centrality, gender, voice in division of labor decision making, and satisfaction with the curren t division of labor). Participants were 126 dual-earner coupl es with small children, and hypotheses were testing using polynomial regression analyses. The results suggested that

PAGE 13

xi congruence between an individual’s own pre-child desires for the division of paid labor and the actual post-child division of paid la bor relates to his/her own career and marital satisfaction, depression, and physical health symptoms. Congruence in the family domain is also important, as desire-divisi on of family labor fit related to affective sentiments toward family and one’s spouse. With the exception of career satisfaction, these relationships were curvilinear, such that deviations in either direction from perfect fit related to poorer well-being. On the othe r hand, there was little evidence for spousal effects, as dual-earner well-being did not relate the congruence between division of labor abilities and spousal demands. Finall y, evidence of moderation was only found in a few cases, and none were consistent with prediction, highlighting the need for future research on the contextual conditions of P-E fit in the dual-earner context.

PAGE 14

1 Chapter One Introduction In the early and mid twentieth century, work and family domains were largely separate entities, normally divided along gender lines. Men were the sole earners, supporting the family financially, whereas wo men maintained primary responsibility on the domestic front. The feminist movement and legislation of the 1960s brought about considerable changes in work and family dynamics, as more and more women joined their husbands in the paid workforce (P adavic & Reskin, 2002). This trend has continued, and current figures estimate that 79% of married/partnered employees are members of dual-earner couples (Galinksy, Aumann, & Bond, 2008). Unlike single-earners, dual-earner couples are faced with the challenge of jointly dividing labor in both work and family domai ns. Although there is considerable variety in attitudes about the way labor should be divided, many men and women endorse egalitarian ideals (Galinsky et al., 2008). However, upon the “launching stage” of life, when children are young, most dual-earner couple s are forced to conf ront the realities that these ideals may not be realized (M oen & Roehling, 2005). Even when couples hold egalitarian attitudes, they tend to a dopt a “neo traditional” division of labor (Clarkberg & Moen, 2001; Moen & Yu, 2000), where women devote more time to family labor and men spend more time in pa id employment (Coltrane, 2000; Moen & Roehling, 2005; U.S. Department of Labor, 2007). The career of the husband is likely to be favored over that of the wife (Pixley & Moen, 2003). This asymmetry is a

PAGE 15

2 persistent finding even when spouses’ have si milar levels of edu cation and occupational prestige (Johnson, Johnson, & Liese, 1991; McNeil & Sher, 1999). Although the asymmetrical division of la bor has clear workplace and societal consequences in terms of women’s ear nings, organizational advancement, and inequality (cf., Padavic & Reskin, 2002), the effects on individual well-being are not well understood. That is, alt hough it is assumed that discre pancies between division of labor preferences and the act ual division of labor result s in detrimental effects on individual well-being (e.g., Singley & Hines, 2005), this relationship has not been comprehensively tested. Thus, the extent th at unmet desires are truly a “problem” for contemporary dual-earners remains largely unknown. The purpose of the present study is to apply the theoretical lens of person-en vironment fit (P-E fit theory, French, Caplan, & Harrison, 1982) to extend the small extant body of research on this topic. Specifically, focusing on both the paid and family labor domains independently, the present study examines how mi sfit between dual-earner couple s’ pre-child division of labor preferences and the postchild actual division of labor relate to affective and health-related well-being. This study makes three major contributions to the literature. First, it examines division of labor, an issue that has received very little empirical a ttention, despite its frequent mention in the industrial or ganizational psychology and organizational behavior literatures as a cruc ial process in work-family interactions. Second, the present investigation is more comprehensive than its few predecessors (Khazan, McHale, & Decourcey, 2008; Milkie, Bianchi, Mattingl y, & Robinson, 2002; Pe rry-Jenkins, Seery, & Crouter, 1992; Ross, Mirowsky, & Huber, 1983) as it examines both paid and family

PAGE 16

3 labor, assesses misfit in an all-encompa ssing manner, includes hypotheses concerning the marital dyad, examines several forms of well-being, and incorporates the role of moderators. Finally, this study is more me thodologically sound than previous work through its use of multi-dimensional division of labor scales, and it is more statistically sound through its use of polynomial regressi on to capture the fit concept. Before introducing the theory and hypot heses for the present study, a review of the general constructs of paid and family labor is provided, followed by a comprehensive summary of previous research sp ecific to desire-division of labor fit and well-being. Lastly, based on this summary, I point out the ga ps and limitations in the extant literature that the present study aims to address and improve. Division of Family and Paid Labor Family labor, or “unpaid work done to maintain family members and/or a home” (Shelton & John, 1996, p. 300), is an integral pa rt of human existen ce (Coltrane, 2000). There are three distinct types of family la bor: household tasks, childcare, and emotion work (Coltrane, 2000; Shelton & John, 1996). Household tasks include routine tasks (e.g., meal preparation, housecleaning, s hopping for groceries and other household items, and laundering clothes) as well as re sidual tasks (e.g., household repairs, yard care, driving other people, or paying bills ) (Blair & Lichter, 1991; Coltrane, 2000; Robinson & Godbey, 1997). Childcare involv es caring for and s upervising children, and emotion work refers to the less tangibl e tasks of maintaining family members’ psychological well-being through the provi sion of emotional support (Erickson, 1993; Hochschild, 1989). Assessment of how family la bor is divided is typically temporal in nature, as most researchers ask participants to indicate the raw or relative amount of

PAGE 17

4 time spent on all tasks (Shelton & John, 1996). Less frequently, researchers focus on who performs specific tasks rather than the amount of time, which allows for weighting of tasks by difficulty or disagr eeableness (Berk & Berk, 1979). Paid labor represents work that is conducted for individual s external to the family in exchange for compensation. There are several features of paid labor that determine the amount of investment and resour ces associated with an individual’s work role (Rapoport & Rapoport, 1969). For example, paid labor roles differ in the extent of commitment, time, and education required, as well as the pay a nd prestige offered. Additionally, there is variety in the connectedness of work roles over time, as one may hold a sequence of disjointed jobs or severa l related jobs that de velop upon one another and lead to a common occupational goal (R apoport & Rapoport, 1969). Lastly, from the individual’s perspective, work differs in the extent that it is personally meaningful and salient to self-identity (Lobel & St. Clair, 1992). Assessment of the division of paid labor is more complicated than that of family labor, due in part to the fact that partners typically have di stinct work roles, whereas the family role is contained in a shared space. In fact, the nomenclature “division of paid labor” is rarely used in the literature, but it is indirectly assessed in other contexts, such as the relative incomes or time spent at wo rk of each spouse (e.g., Barnett, Gareis, & Brennan, 2009; Brines, 1993; Cunningham, 200 7; Deutsch; Roksa, & Meesek, 2003; Jacobs & Gerson, 2001; Kanter, 1977; Raley, Mattingly, Bianchi, 2006). Researchers have also focused on the relative priority of each partner’s career, as couples are often faced with crucial work and family decisions that advance one partner’s career at the expense of the other’s career (Mincer, 1978; Pixley, 2008; Pixley & Moen, 2003).

PAGE 18

5 Over time, this pattern has a cumulative in fluence on work trajectories and the way labor is ultimately divided among couples (Pixley, 2008). Moreover, several theoretical models have been proposed to understand the process by which divisions of family and pa id labor emerge. The four most common explanations are economic, family power, ro le and social constructionist, and time availability theories (Becker, 1993; Sh elton & John, 1996). Economic models suggest that marital partners divide labor in a way that maximizes net family gain, even if it results in personal loss for one partner (e.g., Becker, 1991; Mincer, 1978). Family power theories focus on the relative resources (e.g., income, gender) of each partner as a source of power within social exchange rela tionships. The spouse with the most power has the most weight in making career deci sions, and is likely to use this power to impose outcomes to further his or her own goa ls (cf., McDonald, 1980) Role and social constructionist theories hone in on the socia lization processes that encourage individuals to behave according to gender norms and how institutionalized structures reinforce gendered visions and inequali ties (e.g., Hochschild, 1989). Finally, time availability models (e.g., Kamo, 1988) focus specifical ly on family labor, positing that the amount of time a person has available, largely a function of the amount of time spent in paid work, determines family labor participation. No one perspective is deemed empirically superior to the others (Coltrane, 2000). Some support has been found for all four models, although gender remains a more important determinant of family labor than any other factor (Shelton & John, 1996).

PAGE 19

6 Review of Previous Desire-Division of L abor Fit and Well-Being Research Having defined the broad constructs of family and paid labor divisions and associated theories, I turn to a discussion of previous resear ch that has investigated the congruence between desires for division of paid and family labor and actual division of labor in relation to well-being. It is important to note that this review focuses only on studies that examine desires or preferences toward division of labor, alth ough there are several studies devoted to rela ted constructs, namely division of labor expectations (e.g., Goldberg & Perry-Jenkins, 2004; Kalmu ss, Davidson, & Cushman, 1992; Nicolson, 1990; Ruble, Fleming, Hackel, & Stangor, 1988; Van Egeren, 2004) and gender role ideology (Brennan, Barnett, & Gareis, 2001; Kroska, 2009; Lavee & Katz, 2002; MacDermid, Huston, & McHale, 1990; Mickelson, Claffey, & Williams, 2006; McHale & Crouter, 1992; Pina & Be ngston, 1993; Rochlen, McKelley, Suizzo, & Scaringi, 2008; Sagara, Ito, & Ikeda, 2006). Although these terms are sometimes used in terchangeably in th e literature, there are meaningful differences in the definitions Specifically, desires represent idealistic views about the future, expectations are beliefs about how the future will actually unfold, and gender role ideology is an overa rching philosophy about the extent that gender should determine an individual’s wo rk and family roles. Although desires and expectations can be closely tied, they may al so be entirely distinct, as expectations incorporate external factors, such as spousal desires or societal norms. Similarly, desires and ideologies are not always perfec tly aligned. For example, it is possible to have a general egalitarian ideology but hold a more traditional attitude when thinking about one’s own family life (Hood, 1986; Kroska, 1997; Loscocco & Spitze, 2007).

PAGE 20

7 The focus of the present study is on desires ra ther than on expectat ions or ideologies because desires are more repr esentative of what an indi vidual specifically wants for his/her own life, without regard for external constraints. As such, unmet desires should have greater implications for well-being than unmet expectations or behavior that is inconsistent with gender role ideologi es (Khazan, McHale, & Decourcey, 2008). Only one study was located that investig ated how fit between pre-child division of labor preferences and post-child actual division of labor relates to well-being. However, four other studies we re identified that evaluated the same constructs, without a focus on the transition to parenthood. Instead in these studies, th e fit between current desires and the current division of labor was examined in re lation to well-being. Because of the conceptual similarities, all fi ve studies are included in the review. Of these five studies, three focused on the paid work role, one study explored family labor, and one study examined division of labor in both domains. Details about each study, including the publication outlet, variables, samp le, analyses, and result s, are provided in Table 1. This review reveals several theo retical and methodologi cal opportunities for improvement and extension of this research topic. Gaps and Limitations of Previous Research Theoretical. In many respects, the present lit erature lacks a comprehensive theoretical examination of the link between desire-division of labor fit and well-being. First, the meaning of division of paid and fa mily labor varies across studies and is often narrowly defined and operationalized. With regard to paid labor, studies either examined the division of paid labor based on the wife’s employment status (i.e., PerryJenkins et al., 1992; Ross et al., 1983) or on th e relative incomes of each spouse (i.e.,

PAGE 21

8 Loscocco & Spitze, 2007; Milkie et al., 2002) Each of these va riables addresses a unique aspect of the work domain, although none are alone sufficient to explain the division of paid labor entirely (Gutek, Nakamura, & Nieva, 1981). Family labor division studies (i.e., Khazan et al., 2008; Milkie et al., 2002) also suffer from a limited scope, as they only examined childcare-related tasks, to the neglect of other family labor components, such as household tasks and emotion work (Coltrane, 2000; Shelton & John, 1996). Further exacerbating the issue is the narrow measurement that accompanies these narrow definitions. That is, with such a cons tricted definition of division of labor, most previous researchers have deemed a single ite m adequate to captur e the construct (e.g., "What proportion of childcare-related tasks ar e you responsible for?" or "Is your wife employed?"). From a measurement standpoint, si ngle-item scales are less preferable to multi-item scales, as they have considerable random measurement e rror, poor ability to discriminate among fine degrees of an attribute, and often la ck scope or content validity (Bergkvist & Rossiter, 2007; McIver & Ca rmines, 1981; Nunnally & Bernstein, 1994; Spector, 1992). In summary, a comprehensiv e understanding of th e consequences of desire-division of labor fit is limited by the la ck of previous researchers to incorporate and assess all relevant dimensions of la bor in the paid and family domains. Second previous studies generally lack an all-encompassing examination of the desire-division of labor fit construct. The fit variable or the relationship between desires for division of labor a nd the actual division of labor, may take many forms. That is, within a dual-earner relationship, a pers on may contribute a larger, smaller, or equal proportion of paid or family labor than (s)he desired. Despite this, only two of the five

PAGE 22

9 reviewed studies (i.e., Loscoocco & Spitze, 2007; Ross et al., 1983) included all three forms of congruence in their analyses, and th ey seemed to do so in a cursory fashion, without incorporating th eory or hypotheses relevant to th e different forms of fit. This differentiation is important, as each type of misfit may influence well-being through different processes, resulting in relati onships of various forms and magnitudes (Edwards, 1996). Thus, the examination and differentiation of all forms of fit is a crucial component in fully understanding th e complex fit-well-being relationships. Third, there is also a need for more t horough examination of the consequences of desire-division of labor fit. Although a variety of marital quality and mental health variables have been investigated as well-be ing outcomes, no studies included a measure of well-being in the work domain. This ma y be a function of the publication outlets, as the studies are mostly published in family or gender-focused journals, rather than those geared toward an organizational studies au dience. Nonetheless, because dual-earner couple's division of paid labor has implications fo r the qualities of th e work role that each spouse occupies (Pixley & Moen, 2003), ther e is theoretical reas on to suspect that desire-division of labor fit impacts career-related well-bei ng in addition to health and marital outcomes. As such, a comprehens ive analysis should incorporate well-being specific to the paid work domain in addi tion to well-being in other life domains. Fourth, the process by which couples divide family and paid labor is an issue that inherently involves two people, mean ing two people experience fit (or misfit) between their pre-child desires and post-child division of labor, and the well-being of each person may be affected by their own desire -division of labor fit as well as by the fit of their partner. In order to assess each of these relationships, data from matched

PAGE 23

10 marital dyads is necessary, a condition that was only met in two previous studies (Khazan et al., 2008; Ross et al., 1983). Fu rthermore, only Ross et al. (1983) took full advantage of the dyadic data by examining the fit perceptions of both partners in relation to their own and thei r partner's well-being. However, this study was limited, as it only examined desires toward women's wo rking roles, with no attention paid to women's or men's preferences for men's relative contributions to paid labor. In order to understand the wide-range of consequences th at result from a dyadic process, desiredivision of labor fit and well-being research should include data from matched dyads and incorporate hypotheses that address both pa rtners without any assumptions based on gender. Lastly, one area of inquiry that remains virtually untouched in this stream of research is the role of moderators, or variab les that temper the strength and form of the desire-division of labor fit a nd well-being relationships. W ith the exception of gender, previous research has not in corporated any individual or situational variables into analyses to test for interactive effects. The inclusion of moderators would not only contribute to a more comprehe nsive understanding of the link between desire-division of labor fit and well-being, but it also has the potential to resolve some of the inconsistencies in the results across previous studies. In summary, there is much potential to expand our theoretical understanding of the desire-division of labor fit and we ll-being relationship by including more comprehensive definitions of the division of labor and division of labor fit constructs, examining well-being in the work domain, testing a wide range of hypotheses using

PAGE 24

11 both males and females in dual-earner dyads, and exploring the role of moderators in this relationship. Methodological. A major limitation of the previous desire-division of labor fit and well-being research is the measurement and analysis of the fit variable. Most previous researchers created the fit variable using a difference score, subtracting the actual division of labor from the desired divi sion of labor. As rese arched extensively by Edwards (Edwards, 1994a, 1994b, 2001, 2002, 2007), there are numerous drawbacks to using difference scores as a measure of fit. Five of Edward’s main criticisms are listed below. First, the relationship between fit and an outcome is a three dimensional issue, with three components (the two fit variable s and the outcome variable). Difference scores reduce the two fit variab les into one variable and thus force an inherently three dimensional question into two dimensions (Edwards, 2007). Sec ond, difference scores are not usually equal representations of bot h variables that compose them. Unless the variables have equal varian ce, the difference score will primarily represent the component with the larger variance (E dwards, 1994b, 2002). Third, they confound the effects of their components, such that so metimes the observed relationship between the difference score and outcome is only attri butable to one of the difference score components, not both components as assumed (Edwards, 1994b). Fourth, because difference scores are a simple product of tw o components it is sta tistically impossible for them to explain more variance in an outcome than just looking at the two components separately (Edwards, 1994b), wh ich in many cases negates the entire premise of the research question. Finally, in some instances, difference score are less

PAGE 25

12 reliable than their individual componen ts (Edwards, 1994a, 2001; Johns, 1981; Williams & Zimmerman, 1977). As an alternative to difference scor es, Edwards and colleagues (Edwards, 1991, 1993, 1994b; 2001; 2002; Edwards & Harrison, 1993) recommend the use of polynomial regression, a special case of multiple linear regression. Polynomial regression allows the unique components of th e fit variable to be preserved and treats congruence as it conceptually should be treated, as the extent of correspondence between two components rather than as a single score (Edwards, 1994a, 2002). Additionally, polynomial regressi on uses higher order terms and thus can capture both linear and non-linear effects (Edwards, 1993, 1994b). Given that the effects of different directions of misfit on well-being may vary, bein g able to test for non-linear effects is of utmost importance. One criticism of polynom ial regression is that it is difficult to interpret; however, graphing the equations using response surface modeling provides a three-dimensional image of relationships that greatly aids in interpretation (Edwards, 1993, 2001). In summary, the statistical problems asso ciated with using difference scores to assess fit are well documented, calling into question the accuracy of results based on this method (Edwards, 2001). Polynomial regression, coupled with response surface modeling, is a superior method of analyzi ng the relationship between fit and other variables. As none of the existing desire-division of la bor fit and well-being studies used polynomial regression, there is clear r oom for methodological improvement in this area.

PAGE 26

13 Current study The aforementioned review and critique pinpoints several theoretical gaps and limitations in the extant literature. In the present study, I address each of these shortcomings by examining the congruence between pre-child de sires and post-child actual division of labor fit th rough the lens of P-E fit theo ry. This well-researched theory provides a strong conceptual foundati on for predictions about the nature of the relationship between different forms of (m is)fit and several ty pes of well-being. Additionally, it provides a fr amework for generating hypotheses concerning the direct effects of desire-division of labor fit on one’s own well-being as well as crossover effects within spousal dyads, and allows fo r the test of moderato rs in each of these relationships. Models exhibiting all of the hypotheses that were tested in the current study are presented in Figures 1 and 2. In addition to theoretical advancem ents, the present study improves upon the methodology of the extant literature. Desires for division of labor a nd actual division of labor itself were measured through a multi-ite m scale that includes the three components of division of paid labor (inc ome, work hours, career prioritization) and family labor (household tasks, childcare tasks, emotion work) that have been deemed important by previous researchers (e.g., Coltrane, 2000; Rapoport & Rapoport, 1969; Pixley, 2008). Measurement and analysis of the fit variable will also be improved through the use of polynomial regression and response surface mode ling rather than difference scores. Through these theoretical advancements and methodological improvements, the present study aims to gain a clearer understanding of the well-being cons equences of misfit

PAGE 27

14 between pre-child desires toward division of paid and family labor and the post-child actual division of the labor. In the subsequent sectio ns, the conceptual framework for the present study is outlined, starting with a broad review of P-E fit theory, followed by an application of the theory to the current desire-division of labor context, and concluding with the introduction of specific hypotheses. Person-Environment Fit Theory Fundamental premise. The fundamental ideas behind P-E fit theory were introduced into the psychological literatu re several decades ago (e.g., Lewin, 1935; Murray, 1938, Parsons, 1909) and were refined in more recent times (French et al., 1982). The basic premise of P-E fit theory is that stress occurs when there is a misfit between the person and the environment. Li ke many other theories of stress (e.g., Beehr & Newman, 1978; Cummings & Cooper, 1979, Mc Grath, 1976), P-E fit theory contends that stress results in subse quent strain, which may lead to attitudinal, psychological, and/or behavioral changes (Edwards, Capl an, & Harrison, 1998). On the other hand, chronic periods of fit between the person and the environment can have salutary effects on well-being (Edwards & C ooper, 1988; Harrison, 1978; 1985). With regard to specific processes, the effects of P-E f it on a particular well-being outcome (i.e., attitudes, health, performance) can be inferred by examining P-E fit theory in conjunction with theories relevant to that outcome variable (Edwards & Shipp, 2007). Types of P-E fit. Under the broad umbrella of P-E fit, there are two main distinctions (Edwards, 2009). The first divi des fit into two types: supplementary and complementary (Kristof, 1996; Muchinsky & Mohanan, 1987). Supplementary fit,

PAGE 28

15 which occurs when a person and the environm ent share similar characteristics, is often conceptualized as value congruence (Cable & Edwards, 2004; Kristof, 1996). Complementary fit occurs when the person and environment have something to offer each other, such that the n eeds of the environment are offset or “made whole” by the assets of the individual, and vi ce versa (Muchinsky & Monahan, 1987). Complementary fit is further distinguished based on the source of the requirements (Edwards, 1991). Demands-abilities fit occurs when the environment imposes requirements on the individual, and needs-supp lies fit represents the extent that a person’s needs, desires, or preferences (Kri stof, 1996) are satisfied by the environmental supplies (Edwards, 1996; Fren ch et al., 1982). The second important P-E fit distinction is between the objective and subjective representations of the pers on and the environment. The objective represents the attributes of the person and the environm ent as they actually exist, whereas the subjective refers to an indi vidual’s perceptions of his/ her own attributes and the environment (Edwards et al, 1998; French et al., 1982; Harrison, 1978). The theory stipulates that objective constr ucts affect their subjective counterparts, a process that may be influenced by outside factors such as cognitive distortions, and personal and situational constraints on information acce ss and processing (Harrison, 1978). Thus, subjective perceptions serve as a medi ator between the objective person and environment and strain (Edwards et al., 1998; Kahn et al., 1964; Lazarus & Folkman, 1984). For this reason, many researchers focu s on subjective fit, w ith the understanding that it is a function of the objective com ponents (Edwards et al., 1998; Edwards & Rothbard, 1999).

PAGE 29

16 Measurement. There are three basic approaches to the way P-E is measured (Edwards, Cable, Williamson, Lambert, & Shipp, 2006). The atomistic approach involves measuring the perceived person and environment separately, and then combining them in some way to represen t fit (Cable & Judge, 1996; Edwards, 1996; French et al., 1982). The molecular approach involves direct assessment of the perceived discrepancy between the person and the environment, preserving the direction of difference (i.e., asking an individual whet her their job demands ex ceed or fall short of their abilities) (Beehr, Wa lsh, & Taber, 1976; Lance, Ma llard, & Michalos, 1995; Rizzo, House, & Lirtzman, 1970). The molar approach asks individuals to rate the fit between themselves and their environment, disreg arding the direction of misfit and treating positive and negative discrepancies the same (Cable & DeRue, 2002; Judge & Cable, 1997; Saks & Ashforth, 1997). Edwards et al. (2006) compared each of th ese approaches for needs-supplies fit within a single study. The results suggested th at molecular approach es tend to result in unequally weighted comparisons of the pers on and environment. Molar approaches seem to capture affect more than fit and are limited in the sense that they do not allow for complete tests of P-E fit theory, as pos itive and negative discrepancies in the person and environment are treated the same. Atom istic approaches are burdened with issues of unequivocal referent comparisons in asse ssing person and the environment. Overall, atomistic approaches are preferable, as th is is the only method that does not confound the constructs of the person and environmen t, thereby preventing estimation of their independent effects (Edwards, 1991).

PAGE 30

17 Another important component of P-E f it measurement is that the person and environment dimensions must be commensura te (Dawis & Lofquist, 1984; French et al., 1974; French & Kahn, 1962), meaning they are described in the same terms (e.g., need for autonomy and amount of autonomy envi ronment provides) and are assessed on the same metric (Edwards, 1996). Metric equi valence is best achie ved by using the same response scale for the person and environment with different item stems to distinguish between the two (Edwards & Shipp, 2007). When dimensions are not commensurate, the proximity of the person and environm ent dimension to one another cannot be determined and the idea of fit has li ttle meaning (Edwards et al., 1998). Application of P-E Fit Theory to Desi re-Division of Labor Fit and Well-Being Theoretical framework: needs-supplies fit. The direct effects of desire-division of labor fit on an individual’s own well-b eing fit can be underst ood through a subjective needs-supplies fit framework, wh ere fit is determined by the ex tent that an individual’s needs, desires, or preferences (Kristof, 1996) are satisfied by the individual's perception of supplies in the environment (Edwards, 1996; French et al., 1982) Needs-supplies fit is grounded in the basic premises of need fulfillment theories (Edwards et al., 1998; Harrison, 1978; Kristof, 1996) which argue that a pers on will be satisfied and experience positive mental states if his or her needs are fulfilled by the environment. However, when needs go unmet, stress and s ubsequent strain are likely to occur (e.g., Locke, 1969; Murray, 1938). As such, need s-supplies fit theory assumes that as supplies increase toward needs, stress decreases and well-bein g increases (French et al., 1982; Harrison, 1978). This pattern is predicte d for all relationships, regardless of the content of the specific supplies or needs di mensions (Edwards et al., 1998). However,

PAGE 31

18 when misfit occurs in the other direction, when supplies exceed needs, the relationship becomes more complicated. Depending on th e nature of the fit variables, excess supplies may result in better, worse, or the same level of well-being (French et al., 1982; Locke, 1976). Another component of needssupplies fit theo ry is the relationship between the absolute values of the needs and supplies variables and well-being (Edwards, 1991; Edwards, 1996; Edwards & Rothbard, 1999). Holding the degree of needs-supplies fit constant, the level of well-being may di ffer depending on whether needs and supplies are both high or are both low. Needs-suppl ies fit theory does not predict a universal pattern of relationships, as the impact of absolute valu es on well-being depends on the content of the fit variables (Edwards & Shipp, 2007). Defining needs and supplies. The first step in applying needs-supplies fit theory to dual-earner division of labor is to define needs and supplie s in this context. For the sake of clarity, the nomenclature of Partner A and Partner B is used to refer to each spouse within a marital dyad. Partner A’s environmental labor supplies are determined by the amount of labor that Partner B contri butes relative to the total amount of labor that the environment requires. For example, if Partner B takes on 30% of the family labor, then the environment is supplying 30% of the necessary labor for Partner A. As previously noted, an essential component of P-E fit research is that the person and environment dimensions be commensurate. Accordingly, to match the environmental supplies component, needs can be represented by Partner A’ s preferences for the amount of labor that Partner B contributes to each domain. For example, if Partner A desired that paid work would be divided so th at (s)he contribute d 40% and Partner B

PAGE 32

19 contributed 60%, needs are define d as the 60% of paid work labor that Partner A wanted Partner B to contribute. Because the focus is on subjective fit, both needs and supplies are measured from Partner A’s perspective. It is important to mention that this conceptualization deviates from typical needs-supplies variables, as needs are de fined as preferences for another individual (Partner B) rather than preferences for one’s self (Partner A). Ho wever, when division of labor is measured in a proportional manne r as is done in the present study, Partner A’s desires for Partner B’s labor contributions are a function of Partner A’s self desires. Thus, Partner A’s desires are being assessed, al beit indirectly. With this understanding, hypotheses are generated based on empirical and theoretical research that aligns closely with the research question at hand (the effects of Partne r B’s contributions to labor relative to Partner A’s desires for his/he r contributions on wellbeing) as well as research that is directed at the flip side of this relationship (the effects of Partner A contributing less to paid labor than (s)he desires on Partner A’s own well-being). Additionally, because desires may be different for paid and family labor, needs and supplies are examined separately in each ar ea and hypotheses are test ed independently. Needs-supplies fit and well-being. As previously noted, well-being is conceptualized in a variety of ways in n eeds-supplies fit research. The theoretical nature by which needs-supplies fit influen ces different types of well-being can be inferred by combining the general notions of P-E fit theory with theories about the particular form of well-being (Edwards & Shipp, 2007). In order to gain a thorough understanding of the consequences of desire-l abor division fit, three forms of well-being are examined: satisfaction, depression, and physical health symptoms. Hypotheses are

PAGE 33

20 generated concerning the relationship of shortage of supplies and excess supplies relevant to needs and each form on well -being. However, hypotheses related to the relationship between absolute levels of needs and supplies and well-being are not formed, as there is not strong theoretical ra tionale to expect a di fferential relationship between fit when both desires for one partne r’s labor contributions and his/her actual labor contributions are high than when they are both low. Satisfaction. Satisfaction is a state marked by pl easant, positive feelings (Watson & Tellegen, 1985). Because divisi on of paid labor is a shared process that occurs on the couple level, fit between desire s and reality are likely to influence affective reactions toward the marital partner and the marita l relationship as a whole (i.e., marital satisfaction). Additionally, prev ious research suggests that needs-supplies fit that is specific to a particular life domain may influence satisfacti on about that domain, as well as more global satisfaction (Edwards & Rot hbard, 1999; 2005). Thus, career and family satisfaction are examined when focusing on division of paid and family labor, respectively. The effects of needs-supplies fit on satisfac tion can be inferred from theories of satisfaction and emotion (Edwards & Shi pp, 2007). The Range of Affect Theory (Locke, 1976, 1979) posits that satisfaction is determined by the congruence between one’s values and what is provided by the e nvironment. The theory was created to explain job satisfaction specifically but al so has theoretical relevance in other environments. Similarly, Lazarus’s (1991) th eory of emotions states that positive emotions arise with goal congruence, or the ex tent that a situation is consistent with one’s desires. Thus, both Locke and Lazarus ’s theories align well with the notion of

PAGE 34

21 needs-supplies fit, suggesti ng that as environmental suppl ies increase toward needs, satisfaction improves (Edwards & Shipp, 2007). In the present context, this leads to the prediction that as the propor tional amount of labor Partner B is contributing increases toward the amount of labor Partner A desired Partner B to contribute, Partner A’s wellbeing increases. However, in order to determine the rela tionship between the other direction of misfit, when supplies exceed needs (when Pa rtner B contributes more than Partner A desired him/her to contribute), other theo ries and processes must be considered (Edwards et al., 1998). Self-discrepancy theo ry (Higgins, 1987, 1989) is highly relevant, as it also focuses on discrepanc ies but specifically in regard to an individual’s self concept. The theory posits that people are strongly motivated to maintain a sense of consistency among their beliefs and self-con ceptions. When differences between selfaspirations (ideal selves) a nd actual behaviors (actual selv es) are experienced, negative emotions such as sadness, dissatisfaction, and other depressive states tend to emerge. The theory has generated a good deal of empi rical research, with study results typically lending support to its basic ideas (e. g., Higgins, Bond, Klein, & Strauman, 1986; Newman, Higgins, & Vookles, 1992; Strauman, 1989). When supplies exceed needs, Partner B c ontributes more than Partner A desires him/her to contribute, and by default Partne r A is contributing less than desired for him/herself. The level of contributions that Partner A desires for him/herself represents an ideal self, that is, an ideal conception of one’s family provider and caretaker roles. When supplies are in excess, Partner A is not living up to this ideal, evoking a self-

PAGE 35

22 discrepancy. As predicted by Higgins’ theo ry, such discrepancy results in negative emotions and dissatisfaction. Applying the notion of domain-specific ity with fit (Edwards & Rothbard, 1999, 2005), excess paid labor supplies in the paid domain should influence the ideal work self and thus attitudes in the paid labor domain, namely career satisfaction. Similarly, excess family labor supplies should affect the ideal family self and relate to family satisfaction. Empirical evidence lends some support to the self-concept idea and its relationship to domain satisfaction. Tsaous ides and Jome (2005) experimentally manipulated career compromise, which they vi ewed a form of self -discrepancy toward the ideal worker self. Those in the conditi ons with greater career compromise reported lower anticipated work-related satisfaction and more negative affect than those in conditions with little or no compromise. A lthough not all situations of excess supplies result in career compromise, it is associated with less contribution to paid labor relative to one’s spouse (Pixley & Moen, 2003). With regard to family labor, there are only a few studies that have consider ed self-discrepancies specifi c to family role ideals (e.g., Polasky & Holahan, 1998; Shafer, Wickrama & Keith, 1996). Both studies found a relationship between discrepancies and ne gative emotions, although neither examined satisfaction in specific domains (e.g., family satisfaction). Based on the principles of needs-supplies fit self-discrepancy theories, I predict: Hypothesis 1: Partner A’s career sa tisfaction will increase as paid labor supplies (Partner B’s paid labor contribu tions) increase toward paid labor needs (Partner A’s desires for Partner B’s pai d labor contributions), and will decrease as supplies exceed needs. (see Figure 3)

PAGE 36

23 Hypothesis 2: Partner A’s family satis faction will increase as family labor supplies (Partner B’s family labor contri butions) increase toward family labor needs (Partner A’s desires for Partner B’ s family labor contributions), and will decrease as supplies exceed needs. (see Figure 3) Because desire-division of labor fit is not an individual process, but rather one that occurs within the marital dyadic unit, it should relate to affect toward the marital partner and relationship (i.e., marital satis faction) in addition to domain-specific satisfaction. Based on needs-s upplies fit theory, wh en one’s spouse is not contributing a large enough proportion to paid or family labor (supplies are shor t of needs), stress arises due to lack of need fulfillment. Af fective reactions to stress depend upon source attribution (Perrewe & Zellars, 1999). That is, when an inte rnal source is perceived to be the cause, internally-driven emotions such as guilt and shame are often experienced. However, when the source is external, anger directed at the cause of the stress may arise (Perrewe & Zellars, 1999; Spector & Fox, 2005) Because division of labor supplies are provided by the spouse, (s)he is the source of needs-supplies misfit, meaning negative affective reactions directed toward the spous e and the marital relationship may arise. Previous research is consistent with this idea, as men and wome n who perceive that their partner is contributing too little to paid labor (e .g., Perry-Jenkins et al., 1992; Wilkie, Ferree, & Ratcliff, 1998) and fa mily labor (e.g., Lavee & Katz, 2002; MacDermid et al., 1990; McHale & Crouter, 199 2; Milkie et al, 2002; Pina & Bengston, 1993) experience decreased marital satisfaction. Moreover, the effects of misfit in the ot her direction, excess supplies, on marital satisfaction may be inferred from self-dis crepancy theory. The negative emotions

PAGE 37

24 experienced as a result of ideal and actual self -discrepancies may also be directed at the marital partner, as s(he) is a prominent f actor in creating the di screpancy. Previous research supports this noti on, although it is limited in the sense that studies on family and work role identity threat tend to be se gregated by gender. Speci fically, when wives’ paid labor behavior threatens husband’s pr ovider role identity, husbands experience poorer marital quality (Brennan et al., 2001; Potucheck, 1997). When husbands’ family labor behavior threatens wive s’ maternal identities, wive s report less positive marital relations (MacDermid et al., 1990). Moreover, excess labor supplies may be interpreted as a form of social undermining which ha s been linked to poor marital satisfaction (Westman, Vinokur, Hamilton, & Roziner, 2004). Based on these ideas, the following hypotheses are proposed: Hypothesis 3: Partner A’s marital sa tisfaction will increase as paid labor supplies (Partner B’s paid labor contribu tions) increase toward paid labor needs (Partner A’s desires for Partner B’s pai d labor contributions), and will decrease as supplies exceed needs. (see Figure 3) Hypothesis 4: Partner A’s marital satis faction will increase as family labor supplies (Partner B’s family labor contri butions) increase toward family labor needs (Partner A’s desires for Partner B’ s family labor contributions), and will decrease as supplies exceed needs. (see Figure 3) Depression and physical health symptoms Another commonly studied and theoretically relevant outcome of needs-supplies fit is ment al and physical health, often indicated by depression and physical health symptoms (Edwards & Shipp, 2007). Many theories of the stress process em phasize the link between stressors and

PAGE 38

25 psychological and physiological strains (e.g., Beehr & Newman, 1978; Fletcher & Payne, 1980; Kahn & Byosiere, 1992; Lazarus & Folkman, 1984; Quick, Cooper, Nelson, Quick, & Gavin, 2003). The body’s natural reaction to stress is responsible for this link, as short term adaptive changes in emotions, behavior, hormones, and immune, cardiovascular, and pulmonary functions have destructive and pathogenic effects when chronically activated (Dienstbierm, 1989; Fr ankenhaeuser, 1986; Johnson, Kamilaris, Chrousos, & Gold, 1990; Pearlin, Menagha n, Lieberman, & Mullan, 1981). On the other hand, chronic periods of needs-supplies fit can have salutary effects on depression and physical health (Edwards & Cooper, 1988; Harrison, 1978, 1985). Together, needssupplies fit and other stressor-strain theories suggest that negative health symptoms decrease as environmental supplies (Partner B’s labor contributions) approach personal needs (Partner A’s desires for Part ner B’s labor contributions). However, as mentioned with satisfaction, the effects of excess supplies on health are not universally predicted by needs-suppl ies theory, and self-discrepancy theory becomes important. The rationale previous ly applied to excess supplies and selfdiscrepancy theory extends to depression a nd physical health symptoms. When supplies are in excess, an individual is not contributi ng as much to paid and family labor as (s)he desires and a discrepancy between the actual and ideal self is formed. This discrepancy creates dejection -related emotions, such as sadness and dissatisfaction, which when chronically experienced lead to depre ssion and other mood disorders (APA, 2000). Many studies empirically support the re lationship between ideal-actual selfdiscrepancies and depression (Cornette, Strauman, Abramson, & Busch, 2009; Higgins et al., 1986; Higgins, Vookles, Tykocinski 1992; Orth, Berking, Burkhardt, 2006;

PAGE 39

26 Rodebaugh & Donahue, 2007). Likewise, nega tive emotions have been linked to detrimental immune system alterations, neuroe ndocrine responses, and health behaviors, which in turn contribute to a wide variety of illnesses and diseases (cf., Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002; cf., Kubzansky & Kawachi, 2000). As depression and physical health symptoms permeate all life roles, they are app licable to misfit in both work and family domains. Hypothesis 5: Partner A’s depression will decrease as paid labor supplies (Partner B’s paid labor contributions) increase toward paid labor needs (Partner A’s desires for Partner B’s pai d labor contributions), and will increase as supplies exceed needs. (see Figure 4) Hypothesis 6: Partner A’s depression w ill decrease as family labor supplies (Partner B’s family labor contributions) increase toward family labor needs (Partner A’s desires for Partner B’s family labor contri butions), and will increase as supplies exceed needs. (see Figure 4) Hypothesis 7: Partner A’s physical health symptoms will decrease as paid labor supplies (Partner B’s paid labor contribu tions) increase toward paid labor needs (Partner A’s desires for Partner B’s pai d labor contributions), and will increase as supplies exceed needs. (see Figure 4) Hypothesis 8: Partner A’s physical hea lth symptoms will decrease as family labor supplies (Partner B’s family labor contributions) increase toward family labor needs (Partner A’s desires for Part ner B’s family labor contributions), and will increase as supplies exceed needs. (see Figure 4)

PAGE 40

27 Moderators. In addition to investigating the ma in effects of de sire-division of labor fit on well-being, we can gain a better understanding of desire-division of labor dynamics by examining boundary conditions of the fit process. Three variables are examined as moderators, including domain centrality, gender, and voice in decision of labor decision making. Domain centrality The theoretical rationale behi nd the first moderator, domain centrality, the degree that career or family is considered impor tant to a person’s life as a whole (Gecas & Seff, 1990), stems from early conceptions of P-E fit. French et al. (1974) argued that the effects of n eed-supplies fit on well-being depend on the importance of the dimension to which needs a nd supplies refer. As the importance of a dimension increases, needs-suppl ies misfit poses a greater thr eat to overall self-concept, and negatively affects well-being (Locke, 1976; French et al., 1974). Previous P-E fit research (Edwards & Rothbard, 1999) examin ing needs-supplies fit in various life domains has conceptualized importance as domain centrality, finding that the strength of the relationship between n eeds-supplies fit in a particul ar life domain is dependent upon the importance of that domain to indivi dual. For example, the fit between needs and supplies for autonomy at work related mo re strongly to work well-being for those with higher rather than lower work cen trality (Edwards & Rothbard, 1999). Applying this rationale to the present c ontext, I predict that domain centrality moderates the effects of desi re-division of labor fit on all four types of well-being. However, the form of the moderation is not universal for both dire ctions of misfit. When career centrality, for example, is high, excess supplies are likely to have a large impact on well-being because th e work role is highly releva nt to self-concept. Said

PAGE 41

28 otherwise, someone who identi fies highly with work is mo re likely to be upset when they are not contributing as much as desire d than someone who does not view work as a central part of their identity. When paid labor supplies are too few, career centrality may actually serve as a buffer against negative effects on well-being. To the extent that one highly values one’s individua l contribution in work, one is less likely to be affected by a partner who is contributing less than desi red. A parallel process should occur with family centrality and the division of family labor. Hypothesis 9: Partner A’s career centra lity moderates the relationship between paid labor supplies (Partner B’s paid l abor contributions) and paid labor needs (Partner A’s desires for Partner B’s pai d labor contributions) fit and Partner A’s well-being ((a) career satisfaction,(b) marital satisfaction, (c) depression, and (d) physical health symptoms). The rela tionships between shortage of supplies and well-being will be weaker when career ce ntrality is higher rather than lower. The relationships between excess supplies and well-being will be stronger when career centrality is higher rather than lower. (See Figures 5 and 6.) Hypothesis 10: Partner A’s fa mily centrality moderate s the relationship between family labor supplies (Partner B’s fam ily labor contributi ons) and family labor needs (Partner A’s desires for Partner B’s family labor contributions) fit and Partner A’s well-being ((a) family satis faction, (b) marital satisfaction, (c) depression, and (d) physical health symp toms). The relationships between shortage of supplies and well-being will be weaker when family centrality is higher rather than lower. The relationships between excess supplies and well-

PAGE 42

29 being will be stronger when family centra lity is higher rather than lower. (See Figures 5 and 6.) Gender. Gender is an inevitable considerati on in division of labor research (Coltrane, 2000). Although pr ogress toward more egalitarian practices has been made, husbands still carry greater influence than wi ves in the way paid and family labor is divided (Coltrane, 2000; F ox & Murry, 2000; Gerstel & Gallagher, 2001; Tichenor, 1999). Thus, when a husband’s and wife’s desire s for the division of paid labor conflict, men’s preferences are more likely to prev ail. This asymmetry has implications for gender differences in consequences of desire -division of labor Specifically, women are aware of this situation and app ear to grasp the reality that work and family roles may not live up to their ideals (G erson, 2002; Orrange, 2003; Machung, 1989; Sanders, Lengnick-Hall, Lengnick-Hall, & Steele-Cl app, 1998; Schroeder, Blood, & Maluso, 1993; Spade & Reese, 1991). Perhaps because th e base rate is much lower, men are not as cognizant or do not plan as much for a potential discrepancy betw een their desires in early marriage and post-child reality (Machung, 1989; Maines & Hardesty, 1987; Schroeder et al., 1993; Tinklin, Croxford, Du cklin, & Frame, 2005). This has led some researchers to suggest that, compared to men, women’s greater pr eparedness for desirereality incongruence in labor divisions he lps them develop more effective coping strategies that serve as a buffer against the negative effects of this stressor (Loscocco & Spitze, 2007). Therefore, it is hypothesized: Hypothesis 11: Partner A’s gender mode rates the relationship between paid labor supplies (Partner B’s paid labo r contributions) and paid labor needs (Partner A’s desires for Partner B’s pai d labor contributions) fit and Partner A’s

PAGE 43

30 well-being ((a) career satisfaction,( b) marital satisfaction, (c) depression, and (d) physical health symptoms). The rela tionships between shortage of supplies and well-being and excess supplies and well-be ing will be weaker when Partner A is a female than when Partner A is male. (See Figures 7 and 8.) Hypothesis 12: Partner A’s gender modera tes the relationship between family labor supplies (Partner B’s family labor contributions) and family labor needs (Partner A’s desires for Partner B’s fa mily labor contributions) fit and Partner A’s well-being ((a) family satisfaction,( b) marital satisfaction, (c) depression, and (d) physical health symptoms). Th e relationships between shortage of supplies and well-being and excess supplies and well-being will be weaker when Partner A is a female than when Part ner A is male. (See Figures 7 and 8.) Voice in division of labor decisions. The manner in which couples manage paid and family labor is typically the culmination of several explicit or implicit processes and decisions throughout the marriage. For exam ple, individuals must often make big decisions, such as relocation, promotion, or le ngth of parental leav e, that impact their own and their partner’s work and family situation and thus the division of labor (Wethington, Pixley, & Kavey, 2003). Similarl y, couples face day-to-day decisions, such as who will leave work early to pick up children or who will prepare dinner, that over time culminate to represent the divi sion of labor (Orrange, Firebaugh, & Heck, 2003). In some cases the division of labor emerges as a “silent process,” where there is little discussion or negotiati on between partners, and the partner with more marital power (usually the man) exerts control over labor division dynamics (Kingsbury &

PAGE 44

31 Scanzoni, 1989; Kompter, 1989; Zipp, Prohaska, & Bemiller, 2004; Zvonkovic, Schmiege, & Hall, 1994). In other cases, both marital partners are ac tive players in the division of labor, jointly settli ng on critical career or family decisions (Barnett & Rivers, 1996; Challiol & Mignonac, 2005; Wethington et al., 2003). Research on “voice,” the opportunity to provide input, in decision making finds that pe ople react more positively to decisions when they feel they had a voice in the process, even when the outcome is unfavorable to them (LaTour, 1978; Lind, Kurt z, Musante, Walker, & Thibaut, 1980). This idea has been specifically applied to work and family roles. Spouses who feel they have a say in how their own work and fam ily roles are structured tend to be more satisfied with their roles in both domains a nd with their marriages than people who have no input (Hiller & McCaig, 2007; Madde n, 1987; Thompson & Walker, 1989). Applying this information to the relati onship desire-divisi on of labor fit and outcomes, the amount of voice one has in the processes that led to the current division of labor within the marriage should serv e as a buffer to the negative effects of incongruence. Specifically, when a person’s de sires for the division of labor at the start of marriage are not met post-child, s(he) wi ll be less negatively affected by this discrepancy if s(he) had some input into the decisions that led to the current division of labor. Hypothesis 13: Partner A’s voice in divi sion of paid labor decisions moderates the relationships between paid labor supplies (Partner B’s paid labor contributions) and paid labor needs (Partn er A’s desires for Partner B’s paid labor contributions) fit and Partner A’s well-being ((a) career satisfaction,(b) marital satisfaction, (c) depression, and (d ) physical health symptoms). The

PAGE 45

32 relationships between shortage of supplies and well-being and between excess supplies and well-being will be weaker when voice in division of labor decisions is higher rather than lower. (See Figures 7 and 8.) Hypothesis 14: Partner A’s vo ice in division of family labor decisions moderates the relationships between family labo r supplies (Partner B’s family labor contributions) and family labor needs (Partn er A’s desires for Partner B’s family labor contributions) fit and Partner A’s well-being ((a) family satisfaction,(b) marital satisfaction,(c)depression, and (d ) physical health symptoms). The relationships between shortage of supplies and well-being and between excess supplies and well-being will be weaker when voice in division of labor decisions is higher rather than lower. (See Figures 7 and 8.) In the preceding section, hypot heses were generated concer ning direct effects, in that Partner A’s needs and supplies fit we re related to Part ner A’s well-being. However, the same variables, Partner A’s preferences for Partner B’s labor contributions and Partner B’s actual contributions may also impact Partner B’s wellbeing, in that Partner A’s pr eferences for Partner B repr esent environmental demands and Partner B’s labor contributi ons represent Partner B’s abiliti es or resources to fulfill these demands. With this in mind, in the following section the second type of complementary fit, demands-abilities fit is ap plied to the desire-l abor division context and hypotheses concerning Partner B’ s well-being are generated. Theoretical framework: demands-abilities fit. Demands-abilities fit is based on the match between environmental demands and a person's abilities (Edwards, 1996). Abilities are broadly defined as the knowledge skills, time, energy, and resources that a

PAGE 46

33 person can draw upon to fulfill environmen tal demands, defined as quantitative and qualitative requirements that can be objective or socially constructed (Edwards, 1996; Edwards & Shipp, 2007; French et al., 1982; Kristof, 1996). The fundamental premise of demands-abilities fit is that when a pe rson lacks the abilities to fulfill the demands imposed by the environment, stress is experi enced, which in turn produces strain and decreases well-being (Edwards, 1996; French et al. 1982). Thus, as abilities increase toward demands, well-being also increases. Parallel to needs-s upplies fit, demandsabilities fit theory does not make consistent predictions about the ot her direct of misfit, when abilities exceed demands, and requires ex ternal theory relevant to the specific content of the fit variables (French et al ., 1982; Locke, 1976). Additionally, fit when demands and abilities are low may have a differential impact on well-being than fit when both variables are high, although the natu re of this relations hip also depends on the content of the fit variab les (Edwards & Shipp, 2007). Defining demands and abilities. Within the marital environment, the proportion of family and paid labor that Partner A desires Partner B to contribute acts as an environmental demand for Partner B. Thus, the same variable acting as a need for Partner A also serves as a demand for Partner B. Abilities are represented by the amount of labor Partner B contri butes. The amount of labor that a person contributes may be a function of his/her actual skill level, time and energy available, and resources to do so; thus, abilities in this context ma y be a function of a variety of sources, all consistent with the aforementioned definition of abilities. Furtherm ore, in the present study demands are measured objectively usi ng Partner A’s reports as it is his/her preferences that set the demands for the envi ronment. Abilities are assessed via Partner

PAGE 47

34 B’s reports of his/her own labor contributions, as it is his/he r perception of his/her level of ability that drives the cognitive pro cesses behind the demands-abilities fit – wellbeing relationship (Edwards, 1996; French et al., 1982). To provide a concrete example, if Partner A wanted Partner B to contribu te 60% of the paid labor, demands are for Partner B to contribute 60% to paid labor. If Partner B repor ts that (s)he is contributing 50%, then Partner B’s ability is to assume 50% of the couple’s paid labor requirements. Demands-abilities fi t and well-being. Several P-E fit theorists contend that the effects of demands-abilities fit on satisfaction, ph ysical, and mental health are not direct. Rather, demands-abilities fit affects these fo rms of well-being through needs-supplies fit (Edwards & Shipp, 2007; Harrison, 1978; La wler, 1973; Locke, 1976; Smith, Kendall, & Hulin, 1969). For this reason, I do not hypothesize about the link between demandsabilities fit and each form of well-being separately; inst ead I describe how demandsabilities fit contributes to needs-supplies f it and infer the link from needs-supplies fit to each form of well-being from the previ ously described needs-supplies theory. Moreover, the only form of satisfaction that is examined is marital satisfaction. The domain-specific forms of satisfaction, caree r and family satisfaction are not included because unlike needs, which are directed toward the marriage and specific domains, demands come directly from the spouse. Th erefore, in the present context, it is only theoretically sound to propose hypotheses about demands-abilities fit in relation to satisfaction toward the marital pa rtner. Finally, as stated with needs-supplies fit, there is no theoretical rationale to exp ect that the absolute levels of demands and abilities impact well-being differently when holding the degr ee of fit constant. Thus, no hypotheses are generated with respect to this element of demands-abilities fit theory.

PAGE 48

35 As proposed by Harrison (1978) and exte nded by Edwards and Shipp (2007) and Edwards (2008), there are three main mechanisms by which by which demands-abilities fit translates into needs-supplies fit. First, demands-abilities fit may enhance performance, which by virtue of increased rewards, fulfills needs and increases satisfaction. Second, demands may become in ternalized and accepted as the goals or motives of the individual. When this o ccurs, demands are essentially functioning as needs, and fit between demands and abili ties serves as need satisfaction. Third, demands-abilities fit may create a sense of competence, fulfilling the need for competence, and increasing satisfaction. The first two processes are particularly rele vant in the desire-division of labor fit context. Meeting demands for labor divi sions should contribu te to the one’s performance as a marital partner, enhancing supplies to meet rela tionships needs (i.e., the need to be a high-quality ma rital partner). In addition, re search suggests that marital partners tend to have communal orientations meaning they are cognizant of each other’s needs and respond sympathetically to them (Clark, 1984; Clark, Mi lls, & Powell, 1986; Eisenberg & Miller, 1987). Based on this notion, it is likely that the environmental demands, which are essentially one’s partner’s needs, ar e at least to some degree internalized as one’s own needs. The extent that abilities act as supplies to fulfill those needs determines needs-supplies fit. Take n together, there is substantial reason to assume that demands-abilities fit contributes to an individual’s needs-supplies fit and thus influences satisfaction and mental and physical health. Specifi cally, as Partner B’s labor contributions (abilities) approach envi ronmental demands (Partner A’s desires for Partner B’s labor contributions), satisfacti on and mental and physical health improve.

PAGE 49

36 In hypothesizing about the effects of excess abilities on well-being, selfdiscrepancy theory is again ut ilized. It was previously hypot hesized that for Partner A, excess supplies result in decreased well-b eing because they create a self-discrepancy. When demands are in excess for Partner B, th e same essential process occurs: Partner A is not contributing as much labor as (s)he desired and as a result, suffers from selfdiscrepancy. Thus, by virtue of having exce ss abilities, Partner B c ontributes to lack of need fulfillment for Partner A. Theoretically, not fulfilling one’s partner needs has implications for well-being. It may foster feelings of guilt or frustrati on that contribute to poorer health outcomes and negatively affect marital interacti ons. Grote and Clark (2001) found empirical support for the latter, as not being able to meet one’s partner’s needs contributed to increased martial dissatisfaction. The study wa s longitudinal, giving more credence to the idea that not meeting needs impacted mar ital dissatisfaction, rather than the reverse. Thus, excess abilities harm Partner B’s well -being through their effects on Partner A’s need fulfillment. Combining the ideas of shor tage of abilities and excess abilities, it is predicted that well-being is maximized at perfect demands-abili ties congruence for the division of both family and paid labor. Hypothesis 15: Partner B’s marital satis faction increases as paid labor division abilities (Partner B’s paid labor cont ributions) increase toward demands (Partner A’s desires for Partner B’s pai d labor contributions) and decreases as abilities exceed demands. (See Figure 3.) Hypothesis 16: Partner B’s marital sa tisfaction increases as family labor division abilities (Partner B’s family labor contributions) increase toward

PAGE 50

37 demands (Partner A’s desires for Part ner B’s family labor contributions) and decreases as abilities exceed demands. (See Figure 3.) Hypothesis 17: Partner B’ s depression decreases as paid labor division abilities (Partner B’s paid labor contributions) in crease toward demands (Partner A’s desires for Partner B’s paid labor co ntributions) and increases as abilities exceed demands. (See Figure 4.) Hypothesis 18: Partner B’ s depression decreases as family labor division abilities (Partner B’s fam ily labor contributions) in crease toward demands (Partner A’s desires for Partner B’s fam ily labor contributions) and increases as abilities exceed demands. (See Figure 4.) Hypothesis 19: Partner B’s physical heal th symptoms decrease as paid labor division abilities (Partner B’s paid la bor contributions) increase toward demands (Partner A’s desires for Part ner B’s paid labor contributions) and increase as abilities exceed demands. (See Figure 4.) Hypothesis 20: Partner B’s physical health symptoms decrease as family labor division abilities (Partner B’s family labor contributions) increase toward demands (Partner A’s desires for Part ner B’s family labor contributions) and increase as abilities exceed demands. (See Figure 4.) Gender. Gender was considered as a modera ting factor in the relationship between needs-supplies fit and Partner A’s we ll-being, and it is also relevant to the relationship between demands-a bilities fit and Partner B’s well-being. Compared to men, women are more responsive to the needs of those around them (Gilligan, 1982) and have more empathetic orientations (E isenberg & Lennon, 1983). Applying this idea

PAGE 51

38 to environmental demands, it is probable that women are thus more responsive to demands imposed by their partners (as these re present partner’s needs) and that they are more likely to internalize these needs as their own. The extent that demands are personally meaningful is a strong determinant of the strength of the demands-abilities fit – well-being relationship (French et al ., 1982; Harrison, 1978; Lazarus & Folkman, 1984), such that a shortage of abilities should be more detrimental when demands are deemed more significant. With regard to the effects of excess de mands, there is some evidence that women are more susceptible than men to the impact of stressors affecting their partners (Kessler & McLeod, 1984). There are three proposed m echanisms for these gender differences (Westman, 2004). Because women themselves experience higher levels of distress they are less resilient when facing the stress and strain of their husbands, women are more empathetic to the stress of their husbands and therefore more vul nerable, and women tend to provide more social support and are ther efore more susceptible to spousal stress. Assuming the self-discrepancy that comes with Partner B’s excess abilities (labor contributions) is a source of stress for Part ner A, Partner B should be more negatively affected by this stress when she is a woman than when he is a man. Hypothesis 21: Partner B’s gender modera tes the relationships between demands (Partner A’s desires for Partner B’ s paid labor contributions) and abilities (Partner B’s paid labor contributions) fit and Partner B’s well-being ((a) marital satisfaction, (b) depression, and (c) ph ysical health symptoms). The relationships between shortage of ab ilities and well-being and between excess

PAGE 52

39 abilities and well-being are stronger when Partner B is female rather than male. (See Figures 9 and 10.) Hypothesis 22: Partner B’s gender modera tes the relationships between demands (Partner A’s desires for Partner B’s family labor contributions) and abilities (Partner B’s family labor contributions) fit and Partner B’s well-being ((a) marital satisfaction, (b) depression, and (c ) physical health symptoms). The relationships between shortage of ab ilities and well-being and between excess abilities and well-being are stronger when Partner B is female rather than male. (See Figures 9 and 10.) Satisfaction with division of labor. The aforementioned predictions about the effects of demands-abilities fit on well-being were all based on Partner A’s demands and fulfilling Partner A’s needs, to the neglect of consideration of a Partner’s B’s attitudes toward the current labor division. Because spouses do not always share identical perceptions or attitudes about the division of labor (Jansen & Liefbroer, 2006), it is possible that Partner B’s well-being may be negatively affected due to the stress from not meeting the environmental demands that Partner A has imposed and simultaneously positively affected by Partner B’s satisfac tion with the current division of labor, determined by Partner B’s own preferences, ir respective of Partner A’s position. Said otherwise, satisfaction with current arrang ements should act as a buffer against the negative effects of demands-a bilities misfit. Self-evaluation maintenance theory (Tesser, 1998) applied to marriage (B each & Tesser, 1993; Clark & Bennett, 1992) supports this idea, positing that when situ ations result in positive outcomes for an individual but threaten his/her spouse’s well-being, the individual’s exhibits less

PAGE 53

40 positive reactions than in situations that result in positive outcomes for both. Based on these ideas a moderation effect is proposed: Hypothesis 23: Partner B’s satisfaction with the curre nt division of paid labor moderates the relationships between dem ands (Partner A’s desires for Partner B’s paid labor contributions) and abilities (Partner B’s paid labor contributions) fit and Partner B’s well-being ((a) marita l satisfaction, (b) depression, and (c) physical health symptoms). The relati onships between shor tage of abilities and well-being and between excess abilities and well-being are weaker when satisfaction with current division of paid labor is higher rather than lower. (See Figures 7 and 8.) Hypothesis 24: Partner B’s satisfaction with the current division of family labor moderates the relationships between dem ands (Partner A’s desires for Partner B’s family labor contributions) and ab ilities (Partner B’s family labor contributions) fit and Partner B’s wellbeing ((a) marital satisfaction, (b) depression, and (c) physical health symptoms). The relationships between shortage of abilities and well-being and between excess abilities and well-being are weaker when satisfacti on with the current division of family labor is higher rather than lower. (See Figures 7 and 8.)

PAGE 54

41 Chapter Two Methods Participants and Procedures Due to the nature of the research questi ons, participants were limited to married couples that met the following inclusion crit eria: 1.) each spouse worked at least 10 hours per week in paid employment, 2.) they had at least one child under age 6, 3.) all children were born after the couple was married to each other, and 4.) neither spouse was currently on parental leave from work. Th e first criterion was selected in order to assure that both spouses were participating to some extent in work and family roles. The purpose of the second criterion was to gain consistency across with participants with regard to the timing of their post-ch ild actual division of labor reports. The specific age range was selected to restrict par ticipants to the critical “launching stage” of life when at least one child is pre-school age. The th ird criterion was necessary to ensure that participants we re considering their current spouse when thinking about prechild desires for division of labor. Lastly, the fourth criterion was selected to avoid imbalances in the division of labor that ar e a function of temporary parental leave from work. Spouses were not recruited simultaneousl y, rather one member of the couple was recruited and (s)he was used to recruit his/her spouse. Th e first members of the couples were recruited through a variety of strategies. First, 13,943 alumnae of a large women’s

PAGE 55

42 organization were contacted via individual emails. Second, 845 alumni of a large southeastern university were contacted vi a individual emails. Third, 889 members of a website for first time fathers were contacted via the webmaster’s email listserve. Fourth, a snowball sampling approach was used. Poten tial participants were contacted via email and asked to participate in the study and/or forward the survey link to interested individuals. Specifically, recruitment emails were sent to my friends, family members, and professional acquaintances A total of 184 emails were sent by the principal investigator, but the number of forwarded emails is unknown. With each sampling method, the email described my background, the purpose of the study and provided a link to the online su rvey. Additionally, as an incentive, I offered to donate $2.00 to First Book, a non-pr ofit organization dedicated to childhood literacy, for each completed survey (up to $500.00). The email described the study very generally as a study of work and family issues. The email did not list any of the inclusion criteria; rather, the first question of the survey listed some of the criteria (must work at least 10 hours per week in paid em ployment, be married to someone who works at least 10 hours per week, a nd have at least one child) a nd asked participants whether they met all of these criteria. Those that sele cted “no” were sent to the final page of the survey and did not answer any other survey it ems. This method was utilized to allow for a more specific calculation of response rate. That is, it allowed for calculation of who was willing to participate but was not eligible to do so versus those who were contacted but did not want to participat e. The more specific criteria of children’s age, parental leave, and having children after the current marriage were included as questions in the survey, and participants were later screened during the data anal ysis phase for these

PAGE 56

43 criteria. The number of responses and respons e rates for each method are listed in Table 2. After completing the survey, participants were asked if they were willing to invite their spouse to also participate. T hose who selected yes were presented with two options: to personally provide the survey links to the study to their s pouse or to have the research team contact their spouse directly. Emails with a link to the online spouse survey were then sent to these spouses th at described the study, explained that their spouse had already participated, and asked fo r their participation. An incentive of a $5.00 donation to First Book was offered for eac h completed set of spousal surveys. Spouses were matched for analysis using a unique code system. Each participant was asked to create a code that would be unique onl y to the couple based on the first letter of their residential street an d both spouse’s birthdays (e.g., M822314). This allowed surveys to be matched without gather ing personally identifying information. Information about the number of responses resp onse rates for spouses is listed in Table 3. The majority of matched spousal re sponses were obtained from the women’s organization (85.7%), followed by persona l and extended networks (11.9%), the university alumni (1.4%), and the fi rst time fathers website (.7%). After screening for the aforementioned inclusion criteria, the final sample consisted of 578 individuals. Of these 578 participants, matched spousal data was obtained for 126. Thus, the finale sample wa s 126 couples. Because most of the initial participants were recruited through the women’s organization, the majority of them were female (96%) and the majority of the s pousal respondents were male. In order to obtain a more even distribution of gender for analyses, the spouses were randomly

PAGE 57

44 assigned as Partner A and Pa rtner B. The randomization was done by sorting the couples in alphabetical order a ccording to their unique codes. The first 63 couples were sorted so that the initial respondent was Part ner B rather than A. For the remaining 63, the initial respondent was designated as Pa rtner A and the spouse as Partner B. Of the 126 Partner As, the average age was 35.58 years (SD = 4.43), 50% were female, with a race/ethnicity breakdown of 91.3% White/Caucasian, 2.4% Asian, 1.6% Hispanic/Latino, .8% Black/African America n, and .8% “other.” Four percent of participants did not report th eir race/ethnicity. The sample was highly educated, as the majority of participants ha d a bachelor’s degree (39.7%) or Master’s degree (26.7%). Only 4% had less than a Bachelor’s degr ee, 6.7% had done some graduate work, and 20.8% had an advanced degree (Ph.D., M.D., etc. ). Partner Bs were 50% female, and 97.6% White/Caucasian (.8% Black/Afr ican American, .8% Asian, 3.2% Hispanic/Latino, 2.4% missing). Similar to Pa rtner A, most were highly educated (2.4% less than Bachelor’s degree, 31.7% Bachel or’s degree, 10.3% some graduate work, 35.7% Master’s degree, 19.8% advanced degr ee). The average age was 35.71 years (SD = 4.25). Measures All measures are included in the Appendi ces. Tables 4 and 5 list the descriptive statistics (number if items, mean, standard deviation, minimum score, maximum score, and scale minimum and maximu m) for each measure and demographic information for Partner A and Partner B. Table 6 presents descriptive statistics collapsed across partners but separated by gender. Unless otherwise noted, scores on each scale were obtained by averaging the scores across items, and higher scores

PAGE 58

45 indicate a greater prevalence of the construct. All measures were administered to both members of the marital dyad, but only certain variables were used to test hypotheses for Partner A and Partner B. The source of the report that was used to test each hypothesis is listed in Table 7. Pre-child desires fo r division of labor. Desires for division of labor before children were born were measured using re trospective reports. As no comprehensive measure of this construct exists in the liter ature for the paid or family labor domains, two three-item scales were created for the present study. Each item had a similar stem, asking participants to think back to be fore children were born and indicate how, thinking about the future, they wanted labor to be divided after children were born. For paid labor, each item assesses a different component, including income, work hours, and career prioritization deci sions. For family labor, three co mponents were also assessed, including childcare, household wo rk, and emotional work. The following instructions were given before the set of items in order to make the timeframe clear: “Before having children, many couples think and disc uss what their life will be like 'post-children.’ For the next set of questions, we are interested in knowing what your desires and expecta tions were BEFORE you had child ren with regard to what your life would be like AFTER you did have children.” An example item for paid labor is “Picture the total number of combined hours that you and your spouse spend in paid employment as a pie chart that sums to 100%. Before you had children and were thinking into the future what proportion of that pie chart did you WANT to be YOUR work hours once children were born, and what proportion of that pie chart did you WANT to be YOUR SPOUSE'S work hours onc e children were born? (These numbers

PAGE 59

46 should add up to 100%).” An illustration of an example pie chart was also included to facilitate understandi ng of the question. For family labor, similar questions were asked but the pie chart language was not used because pilot testing revealed that the items were easier to understand without it. Definitions and examples of each form of family labor (childcare, household, and emotion work) were included within the items to ensure that participants adequately understood the terms. The examples were ba sed on the definitions listed in Coltrane (2000) and Shelton and John ( 1996) as well as previous measures for childcare tasks (Bonney, Kelley, & Levant, 1999), household ta sks (Atkinson & Huston, 1984; Blair & Lichter, 1991), and emotion work (Erikson, 1993). An example item in the newly created scale is “Childcare related tasks are activities that involve caring for and raising children. Some examples incl ude supervising, bathing, punis hing, playing with children, and taking children to appointments or play da tes. Picture the total amount of childcare tasks that must be performed in your home Before you had children and were thinking into the future what percen tage of these tasks did you WA NT to perform, and what percentage of these tasks did you WANT your spouse to perform? (These numbers should add up to 100%).” For both measures, the response scale was in the format of percentages, and choices were listed in increments of 5% fr om 0% to 100%. There were two responses for each item, one for self and one for spouse Because these scales were newly create d for the present study, item analysis was conducted prior to deciding on the final set of ite ms to be used in analyses. First, in order to examine the distinctness of the fa mily and paid labor scales, a series of confirmatory factor analyses (CFA) were conducted. A one factor model with all six

PAGE 60

47 items loading onto a single factor was specifi ed, as well as a two f actor model with the paid labor items loading onto one factor and the family labor items loading onto another. The fit statistics for these models ar e listed in Table 8. Th e fit statistics were uniformly better for the 2 factor model and the 2difference test was significant, suggesting that the fit of the tw o factor model was statistically superior to that of the one factor model. Second, reliabili ty analyses were conducted. The reliability analysis for paid labor suggested that the items were all highly correlated with high internal consistency reliabilities ( = .89). The reliability analys is for family labor suggested that the emotion work item was weakly corr elated with the other items. Additionally, the coefficient alpha was low ( = .56). The emotion work item was dropped from the scale and the CFA and internal consistencies we re recomputed. The fit of the two factor model without emotion work was significantly be tter than the original two factor model. Additionally, the internal consistency reliability improved ( = .66). Based on these analyses, all three items we re retained for the pre-child desires for division of paid labor scale. The emoti on work item was dropped from the pre-child desires for division of family labor scale, resulting in a two item measure for this construct. From a theoretical standpoint, emo tion work is considered to be an important component of family labor (Coltrane, 200; Shleton & John, 1996; Erikson, 1993). However, it is reasonable to suspect that be cause it is less concrete than childcare or household work, it is simply not a construct that people consider before they have children. Thus, participants may have had di fficulty accurately answering this question, an idea which was in fact expressed by a participant when the study measures were piloted.

PAGE 61

48 To create the final scales, the percen tages for Partner A’s desired spousal contributions were averaged across the two items to create the family labor needs variable and across the three family labor item s to create the paid labor needs variable for Partner A. These also served as m easures of labor demands for Partner B. Post-child actual division of labor Two three-item scales were created to assess the current division of labor in the paid and family domains. These measures paralleled those used for pre-child for division of labor desires, assessing three components in the paid labor domain (work hours, income, and career prioritization de cisions) and three components in the family labor domain (child care, household work, and emotion work). These items differed in that they asked about the current division of labor rather than asking participants to recall desires for the division of labor before children were born. The following instructions were given before the set of items in order to make the timeframe clear: “For the next set of questi ons, please think about the present time.” An example item for paid labor is “Again, pi cture the total inco me you and your spouse earn from paid labor as a pie chart. Current ly, what proportion of this pie is made up of your income, and what proportion is made of your spouse's income? (These numbers should add up to 100%).” An example item from family labor is “Currently, what percentage of childcare tasks do you and your spouse perform? (These numbers should add up to 100%).” The definitions of each form of family labor were reported earlier in the survey when assessing pre-child desire s for division of labor, and thus were not repeated. For both measures, the response scal e was in the format of percentages, and choices were listed in increm ents of 5% from 0% to 100%. There were two responses

PAGE 62

49 for each item, one for self and one for spous e. The complete scales are listed in Appendix B. Identical procedures as listed above fo r pre-child desires for division of labor were used to examine the items within the ne wly created scales. For Partner A, the CFA suggested that the two factor model fit the data significantl y better than the one factor model. The reliability analyses suggested that the items in paid labor were highly correlated with each other a nd that the items for family labor were fairly highly correlated with each other (l ower correlations were observed for the emotion work item). The coefficient alphas were also acceptable for paid labor ( = .79) for and family labor ( = .74). For Partner B, the two factor model also fit the data significantly better than the one factor model. Internal consistency results sugge sted that items were highly correlated with each other in their respective domains and that internal consistency reliability was acceptable for paid and family labor ( = .84 and .80, respectively). An important condition of P-E fit analys es is that the person and environment variables are commensurate, containing the sa me content measured on the same metric (Edwards et al., 1998). Based on this notion, despite the CFA and reliability analyses suggesting that the three item scale was accep table for post-child actual division of family labor, it was modified to match th e pre-child desires for division of labor measure. When the emotion work item was removed, the two factor model for Partner A and Partner B fit the data better than th e original two factor model, although these differences were not sign ificant according to the 2 difference test. The internal consistency reliability improved with the shortened measure for Partner A ( = .84) and

PAGE 63

50 Partner B ( = .82). All three items were retained for paid labor. The CFA fit statistics are listed in Table 8. To create the paid labor supplies variable for Partner A, the three percentages of Partner A’s spouse’s contributi ons to paid labor were aver aged to create a scale. Similarly, for the family labor supplies vari able for Partner A, the two percentages of Partner A’s spouse’s contributions to family labor were averaged to create the measure. The paid (family) labor abilities variable for Partner B was created by averaging the three (two) percentages of Partner B’s se lf contributions to paid (family) labor. Career satisfaction Career satisfaction was measured using Greenhaus, Parasuraman, and Wormley’s (1990) five item career satisfaction s cale. An example item is “I am satisfied with the success I have achieved in my car eer.” Response options were set on a five point Likert scale that ranges from strongly disagree to strongly agree. The internal consistency reliability was high fo r Partner A (.89). All items are listed in Appendix C. Family satisfaction Family satisfaction was meas ured with a four item scale adapted from Cammann, Fichma n, Jenkins, and Klesh (1979). The scale was created to measure work satisfaction, but the items were ad apted to reflect the family context. An example item is “I am satisfied with my pr esent family situation.” Response options were set on a five point Like rt scale that ranged from st rongly disagree to strongly agree. Coefficient alpha was .87 for Partner A. Items are listed in Appendix D. Marital satisfaction Marital satisfaction was assessed using Norton’s (1983) five item Quality of Marriage Index. An example item is “We have a good marriage.” Response options were set on a on five point Li kert scale ranging fr om strongly disagree

PAGE 64

51 to strongly agree. The scale demonstrated high reliability, as was .97 for Partner A and.96 for Partner B. The full measure is listed in Appendix E. Depression. Depression was measured using Quinn and Shepard’s (1974) 10 item scale. The original scale was set in a wo rk context, but it was modified to refer to life in general over the past 3 months. Th e scale includes a mixture of positively and negatively valenced items, such as “I feel downhearted and blue” and “I still enjoy the things I used to do.” The positively valenced items were reverse coded so that higher scores indicated more depr ession. A five point Likert scale ra nging from strongly disagree to strongly agree was used as the response scale. The internal consistency reliability was acceptable ( = .82 for Partner A and .80 for Partner B). The full measure is listed in Appendix F. Physical health symptoms. A checklist including physical symptoms from the National Study of Daily Experiences (Ryff, 2005) was used to assess this construct. Example symptoms include upset stomach or nausea, headache, and backache. For the current context, the stem of the question wa s altered to ask about general frequencies rather than daily experiences: “Over the past 3 months, how often have you experienced the following symptoms?” Responses were set on a six point scale ranging from never to 5 or more times a week. Rather than averaging, responses to each item were summed to calculate a final score. In order to reduce missing data, in cases where participants answered the majority of items but left a few unanswered (three or fewer), the mean of the remaining items was imputed for that ite m’s value. Because this is a causal indicator scale, coefficient alpha is not m eaningful. Items are listed in Appendix G.

PAGE 65

52 Domain centrality Career centrality was measured using Lobel and St Clair’s (1992) four item adaptation of Lodahl and Ke jner’s (1965) job involvement scale. As has been done in previous research (e.g., Eddleston, Veiga, & Powell, 2006), family centrality was also assessed with the same items, but the word “career” was replaced with “family.” Example items are “A major source of satisfaction in my life is my career” and “Most of the important things that happen to me involve my career.” Response options were set on a five point Likert scale that ranged from strongly disagree to strongly agree. Both scales demonstrated acceptable internal consistency reliabilities (car eer centrality: = .82, family centrality: =.79for Partner A ). Items are listed in Appendix H. Gender. Gender was assessed using a one item question asking participants to indicate their gender. It was dummy coded for analyses (male = 0, female = 1). Voice in division of labor decision making As there were no known existing measures of this construct, a new measure was created based on a review of previous research on general voice in decision maki ng. Four studies were found that included unique measures of the construct (Brockne r, Ackerman, Greenberg, Gelfand, Francesco Chen, et al., 2001; Campion, Medsker, & Higgs, 1993; Denton & Zeytinoglu, 1993; Steel & Mento, 1987). Many of these items were specific to decision making in a particular area, thus the items had to be ad apted to address divisi on of labor decisions. Three items were adapted from Brockner et al. (2001), two items were adapted from Steel and Mento (1987), and one item was adapted from Denton and Zeytinoglu (1993) and Campion et al. (1993), resu lting in a total of seven it ems in the new scale. Two separate scales were created for family and paid labor. Items were identical except the

PAGE 66

53 labels “family” and “paid” labor were inte rchanged (e.g., “In genera l, I have a lot of opportunity to present my views about decisi ons that affect the division of family labor.”) The response scale was a five point Likert scale, with response options ranging from strongly disagree to strongly agr ee. Items are listed in Appendix I. One and two factor CFAs were cond ucted with the 14 items. The two factor model with the family items loading on one f actor and the paid labor items loading on another fit the data significantly better than the one factor model (see Table 8), but the fit statistics, particularly the RMSEA, were lo wer than desirable. Thus, further analysis of the items was conducted via exploratory fact or analysis and examin ation of inter-item correlations. As illustrated in Table 9, the EFA suggested the presence of two factors based on Kaiser’s (1960) rule of eigenvalues over 1.0 and examination of the elbow in the scree plot (Cattell, 1966). The item lo adings on the rotated factor matrix were consistent with expectations with the fa mily and paid items clearly loading onto independent factors. Additionally, all item s were highly correlated with other items within the respective paid or family labor domain. Based on these results and the high internal consistency reliabilities (.95 for family labor and .95 for paid labor), all seven items were retained for each scale. Satisfaction with curre nt division of labor A few studies have previously assessed satisfaction with current family labor divisions (e.g., Rhoades, Petrella, Stanley, & Markman, 2006; Wicki, 1999) with one item measures. In order to avoid problems associated with single item measur es (Bergkvist & Rossiter, 2007; McIver & Carmines, 1981; Nunnally & Bernstein, 1994; Spector, 1992), a thre e-item scale was created using one item adapted from Rhoades et al. (2006) (“In ge neral, how satisfied

PAGE 67

54 are you with the way you and your partner divide the family tasks?”). An example of a new item is “I am pleased with the amount of family labor that I perform relative to my spouse.” No known studies have previously assessed satisfaction with current division of paid labor. Thus, the ne wly created satisfaction with current family labor division scale was adapted for paid labor, substituti ng the work “family” for “paid.” Response options were set on a five point Likert scale, ranging from strongly disagree to strongly agree. The full measures are listed in Appendix J. The results of the CFA suggested that the two factor m odel fit the data significantly better than the one factor model. The CFI a nd TLI fit statistics were acceptable for the two factor model, but the RMSEA was above the conventional .10 cutoff. Thus, a similar procedure as descri bed with voice in divisi on of labor decision making was conducted using EFA and inter-i tem correlations. The EFA suggested a two factor solution and all items loaded hi ghly on the pre-specified paid and family labor factors. Additionall y, the inter-item correlations and internal consistency reliability estimates were good for paid labor ( = .86) and for family labor ( = .88). Based on these findings, all thr ee items were retained for each scale. The CFA results are listed in Table 8 and the EFA results a nd inter-item correlations are listed in Table 10. Control Variables. Due to their associations with the dependent variables, length of marriage, total family income, and family responsibility were included as control variables. Length of marriage was assessed by asking partic ipants the month and year of their current marriage as well as the m onth and year of any previous marriages. Total family income was measured through one item: “Please indicate your total family

PAGE 68

55 income (i.e., the combined salaries of yourself and your spouse).” Family responsibility was measured by asking part icipants to list the month and year of each of their children’s births and each child’s living arrangements. These values were then converted to ages and weighted differe ntially according to Rothausen’s (1999) responsibility for dependents scale (see Appendix K). Demographic information. Demographic information was collected from participants, including age, race/ethnicity, j ob title, education level, and career stage. These measures are listed in Appendix L. Inclusion criteria. As previously mentioned, some of the inclusion criteria were listed at the start of the survey whereas mo re specific information was included only as a question. In order to determine whether par ticipants had a child under age 6, the survey asked them to list the month and year of each child’s birth. This information was used to determine child age (42 couples were eliminated because of this criterion). Participants were also asked to indicate the month and year of all marriages. This information was then compared to the child ’s birthday and any pa rticipant who listed a child being born before the date of their current marriage was excluded from analyses (N = 12). Parental leave was examined by asking participants whether they were currently on parental leave from paid employ ment. Those selecting yes were excluded. (N = 7). These exclusions led to the final sample of 126 dyads.

PAGE 69

56 Chapter Three Results Preliminary Analyses Data screening. Before conducting analyses, the da ta were screened for outliers using boxplots and frequency tables. In all cases where outliers were identified, it was determined to be a case of data entry error ra ther than a valid outli er. These data entry errors were corrected by examining the orig inal data file from the online server. Furthermore, several of the items required participants to manually add percentages up to 100% (self and spouse breakdowns for pre-ch ild desires for paid and family labor and post-child actual divisions for paid and family labor). In approximately 5% of cases for each variable, the total values for self and spouse did not add up to 100%. Every instance where the sum for a variable was not 100% was independently examined. In the majority of cases, the error seemed to be a mathematical one, as the sum was 90% or 110%. In these situations, a proportional scal e was used to convert the 90% or 110% to 100%. For example, if a partic ipant indicated that (s)he co ntributed 60% to paid labor and his/her spouse contributed 30%, these num bers were converted to 60/90 = 67% and 30/90 = 33%. When the sum was less than 90% or greater that 110%, the responses were deleted and were deemed mi ssing for analyses (N = 3). Assumption testing. Data were inspected for assumptions of regression, including independence, normality of dependent variables, normality of residuals, and homoscedascticy of residuals (Hays, 1994). The assumption of independence is largely

PAGE 70

57 a design question, and there is no reason to suspect that couples’ responses are dependent upon other couples’ responses. This notion was further corroborated by inspecting the Durbin-Watson autocorrelation coefficient for each regression equation. In all cases, the values fell within the 1.5 to 2.5 range, wh ich indicate independence of data. Normality of dependent variables was assessed via histograms, boxplots, and calculating skewness and kurtosis values. Ca reer satisfaction and depression appeared to be normally distributed, ma rital and family satisfaction were negatively skewed, and physical health symptoms was positively skew ed. The negative skew in marital and family satisfaction is not surprising, given th e social desirability of these variables as well as the potential for range restriction, as those who are very unsatisfied with their marriages tend to divorce. Similarly, the pos itive skew of physical health symptoms is theoretically meaningful as being sick is a deviation from the healthy norm and many who are experiencing health problems obtain me dical help to correct these issues. Normality of residuals was assessed through histograms and quartile-by-quartile (q-q) plots for each equation. The residuals from equations predicting career satisfaction, depression, and health appeared to be relatively nor mally distributed; however, the equations predicting marital sa tisfaction and family satisfaction did not produce a normal distribution. In both cases, kurtosis was evidence such that the data peaked around the mean and the there were le ss values at the tail s of the distribution than expected. This lack of normality in the residuals is likely a bypr oduct of the lack of normality in the distribution of the variables themselves. Lastly, plots of the actual values versus the predicted regression line were used to examine homoscedasticity of error variance. All errors appear ed to be randomly distribution.

PAGE 71

58 Given that simulation studie s suggest that regression pr ocedures are fairly robust to normality violations and that transformi ng data can create interpretation problems (Norman & Streiner, 2008), th e analyses were conducted us ing the original data. Analysis of Hypotheses Hypotheses 1 through 8 and 15 throu gh 20 were tested using polynomial regression analysis. For each hypothesis, a qua dratic equation was estimated with the well-being variable of interest as the dependent variable, and the person and environment variables as the independent variables. Based on these values, three additional terms were computed (i.e., supplies (abilities) s quared, the product of s upplies (abilities) and needs (demands), and needs (demands) squared) The general form of the equation is Z = b0 + bc1C1 + bc2C2 + bc2C3 + b1X + b2Y+ b3X2 + b4XY + b5Y2 + e, where Z is the dependent variable, C1, C2, a nd C3 are the control variables, X and Y are the two fit components, b0 is the y-intercept, and e is the error term. All independent variables (with the exception of control vari ables) were scale centered to aid in interpretation. In cases where the initial polynomial re gression equation was significant, the slope and curve along the misfit line was tested by setting the Y = -X in the previous equation, so that Z = b0 + bc1C1 + bc2C2 + bc2C3 + (b1 b2 )X + (b3 b4 + b5)X2 + e. This equation indicates that along the misf it line, the curvature of the surface is represented by the quantity (b3 b4 + b5) and the slope of the su rface at the point where X (supplies) = 0 is represented by (b1 b2). When curves or slopes were significant, the data were further examined to determine whet her the nature of the significant slopes and curves were consistent with the predicted shape. This was done via response surface methodology (Edwards & Parry, 1993), where s upplies (abilities) were plotted on the X

PAGE 72

59 axis, needs (demands) on the Y axis, and th e dependent variable on the Z axis. As illustrated in Figure 3, the bold line across the diagonal, the misfit line, is the curve that is interpreted for the hypotheses. As the misf it line moves from the left corner to the right corner, supplies increase toward needs and when the dotted line (the fit line) is crossed, supplies (abilities) exceed needs (demands). Moderator hypotheses (Hypotheses 914 and 21-24) were tested using hierarchical regression analysis (Cohe n, Cohen, West, & Aiken, 2003). For each quadratic equation, the five aforementione d person and environment terms were multiplied by the relevant moderator va riable, resulting in the equation: Z = b0 + bc1C1 + bc2C2 + bc2C3 + b1X + b2Y+ b3X2 + b4XY + b5Y2 + b6M + b7XM + b8YM+ b9MX2 + b10MXY + b5MY2+ e, where M is the moderator va riable. To test for significant moderator effects, the significance of the change in R2 from the original equation without the moderato r terms to the one with the moderator terms was assessed. Coefficients from the regression equations for low, medium, and high levels of the moderating variables were plotting using response surface methodology to determine the nature of the moderating eff ect (Edwards & Rothbard, 1999). A power analysis was conducted and reveal ed that the power was rather low with the alpha level set at .05 (pow er was .53 for a medium effect size). One way to help reduce Type II error when power is low is to increase the alpha to .10 (Aguinis, 1995). Thus, the decision was made to in terpret results as significant if p < .10. This resulted in greater power (.66) to detect a medium effect size. Before running analysis, the variability of the fit variables was examine via two dimensional scatter plots, where the supplies (abilities) variable on the X axis and the

PAGE 73

60 needs (demands) variable on the Y axis for both family and paid labor. All four scatter plots exhibited a similar pattern: the majority of the data fell within the upper left, lower left, and upper right quadrants representing situations wher e supplies (abilities) fell short of needs (demands), supplies (abiliti es) and needs (demands) were congruent at low levels, and supplies (abilities) and need s (demands) were congr uent at high levels, respectively. Less data appear ed in the lower right quadran t, where supplies (abilities) exceeded needs (demands). These patterns sugge st that greater estimation must occur in the needs = supplies (demands = abilities ) curve when supplies exceed needs, as the estimated regression line is based off fewe r actual data points. The response surfaces should thus be interpreted with this in mind. Hypothesis Testing The means, standard deviations, and inte rcorrelations of all study variables were calculated and are presented in Tables 4, 5, 11. Although all variables are presented for both Partner A and B, the shaded variables in these tables represent the variables that were used in hypothesis testing. For ease of interpretation, an abbreviated correlation matrix with the fit variables and moderators on the vertical axis and well-being variables on the horizontal axis is also presented (Table 12). The results for Hypotheses 1, 3, 5, and 7 are presented in Table 13, and results for Hypothesis 2, 4, 6, and 8 are presented in Table 14. Hypothesis 1 predicted that Partner A’s career satisfaction would increase as paid labo r supplies (Partner B’s paid labor contributions) increased toward paid la bor needs (Partner A’s desires for Partner B’s paid labor contributions), and would decrease as supplies exceeded needs. The overall regression equation was significant ( F (8, 114) = 2.50, p < .05), but the curve of

PAGE 74

61 the misfit line was not significant (b3 b4 + b5 = -.003, ns ). The slope of the misfit line was significant (b1 b2 = .0297, p < .01). The response surfac e (see Figure 11) revealed that, consistent with prediction, career sa tisfaction increased as supplies increased toward needs. However, contrary to pred iction, as supplies exceeded needs, the curve remained relatively flat. Therefore, alt hough part of the response surface followed a trend similar to prediction, the full hypothesis was not supported. Hypothesis 2, that Partner A’s family satis faction would increas e as family labor supplies (Partner B’s family labor contribu tions) increase toward family labor needs (Partner A’s desires for Partner B’s family labor contributions), and would decrease as supplies exceed needs was supported. The overall equation was significant ( F (8, 114) = 1.81, p < .09), as was the curve of th e needs = supplies line (b3 b4 + b5 = -.0013, p< .05). As illustrated in Figure 12, the curve closel y resembled that of the predicted curve. Hypothesis 3 predicted that Partner A’s marital satisfaction would increase as paid labor supplies (Partner B’s paid labor contributions) increased toward paid labor needs (Partner A’s desires for Partner B’s pa id labor contributions), and would decrease as supplies exceeded needs. The hypothesis was supported ( F (8, 114) = 2.49, p < .05, b3 b4 + b5 = -.0011, p < .01). The shape of the needs = supplies line was consistent with prediction (see Figure 13). Hypothesis 4 was identical to Hypothesis 3 but focused on family labor. The hypothesis was supported, as the regression equa tion was significant ( F (8, 114) = 2.50, p < .05), as was the curve of the needs = supplies line (b3 b4 + b5 = -.0011, p< .08) .The response surface is illustrate d in Figure 14 and is consistent with prediction (marital satisfaction increased as family labor supplies increased toward needs and decreased as supplies exceeded needs).

PAGE 75

62 Hypothesis 5, that Partner A’s depression would decrease as paid labor supplies (Partner B’s paid labor cont ributions) increase to ward paid labor needs (Partner A’s desires for Partner B’s paid labor contri butions), and would increase as supplies exceeded needs was supported. Both the overall equation ( F (8, 114) = 1.95, p < .06), and the curve of the misf it line were significant (b3 b4 + b5 = -.0009, p < .05). As illustrated in Figure 15, the pattern of the re lationship is consistent with prediction. Hypothesis 6 focused on depression and family labor needs-supplies fit, predicting a curvilinear relationship. Th e hypothesis was not supported ( F (8, 111) = 1.46, ns) Hypothesis 7 focused on the relationship between Partner A’s paid labor needs and supplies and physical health symptoms, pr edicting that Partner A’s physical health symptoms would decrease as paid labor suppl ies (Partner B’s paid labor contributions) increased toward paid labor needs (Partner A’s desires for Partner B’s paid labor contributions), and would increase as supplies exceeded needs. The hypothesis was supported, as the polynomial regression equation ( F (8, 111) = 2.71, p < .05), and curve were significant (b3 b4 + b5 = .0053, p < .09). The graph of the response surface (Figure 16) was similar to that of the pr edicted surface, although the effect was not symmetrical, as the slope of the curve as supplies increased toward needs appeared steeper than the slope as supplies exc eeded needs. No support was found for Hypothesis 8, which was identical to Hypothe sis 7 but focused on family labor needs and supplies. Although the overall re gression equation wa s significant ( F (8, 111) = 32.17, p < .01), the curve of the misf it line was not significant (b3 b4 + b5 =.0001, ns ). Hypothesis 9 involved the moderating eff ect of Partner A’s career centrality on the relationship between paid labor supplies (Partner B’s pa id labor contributions) and

PAGE 76

63 paid labor needs (Partner A’s desires for Pa rtner B’s paid labor contributions) fit and Partner A’s well-being ((a) career satisfact ion,(b) marital satisfaction, (c) depression, and (d) physical health symptoms). Specifical ly, the relationships between shortage of supplies and well-being were hypothesized to be weaker when career centrality was higher rather than lower and that the re lationships between ex cess supplies and wellbeing were hypothesized to be stronger when career centrality was higher rather than lower. As displayed in Table 15, there was no evidence of a significant moderating effect for career satisfaction ( R2 = .038, ns ), depression ( R2 = .013, ns ), or physical health symptoms ( R2 = .018, ns). There was evidence for a moderating effect on marital satisfaction ( R2 = .067, p < .07). As illustrated in Figure 17, the nature of the moderation of career centrality was consistent with prediction for ex cess supplies, as the slope of the curve as supplies exceeded n eeds was steeper when career centrality was higher than when it was lower. It was not c onsistent for shortage of supplies, as the slope of the curve as supplies approached n eeds was steeper, indica ting a stronger rather than weaker relationship to marital satisfaction, when caree r centrality was higher than when it was lower. Although the trend w ith excess supplies was consistent with prediction, Hypothesis 9b as a whole was not supported. Hypothesis 10 stated that Partner A’s family centrality moderates the relationship between family labor supplies (Partner B’s fa mily labor contributi ons) and family labor needs (Partner A’s desires for Partner B’s fa mily labor contributions) fit and Partner A’s well-being ((a) family satisfaction, (b) ma rital satisfaction, (c) depression, and (d) physical health symptoms). The relationshi ps between shortage of supplies and wellbeing were predicted to be weaker when fam ily centrality was higher rather than lower,

PAGE 77

64 and the relationships between excess supplie s and well-being were predicted to be stronger when family centrality is higher rather than lowe r. This hypothesis was not supported for family satisfaction ( R2 = .009, ns ), marital satisfaction ( R2 = .029, ns ), depression ( R2 = .03, ns ), or physical health ( R2 = .032, ns ). The results are presented in Table 16. Hypothesis 11 focused on the moderating role of gender. Specifically, it was predicted that Partner A’s gender would mode rate the relationship between paid labor supplies (Partner B’s paid labor contributions ) and paid labor needs (Partner A’s desires for Partner B’s paid labor contributions) f it and Partner A’s wellbeing ((a) career satisfaction, (b) marital satisfaction, (c) depre ssion, and (d) physical health symptoms). Specifically, the relationships between shor tage of supplies and well-being and excess supplies and well-being were expected to be weaker when Partner A was a female than when Partner A was a male. The results are presented in Table 17. The hypothesis was not supported for career satisfaction ( R2 = .016, ns) marital satisfaction ( R2 = .047, ns ), or depression ( R2 = .069, ns ). Evidence of a significant moderator effect was found with physical health symptoms ( R2 = .115, p < .05). However, inspection of the response surfaces in Figure 18 suggests that the nature of this moderation was not consistent with expectations. That is, for males there was a slight increase in physical health symptoms as supplies approached needs a nd slightly past the poi nt of perfect fit. After this point, physical health symptoms began to decrease. On the other hand, the response surface for females suggested that physical health symptoms decreased as supplies approached needs and increased as supplies exceeded needs. However, the

PAGE 78

65 curve was not symmetrical, as the slope as supplies approached needs seemed to be greater than that of excess supplies. Hypothesis 12, that partner A’s gender would moderate the relationship between family labor supplies (Partner B’s family labor contributions) and family labor needs (Partner A’s desires for Partner B’s family labor contributions) fit and Partner A’s (a) family satisfaction,(b) marital satisfacti on, (c) depression, and (d) physical health symptoms) was not supported. There was not a significant change in R2 in any of the dependent variables (family satisfaction: R2 = .05, ns; marital satisfaction: R2 = .025, ns; depression: R2 = .026, ns; physical health symptoms: R2 = .048, ns ). See Table 18 for results. Hypothesis 13 predicted that Partner A’s voice in division of paid labor decisions would moderate the relationships between pa id labor supplies (Partner B’s paid labor contributions) and paid labor needs (Partner A’s desires for Partner B’s paid labor contributions) fit and Partner A’s well-be ing ((a) career satisfaction, (b) marital satisfaction, (c) depression, and (d) physic al health symptoms), such that the relationships between sh ortage of supplies and well-being and between excess supplies and well-being would be weaker when voice in division of labor decisions was higher rather than lower. As displayed in Tabl e 19, the hypothesis was not supported for career satisfaction ( R2 = .035, ns), depression ( R2 = .052, ns ) or physical health symptoms ( R2 = .024). Evidence of a significant m oderator effect was found for marital satisfaction ( R2 = .097, p < 05); however, the response surface (Figure 19) suggested that the nature of the moderation was opposite that of prediction. Specifically, the relationship between shortage of supplies and marital satisfaction and excess supplies

PAGE 79

66 and marital satisfaction was stronger when voice in division of labor decision making was higher rather than lower. Hypothesis 14 was not supported. Partner A’s voice in division of family labor decisions did not significantly moderate the relationships between family labor supplies (Partner B’s family labor contributions) and family labor needs (Partner A’s desires for Partner B’s family labor contributions) fit and Partner A’s well-being (family satisfaction: R2 = .038, ns ; marital satisfaction: R2 = .047, ns ; depression: R2 = .019, ns ; physical health symptoms: R2 = .032, ns ). The results are displayed in Table 20. The results for Hypotheses 15, 17, and 19 are presented in Table 21, and results for Hypotheses 16, 18, and 20 are presented in Table 22. Hypotheses 15, 17, and 19 focused on the relationship between Partner B’s paid labor demands -abilities fit and marital satisfaction, depression, and physical health symptoms, respectively. None of these hypotheses were supported, as the regression equations were all non-significant (marital satisfaction: F (8, 116) = 1.18, ns ; depression: F (8, 116) = .55, ns ; physical health symptoms: F (8, 111) = 1.15, ns ). Hypotheses 16, 18, and 20 focused on the relationship between Partner B’s family labor demands-abilities fit and marital satisfaction, depression, and physical health symptoms, respectively. Hypotheses 16 and 18 were not supported; the regression equations for marital satisfaction and depression were non-significant ( F (8, 116) = 1.05, ns, F (8, 116) = 1.67, ns, respectively). With regard to Hypothesis 20, the regression equation for physical health symptoms was significant ( F (8, 111) = 2.04, p < 05). The slope of the demands = abilities line was significant (b1 b2 = -.2233, p < .01) but the curv e was not significant (b3 b4 + b5 = -.0021, ns ). The response surface (Figure 20) suggested that as predicted,

PAGE 80

67 physical health symptoms decreased as abi lities approached demands. Contrary to prediction, physical health symptoms con tinued to decrease as demands exceeded abilities. Thus, although the trend for shorta ge of abilities matche d that of prediction, Hypothesis 20 as a whol e as not supported. The results for Hypothesis 21 are pr esented in Table 23. The hypothesis predicted that Partner B’s gender would mode rate the relationship between paid labor demands (Partner A’s desires for Partner B’ s paid labor contributions) and abilities (Partner B’s paid labor cont ributions) fit and Partner B’ s well-being, such that the relationships between shortage of abilities and well-being and between excess abilities and well-being ((a) marital satisfaction, (b) depression, and (c) physical health symptoms) were stronger when Partner B was female rather than male. Hypothesis 21 was not supported. The R2 was significant for marital satisfaction ( R2 = .094, p < .05); however, the response surfaces illustrated in Figure 21 suggest that the nature of the moderation was not consistent with pred iction. Specifically, the misfit curve was relatively flat for males and marital satisfaction generally increased for females for both shortage and excess of supplies. The R2 were not significant for depression ( R2 = .037, ns ) or physical health symptoms ( R2 = .034, ns ). Hypothesis 22 was identical to Hypot hesis 21 but focused on the interaction between gender and family labor demands-abi lities fit rather than paid labor fit for partner B. The results are presented in Table 24. Although evidence for significant moderating effect of gende r was found for depression ( R2 = .079, p < .07) and physical health symptoms ( R2 = .073, p < .08), the nature of the moderation was inconsistent with prediction (that the relati onships between shortage of abilities and excess abilities

PAGE 81

68 and well-being would be stronger for females than males). With regard to depression, the shape of the demands = -a bilities curves were as expect ed but in the reverse manner (see Figure 22 for illustration). That is, th e relationships between both excess supplies and shortage of supplies and depression were stronger for males than females. With regard to physical health symptoms, the response surfaces shown in Figure 23 suggest that demands-abilities fit exhibited a quite di fferent pattern for males and females, with the curve being relatively flat for females and in curving in the opposite direction of males. There was no evidence of gender moderation for marital satisfaction ( R2 = .028, ns ). Thus, Hypothesis 22 was not supported. Hypothesis 23 stated that Partner B’s satis faction with curren t division of paid labor would moderates the relationships be tween demands (Partner A’s desires for Partner B’s paid labor contribu tions) and abilities (Partner B’s paid labor contributions) fit and Partner B’s well-being ((a) marital satisfaction, (b) depre ssion, and (c) physical health symptoms). The nature of the mode ration was expected to be such that the relationships between shortage of abilities and well-being and between excess abilities and well-being were weaker when satisfacti on with current paid labor divisions was higher rather than lower. As shown in Table 25, the R2 was significant for marital satisfaction ( R2 = .136, p < .01), depression ( R2 = .088, p < .05 ), and physical health symptoms ( R2 = .07, p < .09), providing evidence for a mo derator effect. The response surfaces revealing the nature of the moderations are pres ented in Figures 24 (marital satisfaction), 25 (depression), and 26 (physical health symptoms). For marital satisfaction, the relationship be tween shortage of abilities and marital satisfaction and the relationship between excess abilities and marital satisfaction were

PAGE 82

69 both weaker when satisfaction was higher rather than lower. However, the shape of the curve was opposite prediction, such that marita l satisfaction was lowe st at the point of fit. Thus, Hypothesis 23 was not supported fo r marital satisfaction. For depression, the shape of the curves for low and high satisfac tion with current divisi on of labor exhibited a pattern opposite of prediction. Specifical ly, the relationship between shortage of abilities and depression a nd excess abilities and depre ssion were stronger when satisfaction was high rather than low, m eaning Hypothesis 23 was not supported for depression. With regard to physical hea lth symptoms, no support for the hypothesis was found, as the nature of the moderation was not consistent with prediction. The response surface graphs revealed that when satisfacti on with current division of paid labor was low, physical health symptoms decreased as abilities approached demands and increased as abilities exceeded demands. At medium leve ls of satisfaction with current division of paid labor, the relationships were relatively flat; however, at high levels of satisfaction, physical health symptoms increased as abil ities approached demands and decreased as demands exceeded abilities. Hypothesis 24 was similar to Hypothesis 23 but focused on the satisfaction with current division of family labor and family demands-abilities fit. Results were nonsignificant for marital satisfaction ( R2 = .061, ns ) and physical health symptoms ( R2 = .027, ns ). Evidence of a significant modera ting effect was found for depression ( R2 = .08, p < .05); however, the nature of this mode ration was inconsistent with prediction (that the relationships between shortage of abilities and depression and between excess abilities and depression would be weaker wh en satisfaction with current family labor divisions was higher rather than lower). As illustrated in Figure 27, across all levels of

PAGE 83

70 satisfaction with current divi sion of paid labor, the rela tionship between shortage of abilities and depression was relatively consta nt. At low levels of satisfaction with current division of paid labor, the relations hip between excess demands and depression continued to decrease, whereas it increased at medium and high levels of satisfaction. Therefore, Hypotheses 24 was not supported. The results are displayed in Table 26 and Figure 33. Table 27 summarizes the resu lts of all the hypothesis testing. Supplementary Analyses In addition to testing the formally proposed hypotheses, a few subsequent analyses were conducted. First, hypothese s involving only Partner A (Hypotheses 1 – 14) were tested using responses from all initial respondents who completed the survey and met the inclusion criteria, regardless of whether their spouse also responded (N = 578, of whom 126 had spousal data and were us ed to test hypotheses in the previous section). Before conducting the polynomial regression analyses, independent t-tests were used to compare the non-matched portion of the full sample to the matched sample in order to determine if the groups differed on meaningful variables. The results are presented in Table 28 in the column labeled “F (1 vs. 3).” Like ly a function of the breakdown of gender (87.3% female in full sample vs. 50% female in matched sample), the non-matched sample appeared less egalita rian in attitude and behavior than the matched sample. Specifically, they desired th eir spouse to contribute less to family labor and more to paid labor before childre n were born, they reported that their spouses actually contributed less to family labor and mo re to paid labor after children were born, their career centrality was lower, and their fa mily centrality was higher. Thus, there are

PAGE 84

71 some important characteristics that differ between the groups that should be kept in mind when interpreting the following results. The results of the hypothesis testing w ith the full sample of 578 Partner As revealed few differences from the previous hypothesis testing with the smaller sample. There was only one case, Hypothesis 6 (conc erning the relationship between paid labor needs-supplies fit and depr ession), where a hypothesis was supported with the full sample but not with the matched sample. Th e response surface suggested that the curve was consistent with the predicted shape, such that depression decreased as family labor supplies increased toward family labor needs and increased as supplies exceeded needs. Additionally, voice in division of paid la bor decision making significantly moderated the relationship between needssupplies fit and career sati sfaction (Hypothesis 13a), but the shape of the response surface did not support the hypothe sis. Contrary to the prediction that the relationship between shor tage of supplies and career satisfaction and excess supplies and career satisfaction w ould be weaker when voice was higher, the response surface suggested the opposite patter n. That is, the relationship between shortage of supplies and caree r satisfaction was weaker when voice in division of labor decisions was lower versus medium or higher. The relationship between excess supplies and career satisfaction was relatively consistent across all levels of voice in paid labor decision making. Moreover, although no formal hypotheses were proposed about the level of desire-division of labor fit (f it where mutual desires and ac tual contributions are lower versus higher), this issue was examined in an explor atory manner. The main effect analyses for Partner A’s well-being are firs t discussed, followed by Partner B. When

PAGE 85

72 the overall regression equation was significant for a given re lationship, the significance of the slope and curve of the fit line (n eeds = supplies or demands = abilities) was tested. These results are presented in th e last rows of Tables 13, 14, and 22, where b1 + b2 represents the slope and b3 + b4 + b5 represents the curve of th e fit line. The response surfaces that were plotted for the hypotheses are rotated to give a better view of the fit line and are presented again in Figures 28 – 32. Because different rotations were used with different graphs in order to maximize viewing, a line is drawn in the appropriate diagonal for each graph to represent the part of the curve that should be interpreted as the fit line. For paid labor needs-supplies fit and Pa rtner A’s well-being, there was either a significant slope or curve of the fit line for each form of well-being. For career satisfaction, the slope of th e fit line was significant (b1 + b2 = -.0097, p < .05). As shown in Figure 28, the slope is negative, such that as fit increases from fit at mutual low values to mutual higher values, career sa tisfaction decreases. With regard to marital satisfaction, the curve of the fit line was significant (b3 + b4 + b5 = .0004, p < .05). As displayed in Figure 29, the curve generally in creases, such that marital satisfaction increases as the level at which fit occurs increases. The results for depression and physical health symptoms were similar to each other in pattern. In both cases the slope of the fit line was positive a nd significant (b3 + b4 + b5 = .0097, p < .05 for depression; b3 + b4 + b5 = .0688, p < .05 for physical health symptoms), and the response surfaces (Figures 30 and 31), show that depression and physical health symptoms increase as the level at which fit occurs increases.

PAGE 86

73 For family labor needs-supplies fit, the overall polynomial regression equations were significant for family satisfaction, marital satisfaction, and physical health symptoms; thus, the significance of the slope and curve were only tested for those three dependent variables. With regard to family satisfaction, neither th e slope nor the curve of the fit line was significant. With regard to marital satisfaction, both the slope and the curve of the fit lin e were significant (b1 + b2 = -.0117, p < .05; b3 + b4 + b5 = .0006, p < .01). The rotated response surface (Figure 32) shows a Ushaped curve, such that marital satisfaction decreases as the level at which fit occurs incr eases until it reaches medium levels of fit, at which marital satisfaction begins to increase as the level at which fit occurs increases. The depression resu lts suggested that the slope of the fit line was significant (b1 + b2 = .0060, p < .10), and the curve of the fit line was as well (b3 + b4 + b5 =.0048, p < .01). Because neither the slope nor curve of the misfit line was significant, the response surface was not previ ously graphed but is provided in Figure 33. The response surface suggests an inverted U-shaped curve, such that physical health symptoms increase as the level at which fit occurs increases until it reaches medium levels of fit, at which marital satisfaction be gins to decrease as the level at which fit occurs increases. Only one regression equation was signifi cant for relationships involving Partner B’s well-being (family labor needs-supplies fi t and physical health symptoms). Neither the slope nor the curve of the fit line was significant. In order to examine the shape of the f it line in all of th e moderator hypotheses, the response surfaces of relationships wh ere evidence of significant moderation was found were rotated. The rotated curves are not presented; instead, Table 29 summarizes

PAGE 87

74 the shape of the fit line for each level of the moderator variable s. In general, the shapes of the misfit curves did differ across the di fferent levels of the moderators; however, there was no consistent discernable pattern.

PAGE 88

75 Chapter Four Discussion The aim of this study was to examine how misfit between dualearner couples’ pre-child division of labor preferences and pos t-child actual divisions of labor in the paid and family labor domains related to we ll-being. Using P-E f it as the theoretical framework, this question was addressed with in the context of needs-supplies and demands-abilities fit, with a focus on well-being for both members of the marital dyad. Although some previous research exists on th e topic, this study made several unique contributions to the literature by exam ining both the work and family domains, assessing fit in an all-enco mpassing manner, testing for moderator effects, and using more appropriate statistical analyses (i .e., polynomial regression) to examine the research questions. In the subsequent sections, these fi ndings and further summarized and interpreted, followed by a discussion of theo retical and practical implications, study limitations, and future directions. Summary and Interpretation of Results Needs-supplies f it and satisfaction. Based on P-E Fit theory (French et al., 1982) and self-discrepancy theory (Higgins, 1987, 1989), I predicted th at the extent of congruence between Partner B’s post-child la bor contributions (supplies) and Partner A’s pre-child desires for Partne r B’s labor contributions (nee ds), would relate to domainspecific and marital satisfaction. In all cases the relationships were expected to take the

PAGE 89

76 form of an inverted U-shaped curve, such that satisfaction would be maximized at the point of fit and lowest at the points of extreme misfit in either direction. Beginning with paid labor, the results fo r desire-division of paid labor fit and career satisfaction only supporte d part of the hypothesis. Th at is, in situations where supplies fell short of needs (where Partner B contributed less postchild than Partner A desired pre-child), career satisfaction increase d, consistent with expectations. On the other hand, past the point of fit when supp lies exceed needs (Partner B was contributing more to paid labor than Partner A desired), career satisfaction remained constant, rather than decreasing as predicted. This asympt otic relationship is somewhat puzzling, as research suggests that career compromise may relate to less career satisfaction (Tsaousides & Jome, 2005), and it may be difficult to be successful at work, which is often a component of satisfaction (Heslin, 2005), when scaling b ack (Becker & Moen, 1999). One potential explanation for the obser ved pattern of result s lies in the burden that financially providing for one’s family can carry. Research s uggests that there are some notable emotional downsides to greater participation than desi red in paid labor, including feelings of worry, pressure, a nd resentment, as the family’s financial livelihood is more dependent upon the indivi dual’s paid labor contributions (Gerson, 1993; Gilbert, 1985; Meisenbach, 2010). In turn these negative sentiments may have an effect on attitudes at work, fo stering resentment toward th e job. In situations where supplies exceed needs, and Partner A is cont ributing less than (s)he initially desired, the burden of the paid labor role is lifted, as are its negative effects on career satisfaction. This study represents the first known direct inquiry about dual-ear ner couples’ division

PAGE 90

77 of paid labor and career satisfaction. Clear ly, more research is needed to truly understand the observed pattern of effects. With regard to paid labor and marital satisfaction, findings we re in line with expectations. Marital satisfac tion was maximized at the po int of fit and suffered as supplies deviated from needs in either dire ction. Said otherwise, spouses tend to be dissatisfied in their marriages when they feel that their partner is contributing too much or too little to the paid work domain, as it may impact their own need fulfillment or threaten their identity. The finding that th e division of paid labor impacts a family construct, marital satisfaction, further reinfo rces the notion that work and family are highly intertwined (Kanter, 1977). For optim al functioning within a marriage, dualearners should be aware of th eir spouse’s career desires a nd carefully consider how their own career choices affect their spouses’ wo rk situation (Haddock & Rattenborg, 2003). Moving to the division of family labor and satisfaction, results were consistent with the hypothesized relationships for bot h family and marital satisfaction. These findings highlight the delicate nature of dividing family la bor between spouses. Despite the fact that family labor is often view ed as undesirable (Coltrane, 2000), having a spouse who does too much of this disagree able work is associ ated with negative sentiments toward the spouse and family in general. According to self-discrepancy theory, this may be a function of identity. When one’s spouse does more than the ideal share of family labor, one’s ow n family identity is challenged and negative emotions in that domain may ensue. Furthermore, lending greater credence to the marital satisfaction results, they ar e consistent with a previ ous longitudinal study on the childcare labor (Khazan et al., 2008), where women’s marital satisf action was found to

PAGE 91

78 decrease three months post-partum when th eir husbands contributed less to childcare than the wives desired before child birth. Needs-supplies fit and ment al and physical health. Depression and physical health symptoms were also examined as correl ates of needs-supplies fit. A shortage of needs relative to supplies is considered a st ressor, which may trigger a wide variety of psychological and physiological strain re actions in the body (e.g., Beehr & Newman, 1978; Kahn & Byosiere, 1992; Lazarus & Folk man, 1984) with potential chronic effects (Frankenhaeuser, 1986; Johnson et al., 1990). Ba sed on this notion, it was predicted that as Partner B’s labor contribut ions increased toward Partne r A’s desires for Partner B’s labor contributions, depression /physical health symptoms w ould decrease to the point of fit. Building on the ideas that excess supplie s create self-discrepancies which are linked to health through negative emotions (Higgi ns et al., 1986; Kiecolt-Glaser, McGuire, Robles, & Glaser, 2002) it wa s predicted that as Partne r B’s labor contributions surpassed Partner A’s desires for Partner B’ s labor contributions, depression/physical health symptoms would increase. In sum, a U-shaped relationship was hypothesized. The results for desire-division of paid labor fit were consistent with the hypotheses for depression and physical health symptoms. These results are significant, implying that the consequences of desire-div ision of labor fit exte nd beyond attitudes in the work and family domains to health, an area that permeates all functioning. Previous P-E fit research has found that needs-supplies fit at work relates to physical and mental health (Caplan, Cobb, French, Harrison & Pinneau, 1980; Edwards, 1996; Edwards & Harrison, 1993; Edwards & Rothbard, 1999). Based on the present results, it seems that

PAGE 92

79 fit extends to a more macro level, and well-being is maximized when the work situation fits within the context of empl oyee’s broader family life. Contrary to the paid la bor results, the hypotheses con cerning division of family labor and depression and physical health symp toms were not supported. However, when the relationship between paid labor needs-supplies fit and depression was examined in the supplementary analysis with the larger pool of Partner As (N = 578), the equation and curve were significant and the res ponse surface revealed a U-shaped curve consistent with prediction. In order to determ ine if the smaller sample showed a similar, albeit non-significant trend, th e b weights from the initial regression equation were plotted. The trend was consistent with pred iction. Based on these results, it seems likely that there is a relationship between desire-d ivision of paid labor fit and depression, but the effect size is small (and therefore was not detected in the initial analyses with lower power). The non-significant results for physical he alth symptoms do not appear to be a function of power. Although the overall polynomial regression equation was significant for physical health symptoms, further analysis revealed that this was a function of curve in the fit line, rather than misfit line. In fact, the response surface shows that the misfit line is very flat, suggesting that there is not even a small association between paid labor fit and physical health. One potential explana tion for the null results is that the stressorstrain relationship within the desire-division of family labor misfit context is not potent enough to have chronic effects on health-relate d outcomes, particularly those that are physical in nature. Previous research corroborat es this to some exte nt, as Milkie et al.

PAGE 93

80 (2002) found that the relationship between perc eiving that husbands contribute less than the ideal amount to parenting and wives’ stre ss reactions (i.e., strain) was small. As a whole, the health results beg the question of why the stress-strain relationship is apparently weaker with desire -division of family labor than with paid labor. When considering shortage of supplie s, the differences may be a function of the more controllable nature of the family domain compared to work. The division of paid labor tends to be less amenable to change and more dependent upon external factors, whereas family members have greater contro l over their actions at home (e.g., it is more feasible for a spouse to do more housework than to find a higher pa ying job). In turn, control plays an important role in reduci ng the impact of stress on strains (Karasek & Theorell, 1990). With regard to excess suppl ies, the differences revolve around identity permanency. Family identity is one that is somewhat persistent no matter the amount of labor contributed (e.g., being a pare nt is always to some extent part of a parent’s identity no matter how much he or she is contributing to childcare). The work role identity is arguably more highly tied to actual current be havior at work (e.g., identity as upper level manager). In this sense, excess supplies may have some effect on family identity but not a large enough effect to produce severe self -discrepancy and mental and physiological health reactions. The aforemen tioned explanations are tentativ e, and their merit can only be determined through further research on the mediating mechanisms between desiredivision of family labor fit and health. Supplementary analysis of fit line in n eeds-supplies and well-being relationships. In an exploratory manner, the slope and cu rve of the fit line was examined for each significant regression equation from the hypothesis testing. Th e fit line provides

PAGE 94

81 information about variation in well-being that is associated with th e absolute levels of needs and supplies when fit is held co nstant (Edwards & Rothbard, 1999). For desire-division of paid labor fit, there was a signi ficant slope or curve of the fit line for all four well-being variables. For career satisfaction, de pression, and physical health symptoms, the general trend was that as the absolute level of fit increased, wellbeing decreased. In other words, indivi duals who both wanted their partners to contribute a large amount to paid labor before children were born and whose partners did contribute a large amount to paid labor after ch ildren were born were least satisfied with their careers, more depressed, and repor ted experiencing more physical health symptoms. It is not surprising that career satisfaction is lowest at highest absolute levels of fit, given that career satisfaction is dr iven by factors such as hours spent at work, training and development opportunities, and career sponsorship (Ng, Eby, Sorensen, & Feldman, 2005). Those who contribute a small amount to paid labor, as is likely the case when fit occurs at high absolute high levels of the components, are apt to have fewer opportunities to foster career satisfaction. Moreover, the depression and physical he alth results support Barnett and Hyde’s (2001) expansionist theory, which emphasizes th e mental and physical health benefits of multiple role occupation (Simons, 1992;Thoits 1989), particularly when these roles are of high quality. Although the amount that one contributes to paid labor relative to his/her spouse is not a dir ect indication of poorer paid labor role quality, there is arguably an association. Those who work fe wer hours or are in lower paying jobs are less likely to reap the same benefits from the worker role than their more involved counterparts.

PAGE 95

82 The results for division of paid labor f it and marital satisfaction are the opposite of those for the three other forms of well-be ing. Marital satisfaction tends to increase as the absolute value of fit increases, meani ng that those who had high paid labor needs (desired their spouse to cont ribute a great deal) and high labor supplies (the spouse does contribute a large amount) were most satisfied in their marriages. Also, this relationship is not strictly linear, as the rate at which ma rital satisfaction increases is less at low levels of fit than high levels of f it. Because these individuals have higher paid labor needs, they necessarily are placing greater demands on their spouses than those with lower labor needs. When the spous e in turn meets these high de mands, it may generate even more positive affect toward the marital part ner, given the extremity of the needs. Alternatively, there may be fundamental diffe rences in the types of people who have their high paid labor needs met versus those with lower needs. These differences could represent a third variable that also relates to marital satisfaction. Finally, the supplementary analyses with desire-division of family labor fit showed a significant curve of the fit line for marital sati sfaction and physical health symptoms. In both cases, well-being was lowest at medium levels of fit and highest at the extremes. This is surprising, given that previous research suggests that a more equitable division of family labor relates to better marital satisfaction (Cooke, 2006; Stevens, Kiger, & Mannon, 2005) and health (Booth & Johnson, 1994). However, other research suggests that a major disadvantage of such equity is the negative impact it can have on marital relations due to the stressful daily negotiations and compromises (Rosenbluth, Steil, & Whitcomb, 1998). Hence, it is possible that the constant striving for equity that those couples at medium levels of fit face has some negative

PAGE 96

83 repercussions. Furthermore, the expression of gratitude for a spouse’s labor contributions is related to ma rital satisfaction (Hochshild, 1989). When both spouses are contributing a medium amount to labor, the other’s spouse’s contri butions may not be readily apparent and both may be less oblig ed to express gratitude resulting in poorer well-being. In summary, alt hough the absolute leve l of fit was not the focus of the present study, the findings sugge st that it is a ripe ar ea for future research. Moderators of needs-supplies fit and well-bein g. With the exception of gender, previous research has not considered boundary conditions that temper the strength of the relationships between desire-div ision of labor fit and well-bei ng. In an effort to fill this gap, three potential moderators were examined in the present study – domain centrality, gender, and voice in division of labor decision making. Overall, ther e was little evidence for moderators. Evidence of a moderator eff ect was found with three of the paid labor – well-being relationships; however none were fully consistent with the proposed pattern of moderation. No significant interactions were found with desire-division of family labor fit and well-being. The theoretical rationale behind the first moderator, domain centrality, stemmed from French et al.’s (1974) original theory of P-E fit, where the individual importance of the needs and supplies were proposed to have a meaningful impact on the effect of needs-supplies misfit. Greater importan ce of needs should relate to more grave consequences of misfit. In the desire-divisi on of labor context, those who find paid labor central to their identity s hould be more affected by ex cess paid labor supplies (as it translates into less contribution to paid labor for them) and less affected by shortage of paid labor supplies (as it translates into more contribution to paid labor form them)

PAGE 97

84 compared to those with less car eer centrality. The same patte rn as predicted for family centrality and division of family labor. Evidence of significant moderation of career centrality was found only with marital satisfaction. As predicted, excess s upplies (having a partne r that contributed more than desired) related more strongly to marital satisfaction. Contrary to prediction, shortage of supplies also rela ted more strongly to marital sa tisfaction. In summary, those with lower career centrality appear less sens itive to any type of needs-supplies misfit than those with higher career centrality. From a theoretica l standpoint, it is difficult to determine why the results were contrary to ex pectation for shortage of supplies. From a statistical standpoint, there is reason to suspect that the negative correlation between the moderator and dependent variable (career cen trality and marital satisfaction correlation is -.24) could create issues with interpretation of the mode rator effect (Baron & Kenny, 1986). Given that there was no evidence th at career centrality moderated the relationship between fit and ot her forms of well-being (caree r satisfaction, depression, and physical health symptoms), it seems that career centrality only impacts the desiredivision of paid labor fit in one way – desire-division of labor fit matters very little in predicting the marital satisfaction of those who have low career centrality. Family centrality was predicted to mode rate the relationships between desiredivision of family labor fit and family sa tisfaction, marita l satisfaction, depression, and physical health symptoms. No significant m oderating effects were found. Inspection of the descriptive statistics for family centrality reveals that the mean score on this variable was very high with little variance (mean = 4.53; SD = .54). Thus, the moderator analyses comparing those low in family centra lity to those high in family centrality were

PAGE 98

85 in reality comparing two groups with rather high scores on the variable, meaning these hypotheses could not be tested at the level that the theoretical rationale assumed. It is noteworthy that the present study is not the fi rst to encounter this problem, as Edwards and Rothbard (1999) report similarly high means for family centrality. Although it is possible that the true population mean score on this trait is very high, it is likely that scores are inflated to some extent by social desirability bias (an example item is “I am very much involved personally in my famil y.”). Perhaps a more appropriate way to assess this construct is through a measure with work-family conflict scenario-based items that force participants to state whethe r they would engage in the work or family event. A measure of this nature is certainl y not immune to social de sirability bias, but it may be more effective in reducing the bias th an Likert scale items. Nonetheless, future researchers are urged to consider the measur ement of domain centrality carefully. Gender was also investigated as a modera tor of both desire-division of paid and family labor fit. Owing to women’s tendenc y to expect and prepare for struggles in managing work and family (e.g., Orrange, 2003; Sanders et al., 1998; Spade & Reese, 1991), I hypothesized that women’ s well-being would be less affected by needs-supplies misfit in the work and family domains. Ther e was little evidence for a gender effect, as a significant moderator effect was only observed in one instance – the relationship between desire division of paid labor fit and physical health symptoms. The response surfaces revealed that the nature of the significant moderator effect was not consistent with predic tion. Desire-division of paid labor fit mattered relatively little for men, but their well-being was maximi zed at the extreme point of undersupplies, or, in other words, when wive s were contributing much less to paid labor than husbands

PAGE 99

86 had initially desired. The obser ved pattern for women was mu ch different. Women were in the poorest health at the extreme point of undersupplies and in the best health at the point of fit. There was an asymmetrical cu rvilinear relationship, such that as husband’s labor contributions approached pre-child desi res, physical health symptoms decreased. Past the point of fit, where labor contribu tions were in excess, symptoms began to increase again but at a much slower rate. This pattern of results was contrary to prediction but may be explained by the general stress reactivity and gende r socialization literatures. Fi rst, there is some evidence that women experience more acute stress -related somatic symptoms and physical illnesses than men (Jick & Mitz, 1985; Woff ord, Daly, & Jubin, 1999) and that this is particularly marked when the stressors invol ve family issues (Zuckerman, 1989). Thus, although women may be more able to effectivel y cope with desire-div ision of labor misfit than men (Loscocco & Spitze, 2007), this may be offset by their increased propensity to react to stress in general. Second, there are some health benefits to having a partner who contributes relatively little to the paid do main and focuses more on the family domain, in terms of more healthy family meal preparation (Neumark-S tzainer, Hannan, Story, Croll, & Perry, 2003) and less stress surrounding work-family conf lict (Eagle, Icenogle, Maes, Miles, 1998), which could explain men’s tenden cy to be healthiest when spouse’s paid labor supplies are low. However, this does not account for the marked gender difference, as women are least healthy when spouse’s paid labor supplie s are low. What might account for this difference is gender socialization of the br eadwinner role. When husbands contribute significantly less than wives desi red to paid labor, they are minimizing their identity as

PAGE 100

87 breadwinner and placing more of the bread winner burden on the women. While gender roles have become increasingly more egal itarian, the notion of breadwinner is still considered a standard for male but not female identity (Janssens, 1998; Meisenbach, 2010; Warren, 2007). Due to differences in soci alization, the pressure that comes with the breadwinner burden may simply be stronger for women who have not been socialized to take on such a role. Furthermore, research suggests that husbands of female breadwinners may react negatively to wives due to masculine identity threat (Atkinson, Greenstein, & Lang, 2005; Buzzane ll & Turner, 2003; Doucet & Merla, 2007) and often actually reduce their labor at home (Bittma n, England, Folbre, Sayer, & Matheson, 2003), factors which may further ex acerbate wives’ stress and physical health reactions. On the surface, the lack of significant ge nder differences in the present study is surprising, given that most previous desire-d ivision of labor studies found some type of gender effects (e.g., Khazan et al., 2008; Lo scocco & Spitze, 2007; Milkie et al, 2002; Ross et al., 1983). Scrutiny of th ese studies suggests that ther e is great inconsistency in the nature of gender differences across studi es. Although some of this inconsistency may be attributable to the different research contexts, it is reasonable to suspect that some of the variation stems fr om within-gender variation. That is, there is as much variation in most personalit y traits, work tendencies, and family behaviors within genders as there is across genders (Bar nett & Hyde, 2001), making it difficult to accurately predict how a person of a given ge nder will feel or behave. Rather than focusing on the objective characteristic of gender, it would be useful for future researchers to investigate constructs that are more phenomenologically rich, such as

PAGE 101

88 gender role identity and the meaning ascrib ed to work and family roles (Eby, Casper, Lockwood, Bordeaux, & Brinley, 2005). Voice in division of labor decision making was the final variable considered as a moderator in the division of labor needs-s upplies fit and well-being relationships. Based on the general voice in decision making liter ature (e.g., LaTour, 1978; Lind et al., 1980), and a few studies specific to voice in work and family role structure (Hiller & McCaig, 2007; Madden, 1987; Thompson & Walker, 1989) voice in division of labor decision making was predicted to serve as a buffer betw een pre-child division of labor desires and post-child division of labor reality fit and we ll-being. Said otherwis e, both shortage and excess of supplies were predicted to have a stronger relationship with well-being when voice was low rather than high. Results were largely unsupportive of the voice in division of labor decision making moderator hypotheses. There was only once instance when a significant moderating effect was found: paid labor needs-supplies fi t and marital satisfaction. However, the effect was in direct oppositi on to the hypothesis. Thus, low rather than high voice seemed to serve as a buffer agains t the negative effects of misfit. Although laboratory studies suggest that voice improves outcome satisfaction and process appraisal, it seems to function differently in the context of dual -earner’s division of labor. There may be a backlash effect that occurs when voice in decision making is high, but the division of labor st ill does not reflect one’s de sires. Specifically, if an individual has little input into a process and the process ends in desire-division of labor misfit, their marital satisfaction may be minimally affected, as the situation was never discussed. On the other hand, if an indi vidual makes his/her opinion known and the

PAGE 102

89 division of paid labor still e nds up incongruent with his/her desires, the spouse may feel that (s)he was ignored, sparking negative sentimen ts toward the marital partner. In fact, research suggests that the us e of ignore tactics when negotia ting division of labor relates to lower marital satisfacti on (Alafita, 2008). Perhaps a more fruitful approach to understanding the moderating role of the decision making process would be to investigate decision control, or the actual degree of infl uence one has over a decision (Thibaut & Walker, 1975). Deci sion control is distinct from process control (i.e., voice) which refers to the extent to which peopl e have an opportunity to expression their opinions. Theoretically, less backlash shoul d occur with decision control, as by definition it implies influence, rather than just opinion, over the decision. With regard to career satisfaction, ther e was no evidence of vo ice as a moderator effect of paid labor needs-s upplies fit with the main sample but it was significant when tested with the larger sample of Partner As. The pattern of the moderation was relatively consistent with that of marital satisfaction, su ch that the relationship between shortage of supplies and career satisfaction was stronge st when voice was high. There were no differences with excess supplies. It is impo ssible to know whether th e lack of significant results in the smaller sample is a function of power or if the effect of voice were fundamentally different in the two samples. Nonetheless, the fact that the pattern was similar to that of marital satisfaction lends greater credence to the unexpected findings. Demands-abilities fit and well-being. The first set of study hypotheses were focused solely on one member of the marital dyad (“Partner A”). In the next set of hypotheses, the spouse (“Partner B”) was also considered, but in th e context of another form of P-E fit, demands-abilit ies fit. Specifically, Partner A’s needs, or the amount of

PAGE 103

90 labor (s)he desired Partner B to contribute before children were born, were considered environmental demands on Partner B. Partner B’s abilities to fulfill these demands were represented by Partner B’s post-child actual la bor contributions. Based on the core ideas of demands-abilities fit theo ry and self-discrepancy th eories, well-being was again predicted to be highest at the point of fit and lowest at points of extreme misfit. None of the hypotheses were fully supporte d for paid or family labor. Family labor demands-abilities fit was significantly a ssociated with Partne r B’s physical health symptoms, but the relationship was linear. Consistent with prediction, as Partner B’s family labor contributions approached Part ner A’s pre-child desires for Partner B’s contributions, Partner B’s physical health symptoms decreased; however, contrary to prediction, as Partner B’s contributions surp assed Partner A’s desires physical health symptoms continued to decrease. In other wo rds, the more Partner B contributed relative to his/her spouse’s desires, the healthier (s)he was. These findings are puzzling as extreme situations of excess supplies imply that Partner B is doing an disproportional share of family labor, which has been negatively linked to both men and women’s health (Bird & Fremont, 1991) and household strain (Golding, 1990). Future researchers should further investigate this perplexing fi nding to determine th e conceptual link. Speculatively, there is a third variable that is common to situations where abilities exceed demands and good physical health. Regarding this set of hypotheses as a whole, there ar e several potential explanations for the mostly null results. Fi rst, on a conceptual le vel, the effects of demands-abilities fit on well-being are indi rect, operating through needs-supplies fit (Edwards & Shipp, 2007; Harrison, 1978; La wler, 1973; Locke, 1976; Smith et al.,

PAGE 104

91 1969). In this sense, environmental demands must either be intern alized as needs or meet another personal need that enhances we ll-being. Because previous research shows that spouses are typically aw are of and responsive to each other’s needs (Clark, 1984; Clark et al.,1986; Eisenberg & Miller, 1987), it was assumed that Partner A’s pre-child desires for Partner B’s labor contributions acted as demands on Partner B and were internalized as Partner B’s own needs (the need to satisfy Partner B’s desires). This was a theoretical assumption, not a process that was actually measured in the present study. Thus, a lack of demand intern alization as needs may have halted the demands-abilities fit process and contributed to the null results. A second explanation concerns the di stinction between the objective and subjective representation s of the demands and abilities (E dwards et al., 1998; French et al., 1982; Harrison, 1978). Part ner A’s reports of his/her own desires for Partner B’s labor contributions were used as the demands variable for Partner B, representing the objective environment. The subjective envi ronment, or Partner B’s interpretation of these demands, was not assessed. This could ha ve impacted the results in the sense that the objective environment is a more distal predictor, meaning its effects on fit and subsequent well-being were lik ely smaller and more difficult to detect. Additionally, the imperfect relationship between the object ive and subjective environment (Harrison, 1978) may have distorted the relationship. Partner B may have incorrectly interpreted Partner A’s needs, or (s)he may have never have been aware of them in the first place. Research suggests that many couples do discus s their future work-family lives before children are born (Steffy & Ashbaugh, 1986; Wri ght, 2001), but this is certainly not an absolute. One way to better unde rstand the null results is to also measure the subjective

PAGE 105

92 environment, or Partner B’s pe rceptions of Partner A’s prechild desires for the division of labor, in future research. Third, meeting the division of labor demands imposed by one’s partner may simply not be a salient enough demand to impact well-being. Within a marriage, two sets of needs are at play, and spous es’ pre-child desires for paid labor contributions are not always in sync. To illustrate, the correlat ion between Partner A’s pre-child desires for Partner B and Partner B’s pre-child desires fo r him/herself is .35 for family labor and .67 for paid labor. When Partner B’s pre-child de sires for division of la bor are different than Partner A’s, Partner B’s ow n needs-supplies fit (congruen ce between how much Partner B wanted Partner A to contribute before children and Partner A’s actual post-child contributions) may simply be an overriding force. Finally, there may be important moderators of the relationships that are concealing significant main effects. Two moderators that seem to account for so me variation are discussed below. Moderators of demands-ab ilities fit a nd well-being. Personal and contextual factors that should theo retically temper the relationship between demands-abilities fit and well-being were also investigated, namely gender and satisfaction with the current division of labor. Owing to women’s greater responsiveness to ot hers (Gilligan, 1982) and empathetic orientations (Eisenbe rg & Lennon, 1983), both forms of demandsabilities misfit were expected to have a st ronger relationship to well-being for women compared to men. Evidence of a signifi cant moderator effect was found in three instances: paid labor fit and marital satisfaction, family la bor fit and depression, family labor fit and physical health symptoms.

PAGE 106

93 None of the moderating effects were consis tent with prediction and as a whole did not follow a discernable pattern. Beginning with desi re-division of paid labor and marital satisfaction, the moderator effect along the misfit line was quite small. The curve for men and women was similar, although ther e appeared to be a slightly stronger relationship among excess abilitie s and marital satisfaction for women. It is important to note that this relationship was pos itive, rather than negative as predicted, meaning that as Partner B contributed more to paid labor th an Partner A desired him/her to contribute, Partner B’s marital satisfaction increased. There are many benefits of high participation in paid labor, as it is likely to lead to personal career advancement which affords status, independence, and job satisfaction, among ot her things (England & Kilbourne, 1990; Oppenheimer, 1997). Perhaps when Partner B pl aces a greater investment in paid labor than Partner A desires, Partner B recognizes the compromise Partner A had made and as a result is more satisfied with the relations hip. This may be especially marked for women, whose greater career contribu tion is contrary to traditi onal gender norms. Moreover, a stronger moderator effect appear s along the fit line. While th e absolute level of fit did not affect women’s marital satisfaction, ther e was a mostly positive relationship for men, such that marital satisfaction generally increa sed as higher levels of fit (and decreased slightly at extreme high levels). Thus, men seem to be happiest in their marriages when their wives wanted them to contribute a rela tively large amount to pa id labor and they did so. Gender also significantly moderated the re lationships between desire-division of family labor fit and depression and physical h ealth symptoms. In both cases, the shape of the misfit curve for men resembled the prediction in the main effect – a curvilinear

PAGE 107

94 relationship where depression and physical heal th symptoms were lowest at the point of fit. The effect of demands-abilities fit for women was not consistent across the two health outcomes. With regard to depressi on, women were less affected by shortage of abilities. In other words, de pression decreased at a slower rate for women as their labor contributions increased toward their part ner’s desires. This was opposite of the hypothesized effect that women would be more affected. For physical health symptoms, the response surface curve was starkly different for women than men. It was relatively linear, and physical health symptoms decrea sed as Partner B’s contributions approached Partner A’s desires for Partner B’s contri butions and continued to decrease as contributions exceeded desires. Overall, it se ems that men’s health is affected by desiredivision of family labor fit in a reliable manne r, consistent with prediction. On the other hand, the nature of the relationshi ps for women is difficult to in terpret. There tends to be greater variance in women’s amount of family labor contributions than men’s (General Social Survey, 2006), and this lack of hom ogeneity may be part of the reason why the results for women are inconsiste nt and rather inexplicable. The last set of hypotheses examined the moderating role of Partner B’s satisfaction with current divisi on of paid and family labor. Satisfaction with the current division of labor was expected to act as a buffer to the negative effects of demandsabilities misfit, for the nega tive repercussion of not mee ting your partner’s pre-child division of labor demands may be offset by the positive repercussions of personally being happy with the way labor is currently divided. A significant moderator effect was found for a ll three forms of wellbeing desire-division of paid labor fit, but none were consistent with each other, nor w ith prediction. When

PAGE 108

95 satisfaction with current division of paid labor was lower rather than higher, there was in fact a stronger relationship between both directions of demands-abilities misfit and marital satisfaction as predicted. However, the misfit curve was inverted, such that marital satisfaction was lowest at the point of fit, a nd highest at either extreme of misfit. For depression, results were opposite that of prediction, as Partner B was more affected by desire-division of labor misfit when satisf action with the current division of labor was high rather than low, but gain the relationshi p was such that depression was highest rather than lowest at the point of fit. Finally with physical he alth symptoms, the pattern was entirely different. When satisfaction with current division of paid labor was low, the pattern of relationships was as predicted, with the lowest symptoms at the point of fit and the highest at either extreme of misfit. At medium levels of satisfaction with current division of paid labor, there was no relationshi p between fit and healt h. At high levels of satisfaction with current divisi on of labor, the relationship wa s inverted, such that wellbeing was poorest at the point of fit. Finally, satisfaction with the division of current fam ily labor moderated a single family labor relationship: desire-division of family labor fit and de pression. Consistent with expectations, the relati onship between shortage of de mands (Partner B contributing less than Partner A wanted Partner B to cont ribute) and depression was stronger for those with lower satisfaction with the current division of family labor than among those with higher satisfaction. Inconsis tent with expectations, th e relationship between excess demands (Partner B contributing more than Partner A wanted Part ner B to contribute) and depression was stronger for t hose with higher satisfaction.

PAGE 109

96 In summary, although it is clear that satisfa ction with the curren t division of labor impacts the relationship between desire-divisi on of labor fit and wellbeing, the nature of this interaction is varied and in no case consistent with hypotheses. One potential explanation for this diversity of results is that dissatisfaction with the current division of labor can stem from different sources. Sp ecifically, one may be dissatisfied because (s)he is doing too much or dissatisfied because (s)he is doing too little. The nature of the satisfaction with current divi sion of labor measure used in this study did not account for these differences, and perhaps grouping analys es according to the type of dissatisfaction might produce results that are more clearly interpretable. Future researchers are encouraged to further expl ore these relationships. Theoretical Implications As a whole, the findings of the present study have several theoretical implications. Although there have been a few previous stud ies that addressed pre-child desires and post-child division of labor realities, none ha ve employed a strong theoretical framework. This study suggests that P-E fit theory, in particular the n eeds-supplies variant, is an appropriate framework, as the majority of n eeds-supplies fit and we ll-being relationships tested conformed to the theory’s predictions. Moreover, self-discrepancy theory seems to be complementary to needs-supplies fit theor y, providing a basis for the effects of excess supplies. On the other hand, demands-abil ities fit as a framework for understanding desire-division of labor that incorporates both members of the marital dyad received little support. Thus, while there is evidence that pr e-child desires act as a need for one spouse, they do not necessarily cross ove r and act as a demand for the ot her. This study served as a starting point for enhancing the theory behi nd this complicated cons truct, but there is

PAGE 110

97 definitely room for additional theory to help explain how spouses are influenced by each other’s division of labor needs. Likewise, the investigati on of contextual factors was largely unsuccessful, meaning other theoretica lly relevant moderating variables need to be examined before we can fully understand the impact of personal and situational factors on the desire-division of labor fit relationships. One of the primary aims of this study was to comprehensively examine the link between desire-division of labor fit and well-being in an al l-encompassing manner. This was accomplished in a variety of ways. First, three dimensions of the division of paid labor (income, work hours, and career prioritization decisi ons) were assessed, representing both an improvement in measur ement and a more complete understanding of the paid labor domain over past researc h. Second, by using polynomial regression analyses, I was able to capture the full range of misfit and avoid statistical problems associated with other methods of evaluating fit. Not only does this lend greater credence to the accuracy of results, but it also provided theoretical ad vancement. Specifically, in many cases, the effects of one’s partner contri buting too much were not identical to the effects of one’s partner contri buting too little. Thus, the pres ent study solidifies the idea that the different forms of misfit are unique a nd should be considered as such in research. Third, multiple forms of well-being were examined, expanding our understanding of the widespread implications of de sire-division of labor misfit, for the work, family, and general health domains were al l related to paid or family labor misfit. Finally, although few significant findings were found with th e dyadic analyses, the null results are informative and, as noted above, make it clear th at future theorizing is necessary. In sum, this study extended a relatively narrow line of research and provided a richer theoretical

PAGE 111

98 understanding of how lack of congruence betw een dual-earner’s desires for division of labor and actual post-child division of labor relates to well-being. This study also provides in sight about gender. Clearly gender is an important factor in the division of labor as studies consistently fi nd that women do more family labor and men do more paid labor (Coltrane, 2000; Moen & Roehling, 2005). However, there do not seem to be profound differences in the way that desire-d ivision of labor fit relates to men and women’s well-being. Previous research has found some gender differences in some forms of misfit, but ther e is considerably inconsistency within and across studies. As previously noted, there may be other moderators that correlate with gender (i.e., identity) that are actually accounting for this varian ce. In general, research would benefit from a closer in spection of the proce sses that link gender to desire-division of labor misfit, rather than making assumptions largely based on traditional roles. Additionally, organizational researchers tend to pay “lip service” to the division of family labor, frequently mentioning its importance in dual-earner’s work and family functioning but rarely empirically assessing it. The division of paid labor is neglected to the extent that a search of the phrase in psychology research databases produces zero studies. The results of this study suggest th at this is a signific ant omission, given the links between desire-division of labor fit and career satisfaction and health. At the very least, desire-division of labor fit represents an important family structure variable that should be considered, as it sa tiates complaints (Barnett & Hyde, 2001; Eby et al., 2005 Zedeck, 1992) that work-family research is too reliant on objective characteristics of the person to the neglect of role qua lity. Similarly, a key variable in family research, marital satisfaction, was the well-being variable that mo st consistently relate d to desire-division

PAGE 112

99 of labor fit and exhibit severa l significant moderator effects. Thus, family researchers are also encouraged to expand their typical theories to incorporate the quality of paid labor roles research, rather than distilling the work er role to “breadwinne r” as is often done (e.g., Brennan et al., 2001; Loscocco & Spitze, 2007; Potucheck, 1997). To sum up, both organizational and family research ers would benefit from including a comprehensive assessment of desire-division of labor research in th eory and research. Practical Implications In addition to theoretical explanations, th ere are a few practical implications of the present study. The results suggested that needs-supplies fit in the paid labor domain affects career satisfaction. To the extent that organizations wa nt to foster career satisfaction in their employees, they shoul d consider implementing programs that facilitate desire-division of labor fit. As Moen and colleagues (Moen, 2003; Moen & Roehling, 2005; Moen & Orrange, 2002) note, there is a structural lag in most organizations. Career paths are still mode led on the 1950s blueprint, assuming that workers have a full time homemaker to assist them with family res ponsibilities a nd career development. By reducing this lag and r ecognizing the dual-earner context that most workers operate in, organizations may facilitate employees’ ac hievement of a division of labor that is consistent with pre-child desires. Potentia l solutions include making careers more flexible with multiple paths to success, making jobs themselves more flexible by offering flexible work arrangements, and pl acing greater consider ation on spouse’s work situations in relocation decisions. Moreover, if the link between desire-divisi on of labor fit and career satisfaction is not enough to spark organizational interest in the construct, its effects on health should

PAGE 113

100 be. Beyond the added expenses in healthcare insurance, depression costs employers in the United States more than $35 billion annua lly due to reduced performance at work. Physical health symptoms are even more e xpensive, adding up to an estimated $47 billion in productivity losses (Hemp, 2004). Given these statistics, employers should be motivated to foster the mutable variable of employee desire-div ision of labor fit. On the family side, the study’s results ha ve implications for marital counseling. Early in marriages before children are born, couples should be advi sed to discuss their desires regarding the division of labor through va rious stages of their lives. Being aware of each other’s desires may help foster eventual desire-division of labor congruence and can help couples come to early compromises if their desires are in itially incongruent. Additionally, because the division of labor is sometimes a dynamic process, couples who have children might benefit from counseling that promotes consistent goal setting and monitoring to ensure that th eir behaviors stay in line with mutual desires. Limitations Despite its contributions, the present study has several limitations in terms of study design, measurement, sampling, and st atistical power. With regard to study design, retrospective reports were used to a ssess pre-child desires for division of labor, requiring participants to recall their division of labor desires before children were born. There is a great deal of evid ence that retrospective reports are prone to memory biases (cf., Bernard, Killworth, Kronenfeld, & Sailer, 1984). Many cognitive processes influence the recall process, including current emotions (Levine & Safer, 2002; Safer & Keuler, 2002 ) and attitudes (Ross, 1989) source memory errors ( Johnson, Hashtroudi, & Lindsay, 1993), and suggestibil ity effects (Loftus, 1982). Thus, the accuracy of the

PAGE 114

101 pre-child desires measure is unc ertain. The sole way to circumvent this issue is to conduct longitudinal research, assessing attitudes before children and born and postchild division of labor post-partum. Becau se the present study was addressing several novel concepts, the use of retrospective repo rts was deemed a suit able first step, but future researchers are urged to conduct longit udinal studies. Relatedly, a second limitation of the study is its cross-sectional design. Although the hypotheses were all grounded in P-E fit theo ry, which suggests that fit between needs (demands) and supplies (abilities) impact s well-being, the cro ss-sectional design precludes any firm conclusions about dire ctionality. Anothe r concern is common method variance. Even though dyadic data were collected, several of the study hypotheses relied solely on self-re ported variables (i .e., needs-supplies fit and well-being for Partner A). Assessment of common method variance’s influence is difficult (Spector, 2006), and doing so requires considera tion of the constructs of interest as well as the methods by which they were measured (Spector & Brannic k, 1995). In this case social desirability may be especially rele vant, as respondents may have been motivated to exaggerate their marital or family satisfaction as well as their current labor contributions. Beyond study design, there are a few measur ement concerns. Several scales were created for the study (pre-child desi res for division of labor, post-child actual division of labor, voice in division of labor decision making, and satisfaction with current division of labor) because previous measures did not exist or were not comprehensive. Of particular concern are the pre-child desires and post-child actual family division of labor scales. Due to low in ternal consistency reliability and inter item

PAGE 115

102 correlations, the item assessing emotion work was dropped from the pre-child desires scale and thus dropped from the post-child act ual contributions measure for consistency purposes. Even then, the internal consistency reliability of the pre-child desires for division of family labor scale was low ( .66), calling into question the measurement accuracy. The other measures appeared hi ghly internally reliable with clear unidimensional structure. Nonetheless, addi tional tests of their validity in external contexts would bolster conf idence in measurement. Emotion work is considered an important component of family labor (Erikson, 1993), but it has received relatively little em pirical attention (Coltrane, 2000) and no attention in the context of prechild desires. It is unclear whether this concept really exists (i.e., people do not consider their desires for emotion work before children), if it is simply difficult to assess using retrospectiv e reports, or if the problems with the item were due to wording. In genera l, it seems that more research is needed on emotion work, particularly people’s cognitions abou t it before children are born. Another limitation concerns the sample a nd sampling strategy. Participants were recruited through a variety of strategies, all of which i nvolved convenience sampling, rather than random sampling. This resulted in a sample that was not representative of the larger population, as pa rticipants were overwhelmingly White, highly educated, employed in mostly white-collar jobs, and had a high household income. These characteristics could have a profound effect on the division of labor, as past research suggests that blue-collar couples negotiate th e work and family domains differently than white-collar couples (Berardo, Shehan, & Leslie, 1987; Vanfo ssen, 1979) and that racial differences exist in the meaning of the divi sion of labor (Maret & Finlay, 1986; Shelton

PAGE 116

103 & John, 1993; Vega, Patterson, Sallis, Nader, Atkins, & Abramson, 1986). Consequently, the generalization of the results to the larger population is questionable. Moreover, the spouses were randomly selected as Partner A and Partner B for analysis. It is possible that a different randomization process could produce a different pattern of results. A second point of concern with the sampling strategy is the method by which dyadic data was obtained. There may have been meaningful differences in key study variables among participants w hose spouses participated and participants whose spouses did not participate. In an attempt to evaluate this issue empirically, t -tests were conducting comparing the sample of initial respondents whose spouses did not respond to respondents whose spouses did respond. This information is presented in Table 27, in the “F (1 vs. 2)” column. In a few instan ces, significant differenc es on study variables emerged. Initial respondents whose spouses did not respond contributed less to family labor and more to paid labor, reported lower career centrality, and had less voice in family division of labor decisions than those participants who were used in analyses. Therefore, it is possible that the results of the present st udy are biased toward couples with more egalitarian ideals and may not accu rately represent individuals with more traditional marriages. The final limitation is statistical in natu re. A power analysis suggested that the power was lower than ideal for detecting eff ects, particularly when using hierarchical moderated regression. A lack of power c ould have contributed to the erroneous dismissal of effects as null when in fact a m eaningful moderating effect did occur. In an attempt to reduce such Type II error, the al pha level was raised and relationships were

PAGE 117

104 interpreted as significant when p < .10. This however, also raised the possibility of Type I error, which should be kept in mind when interpreting results. Another way to improve power would be to increase the sa mple size, and future researchers are encouraged to do so. Future Directions The present study has paved the way for nume rous future research endeavors. In addition to the ideas for future research ideas presented in previous sections, there are four main areas that remain largely un tapped. First, the present investigation incorporated the division of family and pa id labor, but hypotheses and analyses were conducted independently. In another words, th ere was little consider ation of how desiredivision of labor fit in one domain affect s the other domain. Given that many dualearner work-family management strategies c onsider the work and family domains as a whole (Becker & Moen, 1999; Roehling, 2003), ra ther than each domain independently, research that incorporates both domains simultaneously would be illuminating. One idea is to explore desire-division of labor congruence in one domain as a moderator of the relationship between desire-division of la bor fit in the other domain. This would allow researchers to determine if the association between getting what you wanted in one domain is affected by getti ng what you wanted in the other domain. Theoretically, individuals with high desire-division of labor fit in the work domain may not experience increased well-being if there is very low fit in the family domain. Such a scenario is likely commonplace, as the av erage woman, regardless of employment status, work hours, or income still performs more family labor than her husband (Bianchi, Milkie, Sayer, & Robinson, 2000).

PAGE 118

105 A related idea is to include desire for the division of paid labor as one of the fit components and desire for the division of fa mily labor as the ot her fit component and compare their congruence to well-being. This design would allow researchers to address the research question of where the optimal level of respective desires lies. It may be that desiring 50% in both domains maximizes wellbeing or that a deviation from perfect equality is actually more beneficial. More over, there are several individual difference variables that might moderate this relati onship, such as gender and domain identities. Second, as was mentioned in the introduc tion, several researchers (e.g., Goldberg & Perry-Jenkins, 2004; Kalmuss, Davidson, & Cushman, 1992; Nicolson, 1990; Ruble, Fleming, Hackel, & Stangor, 1988; Van Egeren, 2004) have previously examined a topic that is distinct but related to pre-child desire-post child act ual division of labor fit – met expectations for division of labor. In the organizational l iterature, the concept of the psychological contract (Argyris, 1960; Rousseau, 1989, 1995, 2001) is often used to examine the impact of expectations on empl oyment relationships. In the present study the organizational concept of P-E fit was app lied to the dual-earner context, and may be informative to also apply psychological contracts to dual-earners. Based on psychological contract theories, the non-explic it agreements that couples have before children regarding the future division of la bor may have an important effect on how the actual division of labor pos t-children influences wellbeing. Understanding both how met desires and met expectations function in dual-earners would grea tly contribute to our theoretical understanding of the complexi ty of dual-earner’s work and family management.

PAGE 119

106 A third avenue of future exploration revolves around process variables. The findings of this study and previous res earch (Clarkberg & Moen, 2001; Moen & Roehling, 2005; Moen & Yu, 2000) show that th ere is often a discrepancy between what couples want pre-child and what they get post-ch ild with regard to the division of labor. Less is known about why this discrepancy o ccurs. Some theorists argue that is a function of changing preferences (Hakim 2000), economics (Becker, 1991) or power dynamics (McDonald, 1980), but there is mini mal research on the role of workplace factors in creating this discre pancy. Previous research s uggests that the spouse with greater flexibility may actually experience greater family to work conflict (Hammer, Neal, Newsom, Brockwood, & Colton, 2005). If one spouse’s job is more flexible than the other spouses, (s)he may take on more of the family activities. Over time, this may have an effect on the overall division of labor. Mechanisms linking desires and actual behaviors were not examined in the present study, but future researchers are encouraged to further explore these issues, with a focus on work-re lated processes. Finally, replication of th e research questions addressed in the current investigation would be valuable. Not only w ould replication with a larger sample size help ascertain if the null result s were a function of Type II erro r, but it would also bolster confidence in the polynomial regression an alyses. Although polynomial regression was the appropriate statistical test for the research questions (e.g., Edwards, 2002, 2007), there is some evidence that polynomial regr ession findings are prone to problems with replication (Yang, Levine, Smith, Ispas, & Rossi, 2008). Thus, cross-validation is particularly important in the present context.

PAGE 120

107 Conclusion The question posed by the title of this study is “You can’t always get what you want, but does it matter?” In short, the answer is yes. The results from the study suggest that congruence between an i ndividual’s own pre-child desi res for the division of paid labor and the actual post-child division of paid labor relate s to his/her own career and marital satisfaction, depression, and physical he alth symptoms. Congruence in the family domain, is also important as desire-divisi on of family labor fit related to affective sentiments toward family and one’s spous e. On the other hand, although one’s own desire-division of labor fit impacts their own satisfaction, I found little evidence that an individual’s well-bei ng relates to his/her abilities to meet spousal demands for the division of labor. Finally, moderators of th ese relationships were assessed, including domain centrality, gender, voice in division of labor decision making, and satisfaction with the current division of labor. Ev idence of moderation was only found in a few cases, and none were consistent with prediction, high lighting the need for future research on the contextual condition s of P-E fit in th e dual-earner context. In closing, this study represents an important extension of th e literature through its comprehensive and statistically sophisticated examination of dualearner couples desire-d ivision of labor fit and well-being.

PAGE 121

108 References Aguinis H. (1995). Statistical power problems with moderated multiple regression in management research. Journal of Management 21 1 141-1. Alafita, E.G. (2008). Communication and the division of labor about household tasks: Perceived strategies used to negot iate tasks in the Mexican household. Unpublished doctoral dissertation University of Texas – Austin. American Psychiatric Associa tion (2000). Diagnostic and sta tistical manual of mental disorders (4th ed.). Washington, DC: Author. Anderson, J.C., & Gerbing, D.W. (1988). Stru ctural equation mode ling in practice: A review and recommended two-step approach. Psychological Bulletin 103 (3), 411-423. Argyris, C. (1960). Understanding organisational behaviour Homewood, Illinois: Dorsey Press. Atkinson, M. P., Greenstein, T. N., & La ng, M. M. (2005). For women, breadwinning can be dangerous: Gendered res ource theory and wife abuse. Journal of Marriage and Family, 67 1137 1148. Atkinson, J., & Huston, T.L. (1984). Sex role or ientation and division of labor early in marriage. Journal of Personality and Social Psychology 46 (2), 330-345. Barnett, R.C., & Hyde, J.S. (2001). Women, men, work, and family: An expansionist theory. American Psychologist, 56 (10), 781-796. Barnett, R.C., Gareis, K.C., & Brennan, R.T. (1999). Fit as a mediator of the relationship between work hours and burnout. Journal of Occupational Health Psychology 4 (4), 307-317.

PAGE 122

109 Barnett, R.C. & Rivers, C. (1996). She works/he works: how two-income families are happy, healthy, and thriving San Francisco, CA: Harper Collins. Baron, R.M. & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: Conc eptual, strategic, and statistical considerations. Journal of Personality and Social Psychology 51 1173-1182. Beach, S.R.H. & Tesser, A. (1993) Decision making power and marital satisfaction: a self evaluation main tenance perspective Journal of Socia l and Clinical Psychology 12 471-494. Becker, G.S. (1991). A treatise on the family Cambridge, MA: Harvard University Press. Becker, G.S. (1993). Nobel lecture: Th e economic way of looking at behavior. The Journal of Political Economy 101 385-409. Becker, P. E., & Moen, P. (1999). Scali ng back: Dual-earner couples' work-family strategies. Journal of Marriage and the Family, 61 995-1007. Beehr, T., & Newman, J. (1978). Job stre ss, employee health, and organizational effectiveness: A facet analysis model and literature review. Personnel Psychology 31 665-699. Beehr, T. A., Walsh, J. T., & Taber, T. D. (1976). Relationships of stress to individually and organizationally valued states: Higher-order needs as a moderator. Journal of Applied Psychology, 61, 41-47. Berardo, F. M., Shehan, C. L., & Leslie, G. R. (1987). A residue of tradition: Jobs, careers, and spouses' time in housework. Journal of Marriage and the Family 49 381–390.

PAGE 123

110 Bergkvist, L., & Rossiter, J.R. (2007). The pr edictive validity of multiple-item versus single-item measures of the same constructs. Journal of Marketing Research 44 175–184. Berk, R. A., & Berk, S. F. (1979). Labor and leisure at home: The content and organization of the household day. Beverly Hills, CA: Sage. Bernard, H.R., Killworth, P.D., Kronenfel d, D., Sailer, L. (1984). The problem of informant accuracy: The validity of retrospective data. Annual Review of Anthropology 13 495-517. Bianchi, S.M., Milkie, M., Sayer, L., & Robinson, J. (2000). Is anyone doing the housework? Trends in the gende r division of household labor. Social Forces 79 191-228. Bird C. E., & Fremont, A. M. (1991). Gender, time use, and health. Journal of Health and Social Behavior 32 114-129. Bittman, M., England, P., Sayer, L., Matheson, G. (2003). When does gender trump money? Bargaining and time in household work. American Journal of Sociology ,109, 186-214. Blair, S.L. & Lichter, D.T. (1991). Meas uring the division of household labor: Gender segregation of housework among american couples. Journal of Family Issues 12(1), 91-113. Bonney, J.F.; Kelley, M.L., & Levant, R.F. (1999). A model of fathers' behavioral involvement in child care in dual-earner families. Journal of Family Psychology 13 (3), 401-415.

PAGE 124

111 Booth, A., & Johnson, D.R. (1994). Dec lining health and marital quality. Journal of Marriage & the Family 56 218-223. Brennan, R.T., Barnett, R.C., & Gareis, K.C. (2001). When she earns more than he does: A longitudinal study of dual-earner couples. Journal of Marriage and Family 63 168-182. Brines, J. (1993). The exch ange value of housework. Rationality and Society 5 302-40. Brockner, J., Ackerman, G., Greenberg, J., Gelfand, M.J., Francesco, A.M., Chen, Z.X., Leung, K., Bierbrauer, G. Gomez, C., Kirkman, B.L., & Shapiro, D. (2001). Culture and procedural justice: The infl uence of power distance on reactions to voice. Journal of Experimental Social Psychology 37 (4), 300-315. Buzzanell, P. M., & Turner, L. H. (2003). Emotion work revealed by job loss discourse: Backgrounding-foregrounding of feeli ngs, construction of normalcy, and (re)instituting of traditional masculinities. Journal of Applied Communication Research, 31, 27-57. Cable, D. M., & DeRue, D. S. (2002). Th e convergent and discriminant validity of subjective fit perceptions. Journal of Applied Psychology, 87, 875–884. Cable, D. M., & Edwards, J. R. (2004). Complementary and supplementary fit: A theoretical and empirical integration. Journal of Applied Psychology, 89, 822834. Cable, D. M., & Judge, T. A. (1996). Personorganization fit, job choice decisions, and organizational entry. Organizational Behavior & Human Decision Processes, 67 294-311.

PAGE 125

112 Campion, M. A., Medsker, G. J., & Higgs, A. C. (1993). Relations between work group characteristics and effectiveness: Imp lications for designing effective work groups. Personnel Psychology 46 823–850. Caplan, R.D., Cobb, S., French, J.R.P., Harrison, R., & Pinneau, S.R. (1980). Job demands and worker health Ann Arbour, MI: Institut e for Social Research. Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research 1 629-637. Challiol, H., & Mignonac, K. (2005). Relocation decision-making and couple relationships: a quantitative and qua litative study of dual-earner couples. Journal of Organizational Behavior 26 247-274. Clark, M.S. (1984). Record keeping in two types of relationships. Journal of Personality and Social Psychology 47 (3), 549-57. Clark, M. S., & Bennett, M. E. (1992). Res earch on relationships : Implications for mental health. In D. N. Ruble, P. R. Costanzo, & M. E. Oliveri (Eds.), The social psychology of mental health (pp. 166-198). New York: Guilford Press. Clark M. S. Mills J., & Powell M C. (1986) Keeping track of needs in communal and exchange relationships Journal of Personality and Social Psychology 51 333338 Clarkberg, M., & Moen, P. (2001). Unders tanding the time-squeeze: Married couples' preferred and actual wo rk-hour strategies. American Behavioral Scientist, 44 1115-1136.

PAGE 126

113 Cohen, J., Cohen, P., West, S.G., & Aiken, L.S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd Ed.). Mahwah, New Jersey: Lawrence Erlbaum Associates. Coltrane, S. (2000). Research on household la bor: Modeling and measuring the social embeddedness of routine family work." Journal of Marriage and the Family, 62 1208–1233. Cooke, L.P. (2006). Doing gender in context: Household bargaining and the risk of divorce in Germany and the United States. American Journal of Sociology 112 442-472. Cornette, M.M., Strauman, T.J., Abra mson, L.Y., & Busch, A.M. (2009). Selfdiscrepancy and suicidal ideation. Cognition and Emotion 23 504-527. Cummings, T.G., & Cooper, C.L. (1979) Cybernetic framework for studying occupational stress. Human Relations, 32, 395–418. Cunningham, M. (2007). Influences of women's employment on the gendered division of household labor over the life course: Evidence from a 31-year panel study. Journal of Family Issues 28 (3), 422-444. Dawis, R. V., & Lofquist, L. H. (1984). A psychological theory of work adjustment. Minneapolis: University of Minnesota Press. Denton, M., & Zeytinoglo, U. (1993). Perceive d participation in decision-making in a university setting: The impact of gender. Industrial & Labor Relations Review 46 320-331. Deutsch, F.M., Roksa, J., & Meeske, C. ( 2003). How gender counts when couples count their money, Sex Roles 48 (7-8), 291-304.

PAGE 127

114 Dienstbier, R. A. (1989). Arousal and physiolo gical toughness: Impli cations for mental and physical health. Psychological Review 96 84-100. Doucet, A., & Merla L. (2007) Stay-at homefathering: A strategy for balancing work and home in Canadian and Belgian Families. Community, Work and Family 10 (4), 453-471. Eagle, B.W., Icenogle, M.L., Maes, J.D., & Miles, E.W. (1998). The importance of employee demographic profiles for unders tanding experience s of work–family interrole conflicts. The Journal of Social Psychology 138 (6), 690-709. Eby, L. T., Casper, W. J., Lockwood, A., Bordeaux, C., & Brinley, A. (2005). Work and family research in IO/OB: Content anal ysis and review of the literature (19802002). Journal of Vocational Behavior, 66, 124-197. Eddleston, K.A., Veiga, J.F., & Powell, G. N. (2006). Explaining sex differences in managerial career satisfier preferences: The role of gender self-schema. Journal of Applied Psychology, 91 (2), 437-445. Edwards, J. R. (1991). Person-job fit: A c onceptual integration, literature review, and methodological critique. In C. L. Cooper & I. T. Robertson (Eds.), International review of industrial and organizational psychology (vol. 6, pp. 283-357). New York: Wiley. Edwards, J. R. (1993). Problems with the use of profile similarity in dices in the study of congruence in organizational research. Personnel Psychology 46 641-665. Edwards, J. R. (1994a). Regression analysis as an alternative to difference scores. Journal of Management 20 683-689.

PAGE 128

115 Edwards, J. R. (1994b). The study of congrue nce in organizational behavior research: Critique and a proposed alternative. Organizational Behavior and Human Decision Processes 58, 51-100 Edwards, J. R. (1996). An examination of competing versions of the person-environment fit approach to stress. Academy of Management Journal, 39 292-339. Edwards, J. R. (2001). Ten difference score myths. Organizational Research Methods 4 264-286. Edwards, J. R. (2002). Alternatives to diffe rence scores: Polynomial regression analysis and response surface methodology. In F. Drasgow & N. W. Schmitt (Eds.), Advances in measurement and data analysis (pp. 350-400). San Francisco: Jossey-Bass. Edwards, J. R. (2007). Polynomial regre ssion and response surface methodology. In C. Ostroff & T. A. Judge (Eds.), Perspectives on organizational fit (pp. 361-372). San Francisco: Jossey-Bass. Edwards, J. R. (2008). Person-environment fit in organizations: An assessment of theoretical progress. The Academy of Management Annals, 2, 167-230. Edwards, J. R. (2009). Latent variable modeling in congruence research: Current problems and future directions. Organizational Research Methods, 12, 34-62. Edwards, J. R., Cable, D. M., Williamson, I. O., Lambert, L. S., & Shipp, A. J. (2006). The phenomenology of fit: Linking the person and environment to the subjective experience of pe rson-environment fit. Journal of Applied Psychology, 91, 802-827. Edwards, J. R., Caplan, R. D., & Harrison, R. V. (1998). Person-environment fit theory:

PAGE 129

116 Conceptual foundations, empirical evidence, and directions for future research. In C. L. Cooper (Ed.), Theories of organizational stress (pp. 28-67). Oxford: Oxford University Press. Edwards, J. R., & Cooper, C. L. (1988). Th e impacts of positive psychological states on physical health: A review and theoretical framework. Social Science & Medicine 27 1447-1459. Edwards, J. R., & Harrison, R. V. (1993) Job demands and worker health: Threedimensional reexamination of the rela tionship between person-environment fit and strain. Journal of Applied Psychology 78 628-648 Edwards, J. R., & Parry, M. E. (1993). On th e use of polynomial regr ession equations as an alternative to difference scor es in organizational research. Academy of Management Journal 36, 1577-1613. Edwards, J. R., & Rothbard, N. P. (1999). Work and family stress and well-being: An examination of person-environment fit in the work and family domains. Organizational Behavior and Human Decision Processes, 77 85-129. Edwards, J. R., & Shipp. A. J. (2007). Th e relationship between person-environment fit and outcomes: An integrative theoretical fr amework. In C. Ostroff & T. A. Judge (Eds.), Perspectives on organizational fit (pp. 209-258). San Francisco: JosseyBass. Eisenberg, N., & Lennon, R. (1983). Sex differe nces in empathy and related capacities. Psychological Bulletin 94 (1), 100-131.

PAGE 130

117 Eisenberg, N., & Miller, P.A. ( 1987 ). The relation of empathy to prosocial and related behaviors. Psychological Bulletin 101 91-119. England, P. & Kilbourne, B. S. (1990). Ma rkets, marriages, and other mates: The problem of power. In R. Friedland & A. F. Robertson (Eds.), Beyond the marketplace: Rethinking economy and society (pp. 163-188). New York: Aldine de Gruyter. Erickson, R.J. (1993). Reconceptualizing family work: The effect of emotion work on perceptions of marital quality. Journal of Marriage and the Family 55 888-900. Fletcher, B., & Payne, R.L. (1980) Stress and work: A review and theoretical framework. Personnel Review 9 19-29. Fox, G. L., & Murry, V. M. (2000). Feminist pe rspectives and family research: A decade review. Journal of Marriage and Family 62 1160-1172 Frankenhaeuser, M. (1986). A psychobiological framework for research on stress and coping. In M. Appley & R. Trumbull (Eds.), Dynamics of stress (pp. 101–116). New York: Plenum Press. French, J.R.P., Jr., Caplan, R.D., & Harrison, R.V. (1982). The mechanisms of job stress and strain. London: Wiley. French J. R. P. & Kahn, R. L. (1962). A programmatic approach to studying the industrial environment and mental health. Journal of Social Issues 18 1-47 Galinsky, E., Aumann, K., & Bond, J.T. (2009). Times are changing: Gender and generation at work and at home. Retrieved from http://familiesandwork.org/site/resear ch/reports/ Times_ Are_Changing.pdf

PAGE 131

118 Gecas, V., & Seff, M. A. (1990). Social clas s and self-esteem: Ps ychological centrality, compensation, and the relative effects of work and home. Social Psychology Quarterly, 53 165–173. General Social Survey (2006). National Science Foundation The Roper Center for Public Opinion Research, Un iversity of Connecticut. Gerson, K. (1993). No man's land: Men's changing commitment o family and work New York: Basic Books. Gerson, K. (2002). Moral dilemmas, moral stra tegies, and the transformation of gender: Lessons from two generations of work and family change. Gender & Society 16 8-28. Gerstel, N., & Gallagher, S.K. (2001). Men' s caregiving: Gender and the contingent character of care. Gender & Society, 15 197-217. Gilbert, L.A. (1985). Men in dual-career families: current realities and future prospects. Hillsdale, NJ: Lawrence Earlbuam Associates. Gilligan, C. (1982). In a different vo ice: psychological theory and women's development. Cambridge, MA: Harvard University Press, 1982. Goldberg, A.E., & Perry-Jenkins, M. (2004) Division of labor and working-class women’s well-being across the transition to parenthood. Journal of Family Psychology 18 225–236. Golding, J. M. (1990). Divisi on of household labor, strain, and depressive symptoms among Mexican Americans and Non-Hispanic Whites. Psychology of Women Quarterly 14 103-117.

PAGE 132

119 Greenhaus, J.H., Parasuraman, S., & Wormley, W.M. (1990). Effects of race on organizational experiences, job perfor mance evaluations, and career outcomes. The Academy of Management Journal 33 64 – 86. Grote, N., & Clark, M. S. (2001). Does conf lict drive perceptions of unfairness or do perceptions of unfairn ess drive conflict? Journal of Personality and Social Psychology 80 (2), 281-293. Gutek, B. A., Nakamura, C. Y., & Nieva, V. F. (1981). The interdependence of work and family roles. Journal of Occupational Behavior 2 1-16. Haddock, S. A., & Rattenborg, K. (2003). The bene fits and challenges of dual-earning: Perceptions of successful couples. American Journal of Family Therapy, 31, 325344. Hakim, C. (2000). Work-Lifestyle Choices in the 21st Century: Preference Theory Oxford University Press. Hammer, L. B., Neal, M. B., Newsom, J ., Brockwood, K. J., & Colton, C. (2005). A longitudinal study of the effects of dualearner couples’ utilization of familyfriendly workplace supports on work and family outcomes. Journal of Applied Psychology, 90 799-810. Harrison, R. V. (1978). Person-environment f it and job stress. In C. L. Cooper & R. Payne (Eds.), Stress at work (pp. 175-205). New York: John Wiley and Sons. Harrison, R. V. (1985). The person-environment fit model and the study of job stress. In T. A. Beehr & R. S. Bhagat (Eds.), Human stress and cognition in organizations (pp. 23-55). New York: Wiley. Hays, W. L. (1994). Statistics (5th ed.). Fort Worth: Harcourt College Publishers.

PAGE 133

120 Hemp, P. (2004). Presenteeism: At Work — But Out of It. Harvard Business Review 155 49-58. Heslin, P.A. (2005). Conceptualiz ing and evaluating career success. Journal of Organizational Behavior 26 113-136. Higgins, E. T. (1987). Self-discrepancy: A theory relating self and affect. Psychological Review 94 319-340. Higgins, E. T. (1989). Self-discrepancy theo ry: What patterns of self-beliefs cause people to suffer? In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 22, pp. 93-137). New York: Academic Press. Higgins, E. T., Bond, R. N., Klein, R. and St rauman, T. (1986). Self-Discrepancies and Emotional Vulnerability: How Magnit ude, Accessibility, and Type of Discrepancy Influence Affect. Journal of Personality and Social Psychology 51 5-15. Higgins, E. T., Vookles, J., & Tykocinski, O. (1992). Self and health : How "patterns" of self-beliefs predict types of emotional and physical problems. Social Cognition, 10 125-150. Hiller, H.H., & McCaig, K. (2007). Reasse ssing the role of partnered women in migration decision-making and migration outcomes. Journal of Social and Personal Relationships 24 457-472. Hochschild, A.R. (1989). The Second Shift. Working Parents and the Revolution at Home New York: Viking. Hood, J.C. (1986). The provider role: Its meaning and measurement. Journal of Marriage and Family 48 349-359.

PAGE 134

121 Jacobs, J.A., & Gerson, K. (2001). Overwork ed individuals or overworked families? Explaining trends in work, leisure, and family time. Work and Occupations 28 (1), 40-63. Jansen, M. & Liefbroer, A.C. (2006). Couples ’ attitudes, childbirth, and the division of labor. J ournal of Family Issues 27 1487-1511. Janssens, A. (1998). The rise and decline of the male breadwinner family? An overview of the debate. In A. Jannsens (Ed.), The rise and decline of the male breadwinner family? (pp. 1 – 23).Cambridge: University Press, Cambridge. Jick, T.D., & Mitz, L.F (1985). Se x differences in work stress. Academy of Management Review 10 408-20. Johns, G. (1981). Difference score measures of organizational beha vior variables: A critique. Organizational Behavior and Human Decision Processes 27 (3), 443463. Johnson, M.K., Hashtroudi, S. & Lindsa y, D.S. (1993). Source monitoring. Psychological Bulletin, 114 3-28. Johnson, C. A., Johnson, B. E., & Liese, B. S. (1991). Dual-doctor marriages: The British experience. Journal of the American Me dical Women's Association, 46 (5), 155-159. Johnson, O., Kamilaris, T., Chrousos, G., & Gold, P. (1990). Mechanisms of stress: A dynamic overview of hormonal a nd behavioral homeostasis. Neuroscience and Biobehav ioral Reviews 16 115–130. Judge, T. A., & Cable, D. M. (1997). Appli cant personality, organizational culture, and organization attraction. Personnel Psychology, 50, 359– 394.

PAGE 135

122 Kahn, R. L., & Byosiere, M. (1992). Stress in or ganizations. In M. D. Dunnette & L. M. Hough (Eds.), Handbook of Industrial & Organizational Psychology (2nd ed., Vol. 3, pp. 571-650). Palo Alto, CA: Consulting Psychologists Press. Kahn, R. L, Wolfe, D. M., Quinn, R. P., & Snoek, J. D. (1964). Organizational stress: Studies in role conflict a nd ambiguity. New York: Wiley. Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement 20 141-151. Kalmuss, D., Davidson, A., & Cushman, L. (19 92). Parenting expectations, experiences, and adjustment to parenthood: A test of the violated expectations framework. Journal of Marriage and the Family, 54, 516–526. Kamo, Y. (1988). Determinants of household division of labor. Journal of Family Issues, 9, 177-200. Kanter, R.M. ( 1977). Men and women of the corporation New York: Basic Books. Karasek, R. & Theorell, T. (1990). Healthy work: Stress, productivity and the reconstruction of working life New York: Basic Books. Kessler, R. C. & McLeod, J.D. (1984 ) Sex differences in vulne rability to undesirable life events. American Sociological Review 49 620-631. Khazan, I., McHale, J. & DeCour cey, W. (2008). Violated wish es concerning division of childcare labor predict early coparenting process during stress ful and nonstressful family evaluations. Infant Mental Health Journal 29 343-361. Kiecolt-Glaser, J.K., McGuire, L.M., R obles, T., & Glaser, R. (2002). Emotions, morbidity, and mortality: New pers pectives from psychoneuroimmunology. Annual Review of Psychology, 53, 83-107.

PAGE 136

123 Kingsbury, N.M., & Scanzoni, J. (1989). Pr ocess power and decision outcomes among dual-career couples. Journal of Comparative Family Studies 20 (2), 231-246. Komter, A. (1989). Hidden power in marriage. Gender & Society 3 187-216. Kristof, A. L. (1996). Person-organizati on fit: An integrative review of its conceptualizations, measurement, and implications. Personnel Psychology, 49 149. Kroska, A. (1997). The Division of labor in the home: A review and reconceptualization. Social Psychology Quarterly, 60 304-322. Kroska, A. (2009). Exploring the consequen ces of gender ideology–w ork discrepancies. Sex Roles 60 313–328. Kubzansky, L.D., & Kawachi, I. (2000). Go ing to the heart of the matter: Negative emotions and coronary heart disease. Psychosomatic Research, 48, 323-337. Lance, C. E., Mallard, A. G., & Michalos, A. C. (1995). Tests of the causal directions of global-life facet satisfaction relationships. Social Indicators Research, 34, 69–92. LaTour, S. (1978). Determinants of particip ant and observer satisfa ction with adversary and inquisitorial m odes of adjudication. Journal of Personality and Social Psychology 36 15311545. Lavee, Y., & Katz, K. (2002). Division of la bor, perceived fairness and marital quality: The effect of gender ideology. Journal of Marriage and Family 64 27-39. Lawler, E. E. (1973). Motivation in work organizations Monterey, CA: Brooks/Cole. Lazarus, R. S. (1991). Emotion and adaptation New York: Oxford University Press.

PAGE 137

124 Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer. Levine, L. J., & Safer, M. A. (2002). Sources of bias in memory for emotions. Current Directions in Psychological Science, 11 (5), 169-173. Lewin, K. (1935). A dynamic theory of personality. New York: McGraw-Hill. Lind, E.A., Kurtz, S.K., Musante, L., Walker L., & Thibaut, J.W. (1980). Procedure and outcome effects on reactions to adjudicated resolution of conflicts of interest. Journal of Personality and Social Psychology 39 643-653. Locke, E. A. (1969). What is job satisfaction? Organizational Behavior & Human Decision Processes, 4 309-336. Locke, E. A. (1976). The nature and causes of job satisfaction. In M. Dunnette (Ed.), Handbook of industrial and organizational psychology (pp. 1297-1350). Chicago: Rand McNally. Lobel, S. A., & St. Clair, L. (1992). Effects of family responsibilities, gender, and career identity salience on performance outcomes. Academy of Management Journal, 35, 1057–1069. Lodahl, T. M., & Kejner, M. (1965). Th e definition and measurement of job involvement. Journal of Applied Psychology, 49, 24–33. Loftus, E. F. (1982). Memory and its distortions. In A. G. Kraut (Ed.), The G Stanley hall lecture series (pp. 123-154). Washington, D.C.: American Psychological Association. Loscocco, K., & Spitze, G. (2007). Gender patterns in provider role attitudes and behavior. Journal of Family Issues 28 934-954.

PAGE 138

125 MacDermid, S.M., Huston, T.L., & McHale S.M. (1990). Changes in marriage associated with the transition to parent hood: Individual differences as a function of sex-role attitudes and changes in the division of household labor. Journal of Marriage and Family 52 475-486. Machung, A. (1989). Talking career, thinking job: Gender differences in career and family expectations of Berkeley seniors. Feminist Studies, 15 35–55. Madden, M.E. (1987). Perceived control and power in marriage: A study of marital decision making and task performance. Personality and Social Psychology Bulletin 13 (1), 73-82. Maines, D.R., & Hardesty, M.J. (1987): Temporality and gender: Young adults’ career and family plans. Social Forces, 66 102-120. Maret, E., & Finlay, B. (1984). The dist ribution of household labor among women in dual-earner families. Journal of Marriage and the Family 46 357 – 364. McDonald, G.W. (1980). Family power: The assessment of a decade of theory and research, 1970-1979. Journal of Marriage and Family 42 841854. McGrath, J.E. (1976). Stress and behavior in organizations. In M. Dunnette (Ed.), Handbook of industrial and organizational psychology (pp. 1351–1395). Chicago, IL: Rand McNally. McHale, S. M., & Crouter, A. (1992). You can’t always get what you want: Incongruence between sex-role attitude s and family work roles and its implications for marriage. Journal of Marriage and the Family 54 537–547. McIver, J. P., & Carmines, E. G. (1981). Unidimensional scaling Thousand Oaks, CA: Sage.

PAGE 139

126 McNeil, L., Sher, M. (1999). Th e Dual-career-couple problem. Physics Today 52, 3237. Mickelson, K.D., Claffey, S.T., & Williams, S. L. (2006). The moderating role of gender and gender role attitudes on the link betw een spousal support a nd marital quality. Sex Roles 55 73–82. Milkie, M.A., Bianchi, S.M., Mattingly, M.J., & Robinson, J.P. (2002). Gendered division of childrearing: Id eals, realities, and the rela tionship to parental wellbeing, Sex Roles 47 21 -38. Mincer, J. (1978). Family migration decisions. Journal of Political Economy 86 749773. Meisenbach, R.J. (2010). The female bread winner: Phenomenological experience and gendered identity in work/family spaces. Sex Roles 62 2–19. Moen, P. (2003). Linked lives: Dual careers, ge nder, and the contingent life course. In W.R. Heinz and V.W. Marshall (Eds.), Social Dynamics of the Life Course: Transitions, Institutions, and Interrelations Hawthorne, NY: Aldine de Gruyter. Moen, P., & Orrange, R. (2002). Careers and lives: Socialization, structural lag, and gendered ambivalence. In R. Settersten and T. Owens (Eds.), Advances in Life Course Research: New Frontiers in Socialization (vol. 7, pp. 231-260). London: Elsevier Science. Moen, P. & Roehling, P.V. (2005). The Career Mystique: Cracks in the American Dream Lanham: Rowman and Littlefield. Moen, P. & Yu, Y. (2000). Effective work /life strategies: Working couples, work conditions, gender, and life quality. Social Problems, 47 (3), 291-326. Muchinsky, P.M., & Monahan, C.J. (1987). Wh at is person–environment congruence?

PAGE 140

127 Supplementary versus complementary models of fit. Journal of Vocational Behavior, 31, 268–277. Murray, H.A. (1938). Explorations in personality. Boston, MA: Houghton Mifflin. Neumark-Sztainer, D., Hannan, P. J., Story, M ., Croll, J., & Perry, C. (2003). Family meal patterns: Associations with soci odemographic character istics and improved dietary intake among adolescents. Journal of the American Dietetic Association 103 317–322. Newman, L. S., Higgins, E. T., & Vookles, J. (1992). Self-guide st rength and emotional vulnerability: Birth order as a moderator of self-affect relations. Personality & Social Psychology Bulletin, 18 402-411. Ng, T. W. H., Eby, L. T., Sorensen, K. L ., & Feldman, D. C. (2005) Predictors of objective and subjective career success: A meta-analysis. Personnel Psychology 58 367-408. Nicolson, P. (1990). A brief report of women' s expectations of men's behaviour in the transition to parenthood: C ontradictions and conflicts for counselling psychology practice. Counselling Psychology Quarterly 3 (4), 353-361. Norman, G. R., & Streiner, D. L. (2007). Biostatistics: The bare essentials (3rd ed.), Toronto: B. C. Decker. Norton, R. (1983). Measuring marital quality: A critical look at the dependent variable. Journal of Marriage and the Family 45 141-151. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGrawHill.

PAGE 141

128 Oppenheimer, V. K. (1997). Women's employment and the gain to marriage: The specialization and trading model. Annual Review of Sociology, 23 431-453. Orrange, R.M., Firebaugh, F.M., & Heck, R. K.Z. (2003). Managing households. In P. Moen (Ed.), It's about time: Couples and careers (pp. 153-167). Ithaca, NY: Cornell University Press. Orrange, R. M. (2003). The emerging muta ble self: Gender dynamics and creative adaptations in defining work, family and the future. Social Forces 82 1-34. Orth, U., Berking, M., & Burkhardt, S. (2006) Self-conscious emotions and depression: Rumination explains why shame, but not guilt, is maladaptive. Personality and Social Psychology Bulletin, 32 1608-1619. Padavic, I., & Reskin, B. (2002). Women and men and work New York: Pine Forge Press. Parsons, F. (1909). Choosing a vocation. Boston, MA: Houghton Mifflin. Pearlin, L. L., Leibermann, M. A., Menaghan, E. G, & Mullan, J. T. (1981). The stress process. Journal of Health and Social Behavior, 22, 337-356. Perrew, P.L., & Zellars, K.L. (1999). An exam ination of attributions and emotions in the transactional approach to th e organizational stress process. Journal of Organizational Behavior, 20, 739-752. Perry-Jenkins, M., Seery, B., & Crouter, A. C. (1992). Linkages between women's provider-role attitudes, psychological well-being, and fam ily relationships. Psychology of Women Quarterly 16 311-329.

PAGE 142

129 Pia, D.L. & Bengtson, V.L. (1993). The division of household labor and wives' happiness: Ideology, employment, and perceptions of support. Journal of Marriage and Family 55 901-912. Pixley, J.E. (2008). Life course patterns of car eer-prioritizing decisi ons and occupational attainment in dual-earner couples. Work and Occupations 35 (2), 127-163. Pixley, J. E. & Moen, P. (2003). Prio ritizing careers. In P. Moen (Ed.), It's about time: Couples and careers (pp. 183-200). Ithaca, NY: Cornell University Press. Polasky, L. J., & Holahan, C. K. (1998). Matern al self-discrepancies, interrole conflict, and negative affect among married professional women with children. Journal of Family Psychology, 12 388-401. Potuchek, J. L. (1997). Who supports the family? Gende r and breadwinning in dualearner marriage. Stanford, CA: Stanford University Press. Quick, J.C., Cooper, C.L., Nelson, D.L., Quick, J.D. & Gavin, J.H. (2003). Stress, health, and well-being at wo rk. In J. Greenberg (Ed.) Organizational Behavior: The State of the Science (pp. 53-89). Mahwah, NJ: La wrence Erlbaum Associates Publishers. Quinn, R. P. and Shepard, L. J. (1974). Quality of Employment Survey. Institute for Social Research. University of Michigan, Ann Arbor. Raley, S., Mattingly, M., & Bianchi, S. (2006). How dual are dual income couples? Documentary change from 1970 to 2001. Journal of Marriage and Family 68 (1), 11-28. Rapoport, R., & Rapoport, R.N. (1969). The dua l-career family: a va riant pattern and social change. Human Relations 22 3-30.

PAGE 143

130 Rhoades, G.K., Petrella, J.M., Stanley, S.M., & Markman, H.J. (2006). Premarital cohabitation, husbands' commitment, and wi ves' satisfaction with the division of household contributions. Marriage & Family Review 40 (4), 5–22. Rizzo, J. R., House, R. J., & Lirtzman, S. I. (1970). Role conflict and ambiguity in complex organizations. Administrative Scien ce Quarterly, 15, 150–163. Roehling, P.V. (2003). Dual-earner couples. Sloan Online Work and Family Encyclopedia http://wfnetwork.bc.edu/encycl opedia_entry.php?id=229&area=all Rochlen, A., Suizzo, M, McKelley, R., Scaringi V. (2008). “I’m just providing for my family”: A qualitative study of stay-at-home fathers. Psychology of Men and Masculinity 9 (4), 193-206. Robinson, J. P., & Godbey, G. (1997). Time for Life: The Surprising Ways Americans Use Their Time. University Park, PA: The Pennsyl vania State University Press. Rodebaugh, T. L., & Donahue, K. L. (2007). Could you be more specific, please: Self discrepancies, affect, and variati on in specificity and relevance. Journal of Clinical Psychology, 63, 1193-1207. Rosenbluth, S. C., Steil, J. M., & Whitcomb, J. H. (1998). Marital equality–What does it mean? Journal of Family Issues, 19 227-244. Ross, M. (1989). Relation of imp licit theories to the construction of personal histories. Psychological Review, 96(2), 341-357. Ross, C. E., Mirowsky, J., & Huber, J. ( 1983). Dividing work, sharing work, and inbetween: Marriage patterns and depression. American Sociological Review 48 809-823.

PAGE 144

131 Rothausen, T.J. (1999). “Family” in organiza tional research: A re view and comparison of definitions and measures. Journal of Organizational Behavior 20 817-836. Rousseau, D. M. (1989). Psychological a nd implicit contracts in organizations. Employee Responsibilities and Rights Journal 2 121-139. Rousseau, D. M. (1995). Promises in action: Psychologi cal contracts in organizations Newbury Park, CA: Sage. Rousseau, D. M. (2001). Schema, promise a nd mutuality: The building blocks of the psychological contract. Journal of Occupational and Organizational Psychology 74 (4), 511-542. Ruble, D. N., Fleming, A. S., Hackel, L. S ., & Stangor, C. (1988). Changes in the marital relationship during the transi tion to first time motherhood: Effects of violated expectations concerning divi sion of household labor. Journal of Personality and Social Psychology, 55, 78–87. Ryff, C. (2005). National Study of Daily Experiences Midlife in the United States (MIDUS) II project. National Institute on Aging. Safer, M. A., Keuler, D. J. (2002). In dividual differences in misremembering prepsychotherapy distress: Persona lity and memory distortion. Emotion. 2(2), 162178. Sagara, J., Ito, Y., Ikedda, M. (2006). Gender-ro le attitude and psychological well-being of middle-aged men: Focusing on empl oyment patterns of their wives. Japanese Psychological Research, 48 17–26.lackwell Saks, A. M., & Ashforth, B. E. (2002). Is job search related to empl oyment quality? It all depends on the fit. Journal of Applied Psychology, 87, 646–654. Pu

PAGE 145

132 Sanders, M.M., Lengnick-Hall, M.L., Lengnick -Hall, C.A., & Steele-Clapp, L. (1998). Love and work: Expectations of ne w entrants into the labor force. Journal of Organizational Behavior 19, 603-619. b Schafer, R. B., Wickrama, K. A. S., & Ke ith, P.M. (1996). Self-c oncept disconfirmation, psychological distress and marital happiness. Journal of Marriage and the Family 58 167-77. lishing Asia ORIGINAL ARTICLE Gender-role attitude and psychological well-being Schroeder, K. A., Blood, L. L., & Maluso, D. (1993). Gender differences and similarities between male and female undergraduate students regarding expectations for career and family roles. College Student Journal, 27 237–249. Shelton, B.A., & John, D. (1993). Race, ethnicity and difference: A comparison of White, Black and Hispanic men's household labor time. In J. Hood (Ed.) Men, Work, and Family (pp. 131-150). Beverly Hills, CA: Sage. Shelton, B. A., & John, D. (1996). The division of household labor. Annual Review of Sociology 22 299-322. Simons, R. (1992). Parental st atus strains, salience of parental identity and gender differences in psychological distress. Journal of Health and Social Behavior, 33, 25-35. Smith, P. C., Kendall, L., & Hulin, C. L. (1969). The measurement of satisfaction in work and retirement. Chicago: Rand McNally. Spade, J. Z., & Reese, C. A. (1991). We'v e come a long way, maybe: College students' plans for work and family. Sex Roles 24 309-21 Spector, P.E. (1992). Summated rating scale construction Thousand Oaks, CA: Sage.

PAGE 146

133 Spector, P. E. (2006). Method va riance in organizational resear ch: Truth or urban legend? Organizational Research Methods 9 221-232. Spector, P. E., & Fox, S. (2005). A model of counterproductive work behavior. In S. Fox, & P. E. Spector (Eds.). Counterproductive workplace behavior: Investigations of actors and targets (pp. 151-174). Washington, DC: APA. Steel, R.P., & Mento, A.J. (1987). The pa rticipation-perfor mance controversy reconsidered: Subordinate competence as a mitigating factor. Group & Organization Studies 12 411423. Steffy, B. D., & Ashbaugh, D. (1986). Dual-career planning, ma rital satisfaction and job stress among women in dual-career marriages. Journal of Business and Psychology 1 114–123. Stevens, D.P., Kiger, G., & Mannon, S.E. (2005). Domestic labor and marital satisfaction: How much or how satisfied? Marriage and Family Review 37 4967. Strauman, T.J. (1989). Self-discrepancies in clinical depression and social phobia: Cognitive structures that underlie affective disorders? Journal of Abnormal Psychology 98 14-22. Tesser, A. (1988). Toward a self-evaluation ma intenance model of social behavior. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 21 ) (pp. 181-227). New York: Academic Press. Thibaut, J.W., & Walker, L. (1975). Procedural justice: A psychological analysis Hillsdale, NJ: Erlbaum.

PAGE 147

134 Thoits, P. A. (1986). Multiple identiti es: Examining gender and marital status differences. American Sociological Review 51 259-272 Thompson, L., & Walker, A. J. (1989). Women and men in marriage, work, and parenthood. Journal of Marriage and the Family 51 845-872. Tichenor, V.J. (1999). Status and income as gendered resources: The case of marital power. Journal of Marriage and the Family 61 638-650. Tinklin, T., Croxford, L., Ducklin, A., & Fram e, B. (2005). Gender and attitudes to work and family roles: The views of young people at the Millenium. Gender and Education, 17 (2) 129-142. U.S. Department of Labor. (2007). American time use survey. United States Bureau of Labor Statistics Retrieved from www.bls.gov/ news.release/atus.nr0.htm Van Egeren, L.A. (2004). The development of the coparenting relationship over the transition to parenthood. Infant Mental Health Journal 25 (5), 453–477. Vanfossen, B.E. (1979). Intersexual vi olence in Monroe County, New York. Victimology, 4 299-305. Vega, W.A., Patterson, T., Sallis, J., Nade r, P., Atkins, C.,&Abramson, I. (1986). Cohesion and adaptability in Mexican-American and Anglo families. Journal of Marriage and the Family, 48 857-867. Warren, T. (2007). Conceptualizing breadwinning work. Work, Employment and Society 21 317 – 336. Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood. Psychological Bulletin, 98 219-235.

PAGE 148

135 Westman, M. (2004). Gender asymmetry in cr ossover research. In D.L. Nelson & R.J Burke (Eds.) Gender, work stress, and health (pp. 129-149). Washington, DC, US: American Psychological Association. Westman, M., Vinokur, A., Hamilton, L., & Rozi ner, I. (2004). Crossover of marital dissatisfaction during downsi zing: A study of Russian Ar my officers and their spouses. Journal of Applied Psychology, 89 769-777. Wethington, E., Pixley, J. E., & Kavey, A. (2003). Perceived turning points in work careers. In P. Moen (Ed.), It's about time: Couples and careers (pp. 168-182). Ithaca, NY: Cornell University Press. Wicki, W. (1999). The impact of family re sources and satisfacti on with division of labour on coping and worries after the birth of the first child. International Journal of Behavioral Development 23 (2), 431-456. Wilkie, J. R., Ferree, M. M., & Ratcliff, K. S. (1998). Gender and fairness: Marital satisfaction in two-earner couples. Journal of Marriage and the Family, 60 577594. Williams, R.H., & Zimmerman, D. W. (1977). The reliability of difference scores when errors are correlated. Educational and Psycholog ical Measurement, 37, 679-689. Wofford J.C., Goodwin, V.L., Daly, P.S. (1999). Cognitive-affective stress propensity: a field study. Journal of Organizational Behavior 20 687 – 707. Wright, J.D. (2001). Factors in birt htiming decisions for career women. Unpublished doctoral dissertation Tennessee State University. Yang, L.Q., Levine, E.L., Smith, A. M., Ispas, D. & Rossi, M.E. (2008). Personenvironment fit or person plus enviro nment: A meta-analysis of studies using

PAGE 149

136 polynomial regression analys is. (Special Issue), Human Resource Management Review, 18 311-321. Zedeck, S. (1992). Introduction: Exploring the domain of work and family concerns. In S. Zedeck (Ed.), Work, families and orga nizations (pp. 1–32). San Francisco, CA: Jossey-Bass. Zipp, J.F., Prohaska, A., & Bemiller, A. ( 2004). Wives, husbands, and hidden power in marriage. Journal of Family Issues 25 933-958. Zuckerman, D.M. (1989). Stress, self-esteem, and mental health: How does gender make a difference? Sex Roles 20 (7-8), 429-444. Zvonkovic, A. M., Schmiege, C. J., & Hall, L. D. (1994). Influence strategies used when couples make work-family decisions and their importance for marital satisfaction. Family Relations 43 182–188.

PAGE 150

137 Table 1. Summary of previous desire-division of labor fit and well-being research Authors Journal Sample used in analysis Division of labor variable Well-being variable Type of fit/misfit Analytic method Khazan, McHale, & Decourcey, 2008 Infant Mental Health Journal Husband and wife dyads Congruence between ideal division of childcare labor during pregnancy and actual division measured 3 months post-partum Husbands’ and wives’ marital satisfaction Husband contributing less than he or wife desired to childcare Difference scores Results: When the husband contributed less than the wife desired, her marital satisfaction decreased. When husband was less involved than he had desired, his marital satisfaction did not change. Loscocco & Spitze (2007) Journal of Family Issues Men and women living with a partner (not dyads) Congruence between ideal perceptions of who provides for the family and actual providing situation Anxiety, marital satisfaction, life satisfaction Belief that one is financially providing too much, too little, or right amount Difference scores Results: For men, providing less than desired related to greater a nxiety, but among men who desired to be the main provider, congruence between attitudes and reality related to greater a nxiety. Congruence did not affect men's life or marital satisfaction. For women, providing more than desired was rela ted positively to life satisfaction, and no significant effects were found with regard to martia l satisfaction or anxiety. Milkie, Bianchi, Mattingly, & Robinson (2002) Sex Roles Married men and women with children (not dyads) Discrepancies btw. ideal and actual contributions of men to providing and parenting Men and women’s stress levels Father less involved than ideal; father more involved than ideal Difference scores Results:

PAGE 151

138 Perceptions that fathers contributed less than the ideal am ount to parenting were positively related to stress for both genders, although the effect was only marginally significant. Perceptions that fathers contributed less than the ideal amount to breadwinning also related positively to stress, and th e effect was stronger for men than women. There was no significant association between fathers bei ng more involved than ideal and stress. PerryJenkins, Seery, & Crouter, 1992 Psychology of Women Quarterly Wives Congruence between wives preferences for her provider status and actual provider status Wives’ role overload, depression, marital love, conflict, and satisfaction Wives who are coproviders and want to be; wives who are coproviders but do not want to be; wives who are home-makers and want to be Difference scores Results: Wives who were co-providers but did not want to be reported less marital satisfac tion than wives who desired to be coproviders and were co-providers and wives who desired to be homemakers and were homemakers. No significant effects with role overload, depression, ma rital love, or marital conflict. Ross, Mirowsky, & Huber, 1983 American Sociological Review Husband and wife dyads Congruence between desires toward wives’ employment status and her actual employment status Husbands’ and wives’ depression Husband’s and wives’ desires for her employment (employed or not) and her actual status Interaction terms Results: Husbands were most depressed when they preferred wife to stay home but she had a job. Wives were most depressed when they were staying home but wanted to have a job.

PAGE 152

139 Table 2. Recruitment sources and response rate information for initial respondents Number contacted Number responded (response rate) Number completed (response rate) Number in final sample (response rate) Women’s organization 13943 2379 (17.06%) 993 (7.12%) 498 (3.57%) University Alumni 845 76 ](8.99%) 24 (2.84%) 7 (.82%) First Time Fathers Website 889 16 (1.79%) 10 (1.12%) 3 (.003%) Personal and extended networks unknown 173 129 70 Total 15861 2644 (16.67%) 1134 (7.14%) 578 (3.6%) Table 3. Recruitment sources and response rate information for spouse respondents Total number agreed to spouse participation Number provided spouse email Number agreed to personally contact spouse Number spouse completed (response rate) Number spouses in final sample (response rate) Women’s organization 386 119 267 152 (39.37%) 108 (27.97%) University Alumni 8 5 3 2 (12.5%) 2 (12.5%) First Time Fathers Website 2 2 0 1 (50%) 1 (50%) Personal and extended networks 67 18 49 15 (22.38%) 15 (22.38) Total 463 144 319 170 (36.71%) 126 (27.21%)

PAGE 153

140 Table 4. Descriptive statistics of study variables for Partner A PARTNER A N # of Items M SD Obs. Min. Obs. Max. Scale Min Scale Max Hypotheses Variables Pre-child desires for spouse fam labor (needs for A/demands for B) 125 2 .66 49.77 12.51 20.00 100.00 0 100 Pre-child desires for self fam labor 125 2 .6650.2312.51.00 80.00 0 100 Post-child spouse’s actual fam labor (supplies) 123 2 .84 46.94 17.60 .00 90.00 0 100 Post-child self actual fam labor 123 2 .8453.0617.6010.00100.00 0 100 Pre-child desires for spouse paid labor (needs for A/demands for B) 125 3 .92 50.36 21.16 .00 100.00 0 100 Pre-child desires for self paid labor 125 3 .9249.6321.16.00 100.00 0 100 Post-child spouse’s actual paid labor (supplies) 123 3 .79 48.26 15.83 5.00 88.33 0 100 Post-child self actual paid labor 123 3 .7951.7415.8311.6795.00 0 100 Family satisfaction 125 4 .87 4.33 .62 2.00 5.00 1 5 Career satisfaction 126 5 .89 3.67 .80 1.60 5.00 1 5 Marital satisfaction 125 5 .97 4.46 .64 1.00 5.00 1 5 Depression 120 10 .84 2.41 .64 1.00 3.80 1 5 Physical health symptoms 120 11 n/a 18.64 5.23 11.00 39.00 11 66 Gender* 126 1 .50 .50 0 1 0 1 Career centrality 126 4 .82 3.32 .73 1.50 5.00 1 5 Family centrality 125 4 .79 4.53 .54 2.33 5.00 1 5 Voice in family labor decisions 120 7 .96 4.01 .82 1.71 5.00 1 5 Voice in paid labor decisions 120 7 .95 4.31 .64 2.00 5.00 1 5 Satisfaction with current family labor division 120 3 .903.66 .99 1.33 5.00 1 5

PAGE 154

141 Shaded variables are those of inter est to hypotheses concerning Partner A. *Gender coded male = 0, female = 1 **Family income coded 1 = less than $20,000, 2 = $20,000 $39,000, 3 = $40,000 – 59,999, 4 = $60,000-79,999, 5 = $80,000 – $99,999, 6 = $100,000 $119,999, 7 = $120,000 $139,999, 8 = $140,000 or more ***Career stage coded 1 for earliest career stage to 4 for most established career stage ****Education level coded 1 = Grades 9 through 11 (some high school), 2 =Grade 12 or GED (high school graduate), 3 = Some college, 4 = Asso ciate/two year degree, 5 = Bachelor's degree, 6 = Some graduate work, 7 = Master 's degree, 8 = Advanced degree Satisfaction with current paid labor division 120 3 .873.86 .95 2.00 5.00 1 5 Control variables Family income** 119 1 7.03 1.49 1 8 1 8 Length of marriage 126 1 8.26 3.51 1.83 17.83 Family responsibility 126 1 10.84 4.24 6.00 26.00 Total number children 126 1 1.80 .80 1 5 Number children under age 6 126 1.43 .57 0 3 Number children over age 6 126 .40 .73 0 4 Demographics Age 120 1 35.584.33 27 48 Career Stage*** 126 1 2.27 .57 1 4 1 4 Education Level**** 120 1 6.14 1.38 2 8 1 8 Weekly work Hours 124 1 42.2412.5812 90

PAGE 155

142 Table 5. Descriptive statistics of study variables for Partner B PARTNER B N # of Items M SD Obs. Min. Obs. Max. Scale Min Scale Max Hypotheses Variables Pre-child desires for spouse fam labor 126 2 .61 49.4610.8410.0090.00 0 100 Pre-child desires for self fam labor 126 2 .61 50.5410.8410.0090.00 0 100 Post-child spouse’s actual fam labor 126 2 .82 45.8916.272.50 82.50 0 100 Post-child self actual fam labor (abilities) 126 2 .82 54.11 16.27 17.50 97.50 0 100 Pre-child desires for spouse paid labor 126 3 .90 50.9218.903.33 100.00 0 100 Pre-child desires for self paid labor 126 3 .90 49.0818.90.00 96.67 0 100 Post-child spouse’s actual paid labor 126 3 .84 50.0315.245.00 93.33 0 100 Post-child self actual paid labor (abilities) 126 3 .84 49.96 15.24 6.67 95 0 100 Family satisfaction 126 4 .834.33 .54 2.50 5.00 1 5 Career satisfaction 126 5 .863.67 .72 1.00 5.00 1 5 Marital satisfaction 126 5 .96 4.48 .65 1.80 5.00 1 5 Depression 126 10 .80 2.40 .64 1.00 3.80 1 5 Physical health symptoms 121 11 17.86 5.07 11.00 41.00 11 66 Gender* 126 1 .50 .50 0 1 0 1 Career centrality 126 4 .793.15 .66 1.00 4.50 1 5 Family centrality 126 4 .794.56 .52 3.00 5.00 1 5 Voice in family labor decisions 126 7 .944.07 .70 2.00 5.00 1 5 Voice in paid labor decisions 126 7 .954.25 .60 2.57 5.00 1 5 Satisfaction w/ current family labor division 126 3 88 3.63 .95 1.33 5.00 1 5 Satisfaction w/ current paid labor division 126 3 .86 3.83 .95 1.00 5.00 1 5 Control variables Family income** 119 1 7.03 1.49 1 8 1 8 Length of marriage 126 1 8.26 3.51 1.83 17.83 Family responsibility 126 1 10.84 4.24 6.00 26.00

PAGE 156

143 Shaded variables are those of inter est to hypotheses concerning Partner B. *Gender coded male = 0, female = 1 **Family income coded 1 = less than $20,000, 2 = $20,000 $39,000, 3 = $40,000 – 59,999, 4 = $60,000-79,999, 5 = $80,000 – $99,999, 6 = $100,000 $119,999, 7 = $120,000 $139,999, 8 = $140,000 or more ***Career stage coded 1 for earliest career stage to 4 for most established career stage ***Education level coded 1 = Grades 9 through 11 (some high school), 2 =Grade 12 or GED (high school graduate), 3 = Some college, 4 = Asso ciate/two year degree, 5 = Bachelor's degree, 6 = Some graduate work, 7 = Master 's degree, 8 = Advanced degree Total number children 126 1 1.80 .80 1 5 Number children under age 6 126 1.43 .57 0 3 Number children over age 6 126 .40 .73 0 4 Demographics Age 126 1 35.714.25 27 47 Career Stage*** 125 1 2.26 .56 1 4 1 4 Education Level**** 126 1 6.38 1.21 3 8 1 8 Weekly work hours 126 1 43.1610.0110 72

PAGE 157

144 Table 6. Descriptive statistics of study variables by gender *Career stage coded 1 for earliest career stage to 4 for most established career stage **Education level coded 1 = Grades 9 through 11 (s ome hig school), 2 =Grade 12 or GED (high school graduate), 3 = Some college, 4 = Associat e/two year degree, 5 = Bachelor's degree, 6 = Some graduate work, 7 = Master's degree, 8 = Advanced degree Table 7. Source of reports for hypotheses Hypotheses Needs Supplies Well-being 1 – 14 A A A Hypotheses Demands Abilities Well-being 1524 A B B Males Females M SD M SD Hypotheses Variables Pre-child desires for spouse fam labor (needs A/demands for B) 55.51 10.78 43.76 9.41 Pre-child desires for self fam labor 44.49 10.78 56.23 9.41 Post-child spouse’s actual fam labor (supplies) 56.45 13.72 36.61 13.71 Post-child self actual fam labor (abilities) 43.54 13.72 63.39 13.71 Pre-child desires for s pouse paid labor (needs for A/demands for B) 37.86 14.73 63.32 16.16 Pre-child desires for self paid labor 62.13 14.73 36.68 16.16 Post-child spouse’s actual paid labor (supplies) 43.26 13.62 54.90 15.17 Post-child self actual paid labor (abilities) 56.73 13.62 45.10 15.17 Family satisfaction 4.31 .53 4.36 .63 Career satisfaction 3.67 .73 3.81 .75 Marital satisfaction 4.40 .61 4.56 .66 Depression 4.10 2.40 2.48 .60 Physical health symptoms 17.15 4.80 18.85 5.44 Career centrality 3.23 .71 3.24 .69 Family centrality 4.41 .54 4.68 .49 Voice in family labor decisions 3.99 .71 4.11 .81 Voice in paid labor decisions 4.15 .59 4.40 .62 Satisfaction with current family labor division 3.66 .84 3.63 1.08 Satisfaction with current paid labor division 3.75 .85 3.94 1.03 Demographics Age 35.88 4.17 34.65 4.12 Career Stage* 2.24 .55 2.37 .60 Education Level** 6.18 1.34 6.46 1.24 Weekly work Hours 42.69 11.37 38.52 12.86

PAGE 158

145 Table 8 Results of confirmatory factor analyses CFI TLI RMSE A 2 (df) 2 (df) diff Pre-child desires for division of labor (Partner A) 1 factor .939 .898 .149 33.714(9)** 2 factor .981 .965 .087 15.528(8)* 18.186 (1)** (vs. 1) 2 factor w/o emotion work .939 .987 .063 5.961 (4) 27.753 (5)** (vs. 1) 9.567 (4) (vs. 2) Post-child division of labor (Partner A) 1 factor .889 .816 .171 40.731 (9)** 2 factor .979 .961 .079 13.995 (8) 26.736**(vs. 1) 2 factor w/o emotion work .993 .983 .061 5.793 (4) 34.938** (vs. 1) 8.202 (vs. 2) Post-child division of labor items (Partner B) 1 factor .906 .844 .179 45.108 (9)** 2 factor 1.000 1.021 .000 3.779 (8) 41.329** (vs. 1) 2 factor w/o emotion work 1.000 1.014 .000 2.125 (4) 42.983 (vs. 1) 1.654 (vs. 2) Voice in division of labor decision making (Partner A) 1 factor .906 .844 .179 931.957 (77)** 2 factor .911 .893 .148 267.906 (76)** 664.051 (1)** Satisfaction with current division of labor (Partner B) 1 factor .616 .361 .392 183.598 (9)** 2 factor .958 .921 .138 27.027 (8)** 156.57 (1)** p < .05, ** p < .01

PAGE 159

146 Table 9. Results of exploratory factor analysis and int er-item correlations for voice in division of labor decision making (Partner A) Number eigenvalues > 1 2 Elbow in scree plot 2 Factors Loadings Inter-Item Correlations 1 2 1 2 3 4 5 6 1. Voice paid labor 1 .71 .26 -2. Voice paid labor 2 .86 .28.79-3. Voice paid labor 3 .86 .31.69.83-4. Voice paid labor 4 .86 .31.67.81.87 -5. Voice paid labor 5 .80 .24.56.72.77 .80 -6. Voice paid labor 6 .75 .28.58.69.70 .70 .71 -7. Voice paid labor 7 .84 .30.64.78.78 .82 .78.78 1. Voice family labor 1 .37 .70-2. Voice family labor 2 .28 .86 .76-3. Voice family labor 3 .30 .87 .72.88-4. Voice family labor 4 .30 .85 .72.81.85 -5. Voice family labor 5 .27 .84 .68.77.79 .81 -6. Voice family labor 6 .23 .79 .62.72.72 .74 .74 -7. Voice family labor 7 .31 .85 .70.80.81 .79 .82.80 Table 10. Results of exploratory factor analysis and in ter-item correlations for satisfaction with current division of labor (Partner B) Number eigenvalues > 1 2 Elbow in scree plot 2 Factors Loadings Inter-Item Correlations 1 2 1 2 3 1. Satisfaction w/ current paid labor 1 .28 .86-2. Satisfaction w/ current paid labor 2 .21 .86 .76 -3. Satisfaction w/ current paid labor 3 .20 .68 .65 .63 -1. Satisfaction w/ current family labor 1 .93 .22 -2. Satisfaction w/ current family labor 2 .93 .23.84 -3. Satisfaction w/ current family labor 3 .69 .29.66 .65 -

PAGE 160

147 Table 11. Correlations between study variables Variable 1 2 3 4 5 6 7 8 9 10 11 1. Pre-child desires for spouse’s fam labor (A) -2. Pre-child desires for self fam labor (A) -1.00** -3. Post-child spouse’s actual fam labor (A)b .50** -.50** -4. Post-child self actual fam labor (A) -.50** .50** -1.00** -5. Pre-child desires for spouse’s paid labor (A) -.57** .57** -.56** .56**-6. Pre-child desires for self paid labor (A) .57** -.57** .56** -.56**-1.00** -7. Post-child spouse’s actual paid labor (A)b -.36** .36** -.57** .57**.49** -.49** -8. Post-child self actual paid labor (A) .36** -.36** .57** -.57**-.49** .49** -1.00** -9. Family satisfaction (A) -.08 .09 -.04 .04 .03 -.03 .08 -.08 -10. Career satisfaction (A) -.03 .03 .00 .00 .00 .00 -.27** .27** .19* -11. Marital satisfaction (A) -.17 .1 7 -.17 .17 .02 -.02 .12 -.12 .54** .09 -12. Depression (A) -.22* .22* -.15 .15 .22* -.22* .20* -.20* -.49** -.16 -.34** 13. Phys health symp (A) -.17 .17 -.25** .25**.12 -.12 .21* -.21* -.20* .02 -.09* 14. Gender (A) -.49** .49** -.59** .59**.64** -.64** .35** -.35** .06 .16 .21* 15. Career centrality (A) .24** -.24** .18* -.18* -.33** .33** -.21* .21* -.10 .28** -.24** 16. Family centrality (A) -.26** .26** -.26** .26**.22* -.22* .20* -.20* .44** -.03 .47** 17. Voice fam labor decs (A) -.02 .0 2 .12 -.12 -.02 .02 .05 -.05 .38** .01 .47** 18. Voice paid labor decs (A) .01 -.01 .07 -.07 .02 -.02 .02 -.02 .31** .14 .31**

PAGE 161

148 Table 11. Correlations between study variables (continued) Variable 1 2 3 4 5 6 7 8 9 10 11 19. Satisfaction w/ current division of fam labor (A) .03 -.03 .19* -.19* -.10 .10 .09 .09 .29** .12 .35** 20. Satisfaction w/ current division of p aid labor ( A ) .06 -.06 .03 -.03 .02 -.02 .04 -.04 .42** .22* .39** 21. Age (A) .23* -.23* .17 -.17 -.19*.19* -.15 .15 -.01 -.01 -.04 22. Career stage (A) -.04 .04 -.02 .02 .11 -.11 -.02 .02 .15 .28** .17 23. Education level (A) -.13 .13 -.07 .07 .08 -.08 -.05 .05 .11 .26** .05 24. Weekly work hours (A) .24** -.24** .49** -.49** -.33**.33** -.53** .53**.06 .25** -.09 25. Pre-child desires for spouse’s fam labor (B) -.35** .35** -.50** .50** .50**-.50** .44** -.44**.00 .02 .00 26. Pre-child desires for self fam labor (B) .35** -.35** .50** -.50** -.50**.50** -.44** .44**.00 -.02 .00 27. Post-child spouse’s actual fam labor (B) -.45** .45** -.75** .75** .58**-.58** .66** -.66**.10 -.07 .18* 28. Post-child self actual family laborc (B) .45** -.45** .75** -.75** -.58**.58** -.66** .66**-.10 .07 -.18* 29. Pre-child desires for spouse’s paid labor (B) .51** -.51** .59** -.59** -.67**.67** -.50** .50**-.11 .04 -.11 30. Pre-child desires for self paid labor (B) -.51** .51** -.59** .59** .67**-.67** .50** -.50**.11 -.04 .11 31. Post-child spouse’s actual paid labor (B) .38** -.38** .61** -.61** -.55**.55** -.81** .81**.06 .22* -.01 32. Post-child self actual paid labor c (B) -.38** .38** -.61** .61** .55**-.55** .81** -.8** -.06 -.22* .01 33. Family satisfaction (B) .04 -.04 -.01 .01 -.04 .04 -.01 .01 .23* .12 .24** 34. Career satisfaction (B) .10 -.10 -.06 .06 -.11 .11 .18* -.18* .16 -.05 .11 35. Marital satisfaction (B) .07 -.07 -.08 .08 -.07 007 .11 -.11 .30** .06 .36** 36. Depression (B) -.10 .10 .00 .00 .06 -.06 -.12 .12 -.16 -.08 -.19*

PAGE 162

149 Table 11. Correlations between study variables (continued) 1 2 3 4 5 6 7 8 9 10 11 37. Phys health symp (B) -.20* .20* .03 -.03 .00 .00 -.03 .03 -.03 -.09 -.12 38. Gender (B) .49** -.49** .59** -.59** -.64** .64** -.35** .35** -.06 -.16 -.21* 39. Career centrality (B) .00 -.01 -.17 .17 .12 -.12 .21* -.21* .03 -.11 -.08 40. Family centrality (B) .05 -.05 .17 -.17 -.16 .16 -.14 .14 .03 .06 -.11 41. Voice fam labor decs (B) .10 -.10 -. 05 .05 -.14 .14 -.07 .07 .11 .05 .11 42. Voice paid labor decs (B) .01 -.01 -. 02 .02 -.11 .11 .01 -.01 .10 .00 .11 43. Satisfaction w/ current division of fam labor (B) -.05 .05 -.21* .21* .05 -.05 .14 -.14 .10 .12 .15 44. Satisfaction w/ current division of paid labor (B) .09 -.09 -.03 .03 -.08 .08 .05 -.05 .12 .02 .05 45. Age (B) .09 -.09 -.04 .04 .05 -.05 .05 -.05 -.07 .00 -.05 46. Career stage (B) .11 -.11 .05 -. 05 -.09 .09 .00 .00 .11 -.01 -.11 47. Education level (B) .09 -.09 -.01 .01 .00 .00 -.05 .05 .03 .05 .05 48. Weekly work hours (B) -.42** .42** -.40** .40** .38** -.38** .39** -.39** .04 .06 .08 49. Family income -.09 .09 -.01 .01 .05 -.05 .03 -.03 .08 .14 .07 50. Length of marriage .10 -.10 .17 -.17 -.08 .08 -.19* .19* -.11 .09 -.12 51. Family responsibility .20* -.20* .13 -.13 -.08 .08 -.12 .12 .05 .03 .06 52. Total number children .22* -.22* .15 -.15 -.11 .11 -.16 .16 .05 .05 .05 53. Number children over 6 .30** -.30** .22* -.22* -.23** .23** -.24** .24** .05 .13 .06 54. Number children under 6 .05 -.05 -.04 .04 .10 -.10 .04 -.04 .02 -.06 -.01

PAGE 163

150 Table 11. Correlations between study variables (continued) 12 13 14 15 16 17 18 19 20 21 22 23 12. Depression (A) -13. Phys health symp (A) .37** -14. Gender (A) .06 .22* -15. Career centrality (A) -.03 .10 -.02 -16. Family centrality (A) -.21* -.13 .29** -.29** -17. Voice fam labor decs (A) -.24** -.07 .03 -.15 .30** -18. Voice paid labor decs (A) -.18* -.11 .13 -.02 .03 .61** -19. Satisfaction w/ current division of fam labor (A) -.28** -.03 .02 -.001 .19* .69** .42** -20. Satisfaction w/ current division of paid labor (A) -.25** -.12 .12 -.01 .18* .41** .59** .50**-21. Age (A) .05 -.09 -.23* .12 .00 -.06 -.04 -.01 0114 -22. Career stage (A) -.15 -.09 .17 .03 .07 .14 .25** .08 .14 .29** -23. Education level (A) -.01 -.09 .24** .19* .27** .06 .06 .14 .11 .05 -.04 -24. Weekly work hours (A) -.17 -.15 -.30** .21* -.01 .03 -.08 .01 -.08 .14 .08 .09 25. Pre-child desires for spouse’s fam labor (B) .17 .05 .53** -.14 .20* -.06 -.02 -.06 .01 -.05 .11 .21* 26. Pre-child desires for self fam labor (B) -.17 -.05 -.53** .14 -.20* .06 .02 .06 -.01 .05 -.11 -.21* 27. Post-child spouse’s actual fam labor (B) .13 .13 .59** -.19* .28** .08 .10 -.04 .07 -.20* .05 .08 28. Post-child self actual family laborc (B) -.13 -.13 -.59** .19* -.28** -.08 -.10 .04 -.07 .20* -.05 -.08 29. Pre-child desires for spouse’s paid labor (B) -.14 -.07 -.64** .23** -.32** -.05 -.10 .00 -.06 .16 -.07 -.19* 30. Pre-child desires for self paid labor (B) .14 .07 .64** -.23** .32** .05 .10 .00 .06 -.16 .07 .19*

PAGE 164

151 Table 11. Correlations between study variables (continued) 12 13 14 15 16 17 18 19 20 21 22 23 31. Post-child spouse’s actual paid labor (B) -.21*-.27** -.41** .27** -.16 .04 .03 .08 .00 .24** .04 .06 32. Post-child self actual paid labor c (B) .21*.27** .41** -.27** .16 -.04 -.03 -.08 .00 -.24** -.04 -.06 33. Family satisfaction (B) -.17 -.03 -.03 -.13 .23** -.02 -.02 -.01 .16 -.06 .06 -.10 34. Career satisfaction (B) -.02 -.10 -.17 -.02 .15 .02 -.07 -.02 .04 -.00 -.10 -.02 35. Marital satisfaction (B) -.17 -.20* -.04 -.01 .13 -.01 .03 -.02 .11 .06 .01 -.11 36. Depression (B) .14 .05 -.07 -.01 -.12 -.08 .07 -.05 -.10 -.02 -.04 .09 37. Phys health symp (B) .05 .10 -.11 .04 -.03 .15 .18 .21* -.08 -.13 -.04 .03 38. Gender (B) -.06 -.22** -1.00** .02 -.29** -.03 -.13 -.02 -.12 .23* -.17 -.24** 39. Career centrality (B) .15 -.08 -.04 .00 -.01 -.02 -.07 -.09 -.03 .08 -.09 -.02 40. Family centrality (B) -.14 -.10 -.21* -.05 -.02 -.04 -.04 .06 .01 -.02 .01 -.08 41. Voice fam labor decs (B) .05 -.14 -.13 .1 1 -.07 -.02 .02 -.06 -.06 .09 -.12 -.10 42. Voice paid labor decs (B) .16 .00 -.26** -.03 -.05 .01 .01 .01 .10 .07 -.13 -.14 43. Satisfaction w/ current division of fam labor (B) .04 -.02 .05 .04 .16 .12 .18* .11 .21*-.05 -.12 .08 44. Satisfaction w/ current division of paid labor(B) .07 -.01 -.08 -.03 .04 .03 -. 01 -.01 .17 -.01 -.16 -.04 45. Age (B) .10 .09 .12 .0 8 .05 -.07 .01 -.04 -.18*.79** .27** .09 46. Career stage (B) -.03 .04 -.08 .03 .02 -.07 -.10 .03 -.08 .37** .16 .00 47. Education level (B) -.07 -.26** -.13 .16 -.01 -.02 -.02 .04 .04 .24** .12 .28** 48. Weekly work hours (B) .18 .16 .37** -.04 .13 .00 .01 .00 -.01 -.19* -.02 .27** 49. Family income .03 -.19* -.06 -.05 .03 .05 .00 -.05 -.06 .20* .15 .04 50. Length of marriage .05 .01 .03 .18* -.05 -.02 .04 -.02 -.15 .59** .28** .06 51. Family responsibility -.01 -.09 -.05 -.08 .08 .03 .08 .08 .12 .32** .29** .07 52. Total number children -.02 -.03 -.05 -.05 .06 .03 .06 .07 .09 .40** .31** .07 53. Number children over 6 -.12 -.03 -.10 .12 -.02 .04 .08 .10 .03 .50** .29** .02 54. Number children under 6 .05 -.12 .03 -.19* .08 -.04 .01 -.04 .13 -.12 .06 .09

PAGE 165

152 Table 11. Correlations between study variables (continued) 24 25 26 27 28 29 30 31 32 33 34 24. Weekly work hours (A) -25. Pre-child desires for spouse’s fam labo r (B) -.37** -26. Pre-child desires for self fam labor (B) .37** -1.00** -27. Post-child spouse’s actual fam labor (B) -.50** .63** -.63** -28. Post-child self actual family laborc (B) .50** -.63** .63** -1.00** -29. Pre-child desires for spouse’s paid labor (B) .42** -.67** .67** -.66** .66** -30. Pre-child desires for self paid labor (B) -.42** .67** -.67** .66** -.66** -1.00** -31. Post-child spouse’s actual paid labor (B) .63** -.51** .51** -.64** .64** .59** -.59** -32. Post-child self actual paid labor c (B) -.63** .51** -.51** .64** -.64** -.59** .59** -1.00* -33. Family satisfaction (B) -.04 .15 -. 15 .02 -.02 .06 -.06 .06 -.06 -34. Career satisfaction (B) -.09 -.05 .05 .05 -.05 .13 -.13 -.13 .13 .24** -35. Marital satisfaction (B) .03 .04 -. 04 .06 -.06 .06 -.06 .14 -.14 .59** .16 36. Depression (B) .01 -.15 .15 -.12 .12 .04 -.04 .00 .00 -.37** -.27** 37. Phys health symp (B) .03 -.09 .09 -.09 .09 .00 .00 .03 -.03 -.30** -.20* 38. Gender (B) .30** -.53** .53** -.59** .59** .64** -.64** .40**-.40** .03 .17 39. Career centrality (B) -.06 .17 -.17 .21* -.21* -.12 .12 -.21* .21* -.09 .40** 40. Family centrality (B) .10 -.03 .03 -.17 .17 .10 -.13 .13 -.13 .36** .16 41. Voice fam labor decs (B) .02 .01 -.01 .17 -.17 .10 -.10 .21* -.21* .33** .16 42. Voice paid labor decs (B) -.13 -.02 .02 .08 -.08 .13 -.13 .06 -.06 .21* .23* 43. Satisfaction w/ current division of fam labor (B) -.14 .09 -.09 .40** -.40** -.06 .06 .00 .00 .27** .22*

PAGE 166

153 Table 11. Correlations between study variables (continued) 24 25 26 27 28 29 30 31 32 33 34 44. Satisfaction w/ current division of paid labor (B) -.02 .04 -.04 .07 -.07 .04 -.04 .05 -.05 .38** .43** 45. Age (B) .02 .13 -.13 .05 -.05 -.11 .11 -.03 .03 -.18* -.06 46. Career stage (B) -.05 -.05 .05 -.03 .03 .03 -.03 .01 -.01 .05 .23** 47. Education level (B) .09 .07 -.07 -.02 .02 .05 -.05 .02 -.02 -.04 .12 48. Weekly work hours (B) -.08 .33** -.33** .46**-.46** -.46** .46** -.50** .50** -.13 -.05 49. Family income .22* .04 -.04 .02 -.02 -.04 .04 .12 -.12 .11 .11 50. Length of marriage .13 -.09 .09 -.13 .13 .06 -.06 .18* -.18* -.22* -.12 51. Family responsibility .12 -.06 .06 -. 07 .07 .06 -.06 .17 .17 -.11 .01 52. Total number children .14 -.08 .08 -.10 .10 .08 -.08 .21* -.21* -.11 -.02 53. Number children over 6 .07 -.11 .11 -.11 .11 .13 -.13 .26** -.26** -.02 -.07 54. Number children under 6 .10 .01 -.01 .01 -.01 -.03 .03 -.01 .01 -.06 .04

PAGE 167

154 Table 11. Correlations between study variables (continued) 35 36 37 38 39 40 41 42 43 44 45 35. Marital satisfaction (B) -36. Depression (B) -.38** -37. Phys health symp (B) -.30** .50** -38. Gender (B) .04 .07 .11 -39. Career centrality (B) -.06 .01 -.09 .04 -40. Family centrality (B) .28** -.18* -.02 .21* -.21* -41. Voice fam labor decs (B) .46** -.31** -.17 .13 .20* .03 -42. Voice paid labor decs (B) .25** -.16 -.07 .26** .12 .11 .53** -43. Satisfaction w/ current division of fam labor (B) .34** -.36** -.15 -.05 .11 .03 .60** .35** -44. Satisfaction w/ current division of p aid labor ( B ) .28** -.31** -.30**.08 .16 .12 .38** .50** .40** -45. Age (B) -.04 .02 -.10 -.12 .0 7 -.08 -.04 -.03 -.05 -.04 -46. Career stage (B) -.05 -.04 -.19*.08 .02 .11 -.04 -.01 -.01 .19* .41** 47. Education level (B) .09 .07 -.04 .13 .28** -.04 .06 .02 .02 .02 .14 48. Weekly work hours (B) -.15 .06 .07 -.37** .20* -.19* -.18* -.20* -.04 -.17 .05 49. Family income .14 -.05 -.03 .06 .13 .08 .11 -.01 -.04 .00 .23* 50. Length of marriage -.10 -.07 -.13 -.03 -.08 -.02 -.06 -.01 -.05 -.06 .62** 51. Family responsibility .06 -.07 -.11 .05 -.14 .08 -.09 .04 -.03 .07 .33** 52. Total number children .07 -.10 -.12 .05 -.16 .08 -.07 .03 -.02 .05 .40** 53. Number children over 6 .14 -.15 -. 16 -.10 -.12 .13 .05 .03 .08 .02 .41** 54. Number children under 6 -.03 .08 -.03 -.03 -.07 -.05 -.13 .03 -.10 .09 -.01

PAGE 168

155 Table 11. Correlations between study variables (continued) p < .05 ** p < .01 In P-E fit language this represents Partner A’ s labor needs and Partner B’s labor demands b In P-E fit language this represents Partner A’s labor supplies c In P-E fit language this represents Partner B’s labor abilities N for correlations vary from 117 to 126 due to missing data. Letter in parentheses (A or B) indicates source of the variable. Variables without parentheses are common to both A and B. Shaded variables are those u sed in hypothesis testing. 45 46 47 48 49 50 51 52 53 54 45. Age (B) -46. Career stage (B) .41** -47. Education level (B) .14 -.09 -48. Weekly work hours (B) .05 .08 .06 -49. Family income .23* .18 .09 .26**-50. Length of marriage .62** .27** .15 -.08 .04 -51. Family responsibility .33** .21* -.02 -.16 -.03 .52** -52. Total number children .40** .24** .00 -.18*-.02 .61** .98**-53. Number children over 6 .41** .26** .11 -.22*-.07 .64** .57**.69**-54. Number children under 6 -.01 .05 -.15 .04 .09 -.02 .59**.47**-.27** -

PAGE 169

156 Table 12. Correlations between study variables used in hypothesis testing p < .05, ** p < .01 In P-E fit language this represents Partner A’ s labor needs and Partner B’s labor demands b In P-E fit language this represents Partner A’s labor supplies c In P-E fit language this represents Partner B’s labor abilities Letter in parentheses (A or B) indicates source of the variable. Career sat (A) Family sat (A) Marital sat (A) Deprs (A) Health (A) Marital sat (B) Deprs (B) Health (B) Pre-child desires for spouse’s fam labor (A) -.03 -.08 -.17 -.22* -.17 .07 -.10 -.20* Pre-child desires for spouse’s paid labor (A) .00 .03 .02 .22* .12 -.07 .06 .00 Post-child spouse’s actual fam labor b (A) .00 -.04 -.17 -.15 -.25**-.08 .00 .03 Post-child spouse’s actual paid labor b (A) -.27** .08 .12 .20* .21* .11 -.12 -.03 Post-child self actual family laborc (B) .07 -.10 -.18* -.13 -.13 -.06 .12 .09 Post-child self actual paid laborc (B) -.22* -.06 .01 .21* .27**-.14 .00 -.03 Career centrality (A) .28** -.10 -.24** -.03 .10 -.01 -.01 .04 Family centrality (A) -.03.44**.47**-.21* -.13.13-.12-.03 Gender (A) .16 .06 .21 .06 .22* .04 -.07 -.11 Gender (B) -.16 -.06 -.21* -.06 -.22* .04 .07 .11 Voice in division of family labor (A) .01 .38** .47** -.24** -.13 -.01 -.08 .15 Voice in division of paid labor (A) .14 .31** .31** -.18* -.07 .03 .07 .18* Satisfaction w/ current division of family labor (B) .12 .29** .35** .04 -.02 .34** -.36**-.15 Satisfaction w/ current division of paid labor (B) .22* .42** .39** .07 -.01 -.28** -.31**-.30**

PAGE 170

157 Table 13. Polynomial regression equations regressing Partner A’s well-being on Partner A’s paid labor needs and supplies (Hypotheses1, 3, 5, 7) † p < .10, p < .05, ** p < .01. The coefficients listed are the unstandardized b weights. A’s needs represent Partner A’s pre-child desi res for spouse’s paid labor contributions A’s supplies represent Partner A’s spouse’s pos t-child actual paid labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). b1 b2 b3 b4 and b5 are the coefficients of needs, supplies, needs2, needs x supplies, and supplies2 Partner A’s well-being Career Satisfaction Marital satisfaction Depression Phys health Symptoms Control variables Family responsib -.0144 .0115 -.0012 -.1542 Length of marriage .0247 -.0233 .0173 .2096 Total family income .0681 .0359 .0154 -.7598* Needs and supplies Partner A’s paid labor needs(A) .0010 .0054 .0006 -.0128 Partner A’s paid labor supplies(A) -.0197** .0005 .0092 .0816* Partner A’s paid labor needs2 -.0002 -.0002 .0000 -.0003 Partner A’s paid labor needs x Partner A’s paid labor supplies .0002 .0007* -.0005* -.0034* Partner A’s paid labor supplies2 .0002 -.0002 .0003 0.0023 Intercep t 3.1674.2952.13724.167 F Df 2.50* 8, 114 2.49* 8, 114 1.95† 8, 111 2.71* 8, 111 R2 .148 .149 .123 .164 Needs = Supplies (misfit) line shape b1 b2 .0297** .0049 -.0086 -.0944 b3 b4 + b5 -.0003 -.0011** .0009* .0053† Needs = Supplies (fit) line shape b1 + b2 .0097* .0059 .0097* .0688* b3 + b4 + b5 .0002 .0004* -.00 01 -.0014

PAGE 171

158 Table 14. Polynomial regression equations regressing Partner A’s well-being on Partner A’s family labor needs and supplies (Hypotheses 2, 4, 6, 8) † p < .10, p < .05, ** p < .01. The coefficients listed are the unstandardized b weights. Partner A’s needs represent Partner A’s pre-child desires for spouse’s family labor contributions; Partner A’s supplies represent Partner A’s spouse’s post-child actual family labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). b1 b2 b3 b4 and b5 are the coefficients of needs, supplies, needs2, needs x supplies, and supplies2 Partner A’s well-being Family satisfaction Marital satisfaction Depression Phys health symptoms Control variables Family responsib .0278 .0255 -.0017 -.1494 Length marriage -.0357 -.0363 .0176 .1947 Total family income .0442 .0447 -.0027 -.7871* Needs and supplies Partner A’s family labor needs(A) -.0063 -.0081 -.0069 .0019 Partner A’s family labor supplies(A) -.0005 -.0036 -.0023 -.0679* Partner A’s family labor needs2 -.0001 -.0002 -.0001 -.0021 Partner A’s family labor needs x Partner A’s family labor supplies .0008* .0008* -.0004 -.0023 Partner A’s family labor supplies2 -.0004* -.0001 .0001 -.0004 Intercept 4.064 4.137 2.315 24.695 F df 1.81† 8, 114 2.50* 8, 114 1.46 8,111 3.17** 8, 111 R2 .113 .149 .095 .186 Needs = Supplies (misft) line shape b1 b2 -.0058 -.0046 .0698 b3 b4 + b5 -.0013* -.0011† -.0001 Needs = Supplies (fit) line shape b1 + b2 -.0068 -.0117* -.0060† b3 + b4 + b5 .0003 .0006** -.0048**

PAGE 172

159 Table 15. Hierarchical moderated regression involving the regression of Partner A’s well-being on the interaction between career centrality and Partner A’s paid labor fit (Hypothesis9) † p < .10, p < .05, ** p < .01. The coefficients listed are the unstandardized b weights. A’s needs represent Partner A’s pre-child desi res for spouse’s paid labor contributions A’s supplies represent Partner A’s spouse’s pos t-child actual paid labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). Partner A’s well-being Career satisfaction Marital satisfaction Depression Phys health symptoms Control variables Family responsibility .0031 -.0023 .0003 -.1217 Length of marriage .0036 -.0125 .0128 .2057 Total family income .0890 .0172 .0236 -.6494 Needs and supplies Partner A’s paid labor needs (A) .0246 .0006 .0230 -.029 Partner A’s paid labor supplies (A) -.0053 -.0083 -.0008 .2421 Partner A’s paid labor needs2 -.0009 .0003 -.0003 .0012 Partner A’s paid labor needs x Partner A’s paid labor supplies .0009 -.0012 -.0005 -.0024 Partner A’s paid labor supplies2 .001 .001 .0007 -.0069 Career centrality .2771* -.1526 .0277 .2172 R2 .195 .196 .124 .206 Interaction Terms Partner A’s needs x career cent -.0039 .0011 -.0068 .0052 Partner A’s supplies x career cent -.0041 .0030 .0033 -.0472 Partner A’s needs2 x career cent .0002 -.0002 .0001 -.0005 Partner A’s needs x Partner A’s supplies x career cent -.0002 .0006 -.0000 -.0005 Partner A’s supplies2 x career cent -.0003 -.0004 -.0001 .0029 Intercept 2.069 4.995 2.007 22.281 R2 .233.263.137.188 R2 .038 .067† .013 .018 F df 3.05** 14, 108 3.06** 14, 108 1.73 14, 105 1.74 14, 105

PAGE 173

160 Table 16. Hierarchical moderated regression involving the regression of Partner A’s well-being on the interaction between family centrality and Partner A’s family labor fit (Hypothesis10) p < .05, ** p < .01. The coefficients listed are the unstandardized b weights. A’s needs represent Partner A’s pre-child desi res for spouse’s family labor contributions A’s supplies represent Partner A’s spouse’s post-child actual family labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). Partner A’s well-being Family satisfaction Marital satisfaction Depression Phys health symptoms Control variables Family responsibility 0175 .0146 .0087 -.0973 Length of marriage -.0284 -.0322 .0092 .1384 Total family income .0362 .0416 -.0053 -.7517* Needs and supplies Partner A’s family labor needs (A) .0215 -.0111 .0954 -.7274 Partner A’ family labor supplies (A) -.0089 -.0806* -.0299 -.3008 Partner A’s family labor needs2 -.0002 -.0004 .0006 .0204 Partner A’s family labor needs x Partner A’s family labor supplies -.0011 .0038 -.0069 .0103 Partner A’s family labor supplies2 .0001 .0005 .0028 .0184 Family centrality .4939 .4916** -.2655 -.3166 R2 .249 .285 .163 .199 Interaction Terms Partner A’s needs x fam cent -.0050 .0002 -.0222 .1442 Partner A’s supplies x family cent .0024 .0175* .0058 .0453 Partner A’s needs2 x family cent .0001 .0001 -.0003 -.0052 Partner A’s needs x Partner A’s supplies x family cent .0004 -.0007 .0015 -.0014 Partner A’s supplies2 x family cent -.0003 -.0001 -.0006 -.0043 Intercept 1.946 2.028 3.499 25.853 R2 .257 .314 .194 .231 R2 .009 .029 .03 .032 F df 4.16** 14, 108 5.01** 14, 108 2.29* 14, 105 3.03** 14, 105

PAGE 174

161 Table 17. Hierarchical moderated regression involving the regression of Partner A’s well-being on the interaction between gender and Partner A’s paid labor fit (Hypothesis11) p < .05, ** p < .01. Gender is coded male = 0 female = 1 The coefficients listed are the unstandardized b weights. A’s needs represent Partner A’s pre-child desi res for spouse’s paid labor contributions A’s supplies represent Partner A’s spouse’s pos t-child actual paid labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). Partner A’s well-being Career satisfaction Marital satisfactio Depression Physical health symptoms Control variables Family responsibility -.0049 .0191 .0013 -.0825 Length of marriage .0126 -.0317 .0154 .1173 Total family income .0838 .0295 .0253 -.623* Needs and supplies Partner A’s paid labor needs (A) -.0009 -.0025 .0122 .1417 Partner A’s paid labor supplies (A) -.0178 .0016 .0074 .0092 Partner A’s paid labor needs2 -.0005 .0001 -.0001 .0018 Partner A’s paid labor needs x Partner A’s paid labor supplies .0008 .0001 .0001 .0013 Partner A’s paid labor supplies2 -.0003 -.0001 .0001 -.0038 Gender .3474 .3948* -.2170 .8958 R2 .179 .209 .137 .182 Interaction Terms Partner A’s needs x Gender -.0076 .0069 -.0323 -.4181** Partner A’s supplies x Gender .0018 -.0154 .0137 .1576 Partner A’s needs2 x Gender .0006 -.0004 .0007 .0029 Partner A’s needs x A’s supplies x Gender -.0008 .0011 -.0001 -.0079 Partner A’s supplies2 x Gender .0005 -.0001 .0002 .0071 Intercept 2.914 4.104 2.248 23.732 Total R2 .195 .255 .207 .297 R2 .016 .047 .069 .115** F df 1.87* 14, 108 2.65** 14, 108 1.96* 14, 105 3.17** 14, 105

PAGE 175

162 Table 18. Hierarchical moderated regression involving the regression of Partner A’s well-being on the interaction between gender and Part ner A’s family labor fit (Hypothesis12) p < .05, ** p < .01. The coefficients listed are the unstandardized b weights. Gender is coded male = 0 female = 1 A’s needs represent Partner A’s pre-child desi res for spouse’s family labor contributions A’s supplies represent Partner A’s spouse’s post-child actual family labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). Partner A’s well-being Family satisfaction Marital satisfaction Depression Phys health symptoms Control variables Family responsibility .0230 .0236 -.0004 -.1759 Length of marriage -.0321 -.0391 .0186 .2265 Total family income .0631 .0635 -.0148 -.6557 Needs and supplies Partner A’s family labor needs (A) -.0255 -.0215 .0090 -.1519 Partner A’ family labor supplies (A) -.0026 -.0028 -.0139 .0207 Partner A’s family labor needs2 .0002 .0001 -.0003 -.0002 Partner A’s family labor needs x Partner A’s family labor supplies .0012 .0011* -.0010 .0032 Partner A’s family labor supplies2 -.0003 -.0000 .0007 -.0045 Gender -.0349 .1809 -.033 -.5695 R2 .113 .16 .101 .189 Interaction Terms Partner A’s needs x Gender .0239 .0319 -.0155 -.0948 Partner A’s supplies x Gender -.0120 -.0102 .0112 -.1874 Partner A’s needs2 x Gender -.0003 .0001 .0003 -.0118 Partner A’s needs x A’s supplies x Gender -.0005 -.0001 .0008 -.0085 Partner A’s supplies2 x Gender -.0007 -.0005 -.0006 .0025 Intercept 3.944 3.971 2.368 23.91 Total R2 .163 .185 .127 .237 R2 .05 .025 .026 .048 F df 1.51 14, 108 1.76 14, 108 1.09 14, 105 2.33** 14, 105

PAGE 176

163 Table 19. Hierarchical moderated regression involving the regression of Partner A’s well-being on the interaction between voice in paid labor decision making and Partner A’s paid labor fit (Hypothesis 13) p < .05, ** p < .01. The coefficients listed are the unstandardized b weights. A’s needs represent Partner A’s pre-child desi res for spouse’s paid labor contributions A’s supplies represent Partner A’s spouse’s pos t-child actual paid labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). Partner A’s well-being Career satisfaction Marital satisfaction Depression Physical health symptoms Control variables Family responsibility -.0060 .0046 .0003 -.134 Length of marriage .0172 -.0205 .0244 .1933 Total family income .0925 .0126 -.0019 -.7751 Needs and supplies Partner A’s paid labor needs(A) .0095 .0301 .0070 .1575 Partner A’s paid labor supplies (A) -.0061 .0010 -.0164 .2174 Partner A’s paid labor needs2 .0002 .0014 -.0020* -.0006 Partner A’s paid labor needs x Partner A’s paid labor supplies .0013 -.0018 .0043* .0101 Partner A’s paid labor supplies2 -.0013 .0022 .0002 .0034 Voice paid labor decs .1036 .5145 -.1492 .2465 R2 .175 .209 .140 .164 Interaction Terms Partner A’s needs x Voice .0001 -.0062 -.0013 -.0368 Partner A’s supplies x Voice -.0033 .0004 .0060 -.0331 Partner A’s needs2 x Voice -.0001 -.0004 .0005* .0000 Partner A’s needs x Partner A’s supplies x Voice -.0003 .0006 -.0011* -.0030 Partner A’s supplies2 x Voice .0004 -.0006 .00003 -.0001 Intercept 2.496 2.289 2.825 23.103 Total R2 .210 .306 .193 .188 R2 .035.097*.052.024 F df 1.99* 14, 105 3.30** 14, 105 1.79* 14, 105 1.74 14, 105

PAGE 177

164 Table 20. Hierarchical moderated regression involving the regression of Partner A’s well-being on the interaction between voice in family labor decision making and Partner A’s paid labor fit (Hypothesis 14) p < .05, ** p < .01. The coefficients listed are the unstandardized b weights. A’s needs represent Partner A’s pre-child desi res for spouse’s family labor contributions A’s supplies represent Partner A’s spouse’s post-child actual family labor contributions Letter in parentheses (A) indicates source of the variable (Partner A). Partner A’s well-being Family satisfaction Marital satisfaction Depression Physical health symptoms Control variables Family responsibility 0264 .0161 .0016 -.0899 Length of marriage -.0349 -.0261 .0153 .1301 Total family income .0307 .0153 .0007 -.8635 Needs and supplies Partner A’s family labor needs (A) -.0115 .0276 .0374 -.0124 Partner A’s family labor supplies (A) -.0278 -.0213 -.0022 -.2905 Partner A’s family labor needs2 -.0000 -.0054 .0007 .0298 Partner A’s family labor needs x Partner A’s family labor supplies .0019 .0079* -.002 -.0340 Partner A’s family labor supplies2 -.0013 -.001 .0004 .0034 Voice paid labor decs .216 2* .3528** -.1719 .1741 R2 .199 .348 .136 .189 Interaction Terms Partner A’s family labor needs x Voice .0012 -.0068 -.0112 -.0076 Partner A’s family labor supplies x Voice .0068 .0040 .0004 .0613 Partner A’s family labor needs2 x Voice -.0000 .0013 -.0002 -.0077 Partner A’s family labor needs x Partner A’s family labor supplies x Voice -.0004 -.0019* .0005 .0079 Partner A’s family labor supplies2 x Voice .0003 .0003 -.0001 -.0012 Intercept 3.265 2.898 2.971 24.494 Total R2 .237 .395 .155 .221 R2 .038 .047 .019 .032 F df 2.34** 14, 105 4.90** 14, 105 1.38 14, 105 2.13* 14, 105

PAGE 178

165 Table 21. Polynomial regression equations regressing Partner B’s well-being on Partner B’s paid labor demands and abiliti es (Hypotheses 15, 17, 19) p < .05 The coefficients listed are the unstandardized b weights. B’s demands represent Partner A’s pre-child desires for spouse’s family labor contributions B’s abilities represent Partner B’s post-child actual family labor contributions Letter in parentheses (A or B) indicates source of the variable (Partner A or B). Partner B’s well-being Marital satisfaction Depression Phys health symptoms Control variables Family responsibility .0218 -.0013 -.0542 Length of marriage -.0322 -.0149 -.2504 Total family income .0534 -.0291 -.0272 Demands and abilities Partner B’s paid labor demands (A) .0002 .0021 -.0061 Partner B’s paid labor abilities (B) -.0054 -.0032 -.0202 Partner B’s paid labor demands 2 -.0001 .0001 .0022* Partner B’s paid labor x Partner B’s paid labor abilities .0003 -.0004 -.0032 Partner B’s paid labor abilities2 -.0002 .0001 .0011 Intercept 4.175 2.798 19.665 F df 1.18 8, 116 .55 8, 116 1.47 8, 116 R2 .075 .037 .077

PAGE 179

166 Table 22. Polynomial regression equations regressing Partner B’s well-being on Partner B’s family labor demands and abilities (Hypotheses 16, 18, 20) † p < .10, p < .05, ** p < .01 The coefficients listed are the unstandardized b weights. B’s demands represent Partner A’s pre-child desires for spouse’s family labor contributions B’s abilities represent Partner B’s post-child actual family labor contributions Letter in parentheses (A or B) indicates source of the variable (Partner A or B). b1 b2 b3 b4 and b5 are the coefficients of demands, abilities, demands2, demands x abilities, and abilities2 Partner B’s well-being Marital satisfaction Depression Phys health symptoms Control variables Family responsibility .0214 .0039 .0287 Length of marriage -.0302 -.0114 -.2232 Total family income .0669 -.0328 -.1844 Demands and abilities Partner B’s family labor demands (A) .0064 -.0054 -.1224** Partner B’s family labor abilities (B) -.0061 .0078 .1009* Partner B’s family labor demands 2 .0001 .0002 -.0002 Partner B’ family labor x Partner B’s family labor abilities -.0001 -.0007* .0002 Partner B’s family labor abilities2 .0001 .0001 -.0016 Intercept 4.015 2.737 20.366 F df 1.05 8, 116 1.67 8, 116 2.04* 8, 116 R2 .067 .103 .123 Shape of Demands = Abilities line b1 b2 -.2233** b3 b4 + b5 -.0021 Shape of Demands = Abilities line b1 + b2 -.0195 b3 b4 + b5 -.0016

PAGE 180

167 Table 23. Hierarchical moderated regression involving the regression of Partner B’s well-being on the interaction between gender and Partner B’s paid labor fit (Hypothesis 21) p < .05 The coefficients listed are the unstandardized b weights. Gender is coded male = 0 female = 1 B’s demands represent Partner A’s pre-child desires for spouse’s paid labor contributions B’s abilities represent Partner B’s post-ch ild actual paid labor contributions Letter in parentheses (A or B) indicates source of the variable (Partner A or B). Partner B’s well-being Marital satisfaction Depression Phys health symptoms Control variables Family responsibility .0153 .0053 -.0002 Length of marriage -.0204 -.0173 -.2908 Total family income .0651 -.0414 -.2219 Demands and abilities Partner B’s paid labor demands (A) .0079 -.0043 -.0728 Partner B’s paid labor abilities (B) .003 -.0076 -.0556 Partner B’s paid labor demands2 -.0002 .0003 .0036* Partner B’s paid labor x Partner B’s paid labor abilities -.0004 -.0002 .0001 Partner B’s paid labor abilities2.0005 .0000 .0005 Gender (B) .106 .0932 .8047 R2 .077 .05 .103 Interaction Terms Partner B’s demands (A) x Gender .0016 -.0095 .0012 Partner B’s abilities (B) x Gender -.022 .0097 -.0024 Partner B’s demands2 x Gender .0005 -.0007 -.0031 Partner B’s demands x B’s abilities x Gender .0005 -.0001 -.0063 Partner B’s abilities2 x Gender -.001* .0001 .0013 Intercept 3.960 2.782 20.317 Total R2 .171 .087 .138 R2 .094* .037 .034 F df 1.62 14, 110 .75 14, 110 1.25 14, 110

PAGE 181

168 Table 24. Hierarchical moderated regression involving the regression of Partner B’s well-being on the interaction between gender and Partner B’s family labor fit (Hypothesis 22) † p < .10, p < .05 The coefficients listed are the unstandardized b weights. Gender is coded male = 0 female = 1 B’s demands represent Partner A’s pre-child desires for spouse’s family labor contributions B’s abilities represent Partner B’s post-child actual family labor contributions Letter in parentheses (A or B) indicates source of the variable (Partner A or B). Partner B’s well-being Marital satisfaction Depression Physical health symptoms Control variables Family responsibility .0185 .0094 .0010 Length of marriage -.0295 -.0154 -.1971 Total family income .0593 -.0518 -.3756 Demands and abilities Partner B’s family labor de mands (A) .0327 .0073 -.0019 Partner B’s family labor ab ilities (B) -.0099 .0192* .0571 Partner B’s family labor demands 2 .0009 .0010 .0123 Partner B’s family labor x Partner B’s family labor abilities .0005 -.0006 -.0118 Partner B’s family labor abilities2-.0003 .0004 .0037 Gender (B) -.0931 .1076 2.5929 R2 .069 .104 .140 Interaction Terms Partner B’s demands (A) x Gender -.0254 .0127 -.0469 Partner B’s abilities (B) x Gender .0119 -.0306 .1294 Partner B’s demands 2 x Gender -.0008 -.0012 -.0159 Partner B’s demands x Partner B’s abilities x Gender -.0008 -.0007 .0136 Partner B’s abilities2 x Gender .0003 .0002 -.0088* Intercept 4.134 4.072 20.796 Total R2 .097 .183 .213 R2 .028 .079† .073† F df .84 14, 110 1.76 14, 110 2.13* 14, 110

PAGE 182

169 Table 25. Hierarchical moderated regression involving the regression of Partner B’s well-being on the interaction between satisfaction with curren t division of paid labor and Partner B’s paid labor fit (Hypothesis 23) † p < .10, p < .05, ** p < .01 The coefficients listed are the unstandardized b weights. B’s demands represent Partner A’s pre-child desires for spouse’s paid labor contributions B’s abilities represent Partner B’s post-ch ild actual paid labor contributions Letter in parentheses (A or B) indicates source of the variable (Partner A or B). Partner B’s well-being Marital satisfaction Depression Phys health symptoms Control variables Family responsibility .0163 .0094 .0500 Length of marriage -.0263* -.0204 -.3835* Total family income .0648 -.0479 -.1353 Demands and abilities Partner B’s paid labor demands (A) -.0149 .0132 .0103 Partner B’s paid labor abilities (B) -.0361 -.0000 .0043 Partner B’s paid labor demands 2 .0000 .0007* .0095** Partner B’s paid labor x Partner B’s paid labor abilities -.0021** .0003 -.0118 Partner B’s paid labor abilities2-.0001* .0003 .0009 Satisfaction with current division of paid labor (B) .1184 -.0762 -.7842 R2 .132 .119 .154 Interaction Terms Partner B’s demands (A) x Satisfaction .0038 -.0035 -.0087 Partner B’s abilities (B) x Satisfaction .0085 -.0005 -.0059 Partner B’s demands 2 x Satisfaction .0000 -.0002* -.0023** Partner B’s demands x Partner B’s abilities x Satisfaction .0005** -.0001 .0033 Partner B’s abilities2 x Satisfaction .0001 -.0001 -.0006 Intercept 3.582 3.207 23.688 Total R2 .268 .207 .224 R2 .136** .088* .07† F df 2.87** 14, 110 2.05* 14, 110 2.27** 14, 110

PAGE 183

170 Table 26. Hierarchical moderated regression involving the regression of Partner B’s well-being on the interaction between satisfaction with cu rrent division of family labor and Partner B’s family labor fit (Hypothesis 24) p < .05, ** p < .01 The coefficients listed are the unstandardized b weights. B’s demands represent Partner A’s pre-child desires for spouse’s family labor contributions B’s abilities represent Partner B’s post-child actual family labor contributions Letter in parentheses (A or B) indicates source of the variable (Partner A or B). Partner B’s well-being Marital satisfaction Depression Phys health symptoms Control variables Family responsibility .0216 .0018 .0281 Length of marriage -.0395* -.0045 -.2489 Total family income .0915* -.0404 -.3298 Demands and abilities Partner B’s family labor dema nds (A) .0751* -.0614* -.5681* Partner B’s family labor abilities (B) -.0134 .0080 .1627 Partner B’s family labor demands 2 -.0006 .0007 .0086 Partner B’s family labor x Partner B’s family labor abilities -.0027* .0009 .0040 Partner B’s family labor abilities2.0007 .0003 -.0028 Satisfaction with current division of family labor (B) .2277** -.1495* -.2362 R2 .199 .214 .131 Interaction Terms Partner B’s demands (A) x Satisfaction -.0194* .0161* .1222 Partner B’s abilities (B) x Satisfaction .0031 -.0015 -.0175 Partner B’s demands 2 x Satisfaction .0002 -.0001 -.0027 Partner B’s demands x Partner B’s abilities x Satisfaction .0007 -.0004 -.0007 Partner B’s abilities2 x Satisfaction -.0001 -.0002 .0003 Intercept 3.04 3.350 22.432 Total R2 .260 .294 .051 R2 .061 .080* .027 F df 2.76* 14, 110 3.27** 14, 110 1.47 14, 110

PAGE 184

171 Table 27. Summary of hypothesis testing Hypoth. Variables Supported 1 Paid needs-supplies and career satisfaction Partial 2 Family needs-supplies and family satisfaction 3 Paid needs-supplies and marital satisfaction 4 Family needs-supplies and marital satisfaction 5 Paid needs-supplies and depression 6 Family needs-supplies and depression X 7 Paid needs-supplies and physical health symptoms 8 Family needs-supplies and physical health symptoms X 9a Paid needs-supplies x career centrality and career satisfaction X 9b Paid needs-supplies x career centrality and marital satisfaction Partial 9c Paid needs-supplies x career centrality and depression X 9d Paid needs-supplies x career centrality and physical health symptoms X 10a Family needs-supplies x family centrality and family satisfaction X 10b Family needs-supplies x family centrality and marital satisfaction X 10c Family needs-supplies x family centrality and depression X 10d Family needs-supplies x family centrality and phy health symptoms X 11a Paid needs-supplies x gender and career satisfaction X 11b Paid needs-supplies x gend er and marital satisfaction X 11c Paid needs-supplies x gender and depression X 11d Paid needs-supplies x gender and physical health symptoms X~ 12a Family needs-supplies x gender and family satisfaction X 12b Family needs-supplies x gender and marital satisfaction X 12c Family needs-supplies x gender and depression X 12d Family needs-supplies x gender and physical health symptoms X 13a Paid needs-supplies x voice and career satisfaction X 13b Paid needs-supplies x voice and marital satisfaction X~ 13c Paid needs-supplies x voice and depression X 13d Paid needs-supplies x voice and physical health symptoms X 14a Family needs-supplies x voice and family satisfaction X 14b Family needs-supplies x voice and marital satisfaction X 14c Family needs-supplies x voice and depression X 14d Family needs-supplies x voice and physical health symptoms X 15 Paid demands-abilities fit and marital satisfaction X 16 Family demands-abilities and marital satisfaction X 17 Paid demands-abilities and depression X 18 Family demands-abilities and depression X 19 Paid demands-abilities and physical health symptoms X 20 Family demands-abilities and physical health symptoms Partial 21a Paid demands-abilities fit x gende r and marital satisfaction X~ 21b Paid demands-abilities f it x gender and depression X 21c Paid demands-abilities fit x gender and physical health symptoms X 22a Family demands-abilities fit x gender and marital satisfaction X 22b Family demands-abilities f it x gender and depression X~ 22c Family demands-abilities fit x gender and physical health symptoms X~ 23a Paid demands-abilities fit x satisfa ction and marital satisfaction X~ 23b Paid demands-abilities fit x sa tisfaction and depression X~

PAGE 185

172 23c Paid demands-abilities fit x sa tisfaction and physical health symptoms X~ 24a Family demands-abilities fit x satis faction and marital satisfaction X 24b Family demands-abilities fit x satisfaction and depression X~ 24c Family demands-abilities fit x satisfaction and phys health symptoms X indicates full support, X indicates no support, X~ indicates significant results but nature of moderation different than predicted.

PAGE 186

173 Table 28. T-test results comparing Partner As that not are not matched to partner As that are matched preand postrandomization. Column Number 1 2 3 Hypotheses Variables Non-matched (N = 452) Prerandomized matched (N = 126) F (1 vs. 2) Randomized matched (N = 126) F (1 vs. 3) Pre-child desires for spouse fam labor (needs for A/demands for B) 43.68 (11.22) 44.47 (11.31) .71 49.77 (12.51) 5.23** Post-child spouse’s actual fam labor (supplies) 33.31 (15.06) 37.66 (15.14) 2.88** 46.94(17.06) 8.57** Pre-child desires for spouse paid labor (needs for A/demands for B) 64.94 (17.74) 62.11 (17.64) 1.60 50.36 (21.16) 7.05** Post-child spouse’s actual paid labor (supplies) 59.65 (17.74) 54.45 (15.69) 3.22** 48.26 (15.83) 6.89** Family satisfaction 4.34 (.60) 4.34 (.62) .82 4.33 (.62) 1.00 Career satisfaction 3.75 (.75) 3.81 (.68) .73 3.68 (.80) .95 Marital satisfaction 4.43 (.73) 4.54 (.66) 1.47 4.46 (.64) .45 Depression 2.41 (.61) 2.47 (.60) .99 2.41 (.64) .06 Physical health 19.04 (5.56) 18.78 (5.27) .53 18.32 (4.79) 1.38 Gender .98 (.15) .94 (.24) 1.78 .50 (.50) 10.56** Career centrality 3.12 (.72) 3.27 (.69) 2.07* 3.32 (.73) 2.73** Family centrality 4.73 (.40) 4.71 (.43) .53 4.53 (.54) 2.96** Voice in family labor decisions 3.97 (.80) 4.08 (.80) 1.38 4.02 (.82) Voice in paid labor 4.21 (.68) 4.39 (.62) 2.63** 4.31 (.64) 1.40 Satisfaction with current family labor division 3.43 (1.05) 3.61 (1.05) 1.73 3.66 (.99) 2.19 Satisfaction with current paid labor division 3.75 (.96) 3.92 (1.02) 1.72 3.87 (.95) 1.20 Control variables Family income 6.84 (1.48) 7.03 (1.49) 1.22 7.03 (1.49) 1.22 Length of marriage 8.26 (3.81) 8.19 (3.51) .20 8.19 (3.51) .20 Family responsibility 10.5 (3.68) 10.83 (4.24) .85 10.83 (4.24) .85 Total number children 1.75 (.72) 1.80 (.80) .69 1.80 (.80) .69 Number children under age 6 1.37 (.54) 1.43 (.57) 1.16 1.43 (.57) 1.16

PAGE 187

174 p < .05, ** p < .01 Table 29. Description of shape of fit line for significant moderator relationships Hyp Variables Low Med/Male High/Female 9b Paid needs-supplies x career centrality and marital satisfaction Linear increase U shaped U shaped 11d Paid needs-supplies x gender and physical health symptoms Linear increase Slight linear decrease 13b Paid needs-supplies x voice and marital satisfaction Linear increase Linear increase Flat 21a Paid demands-abilities fit x gender and marital satisfaction Sharp increase with slight decrease at high levels of fit ( > 80) Relatively flat; slight inverted U shape 22b Family demands-abilities fit x gender and depression Linear increase Increase with slight decrease at high levels of fit ( > 80) 22c Family demands-abilities fit x gender and physical health symptoms Ushaped Relatively flat; slight inverted U shape 23a Paid demands-abilities fit x satisfaction and marital satisfaction Relatively flat; slight inverted U shape Flat Slight decrease to medium levels, sharp increase past 23b Paid demands-abilities fit x satisfaction and depression Flat Flat Inverted U shape 23c Paid demands-abilities fit x satisfaction and physical health symptoms Flat Flat Flat 24b Family demands-abilities fit x satisfaction and depression Linear decrease Increase with slight decrease at high levels of fit (> 80) Increase with slight decrease at high levels of fit (> 70) The shape described is the pattern of the fit line as the level at which fit occurs increases. Number children over age 6 1.43 (.57) .40 (.73) 1.12 .40 (.73) 1.12 Demographics Age 35.22 (4.21) 35.02 (4.03) .49 35.58 (4.34) .84 Career Stage 2.38 (.62) 2.30 (.58) 1.29 2.27 (.57) 1.94 Education Level 6.26 (1.18) 6.49 (1.18) 1.99* 6.14 (1.38) .84 Weekly work hours 37.09 (13.47) 38.65 1.19 42.24 (12.58) 3.82

PAGE 188

175 Figure 1. Model of division of family labor hypotheses SUPPLIES Partner B’s post-child actual family labor contributions NEEDS Partner A’s pre-child desires for Partner B’s family labor contributions DEMANDS ABILITIES Partner B’s post-child actual family labor contributions Needs-supplies fit for Partner A Demands-abilities fit for Partner B Physical health symptoms Family Satisfaction Marital Satisfactio n Depression MODERATORS Gender Family centrality Voice in division of family labor decisions MODERATORS Gender Satisfaction with current division of family labor

PAGE 189

176 Figure 2. Model of division of paid labor hypotheses SUPPLIES Partner B’s post-child actual paid labor contributions NEEDS Partner A’s pre-child desires for Partner B’s paid labor contributions DEMANDS ABILITIES Partner B’s post-child actual paid labor contributions Needs-supplies fit for Partner A Demands-abilities fit for Partner B Physical health s y m p toms Career Satisfaction Marital Satisfact ion Depression MODERATORS Gender Satisfaction with current division of paid labor MODERATORS Gender Career centrality Voice in division of paid labor decisions

PAGE 190

F F igure 3. Pre d igure 4. Pre d 0 2.4 2. 4 d icted pattern d icted patter n -1.2 0 .6 3 -1.2 0.6 4 of relations h n relationshi p 3 -3 18 3 -3 -1.8 177 h ips in Hypot h p s in Hypothe s 1 8 -0.6 X -0.6 0.6 X h eses 1, 2, 3, 4 s es 5, 6, 7, 8, 0.6 1.8 X supplies/a b 0 0 1 1 2 2 3 3 4 4 5 1.8 3 X supplies/ab i 4 15, and 1 6 17, 18, 19, 2 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 3 Z well being b ilities 5 .5 .5 3 3 .5 4 4 .5 5 Z well-being i lities 6 0 Zwell being Z

PAGE 191

F N c e A a n F N c e A a n igure 5. Pre d N ote: The fig u e ntrality is lo w A ’s domain c e n d the Z axis igure 6. Pre d N ote: The fig u e ntrality is lo w A ’s domain c e n d the Z axis -1.2 0.6 2.4 -1.2 0.6 2.4 Y Low Low d icted pattern u re labeled L o w e r and the f e ntrality is hi g represents th d icted pattern u re labeled L o w e r and the f e ntrality is hi g represents th -3 -3 -1.8 -3 -3 -1.8 of relations h o w represent s f igure labele d g he r The X a e dependent w of relations h o w represent s f igure labele d g he r The X a e dependent w -0.6 0.6 1.8 3 X -0.6 0.6 1.8 3 X 178 h ips for parts s the predicte d High repres xis represent s w ellb eing v a h ips for part s s the predicte d High repres xis represent s w ellb eing v a 0 1 2 3 4 5 3 Z 2 0 1 2 3 4 5 3 Z 2 H a and b of H y d relationshi p ents the pred i s abilities, th e a riable. c and d of H y d relationshi p ents the pred i s abilities, th e a riable. -3 -1.2 0.6 2 .4 Y -3 -1.2 0.6 .4 Y High H i g h y potheses 9 a n p when partn e i cted relation s e Y axis repr e y potheses 9 a n p when partn e i cted relation s e Y axis repr e -3 -1.8 -0.6 06 X -3 -1.8 -0.6 n d 10 e r A’s domai n s hip when p a e sents deman d n d 10 e r A’s domai n s hip when p a e sents deman d 0 1 2 3 4 5 0 6 1.8 3 Z X 0.6 1.8 3X n a rtner d s, n a rtner d s, 4 5 Z 0 1 2 3 4 5 Z

PAGE 192

F pa N ( H ( H ( H p a d e d i r e igure 7. Pre d a rt a of Hyp o N ote: The fig u H ypotheses 1 H ypotheses 1 H ypotheses 2 a rtner A is a f e cisions is hi g i visions is hi g e presents nee d -1. 8 -0.6 0.6 1.8 3 Low d icted pattern o theses 23, a n u re labeled L o 1 and 12), w h 3 and 14), an 3 and 24). T h f emale (Hyp o g her (Hypot h g her (Hypoth e d s/demands, a -3 8 -3 -1.8 X of relations h n d 24 o w represent s h en partner A d when part n h e figure labe l o theses 11 an d h eses 13 and 1 e ses 23 and 2 a nd the Z axi -0.6 0.6 1.8 3 X 179 h ips for parts s the predicte A ’s voice in d i n er B’s satisf a l ed High rep r d 12), when p 1 4), and whe n 2 4). The X a x s represents t 0 1 2 3 4 5 3 Z 1 a and b of H y d relationshi p i vision of lab o a ction with c u r esents the pr e p artner A’s v o n partner B’s x is represents t he dependen t -3 -1.8 -0.6 0.6 1 .8 3 Y Hi g h y potheses 11, p when partn e o r decisions i u rrent labor d i e dicted relati o o ice in divisi o satisfaction w supplies/abi l t wellb eing v -3 -1.8 -0.6 12, 13, 14, a e r A is a mal e s lower i visions is lo w o nship when o n of labor w ith current l a l ities, the Y a x v ariable. 0.6 1.8 3X nd e w er a bor x is 0 1 2 3 4 5 Z

PAGE 193

F pa N ( H ( H ( H p a d e d i r e igure 8. Pre d a rts b and c o N ote: The fig u H ypotheses 1 H ypotheses 1 H ypotheses 2 a rtner A is a f e cisions is hi g i visions is hi g e presents nee d -1.2 0.6 2.4 Low d icted pattern o f Hypothese s u re labeled L o 1 and 12), w h 3 and 14), an 3 and 24). T h f emale (Hyp o g her (Hypot h g her (Hypoth e d s/demands, a -3 -3 -1.8 of relations h s 23, and 24 o w represent s h en partner A d when part n h e figure labe l o theses 11 an d h eses 13 and 1 e ses 23 and 2 a nd the Z axi -0.6 0.6 1.8X 180 h ips for parts s the predicte A ’s voice in d i n er B’s satisf a l ed High rep r d 12), when p 1 4), and whe n 2 4). The X ax i s represents t 0 1 2 3 4 5 3Z c and d of H y d relationshi p i vision of lab o a ction with c u r esents the pr e p artner A’s v o n partner B’s i s represen t s t he dependen t 3 -1.2 0.6 2.4 Y Hi g h y potheses 11, p when partn e o r decisions i u rrent labor d i e dicted relati o o ice in divisi o satisfaction w supplies/abil i t wellb eing v 3 -3 -1.8 06 12, 13, 14, a n e r A is a mal e s lower i visions is lo w o nship when o n of labor w ith current l a i ties, the Y a x v ariable. 0 6 0.6 1.8 3X n d e w er a bor x is 0 1 2 3 4 5 Z

PAGE 194

F N a n a x w F N a n a x w igure 9. Pre d N ote: The fig u n d the figure x is represent s w ellb eing va r igure 10. Pr e N ote: The fig u n d the figure x is represent s w ellb eing va r -1.8 -0.6 0.6 1.8 3 -1.2 0.6 2.4 Y Female Female d icted pattern u re labeled F e labeled Male s abilities, th e r iable. e dicted patter n u re labeled F e labeled Male s abilities, th e r iable. 3 -3 -1.8X -3 -3 -1.8 of relations h e male repres e represents t h e Y axis repre n of relation s e male repres e represents t h e Y axis repre -0.6 0.6 1.8 3 -0.6 0.6 1.8X 181 h ips for part a e nts the predi c h e predicted r e sents deman d s hips for part s e nts the predi c h e predicted r e sents deman d 0 1 2 3 4 5Z 1 0 1 2 3 4 5 3Z a Hypotheses c ted relation s e lationship w d s, and the Z a s c and d of H c ted relation s e lationship w d s, and the Z a -3 -1.8 -0.6 0.6 1 .8 3 Y -3 -1.2 0.6 2.4 Y Male Male 21 and 22 s hip when pa r w hen partner B a xis represen H ypotheses 2 1 s hip when pa r w hen partner B a xis represen -3 -1.8 -0.6 06 X -3 -1.8 -0.6 X r tner B is a f e B is a male. T ts the depen d 1 and 22 r tner B is a f e B is a male. T ts the depen d 0 1 2 3 4 5 0 6 1.8 3 Z X 0 1 2 3 4 5 0.6 1.8 3 X e male, T he X d ent e male, T he X d ent 5 Z 0 1 2 3 4 5 Z

PAGE 195

F A F A igure 11. R e s (Hypothesis 1 igure 12. R e s (Hypothesis 2 Nee 4 Ne e s ponse surfa c 1 ) s ponse surfa c 2 ) 5 -20 10 40 ds 5 -20 10 4 0 e ds c e for paid la b c e for f amily l 5 0 -50 30 5 0 -50 30 182 b or needs up p l abor needsu 30 -10 10 S 30 -10 10 p lies fit and c u pplies fit an d 1 1 2 2 3 3 4 4 5 30 50 S upplies 1 30 50Supplies c a r eer satisfa c d family s atis f .5 2 2 .5 3 3 .5 4 4 .5 5 Career Satisfaction 1 1.5 2 2.5 3 3.5 4 4.5 5 FamilySatisfaction c tion f or Par t f action f or P a Family Satisfaction t ner a rtner

PAGE 196

F A F P igure 13. R e s (Hypothesis 3 igure 14. R e s P artner A (Hy p 4 Ne e 4 Nee d s ponse surfa c 3 ) s ponse surfa c p othesis4) -20 10 4 0 e ds 5 -20 10 4 0 d s c e for paid la b c e for family l 50 -50 30 5 0 -50 30 183 b or needs up p l abor needsu 30 -10 10 30 -10 10 p lies fit and m u pplies fit an d 0 30 50Suppli e 10 30 50Supplies m arital satis fa d marital sati s 0 1 2 3 4 5 6 e s 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 MaritalSatisfaction fa ction for Pa r s faction for Marital Satisfaction Marital Satisfaction r tner

PAGE 197

F ( H F P F igure 15. R e s H ypothesis5) igure 16. R e s P artner A (Hy p 4 0 Ne e 4 Ne e s ponse surfa c s ponse surfa c p othesis7) -5 0 -20 10 0 e ds 5 -20 10 4 0 e ds c e for paid la b c e for paid la b 0 -50 30 5 0 -50 30 184 b or needs up p b or needs up p 30 -10 10 S 30 -10 10 p lies fit and d p lies fit and p 0 0 1 1 2 2 30 50 S upplies 30 50Supplies d epression fo r p hysical heal t 0 0 .5 1 1 .5 2 2 .5 3 3.5 4 4.5 5De p ression 11 16 21 26 31 36 41 46 PhysicalHealthSymptoms r Partner A t h symptoms f p Physical Health Symptoms f or

PAGE 198

F pa igure 17. R e s a id labor ne e 5 -20 10 40 Needs -2 0 10 40Needs Low High s ponse surfa c e ds upplies fi t 5 0 -50 -30 10 -50 0 -50 c e for moder a t and marital 10 10 30 50Supplies -30 -10 10Sup p 185 a ting role of c satisfaction f 1 2 3 4 5 6Marital Satisfaction 4 1 2 3 4 5 6 7 8 30 50 MaritalSatisfaction p lies c areer centra l f or Partner A -50 -20 10 4 0Needs Marital Satisfaction Med l ity between t h (Hypothesis 9 -50 -30 -10 h e relationsh 9 b) 10 30 50Supplies ip of 1 2 3 4 5 6 7 8 9Marital Satisfaction

PAGE 199

F l a igure 18. R e s a bor needs u p 2 10 40Ne e 2 10 40Needs Male Female s ponse surfa c p plies fit and p -50 2 0 -50 e ds -50 2 0 -50 c e for moder a p hysical hea l -30 -10 -30 -10 186 a ting role of g l th symptoms f 10 30 50 Supplies 10 30 50 Supplies g ender betwe e f or Partner A 11 21 31 41 51 61 50 Physical Health Symptoms 11 21 31 41 51 61 50 Physical Health Symptoms e n the relatio n A (Hypothesi s n ship of paid s 11d)

PAGE 200

F m P igure 19. R e sp m aking betwe e P artner A (Hy p 5 -20 10 40Need -20 10 40Nee d Low High sp onse surfac e e n the relatio n p othesis13b) 5 0 -50 -30s -50 -50 30 d s e for modera t n ship of paid -10 10 30Supplies 30 -10 10 30 Suppl 187 t ing role of v o labor needs 0 1 2 3 4 5 6 50 Marital Satisfaction 4 -1 0 1 2 3 4 5 30 50Marital Satisfactionies o ice in divisi o upplies fit an -50 -20 10 4 0Needs Med o n of paid la b d marital sat i -50 -30 10 b or decision i sfaction for 10 10 30 50Supplies 0 1 2 3 4 5 6 7Marital Satisfaction

PAGE 201

F fo igure 20. R e sp o r Partner B ( 1 40 Dem a sp onse surfac e ( Hypothesis2 0 -5 0 -20 0 a nds e for family l a 0 ) 0 -50 188 a bor deman ds -30 -10 ds -abilities fit 10 30 A biliti e and physical 0 5 10 15 20 25 30 50 PhilHlhS e shealth symp t Ph ys i ca l H ea l t h S ymptoms t oms

PAGE 202

F d e igure 21. R e sp e mands-abili t 1 40 De 10 40 Deman d Male Female sp onse surfac e t ies fit and m a -50 -20 1 0 mands -50 -20 d s e for modera t a rital satisfa c -50 -30 -50 -40 -30 20 189 t ing role of g e c tion for Par t -10 10 30 Ab 20 -10 0 10 20 Ab e nder betwee n t ner B (Hypo t 1 1.5 2 2. 5 3 3. 5 4 4. 5 5 5. 5 6 30 50 b ilities 20 30 40 50 b ilities n the relatio n t hesis21a) 5 5 5 5 Marital Satisfaction 1 2 3 4 5 6 7 Marital Satisfaction n ship of paid l l abor

PAGE 203

F l a igure 22. R e sp a bor demand s 1 40 Dema n 10 40 Deman d Female Male sp onse surfac e s -abilities fit a -50 -20 1 0 n ds -50 -20 d s e for modera t a nd depressi o -50 -30 -50 -30 190 t ing role of g e o n for Partne r -10 10 30 A b -10 10 A bi l e nder betwee n r B (Hypothe s 1 2 3 4 5 6 7 8 9 50ilities 1 1. 2 2 3 3 4 4 5 5 6 30 50 l ties n the relatio n s is22b) Depression 5 .5 .5 4 4 .5 5 5 .5 6 Depression n ship of famil y y

PAGE 204

F l a igure 23. R e sp a bor demand s 1 40 Dem 1 0 40 Dema n Female Male sp onse surfac e s -abilities fit a -50 -20 1 0 ands -50 -20 0 n ds e for modera t a nd physical -50 -30 -50 -30 191 t ing role of g e health sympt o -10 10 A -10 10 A b i e nder betwee n o ms for Part n 1 1 1 6 2 2 3 3 4 4 5 30 50 A bilities 2 0 5 30 50 i lities n the relatio n n er B (Hypot h 1 6 1 6 3 1 3 6 4 1 4 6 5 1Physical Health Symptoms 2 0 15 10 5 0 5 10 15 20 25 30Physical Health Symptoms n ship of famil y h esis22c) y

PAGE 205

F l a P igure 24. R e sp a bor between P artner B (Hy p 5 -20 10 40Demand s -20 10 40Demands Low High sp onse surfac e the relations p othesis23a) 5 0 -50 -30 -50 -50 30 e for modera t hip of paid l a -10 10 30 50 Abilities 30 -10 10 A bilit i 192 t ing role of s a a bor demand s 0 1 2 3 4 5 6 7 50 Marital Satisfaction 4 0 1 2 3 4 5 6 7 30 50Marital Satisfaction i es a tisfaction wi s -abilities fit a -50 -20 10 4 0Demands Med th current di v a nd marital s -50 -30 -10 v ision of pai d atisfaction f o 0 0 1 1 2 2 3 3 4 4 5 10 30 50Abilities d o r 0 0 .5 1 .5 2 2 .5 3 3 .5 4 4 .5 5 Marital Satisfaction

PAGE 206

F l a ( H igure 25. R e sp a bor between H ypothesis23 -5 0 -20 10 40Deman ds 5 -20 10 40Demand s High Low sp onse surfac e the relations b) 0 -50 -30 10 5 0 -50 -30 e for modera t hip of paid l a 10 10 30 50Abilities -10 10 30 50 A bilitie s 193 t ing role of s a a bor demand s 1 1.5 2 2.5 3 3.5 4 4.5 5Depression 4 D 1 1.5 2 2.5 3 3.5 4 4.5 5 50 Depression s a tisfaction wi s -abilities fit a -50 -20 10 4 0 D emands Med th current di v a nd depressi o -50 -30 -10 10 v ision of pai d o n f or Partne r 1 1 2 2 3 3 4 4 5 10 30 50Abilities d r B 5 .5 .5 4 4 .5 5 Depression

PAGE 207

F l a fo igure 26. R e sp a bor between o r Partner B ( 5 -20 10 40Demand s -20 10 40Demands Low High sp onse surfac e the relations ( Hypothesis2 3 5 0 -50 -30 50 -50 30 e for modera t hip of paid l a 3 c) -10 10 30 50 Abilities 30 -10 10 30 A biliti e 194 t ing role of s a a bor demand s 11 16 21 26 31 36 41 46 50 Physical Health Symptoms 2 7 12 17 22 27 32 37 42 47 30 50Physical Health Symtptoms e s a tisfaction wi s -abilities fit a -50 -20 10 40Demand s Med th current di v a nd physical -50 -30 -10 v ision of pai d health sympt o 10 30 50Abilities d o ms 11 16 21 26 31 36 41 46Physical Health Symptoms

PAGE 208

F l a ( H igure 27. R e sp a bor between H ypothesis24 -5 -20 10 40Deman ds -20 10 40Demands Low High sp onse surfac e the relations b) 0 -50 -30 10 -50 -50 30 e for modera t hip of family 10 10 30 50Abilities 30 -10 10 30 A bili t 195 t ing role of s a labor deman 1 2 3 4 5Depression 1 4 0 D -1 0 1 2 3 4 5 30 50Depression t ies a tisfaction wi ds-abilities f i -50 -20 1 0 0 emands Med th current di v i t and depres s -50 -30 -10 v ision of f am i s ion f or Part n 1 1 2 2 3 3 4 4 5 10 30 50Abilities i ly n er B .5 2 2 .5 3 3 .5 4 4 .5 5 Depreression

PAGE 209

F sa F sa igure 28 (Ro t a tisfaction fo r igure 29. (R o a tisfaction fo r Need s t ated Figure 1 r Partner A ( H o tated Figure r Partne r A ( H 40 -50 -20 10 40 -50 s 1 1). R espons e H ypothesis1) 13). R espon s H ypothesis3) 10 Suppli e -30 -10Sup p 196 e surface for p s e surface fo r 20 e s 10 30 p lies p aid labor n e r paid labor n 50 1 1.5 2 2.5 3 3.5 4 4.5 5 5 0 3 0 1 0 1 0 3 0 5 0NeCareer Satisfaction 0 1 2 3 4 5 6 50 e eds upplies f eeds upplies f eds Marital Satisfaction f it and caree r f it and marit a r a l

PAGE 210

F fo F h e igure 30. (R o o r Partner A ( igure 31 (Ro t e alth sympto m Needs Needs o tated Figure ( Hypothesis5 ) t ated Figure 1 m s for Partn e 50 -20 10 40 -50 -50 -20 10 40 -50 15). R espon s ) 1 6). R espons e e r A (Hypoth e -30 -10Supp -30 -10Suppl 197 s e surface fo r e surface for p e sis7) 10 30lies 10 30 50 ies r paid labor n p aid labor n e 0 0.5 1 1.5 2 2. 5 3 3. 5 4 4. 5 5 50 11 16 21 26 31 36 41 46 50 eeds upplies f e eds upplies f 5 5 5 Depression Physical Health Symptoms f it and depre s f it and physic s sion al

PAGE 211

F sa F P igure 32 (Ro t a tisfaction fo r igure 33. R e s P artner A Need s -5 0 -2 0 1 4 Need s t ated Figure 1 r Partner A ( H s ponse surfa c 50 -20 10 40 -50 s 0 0 1 0 4 0 -50 30 s 1 4). R espons e H ypothesis4) c e for family l -30 -10Su p 30 -10Suppl 198 e surface for f l abor needsu 10 30 p plies 10 30ies f amily labor u pplies fit an d 1 1.5 2 2.5 3 3.5 4 4.5 5 5. 5 50 0 5 10 15 20 25 30 3 5 50 needs upplie s d physical he a 5 Marital Satisfaction 5 Physical Health Symptoms s fit and mar i a lth symptom s i tal s for

PAGE 212

199 Appendices

PAGE 213

200 Appendix A Pre-child desires for division of paid labor Before having children, many couples think and discuss what their life will be like 'postchildren.' For the next set of questions, we are in terested in knowing wh at your desires and expectations were BEFORE you had children wi th regard to what your life would be like AFTER you did have children. 1. Picture the total number of combined hour s that you and your spouse spend in paid employment as a pie char t that sums to 100%. Before you had children and were thinking into the future... ...what proportion of that pie chart did you WANT to be YOUR work hours once children were born? ..what proportion of that pie chart did you WANT to be YOUR SPOUSE'S work hours once children were born? These percentages should add up to 100%. 2. Picture the total income that you and your spouse earn from paid employment as a pie chart that sums to 100%.

PAGE 214

201 Before you had children and were thinking into the future... ...what proportion of that pie chart did you WANT to be YOUR income once children were born? ..what proportion of that pie chart did you WANT to be YOUR SPOUSE'S income once children were born? These numbers should add up to 100% 3. Career favoring decisions are decisions th at require one spouse's career to be given priority over the other spouse' s career. Picture the total num ber of career prioritization decisions in your marriage as a pie chart that sums to 100%. Before you had children and were thinking into the future... ...what proportion of that pie chart did you WANT to be decisions that favored YOUR career once children were born? ..what proportion of that pi e chart did you WANT to be decisions that favored YOUR SPOUSE'S career once children were born? These numbers should add up to 100% Pre-child desires for division of family labor 1. Childcare related tasks are activities that involve caring for and raising children. Some examples include supervising, bathing, puni shing, playing with children, and taking children to appointments or play dates. Think about the total amount of childcare tasks that must be performed in your home. Before you had children and were thinking into the future... Family Labor Desires and Expectations Before Chdren ...what percentage of these ta sks did YOU WANT to perform? ..what percentage of these tasks did you WANT YOUR SPOUSE to perform? These numbers should add up to 100%

PAGE 215

202 2. Maintaining a household requ ires several household tasks to be completed, such as meal preparation, cooking, housecleaning, yard work, shopping for groceries and household goods, washing dishes or cleaning up after meals, doing laundry, paying bills, and taking out the trash. Think about the total amount of household task s (not including childcare) that must be performed in your home. Before you had child ren and were thinking into the future... ...what percentage of these tasks did YOU W ANT to perform once children were born? ..what percentage of these tasks did you WANT YOUR SPOUSE to perform once children were born? These numbers should add up to 100% 3. Emotion work is work that enhances othe r family member’s well-being and creates a positive emotional atmosphere. Some examples include offering encouragement and advice, paying attention to important events in other family member’s lives, giving compliments, and expressing concern for family member’s well-being. Think about the total amount of emotion work that is performed in your home. Before you had children and were thinking into the future... ...what percentage of this work did YOU WA NT to perform once children were born? ...what percentage of this work did you WANT YOUR SPOUSE to perform once children were born? These numbers should add up to 100%

PAGE 216

203 Appendix B Post-child actual division of paid labor For the next set of questions, pl ease think about the present time. 1. Again, picture the total WORK HOURS you an d your spouse spend in paid labor as a pie chart. Currently, what proportion of this pie is made up of your work hours, and what proportion is made of your spouse's work hours? __self __spouse 2. Again, picture the total INCOME you and your spouse earn from paid labor as a pie chart. Currently, what proportion of this pie is made up of your income, and what proportion is made of your spouse's income? __self __spouse 3. Again, picture the total CAREER FAVORING DECISIONS that have been made throughout your marriage as a pie chart. Currently, what proportion of this pie is made up of decisions that have favored your career, and what proportion is made of decisions that have favored your spouse's career? __self __spouse Paid Labor Current Post-child actual division of family labor For the next set of questions, pl ease think about the present time. 1. Currently, what percentage of CHIL DCARE tasks do you and your spouse each perform? __self __spouse 2. Currently, what percentage of HOUS EHOLD tasks do you and your spouse each perform? __self __spouse 3. Currently, what percentage of EM OTION WORK do you and your spouse each perform? F __self __spouse

PAGE 217

204 Appendix C Career Satisfaction (G reenhaus et al., 1990) 1. I am satisfied with the succe ss I have achieved in my career. 2. I am satisfied with the progress I have ma de toward meeting my overall career goals. 3. I am satisfied with the progress I have made toward meeting my goals for income. 4. I am satisfied with the progress I have made toward meeting my goals for advancement. 5. I am satisfied with the progress I have made toward meeting my goals for the development of new skills.

PAGE 218

205 Appendix D Family Satisfaction (adapted from Cammann et al., (1979) 1. I am happy with my progress toward the goals I have for my family. 2. I am satisfied with my present family situation. 3. Overall, I am pleased with th e state of my family life. 4. In general, I like my family life.

PAGE 219

206 Appendix E Marital Satisfaction (Norton, 1983) Think about your marriage and your spouse when responding to the following questions. (Response scale: strongly disagree, disagree, neither disagree nor agree, agree, strongly agree) 1. My relationship with my pa rtner is very stable. 2. We have a good marriage. 3. My relationship with my partner makes me happy. 4. I really feel like part of a team with my partner. 5. Our marriage is strong.

PAGE 220

207 Appendix F Depression (Quinn & Shepard, 1974) Considering the past 3 months, indicate your agreement with the following statements. (Response scale: strongly disagree, disagree, neither disagree nor agree, agree, strongly agree) Health I feel downhearted and blue. I get tired for no reason. I find myself restless and can’t keep still. My mind is as clear as it used to be. I find it easy to do the things I used to do. I feel hopeful about the future. I find it easy to make decisions. I am more irritable than usual. I still enjoy the things I used to. I feel that I am useful and needed.

PAGE 221

208 Appendix G Physical Health Symptoms (adapted from National Study of Daily Experiences) Over the past 3 months, how often have you experienced the following symptoms? (never, less than once per month, 2-3 times a month, 1-2 times a week, 3-4 times a week, 5 or more times a week) 1. Upset stomach or nausea 2. Backache 3. Headache 4. Acid indigestion or heartburn 5. Diarrhea 6. Stomach cramps (Not menstrual) 7. Loss of appetite 8. Shortness of breath/difficulty breathing 9. Dizziness 10. Chest pain 11. Flu or cold symptoms (fev er, sore throat, chills)

PAGE 222

209 Appendix H Career centrality (Lobel & St Clair, 1992) 1. A major source of satisfacti on in my life is my career. 2. Most of the important things that happen to me involve my career. 3. I am very much involved personally in my career. 4. Most of my interests ar e centered around my career Family centrality (Eddleston et al., 2006) 1. A major source of satisfaction in my life is my family. 2. Most of the important things that happen to me involve my family. 3. I am very much involved personally in my family. 4. Most of my interests ar e centered around my family.

PAGE 223

210 Appendix I Voice in division of paid labor decisions 1. In general, I have a lot of opportunity to present my views about decisions that affect the division of paid labor in my family.1 2. My views are considered and taken into account in decision-m aking related to the division of paid labor among my spouse and I.1 3. What I want is considered when my part ner and I arrive at d ecisions concerning the division of paid labor.1 4. I feel my voice is heard in family d ecisions about the division of paid labor.2 5. I am granted a significant de gree of influence in decisions that affect the division of paid labor.3 6. My spouse usually asks for my opinions and thoughts about decisions affecting our division of paid labor.3 7. In my marriage, I have a real say in the im portant decisions that im pact the division of paid labor. 4 Voice in division of family labor decisions 1. In general, I have a lot of opportunity to present my views about decisions that affect the division of family labor.1 2. My views are considered and taken into account in decision-m aking related to the division of family labor among my spouse and I.1 3. What I want is considered when my part ner and I arrive at d ecisions concerning the division of family labor.1 4. I feel my voice is heard in decision s about the division of family labor.2 5. I am granted a significant de gree of influence in decisions that affect the division of family labor.3 6. My spouse usually asks for my opinions and thoughts about decisions affecting our division of family labor.3 7. In my marriage, I have a real say in the im portant decisions that im pact the division of family labor. 4 Previous research that the new scales were based on (* indicates item that was adapted): 1Brockner et al (2001) *1. I had a lot of opportunity to present my views about how this dispute should be resolved *2. My views were considered and taken into account *3. What I wanted was considered in arriving at a solution. 2Denton & Zeytinoglu (1993 ) 1. I have been a member of important decision making committees in the Department. *2. I feel my voice is heard in Department and Committee meetings of the Department. 3. I have been a member of important decision making committees in the Faculty.

PAGE 224

211 4. I feel my voice is heard in F aculty level committee meetings. 5. I have been a member of important decision making committees at the University level. 6. I feel my voice is heard in Univ ersity level committee meetings. 3Steel & Mento (1987) 1. Within my work group the people most aff ected by decisions frequently participation in making the decisions. 2. In my work group there is a great deal of opportunity to be involved in resolving problems which affect the group. 3. I am allowed to participate in decisions regarding my job. *4. I am allowed a significant degree of infl uence in decisions regarding my work. *5. My supervisor usually asks for my opini ons and thoughts in decisions affecting my work. 4Campion, Medsker, & Higgs (1993) *1. As a member of a team, I have a real sa y in how the team carries out its work. 2. Most members of my team get a chan ce to participate in decision making. 3. My team is designed to let everyo ne participate in decision making.

PAGE 225

212 Appendix J Satisfaction with current division of family labor 1. I am satisfied with the way that my partner and I divide family labor. 2. I am pleased with the amount of family la bor that I perform relative to my spouse. 3. I am unhappy with the current division of family labor in my home. Satisfaction with current division of paid labor 1. I am satisfied with the way that my partner and I divi de paid labor. 2. I am pleased with the amount of paid labo r that I perform relative to my spouse. 3. I am unhappy with the current divisi on of paid labor in my home. reverse scored item.

PAGE 226

213 Appendix K Family Responsibility 1. Please indicate the month and year your children were born (e.g., February 2005). Leave blanks empty that are not applicable. First Born _______ Second Born _______ Third Born ________ Fourth Born_________ Fifth Born ________ Sixth Born _______ 2. Approximately what percent of the time does your child live in your home? (Response options are 0%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%). First Born _______ Second Born _______ Third Born ________ Fourth Born_________ Fifth Born ________ Sixth Born _______ Item Weights Living with Individual No t living with Individual Each child ages < 1 year 7.0 4.0 Each child aged 1 -2 years 6.5 3.5 Each child aged 3 -5 years 6.0 3.5 Each child aged 6 – 9 years 5.0 3.0 Each child aged 10 – 14 years 5.0 3.0 Each child aged 15 – 18 years 4.5 2.5 Each child aged 18 and older 3.0 1.0

PAGE 227

214 Appendix L Age What is your current age in years? Ethnicity Please select the answer that best describe s your ethnicity (select all that apply). White/Caucasian Black/African-American Asian Hispanic/Latino Native American Other, please describe: ____________ Job title What is your curren t job title? _________ Education level What is the highest grade or year of school you completed? Grades 9 through 11 (some high school) Grade 12 or GED (high school graduate) Some college Associate/two year degree Bachelor's degree Some graduate work Master's degree Advanced degree (e.g., PhD, MD) Career Stage Please select the option below that most accurately describes your current career situation: -I have recently started my car eer and am just beginning to explore my career options. I am involved in self-examination and trying to discover the kind of work and career that will best suit me. -I am mostly concerned with securing my place in my organization/vocation, demonstrating outstanding performance, esta blishing relationships with others, and advancing to new levels of responsibility. I feel relatively stab le in my career. -I am focused on preserving my career achieveme nts already attained and my self image. I have a strong personal iden tification with my career and organization/vocation. -I am approaching retirement and beginning to detach from my job, organization, and occupation.

PAGE 228

About the Author Kristen M. Shockley received a Bachelor’s Degree in Psychology from the University of Georgia in 2004 with a minor in French. She entered the Industr ial and Organizational Psychology Ph.D. program at the University of South Florida in 2005 and received her Master’s Degree in Industrial and Or ganizational Psychology in 2007 under the advisement of Dr. Tammy D. Allen. While enro lled at the University of South Florida, Ms. Shockley conducted research in various topics, including managing work and family roles, dual-career couples, mentoring, and career success. She has co-authored five published articles in the Journal of Vocational Behavior as well as three chapters in edited books. She has presented at severa l professional conferences, including the Society for Industrial and Organizational Psychology (SIOP) and the Society for Occupational Health Psychology. Ms. Shockley also received a p ilot grant through the National Institute for Occupational Safety and Health. She has been formally recognized for her contributions in both research (Mary L. Tenopyr award granted through SIOP; Top Poster award granted at the 2009 SIOP conference) and teaching (University of South Florida Provost’s commendation for outst anding teaching by a graduate student).


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004543
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Shockley, Kristen.
0 245
You can't always get what you want, but does it matter? the relationship between pre-child preferences and post-child actual labor division fit and well-being
h [electronic resource] /
by Kristen Shockley.
260
[Tampa, Fla] :
b University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Dissertation (PHD)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: Significant shifts in social ideology and legislation have brought about considerable changes in work and family dynamics in the Western world, and the male as breadwinner-wife as homemaker model is no longer the norm. However, despite increasingly gender egalitarian ideals, the division of labor among dual-earner couples tends to adopt a "neo traditional" once children are born, where women devote more time to family labor and men spend more time in paid employment Although asymmetrical divisions of labor have clear workplace and societal consequences in terms of women's earnings, organizational advancement, and inequality, the effects on individual well-being are not well understood. The purpose of the present study was to apply the theoretical lens of person-environment fit to examine how misfit between dual-earner couples' pre-child division of labor preferences and post-child actual divisions of labor relate to affective (career, marital, and family satisfaction) and health-related (depression and physical health symptoms) well-being. Additionally, several conditions were posited to temper the strengths of these relationships (domain centrality, gender, voice in division of labor decision making, and satisfaction with the current division of labor). Participants were 126 dual-earner couples with small children, and hypotheses were testing using polynomial regression analyses. The results suggested that congruence between an individual's own pre-child desires for the division of paid labor and the actual post-child division of paid labor relates to his/her own career and marital satisfaction, depression, and physical health symptoms. Congruence in the family domain is also important, as desire-division of family labor fit related to affective sentiments toward family and one's spouse. With the exception of career satisfaction, these relationships were curvilinear, such that deviations in either direction from perfect fit related to poorer well-being. On the other hand, there was little evidence for spousal effects, as dual-earner well-being did not relate the congruence between division of labor abilities and spousal demands. Finally, evidence of moderation was only found in a few cases, and none were consistent with prediction, highlighting the need for future research on the contextual conditions of P-E fit in the dual-earner context.
590
Advisor: Tammy Allen, Ph.D.
653
Division of household labor
Paid labor
Person-environment fit
Dual-earner couples
Dual-career couples
Work-family management
Career prioritization
690
Dissertations, Academic
z USF
x Psychology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4543