MALLEABILITY IN GENDER ROLE ATTITTUDES DURING THE TRANSITION TO DIVORCE AND REMARRIAGE: A LONGITUDINAL STUDY OF MARITAL EXPERIENCES Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. ____________________________________________ Mallory J. Lucier Certificate of Approval: _________________________________ Jennifer Kerpelman Professor Human Development & Family Studies _________________________________ Francesca Adler-Baeder, Chair Associate Professor Human Development & Family Studies _________________________________ Thomas Smith Associate Professor Human Development & Family Studies _________________________________ George T. Flowers Dean Graduate School MALLEABILITY IN GENDER ROLE ATTITTUDES DURING THE TRANSITION TO DIVORCE AND REMARRIAGE: A LONGITUDINAL STUDY OF MARITAL EXPERIENCES Mallory J. Lucier A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama August 9, 2008 iii MALLEABILITY IN GENDER ROLE ATTITTUDES DURING THE TRANSITION TO DIVORCE AND REMARRIAGE: A LONGITUDINAL STUDY OF MARITAL EXPERIENCES Mallory J. Lucier Permission is granted to Auburn University to make copies of this thesis at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. ____________________________________ Signature of Author ____________________________________ Date of Graduation iv VITA Mallory J. Lucier, daughter of Geoffrey Lucier and Laura Bennett, was born August 29, 1984 in Columbus, Georgia. She graduated Columbus High School Liberal Arts Magnets in 2002. She attended The University of Georgia and graduated summa cum laude with a Bachelor of Science in Psychology and a Bachelor of Arts in Speech Communication in December 2005. She entered Graduate School at Auburn University in August 2006. v THESIS ABSTRACT MALLEABILITY IN GENDER ROLE ATTITTUDES DURING THE TRANSITION TO DIVORCE AND REMARRIAGE: A LONGITUDINAL STUDY OF MARITAL EXPERIENCES Mallory J. Lucier Master of Science, August, 9, 2008 (B.S., The University of Georgia, 2005) (B.A., The University of Georgia, 2005) 81 Typed Pages Directed by Francesca Adler-Baeder Gender roles affect many aspects of individual development and daily life including career goals, one?s decision to parent, familial contact, and relationships, yet most of the research on gender roles and related gender role attitudes across time occurs at the sociological level. Gender role attitudes have been shown to influence marital satisfaction and marital discord and is theorized to be malleable, yet little research exists related to how gender role attitudes change during the course of individual development particularly in the context of divorce and remarriage. The primary purpose of the present study was to examine gender role attitudes longitudinally among groups of individuals with differing marital histories. vi The data for this study come from the Inter-University Consortium for Political and Social Research (ICPSR). Specifically, the data set ? Marital Instability over the Life Course ? is a nationally representative longitudinal study of marital instability collected between 1980 and 2000. The subsample utilized for the current study consists of 590 individuals all of whom were in first marital relationships at the beginning of data collection. Results indicate a significant decrease in traditional gender role attitudes for the group across the 20 year period. Results also show no significant between group differences based on changes in gender role attitudes over time based on age, yet significant between group differences in reported gender role attitudes over time based on sex were found, wherein women reported greater increase in egalitarian attitudes across time compared to men. Results reveal significant time X marital status group interaction effects such that continually married individuals developed more egalitarian gender role attitudes gradually over time; individuals who divorced and remained divorced showed a sharper decrease in traditional gender role attitudes across time; while, those who divorced and remarried alternated significant decreases and significant increase in more traditional gender role attitudes across time. This suggests that while change over time occurred for the entire group, individuals in each marital status group experienced distinct patterns of change in gender role attitudes. These differences remained significant even when controlling for age, sex, and ethnicity. Overall, results indicate that gender role attitudes appear to be malleable not only at the broad sociological level, but also over the course of individual life course events vii and that experiences in marital relationships affect the pattern of change. The implications of these findings and directions for future research are discussed. viii ACKNOWLEDGEMENTS I would like to thank my parents, Geoffrey Lucier and Laura Bennett, for two decades worth of support and encouragement through my academic journey, but primarily for continuing to remind me that the most important things in life cannot found in a textbook. I want to thank my sister, Jessica Lucier, for knowing how to make me laugh at the most stressful moments and for never being too far away. I would like to thank Dr. Jennifer Kerpelman and Dr. Tom Smith for serving in an advisory capacity and graciously offering their time and expertise. Finally, a special thanks to Dr. Francesca Adler-Baeder, my major professor, who always seems to be able to see the light at the end of the tunnel; her spirit of calmness and endurance is a continual encouragement to me personally and professionally. ix Style manual or journal used American Psychological Association, Fifth Edition Computer software used Microsoft Word 2007 for Windows, SPSS 16.0 for Windows x TABLE OF CONTENTS LIST OF TABLES ............................................................................................................ xi LIST OF FIGURES ......................................................................................................... xii INTRODUCTION ............................................................................................................. 1? REVIEW OF THE LITERATURE ................................................................................... 7? Theory ............................................................................................................................ 8? The Study of Gender Role Attitude Changes from a Sociological Perspective .......... 14? Studying the Malleability of Gender Role Attitudes over the Life Course from a Family Science Perspective ......................................................................................... 15? Transition to Divorce, Singlehood, and Remarriage ................................................... 22? METHODS ...................................................................................................................... 29? Participants ................................................................................................................... 29? Measures ...................................................................................................................... 32? Plan of Analysis ........................................................................................................... 34? RESULTS ........................................................................................................................ 37? DISCUSSION .................................................................................................................. 46? REFERENCES ................................................................................................................ 58? APPENDIX A .................................................................................................................. 66 APPENDIX B .................................................................................................................. 68 xi LIST OF TABLES Table 1. Respondent Attrition 30 Table 2. Demographic Information for all Participants and Relevant Subsamples 32 Table 3. Means, Standard Deviations, and Reliabilities for Survey Scales (n = 590) 37 Table 4. Between-Groups Analysis of Variance of Gender Role Attitudes 44 xii LIST OF FIGURES Figure 1. RMANOVA of Gender Role Attitudes 42 Figure 2. RMANOVA of Gender Role Attitudes Controlling for Age, Sex, and Race 43 Figure 3. RMANOVA of Gender Role Attitudes 44 Figure 4. RMANOVA of Gender Role Attitudes Controlling for Age, Sex, and Race 45 1 I. INTRODUCTION An abundance of literature exists dedicated to understanding how gender roles are socially constructed beginning in childhood, and how they impact interpersonal relationships throughout the course of life. This body of literature explores how personal conceptualizations of gender roles are influenced by such factors as one?s family of origin, race, sex, and religion. Gender roles, in turn, affect many aspects of individual development and daily life including career goals, one?s decision to parent, familial contact, and relationships (Fan & Marini, 2000; Softas-Nall & Sukhodolsky, 2006; Stone & Lovejoy, 2004). Within the category of relationships, gender roles influence peer, occupational, and even parent/child relationships, but it can be argued that gender based roles are most clearly evident and influential within marital relationships (Sullivan & Lewis, 2006). Specifically, gender roles are multidimensional, multifaceted expectations for men and women based on societal norms. Gender roles encompass patterns of behavior, personality attributes, and attitudes about social norms and interpersonal relationships (American Psychiatric Association [DSM-IV-TR], 2000). Research demonstrates that often there is a complex and variable relationship between attitudes and behaviors (Seitz, Lord, & Taylor, 2007; Wallace, Paulson, Lord, & Bond, 2005). This study specifically 2 addressed gender role attitudes and perceptions of appropriate roles as this appears to be a distinct area of focus in the gender role literature. The majority of this research addressing gender roles in marital relationships utilizes a sociological framework in which the researchers gauge change in gender role attitudes across age-related cohorts. Studies have found that gender role attitudes are becoming more egalitarian across younger generational cohorts (Brooks & Bolzendahl, 2004; Rogers & Amato, 2000). These younger cohorts also report increased equality in decision making, household responsibilities, and labor force participation when compared to generations past, but on the whole, these more egalitarian cohorts also experience greater levels of marital distress. Accordingly, there appear to be macro- level changes in gender role attitudes that have an influence on marital relationships. Macro-level influences refer to sociological variables including historical and cultural events and changes. Because most studies compare cohort differences, there seems to be an implicit assumption that gender role attitudes are either static belief systems at the individual level or perhaps indicative of a particular group. There is a dearth of information on the malleability of gender roles at the micro- or individual level of development. That is, do gender role attitudes change significantly during adult development based on micro-level processes? Micro-level influences include interpersonal relationships, intrapersonal thought processes about self, and life course experiences. Although, gender role attitudes appear malleable during childhood years (Ruble, Taylor, Cyphers, Greulich, Lurye, & Shrout, 2007) once formed, little research investigates how development 3 throughout adulthood and life course events affect or alter one?s gender role attitudes later in life. With a life course developmental perspective, one would assume that significant life events are opportunities for development, growth, and change. Therefore, it would follow that gender identities and related attitudes may vary not only according to one?s age-based cohort, but also based on interpersonal experiences. Theoretically, role acquisition, in general, and related beliefs about those roles continue to develop in the context of personal and relational development (Thornton & Nardi, 1975). Some evidence has been found to support the notion that gender role attitude malleability occurs at the individual or micro-level of development in young adult, collegial, and parent/child samples (e.g. Blee & Tickamyer, 1995; Bryant, 2003; Fan & Marini, 2000). These studies are among the first to document that existing gender role attitudes may be modified, but they are also limited in terms of the use of retrospective data accounts, and/or a limited timeframe to observe changes in gender role attitudes. The purpose of this study is to substantiate and expand these findings by using a prospective approach with data gathered from a 20 year longitudinal study to explore marital contexts in which gender role attitude change may occur. Marital formation and dissolution are significant life course events that often influence an individual?s economic standing, parenting decisions, behavioral roles, relational expectations, level of stress, mental health, and presumably one?s concept of self. It would follow ? yet still remains to be determined ? that the experiences of 4 divorce and remarriage may also affect gender role attitudes as individuals experience new circumstances and novel role sets. The literature on gender role attitudes suggests that malleability occurs in relation to specific events or experiences the individual undergoes (Fan & Marini, 2000). Divorce and remarriage serve as relational and life transitions that create new interpersonal interactions and intrapersonal experiences. Presumably this change in formal marital roles coupled with novel relational experiences would set the stage for individuals to adapt their gender role beliefs according to their new positions within society and conceptualization of self. The literature examining remarriages and stepfamily relationships has developed over the past 50 years as the United States has experienced considerable changes in the composition of families, living arrangements, and family structure (Fields, 2003). Scholars introduced and continue to address the practical and theoretical implications of studying remarital dyads and the factors that make remarriages unique (Bernard, 1956). Unlike first marriages, one or both individuals have already experienced the creation and termination of a marriage. This sets the stage for growth and development as one enters a new relationship where new patterns may be created that prove to be more functional and/or desirable. Most of the research addressing gender roles in remarital relationships posits that remarried couples will report more egalitarian attitudes than couples in first marital unions and will prefer a nontraditional division of labor and power (Clarke, 2005; Smith, Goslen, Byrd, and Reece, 1991), but most of this research consists of retrospective 5 accounts by individuals comparing perceived gender role attitudes in their first marriages and second marriages. For example, Smith et al. (1991) studied a convenience sample of 64 divorced and remarried individuals that reported significant changes in gender role orientation scores ? reflecting more egalitarian beliefs ? for both men and women when cross-sectionally assessing the participants? perceptions of their first marriages and second marriages. The authors speculated that the reasons for these changes differ between the sexes whereas women wanted to avoid the self-sacrificing behaviors that they exhibited in traditional first marriages and men realized that companionship would not survive without these shared roles. The handful of studies questioning change in gender role attitudes due to marital status transitions have provided some indications that gender role attitudes are malleable over the course of divorce and remarriage. The importance of the current study is based on, first, the considerable number of remarriages that occur in the United States each year, and second, evidence that gender role attitudes influence many aspects of daily life. Within the context of marriage, gender role attitudes have been found to impact relationship satisfaction and distress. There are even implications for marital instability when examining partners? gender role attitudes. Thus, there is a need to more fully understand if and how gender role attitudes change during periods of marital transition. To date, no published studies exist that prospectively examine gender role attitudes in the context of divorce and remarriage over a longitudinal timeframe. As this is a new area of inquiry, it is also important to begin to examine how characteristics of the individual may affect gender role attitude change, because it would 6 follow that differences in gender role malleability cannot be predicted in universal terms. Factors that have previously been found to influence and predict gender role attitude orientations will be examined including participants? age-based cohort and sex. Accordingly, the primary purpose of this study was to assess changes in gender role attitudes in the context of marital status transitions. Specifically, gender role attitudes over the course of marital experiences were examined among groups of individuals with differing marital experiences. This information allowed the researcher to compare profiles of continuously married individuals with those who experienced divorce and remarriage. The secondary purpose of this study was to explore how individual characteristics including age and sex affected potential change in gender role attitude scores across time. Developing a greater understanding of lifecourse factors related to changes in gender role attitudes across time opens the door to further explorations of functionability and dynamics in different marital contexts, particularly remarital relationships, and the contributions of gender role attitudes to remarital quality and stability. 7 II. REVIEW OF THE LITERATURE In this chapter, gender role definitions and the theoretical approaches to the study of gender roles and, specifically, gender role attitudes will be introduced. Second, a synthesis of the literature on gender role attitudes will be presented to include an examination of empirical studies of gender role attitudes at the various stages of marital status transition from first marriage relationships into the transitions of divorce/singlehood and remarriage. Gender roles are multidimensional, multifaceted expectations based on societal norms prescribed for men and women. These roles consist of attitudes, patterns of behavior, and personality attributes based on one?s perception of gendered expectations via societal norms and interpersonal relationships (American Psychiatric Association [DSM-IV-TR], 2000). Gender role attitudes are conceptualized in the literature on a continuum ranging from traditional to egalitarian. Individuals reporting more traditional attitudes frequently respond to others based on stereotypical characteristics connected to their sex; conversely, individuals with egalitarian attitudes tend to respond to others independent of their sex (King, Beere, King, & Beere, 1981). Those who maintain traditional gender role attitudes tend to view marital roles as specialized and distinct often based on a husband-breadwinner/wife-mother or homemaker system with an implied differential power structure. Individuals who identify with nontraditional or 8 egalitarian gender role attitudes, however, tend to emphasize shared capacities for economic output, familial nurturance, and power relations within the relationship (King et al., 1981). Theory Symbolic Interactionism: Role Acquisition and Development Roles can be described as transactional or negotiated concepts between individuals and their social structure; a role is a ?set of behaviors which are based on expectations, obligations, and prescriptions?expressed in a relationship? based on societal norms (Kelley, 1995, p. 39). Gender roles are prescriptions of behavior based on gendered expectations. While sex is a biological attribute, gender is a social construction created by scripts in our society. Roles can essentially be thought of as the link between society and self. More specifically, symbolic interactionism conceptualizes roles as shared norms between individuals that create a system of meaning within society. These shared meanings allow people to predict the ways in which others will behave in certain settings such as on a first date or in marriage. At a more proximal level, this theory emphasizes the ?looking glass self? (Cooley, 1902), whereas one?s sense of self is not innate or revealed through introspection. Rather, we develop our sense of self through interactions with others and by taking the perspective of others within our primary group to imagine how we look from their perception. As such, roles can only be conceptualized in relation to complementary and/or counter roles which change during the life course (LaRossa & Reitzes, 1993). 9 Symbolic interactionism assumes that meaning is generated and altered through the process of interacting with others (LaRossa & Reitzes, 1993); therefore, scholarly opinion would expect that each marital relationship conceptualizes roles differently as partners may have preexisting definitions/expectations of roles based on social prescriptions that are then edited into personalized roles and role sets through dyadic interactions. Role acquisition or role making is a developmental process made up of four stages ? the anticipatory stage, formal stage, informal stage, and personal stage (LaRossa & Reitzes, 1993). The anticipatory stage is a knowledge gathering phase in which the individual learns about role expectations from family groups, media sources, and reciprocal-role others before entering that role (Thornton & Nardi, 1975). For example, information is given through various outlets to ?educate? individuals on socially acceptable gender roles in marriage (i.e. bridal magazines, advice from parents, messages from the media, religious expectations, etc.). At this stage, role conception is thought to be presented as stereotyped in ideal expectations and often lacks depth. The formal stage of role creation is hypothesized to include formalized or normative expectations and behaviors prescribed by society (i.e. obtaining a marriage license, moving in together, combining bank accounts, etc.). While formal roles tend to refer to behavioral aspects of role creation, the informal stage of role acquisition often refers to the development of personal attitudes and cognitions surrounding that particular role. In this stage, individuals also discover hidden role expectations transmitted by reciprocal-role others. To this point, the 10 individual has been socially educated as to their new role, has formally accepted the new role, and has begun to develop meaning within that role. The final stage, the personal stage, is conceptualized as the process of merging one?s self with the role and with the social meanings associated with that role (LaRossa & Reitzes, 1993). The role becomes internalized or adopted; behaviors, thought processes, and attitudes about that role become second nature. While there is not a stage specifically established to address times when individuals re-examine their roles and role sets, theoretically the informal stage and personal stage of development can be revisited when new or contrary information is presented. Theoretical opinion would suggest that attitudes and cognitions about roles become second nature, but they are not permanent. Importantly, it is theorized that role acquisition continues throughout personal and relational development as individuals? process new information from societal, familial, and dyadic levels (Thornton & Nardi, 1975). Symbolic interactionism posits that role-making is an interactive, ever-evolving process that allows individuals to personalize new information into behaviors that they deem most appropriate. According to this theoretical approach, roles ? including gender roles ? are not fixed. Rather, they are conceptualized as being continuously negotiated between an individual and his or her social structure beginning in childhood. Symbolic interactionism theorists would likely posit that gender roles and related gender role attitudes are reformulated throughout our life course as we encounter new social realities and contexts such as divorce and remarriage as these new social circumstances provide a context for redefining aspects of 11 the self. Accordingly, these theorists would likewise predict that these attitudes would be more stable when transitions that bring new social realities are not encountered. Gender Roles: A Component of Identity Identity theory has evolved from a symbolic interactionistic framework with a goal of understanding and explaining ?how social structures affect self and how self affects social behaviors? (Stryker & Burke, 2000, p. 285). The relationship between individuals and their contextual environment is described and studied in more detail by identity theorists (Stryker, 1968; Stryker & Burke, 2000). This theoretical perspective synthesizes the ideas of George Herbert Mead and William James ? American philosophers and psychologists ? while incorporating theoretical assumptions from modern sociologists and role theorists. One branch of identity research, known as Stryker?s theory, emphasizes how societal factors influence the individual (Stryker, 1980), while a complementary line of identity research, formulated primarily by Peter Burke, underlines the ways in which the individual affects societal processes (Burke & Reitze, 1991). Most identity theorists tend to espouse a sociological perspective as they conceptualize society as a relatively stable organization of diverse social networks. Social boundaries such as class, race, religion, education, and even marital status define as well as restrict an individual?s social network. Identity theory maintains that individuals have as many identities as distinct networks to which they are engaged (James, 1890, as cited in Stryker & Burke, 2000), and individuals tend to engage in roles that sustain their position within that network (Stryker & Burke, 2000). Therefore, the 12 self is viewed as a ?structure of identities reflecting roles played in differentiated networks of interaction? (Stryker, 1987, p. 91). These identities are then ranked by the individual in a hierarchal order based on personal salience and commitment to that identity and role set (Stryker, 1968). Salience refers to the probability that an identity will be invoked across different situations, while commitment refers to the affective importance of social relationships in which the identity is linked (Stryker & Serpe, 1982). Identity salience and commitment are typically characterized as stable constructions (Stryker & Burke, 2000), however, some identity theorists concurrently posit that these constructs can be malleable (Burke, 1991; Pasley, Kerpelman, & Guilbert, 2001). Identity disruption and role reconstruction may occur as a result of altered interpersonal interactions and/or intrapersonal perceptions (Pasley et al., 2001). Similar to Cooley?s looking glass self, these identity theorists assume that social interactions provide vital feedback to the individual. While Cooley and symbolic interactionism as a whole emphasize the importance of societal and/or cultural feedback to the individual, identity theory underscores the impact of dyadic and interpersonal relationships on the individual and his/her perception of self. This theoretical evolution recognizes that both macro- and micro-level experiences influence the individual. Sometimes referred to as a microsociological theory (Desrochers, Andreassi & Cynthia, 2002), identity theory places emphasis on how dyadic and relational interactions create unique experiences at the individual level. Hence, this theory emphasizes that not all meaning is derived and shared uniformly between individuals (Lindesmith & Strauss, 13 1956, as cited in Stryker & Burke, 2000); thus, it is important to examine micro-level characteristics of the individual, including but not limited to marital status, to more fully understand how interpersonal relationships and intrapersonal factors influences attitudes. The current research assumes that life transitions ? such as the transition from marriage to divorce and singlehood with the possibility of remarriage ? cause individuals to experience new social interactions, thus creating new positions within their social context as they move into a different phase of their life. According to identity theory, these life transitions coupled with new social interactions would likely set the stage for individuals to alter their role hierarchies, in general, and specifically, to adapt their gender role beliefs in response to new positions within society and their new conceptualization of self. Roles over the Life course Lifespan and family developmental perspectives predict and account for systematic changes the individual or family is likely to undergo; these theories tend to address cohort and period changes (Aldous, 1990). Conversely, life course developmental theories assume that significant events may alter the trajectory of systematic family development, whereas families and individuals have unique developmental patterns based on interpersonal experiences. A life course developmental framework would consider the events of a person?s life as opportunities for development, growth, and change. Significant family events ? such as marital status transitions ? are likely to impact the roles and outcomes of family members. In the context of lifecourse changes, gender identities and gender role attitudes may be altered 14 as an individual navigates and adapts to new experiences throughout his/her life. This study utilized a microsociological identity framework coupled with a life course development perspective in order to view gender role attitudes within the context of societal influence as well as individual and family development over the course of 20 years. The Study of Gender Role Attitude Changes from a Sociological Perspective The malleability of gender role attitudes across one?s life course has yet to be thoroughly examined, but aggregate findings support the notion that gender role attitudes in the United States are becoming significantly less conservative among younger age- based cohorts (Ciabattari, 2001; Rogers & Amato, 2000). Ciabattari (2001) explored this trend toward more egalitarian gender role attitudes from two sociological perspectives: period and cohort effects. Period effects influence all of society at one time, whereas cohort effects emphasize distinct characteristics of individual groups based on their particular life stage. The author notes that the feminist movement of the 1970?s was a period effect that all people experienced, but cohorts (or age groups) differed in the way they experienced and responded to the feminist movement. A national, cross-sectional sample by the National Opinion Research Center (NORC) between the years of 1974 and 1998 indicated that the oldest cohort of men held the most traditional beliefs when compared to the more egalitarian beliefs with each younger cohort. The researchers concluded that, on average, in the U.S. there has been a notable decline in gender role conservatism for both men and women and an increase in egalitarian beliefs. 15 Similarly, Rogers and Amato (2000) compared longitudinal data from two distinct generations of marital couples. The first cohort data was collected between 1964 and 1980, and the data on the second age-based cohort was collected from different couples between 1981 and 1997. Identical questions related to gender role attitudes, marital power, and marital quality were asked to both age-based cohorts. Rogers and Amato (2000) reported that marital relationships have changed in their dynamics and structure over the course these two generations as the younger cohort reported more egalitarian beliefs and more equality in terms of power within the marital relationship. These cultural changes at the macro-level were taken into consideration in the current study by examining period and cohort effects between 1980 and 2000. Studying the Malleability of Gender Role Attitudes over the Life Course from a Family Science Perspective Blee and Tickamyer (1995) originally sought to identify how demographic and life course factors impact the formation of gender role attitudes. The data ? gathered by the National Longitudinal Surveys of the Labor Market Experience (NLS) ? utilized a multistage sampling procedure to ensure a representative sample of the non-incarcerated U.S. population. One hundred and thirty-six African American men and 381 White men were followed over a ten year period and surveyed in 1971, 1976, and 1981. At each of the three time points, the men reported general demographic information as well as their gender role attitudes using two scales: Working Wives Scale (? range from .68 to .69) and the New Gender Role Scale (? = .76). On average, both subgroups reported significant overall change in their gender role attitudes towards a more egalitarian 16 direction. Interestingly, the authors report that there were no significant between group differences between African American and White men except on items measuring attitudes about women in the workforce. On these items, African American men reported more egalitarian views when compared to White men at all points. This finding is congruent with a socio-historical perspective whereas there was less of an option for African American women not to work. While Blee and Tickamyer (1995) found that factors such as historical structural factors influence existing gender role attitudes and beliefs, their longitudinal findings are among the first to exhibit evidence that gender role attitudes change over time at the individual level. Additional exploration is needed to confirm these results, because it is unclear whether individual life course events altered gender role attitudes or if period effects influenced the participants? reported gender role attitudes. Continued research is needed to address this issue particularly because this study occurred over a quarter of a century ago during a time of pronounced cohort effects resulting from social discourse focused on equality and empowerment of women. Such events as the Equal Rights Amendment (1972), Title IX (1972), Roe v. Wade (1973), and the Pregnancy Discrimination Act banning employment discrimination against pregnant women (1978) were period markers of societal shifts in views of women. Further research is needed to longitudinally examine individual change during a time comparatively void of significant period events that could impact gender role changes at the societal level. Another study reports that across specific life events gender role attitudes ? specifically related to perceived marital roles ? become more liberal for both males and 17 females (Bryant, 2003). This national sample of college students collected by the Cooperative Institutional Research Program at UCLA gathered information from 14,973 freshmen from 127 U.S. colleges and universities beginning in 1996. This study examined detailed demographic questions such as high school experiences, religious views, family background, political views, and GPA. Gender role attitudes were assessed in one global measure (?Activities of married women are best confined to home and family?) where participants responded either: Agree Strongly, Agree Somewhat, Disagree Somewhat, or Disagree Strongly. In 2000, they reevaluated this sample to assess for changes in their perceived gender and marital role beliefs. The researchers examined the relationship between preexisting factors (i.e. gender role beliefs at time one), the environment (i.e. college & collegiate experience), and the outcome (i.e. gender role attitudes at time two). From this data, the researcher concluded that gender role attitudes became more liberal or nontraditional over the course of college experiences for both men and women. As expected, women started out with more egalitarian views of gender roles in marriage and remained more egalitarian at the end of their collegiate experience. These findings help support the hypothesis that gender role attitudes are not stable belief systems and suggest that specific experiences may influence and shape gender role traditionalism in men and women, but these findings are limited to a college population. The author notes that White students, high-income families, and private institutions are over-represented in this sample. The generalizability of this study is difficult to determine due to the narrow time span of development covered and the 18 homogenous nature of participant characteristics. Although, this study proposes to examine environmental factors in relation to gender role attitude malleability, the environmental factors and contextual stressors are limited to college experience as a whole. Other experiences and stressors at the individual level are not examined such as marital status transitions. Another study assessing the malleability of gender role attitudes recognizes that attitudinal change has occurred at the cultural or macro-level, and therefore the researchers hypothesized that gender role attitudes may also fluctuate at the micro-level of development. Fan and Marini (2000) examined data from a representative sample of 8,822 youth aged 14-22 with an over sample of Black and Hispanic youth. This sample from the National Longitudinal Survey of Youth (NLSY) held a 90.4% retention rate over the three measurement points in 1979, 1982, and 1987 as the youth transitioned into adulthood. At each interval, the participants reported their attitudes towards gender roles using a construct that was created for this study. The scale is an 8-item construct that assesses attitudes toward female employment, men engaging in house work, child rearing, and appropriate roles. The items were totaled to indicate overall attitudes towards gender roles. At each evaluation point, the researchers additionally recorded educational achievement, current employment, entry into marriage, and whether the participant had a child. Findings suggest that changes in gender role attitudes were a function of ?specific socializing experiences during the transition to adulthood? (Fan & Marini, 2000, p. 277). Specifically, for women, entry into marriage and leaving the labor force 19 were associated with an increase in traditional gender role attitudes; no significant effects were found for men related to these events. Parenthood was associated with a shift in gender role attitudes toward a more traditional direction for both men and women. Conversely, increased education correlated, on average, with shifts toward more egalitarian attitudes for both men and women. Consistent with previous research, African American men and women tended to have more egalitarian attitudes throughout the study. Fan and Marini (2000) presume that there is both stability and malleability in gender role attitudes to be determined as a function of specific events or experiences the individual undergoes. The authors report that there did not appear to be systematic sex or maturation related changes in the malleability of gender role attitudes during this eight year period. While the data were gathered over 20 years ago, the conclusion that gender role attitudes tend to be rather stable except under the influence of new social experiences is a noteworthy finding. Gender Role Attitudes as a Function of Sex Research examining the impact of sex on perceived marital satisfaction and perceived role expectations has flourished following a key publication by Jessie Bernard (1956) who introduced and investigated the concept of ?his? and ?hers? marriages. This research has led contemporary scholars to study the diverse experiences of men and women in marital relationships, and researchers have begun to conceptualize how sex and gender identity impact marital relationships (Fowers, 2004). 20 Entry into marriage is a specific life experience that has been shown to have a traditionalizing effect on the behavior and attitudes of men and women (Fan & Marini, 2000; Schuchts & Witkin, 1989). However, in general, both married and non-married women appear to espouse more egalitarian gender role attitudes than men (Cassidy & Warren, 1992; Fan & Marini, 2000). Roper and Labeff (1977) reported that women hold more favorable views towards feminism than their male counterparts in terms of economic, domestic, and political equality. Larson (1974) found that, on average, husbands and wives tended to view power within the family as egalitarian, but each spouse ultimately tended to ascribe more power to themselves. Thornton (1989) reviews the literature from 1960 through the mid-1980?s and concludes that both men and women report a decrease in traditional assumptions including decreased urgency to marry, to stay married, to have children, and to maintain traditional division of labor within the home. Guilbert, Vacc, and Pasley (2000) report that married women often perceive themselves to be more egalitarian than men, in general, and subsequently found that women with the most egalitarian beliefs reported higher levels of marital instability. Diekman, Goodfriend, and Goodwin (2004) conclude that gender role attitudes may in fact be changing at the macro-level toward more egalitarian roles, but women continue to have less power in political, economic, occupational, and relational roles than men. These findings demonstrate the complex influences that one?s sex has on impacting gender role attitudes. In light of this research, findings are still rather inconsistent as to how gender role attitudes relate to marital quality, stability, and satisfaction. Kaufman (2000) 21 attempts to explore how traditional and egalitarian beliefs influence family formation and functioning from the perspective of men and women. This study was a secondary data analysis from the National Survey of Families and Households. As seen in other studies (e.g. Fan & Marini, 2000; Schuchts & Witkin, 1989), Kaufman (2000) found that gender role attitudes had differing effects for husbands and wives. Specifically, egalitarian men were more accepting of their wife?s work, shared more household chores, and were seen as more attractive partners as evidenced by lower percentages of divorce. Egalitarian women were more likely to find that their beliefs about equality were not mirrored in their marital interactions. Kaufman (2000) concludes that there were lower rates of divorce for men who maintained egalitarian gender role attitudes, and higher rates of divorce among women who promoted egalitarian gender role attitudes. This suggests that ?egalitarian women need the support of egalitarian men? (Kaufman, 2000, p. 141). While gender role congruence falls outside the scope of the current study, it is important to note that sex differences are prominent within the literature on gender role attitudes and these differences likely have implications for marital stability and satisfaction. In their chapter on ?multiple identities,? Chisholm and Greene (2008) contend that all individuals have more than one identity as a function of their race/ethnicity, age, class, sex, and so on; researchers are encouraged to embrace these meaningful human dimensions to better understand American psychology. Thus, further research is needed to clarify these patterns and differentiate how demographic factors and life course experiences influence gender role attitudes. 22 Transition to Divorce, Singlehood, and Remarriage Whereas marriage produces more traditional roles and attitudes toward gender related dynamics (Fan & Marini, 2000; Schuchts & Witkin, 1989), divorce can lead to role overload and role strain for both men and women. Hill & Hilton (1999) observed individuals as they coped with new roles following marital dissolution. They used role theory to predict the experiences and functioning of single, custodial parents post divorce. From their analysis on the NSFH (National Survey of Families and Households) dataset, Hill and Hilton found that during the transition to divorce, single parents had to adopt behavioral roles of both ?provider? and ?caretaker? as a matter of necessity for their family regardless of their sex or their previous roles in the marital relationship. Diedrick (1991) notes that during this transition women acquire new roles (i.e. provider roles) which often provide a sense of independence and increased self- esteem, whereas men adjust to roles that are comparatively less valued in our society (i.e. caretaker roles). This process of role acquisition and behavioral change appears to differ for men and women post divorce, but this study did not assess if or how gender role attitudes changed amidst behavioral role change. In an empirical article relevant to the current study, Walzer (2008) predicted that men and women likely do not recognize the influence that gendered processes have on their marriage. This study posited that during the transition to divorce gender-based attitudes would be revised by the individual. Walzer (2008) completed secondary data analysis on qualitative interviews from 135 mothers and 57 fathers. During the first interview in 1980, the parents having been separated less than 6 months were asked to 23 describe their marital relationship. Each individual was then re-interviewed one year later and asked to again describe their previous marital relationship in terms of norms, roles, parenting, and the relationship in general as well as current beliefs about gender role attitudes. From these qualitative results, Walzer (2008) discusses themes related to ?redoing? or reevaluating gender-specific roles. While not all participants described themselves as actively altering their gender roles and/or gender-related attitudes, many of the participants affirm that their views on paid work, family work, parental roles, and relational needs had changed within the year since their separation. This study maintains that ?marriage is a site of ?doing? gender, and for some ex-spouses, divorce is a site for ?redoing? gender? (Walzer, 2008, p. 18). Walzer (2008) suggests that further research is needed to confirm these results with more recent and representative samples. Another qualitative study specifically addresses the experiences of women during mid- and later life transitions to divorce and remarriage (Sakraida, 2005). Sakraida (2005) conducted interviews with 24 women between the ages of 34 and 53 to describe their coping responses during the transition to divorce. Open-ended interview questions sought to understand the terms of the divorce, the amount and type of support received during the transition to divorce, and strategies used to cope. Sakraida (2005) transcribed each interview and used qualitative coding software for data analysis. The study concludes that ?adapting to the role of being a divorcee and/or single parent involved new self identities and assuming a greater burden of responsibilities? (Sakraida, 2005, p. 82). 24 Self identities are again recreated if/when an individual decides to remarry. Remarriage creates a unique marital dyad wherein one or both individuals have already experienced the creation and termination of a marriage and experienced post-marital singlehood. Remarried dyads differ from first marriage relationships in a variety of ways. The remarried system often creates a stepfamily, thus creating connections to other individuals and households (i.e. former partners and children residing with former partners). Remarriages establish new formal and informal roles that shape the identity of an individual (i.e. formally becoming a stepdad, thinking of yourself as someone?s 2 nd wife, etc.). A life course perspective would presume that significant family events ? such as divorce, the transition to being single and/or a single parent, and remarriage ? are likely to impact the roles and role related attitudes of family members. One of the few studies specifically addressing gender role attitudes in remarital relationships posited that remarried couples would report more egalitarian attitudes and prefer a nontraditional division of labor and power when compared to experiences in their first marriage (Smith et al., 1991). The data consisted of semi-structured interviews from 64 divorced men and women who had subsequently remarried within the past three years. Smith et al. (1991) assessed participants? self-other orientation as well as sex-role orientation through self administered questionnaires followed by open ended interview questions. They concluded that there were significant changes in the gender role orientation scores of participants when retrospectively comparing their first marriages and their second marriages (Smith et al., 1991). No comments were included about the participants? gender role attitudes during the time between these relationships. 25 The trend towards more egalitarian roles was reported for both women and men, but the authors speculate after performing in-depth interviews with the participants that the reasons for these attitudinal shifts differ between the sexes. Women reported that they were looking for more sharing in remarriages and wanted to avoid the self- sacrificing behaviors that they exhibited in their more traditional first marriages. Men in remarriages appeared to be looking for sharing and more egalitarian gender roles, because they realized that companionship would not survive without these shared roles. Based on this evidence, roles in remarriages seem to be based less on traditional role sets and more on negotiated agreements. Past experience in first marriages seem to influence this change. Smith et al. (1991) provide an indication that remarriage may affect gender role attitudes, but the generalizability of this study is limited due to the retrospective and the qualitative nature of the study. Similarly, Clarke (2005) conducted interviews with women aged 52 to 90 in order to examine their experiences of remarriage in terms of the power structure, access to resources, and the division of household labor. Based on these interviews, Clarke (2005) finds evidence that remarriages were more likely to espouse egalitarian gender- related attitudes and behaviors that emphasize sharing, teamwork, and companionship when compared to the participants? first marital relationships. While this retrospective, qualitative study is limited to women over the age of 35, it does offer further indication that marital status transitions may influence gender roles and attitudes about self. Currently, we know that, on average, individuals in remarried relationships report more liberalized or egalitarian gender role attitudes when compared to their first 26 marriage via retrospective analysis (Clarke, 2005; Sakraida, 2005; Smith et al. 1991). Likewise, a review of the literature on remarriage and couplehood suggests that, on average, couples in remarried relationships assume more liberalized gender role attitudes when compared to group means of couples in first marriages (Bernstein, 2000). In order to increase the generalizability of previous findings, quantitative data can be used to complement and affirm this process of recreating gender role accountability at the individual and relational level. A larger dataset may also be able to examine other factors that contribute to gender role attitude malleability following marital status transitions. Furthermore, a comparison group may be beneficial to contrast the different experiences and reported gender role attitudes of those who divorce and remarry versus those who remain in first marriage relationships. After an extensive review of the literature, no empirical evidence could be found that prospectively examines the malleability of gender role attitudes across a longitudinal course with emphasis on marital status transitions. An abundance of research exists documenting sociological change in gender role attitudes in the United States, but few studies address attitudinal change at the individual or micro-level of development. Theoretically, scholarly opinion would predict that gender role attitudes are malleable constructs, and while some evidence can be found to support this notion (e.g. Blee & Tickamyer, 1995; Fan & Marini, 2000; Walzer, 2008) findings are often difficult to generalize. This study addressed this limitation by examining between group differences in the malleability of gender role attitudes in relation to those who experiences marital status transitions and those who are continuously married. 27 Based on the findings from Fan and Marini (2000) and in accordance with identity theory, this study assumed that gender role attitudes are not fixed beliefs. Conversely, these attitudes are thought to be modified and refined within the dyadic interplay of intra- and interpersonal experiences. Therefore, it is hypothesized that certain significant experiences including divorce/singlehood and remarriage alter an individual?s beliefs about gender related roles. Because individuals are impacted by their contextual environment, it is also posited that general exposure to societal influences will impact one?s sense of identity and roles. Social and cultural trends were predicted to have a distal effect on gender related attitudes; thus, some systematic cohort and/or period effects were expected to occur over the course of the study. Just as entry into marriage and other life course experiences have been found to influence gender role attitudes at the individual level (Fan & Marini, 2000), the focus of this study is on divorce and remarriage as significant life experiences that serve as proximal influences to the individual and his/her gender role attitudes. This study posited that greater malleability would be observed in the subsample of participants who experienced divorce and remarriage, but period effects over the course of the study were also expected to influence gender role attitudes. This study tested the following research questions and hypotheses: Hypothesis 1: There have been significant period effects on gender role attitudes between 1980 and 2000. Research Question 1: Does age group and/or sex affect gender role attitudes across time? 28 Hypothesis 2: Individuals who divorce and remarry will, on average, report more egalitarian attitudes in their remarried relationship when compared to their attitudes in their first marital relationship. Hypothesis 3: Individuals who divorce will, on average, report more egalitarian attitudes when they are divorced compared to their gender role attitudes in their marriage. Hypothesis 4: The experience of marital status change will be related to greater gender role malleability, such that those who experience marital status change will demonstrate significantly greater change in gender role attitudes across time compared to individuals who remain in first marital relationships. 29 III. METHODS Participants Secondary data analysis of a 20-year longitudinal investigation entitled Marital Instability over the Life Course was conducted. In 1980, investigators Booth, Johnson, Amato, and Rogers recruited a large, nationally representative sample through funding, in part, by the National Institute on Aging within the United States Department of Health and Human Services. The initial criterion for the sample maintained that participants must be married and between the ages of 18 and 55 in households in the continental United States. Using a simple random-digit dialing procedure, the investigators utilized a clustering technique to ensure that the sample represented families in diverse geographical locations. When a household had been identified as having met the criteria of the study, a second random selection process was used to select either the husband or wife to complete the interview. The interview consisted of a series of questions assessing individual and family functioning, beliefs/attitudes, and information about the respondents? health, life style, and relationships. Of those individuals eligible to participate in the survey, 65% initially gave complete interviews. The sample consisted of 2,033 married persons; their spouses were not interviewed. When compared with U.S. Census data, the investigators reported that the sample was representative of married individuals with respect to age, race, household 30 size, housing tenure, presence of children, and region of the country (Booth, Johnson, Amato, & Rogers, 2003). Following the initial interview, five additional interviews were completed in 1983, 1988, 1992-1994, 1997, and 2000. Attrition rates varied between 10-21% between each interview period (Table 1), which the principal investigators ascribed as a normative attrition rate compared to similar longitudinal studies. If participants refused to complete the survey, they generally were not contacted again in the next interview year. Attrition occurred in predictable categories including African Americans, renters, households in which husbands had no college education, and households who moved/could not be located (Myers, 2004). Of the 2,033 who began the study in 1980, 962 participants completed each of the six re-interviews. Notably, participants were not given any form of compensation by the project investigators. Table 1. Respondent Attrition Time 1983 1988 1992-1994 1997 2000 Completed n = 1,593 n = 1,341 n = 1,193 n = 1,066 n = 962 Attrition Refusal 133 153 111 54 43 Deceased 13 15 18 25 18 Not located 264 50 15 25 12 20 callbacks 22 32 7 4 Illness 1 1 1 8 5 Other 7 1 1 11 26 Total 2,033 1,593 1,346 a 1,193 1,066 a Contacted participants who completed partial interviews in 1988 The current study examined gender role attitude malleability with a focus on the transition from first marital relationships into divorce/singlehood and remarriage. Based 31 on the previous review of the literature and national data trends for divorce and remarriage, the researcher elected to examine participants at the beginning and conclusion of the study ? 1980 and 2000 ? as well as utilize a mid-point time in 1988 to explore patterns of gender role attitude change. National data trends demonstrate increased divorce rates from 1950 into the 1980?s followed by a period of declining divorce rates through the year 2000 (Fields, 2003); therefore, the 1988 data appears to be the most appropriate for examining this timeframe of increased marital status transitions in the United States. Further criterion was established in the current study to ensure that all participants reported similar marital statuses at the commencement of the study. Those participants who were already in a remarital relationship at Time 1 (1980) were not included in this secondary data analysis resulting in 855 eligible participants. Additionally, due to the nature of the analyses, which will be discussed more fully in a later section, a repeated measures design ?requires complete data on all trails from all participants? as one of the primary assumptions of validity (Osborne, 2008). A total of 590 participants met all aforementioned criteria. For the purposes of this study, no weighting system was used. The demographic profiles of the current study compared to the original dataset and final sample in 2000 are summarized in Table 2 (Booth et al., 2003). The majority of the participants were female (males, n = 223; females, n = 367). The average length of years married in 1980 among the participants was 12.71 years (SD = 9.41). At time 32 three, 20 years after the initial sample, the participants ranged in age from 38 to 75 with a mean of 54.17 years (SD = 9.0). Table 2 Demographic Information for all Participants and Relevant Subsamples 1980 Sample (n = 2,033) 2000 Sample (n = 962) Current Study (n = 590) Sex Male 821 (40.4%) 351 (36.5%) 223 (37.8%) Female 1212 (59.6%) 611 (63.5%) 367 (62.2%) Age in 1980 Male M = 36.58 SD = 9.36 M = 36.37 SD = 9.10 M = 35.05 SD = 8.98 Female M = 34.86 SD = 9.64 M = 34.81 SD = 9.04 M = 33.63 SD = 8.98 Ethnicity European American 1,789 (88%) 898 (93.3%) 550 (93.2%) African American 98 (4.8%) 26 (2.7%) 13 (3.3%) Latino American 106 (5.2%) 28 (2.9%) 20 (3.4%) Other 37 (1.8%) 10 (1.0%) 7 (1.2%) Measures To assess the malleability of gender role attitudes within the context of marital status transitions, the current study utilized quantitative data from surveys in 1980, 1988, and 2000 and participants were differentiated into marital status groups based on their marital experiences at the conclusion of the study. Demographic variables were also examined in relation to gender role attitude change. Gender Role Attitudes. Gender role attitudes were measured by a seven-item summated scale in which the participants were able to rate each statement using a 4- point scale (strongly agree to strongly disagree). Scores were converted back to the 33 original scale so that scores range from one to four. Four of the questions were reverse coded, so that higher scores indicated more traditional values (i.e. distinct roles for men and women; unequal power). The scale was created by the original investigators based on information from the Sex Role Egalitarianism Scale (SRES) as well as prominent themes in the gender role literature pertaining to appropriate/desired roles in the home and workplace for husbands and wives. Participants were also able to refuse to answer a question or indicate if they do not know the answer to any of the statements. Because participants were given this option, the original investigators chose to add the value of the items the participant did answer, and then weight that total for the number of items with ?I do not know? or ?refuse? responses (Booth et al., 2003). The current study opted not to utilize imputation or substitution methods, but rather to employ a casewise deletion technique to ensure that all scores obtained represent the same set of survey questions to more fully and equally assess gender role attitudes among participants. In this study, internal consistency among the seven items was reported at ? = .70 in 1980, ? = .72 in 1988, and ? = .60 in 2000. Items can be found in Appendix A. Age Cohort. Because of the extended age range of the current sample, two age- based groups were created to decipher potential cohort effects on gender role malleability. At time one, the sample ranges from age 19 to 55 with a median of 33 years. Univariate analysis reveals a fairly symmetric distribution (M = 34.14, SD = 8.99). From these analyses, two categorical variables ? the younger cohort and the older cohort ? have been derived from this sample. The younger cohort consists of 34 participants from age 19 to 34 representing approximately 57.1% of the sample. The older cohort is comprised of participants aged 35 to 55. This second cohort includes a larger range of ages, but slightly less than half of the sample size (42.9%). Marital Experience. At time one, all of the respondents (n = 590) were in first marital relationships. Marital trajectories of the participants were examined, and participants were then grouped according to their marital status at the conclusion of the study (i.e. still in first marriage, divorced from first marriage, or remarried). The current investigator used the data from the interviews in 1983, 1992-1994, and 1997 to ensure that any changes in marital status were recorded (Appendix B). For analysis purposes, if participants had experienced a 2 nd divorce or a higher order union than a 2 nd marriage, they were selected out of the current study. Plan of Analysis The purpose of this study was to explore gender role attitude change over the course of individual level experiences particularly related to marital experiences. Period effects were examined in the first hypothesis to assess gender role malleability for the entire sample. To test this, a paired samples t-test was conducted using gender role attitude scores in 2000 and comparing them to gender role attitude scores in 1980. Subsequent research questions were addressed to understand how participant characteristics were related to gender role attitude change over the course of the study. Interaction effects of time X characteristic on gender role attitudes were examined using mixed between-within subjects repeated measures analyses of variance (RMANOVA) for each of the research questions. The gender role attitude mean score in 1980, 1988, 35 and 2000 served as the repeated measure, and the between-subjects variable groups are based on age group and sex. The second hypothesis stated that individuals would, on average, report more egalitarian attitudes in their remarried relationship when compared to their attitudes in their first marital relationship. To assess this hypothesis a paired samples t-test was conducted using the data only from remarried participants (males, n = 24; females, n = 30) comparing their gender role attitude scores from their first marital unions (1980) and gender role attitude scores from their remarital unions (2000). Similarly, the third hypothesis stated that individuals would, on average, report more egalitarian attitudes following divorce when compared to their attitudes in their first marital relationship. To assess this hypothesis a paired samples t-test was conducted using the data only from participants who had divorced and remained single (males, n = 14; females, n = 34) by comparing their gender role attitude scores from their first marital unions (1980) and gender role attitude scores from their time of singlehood following divorce (2000). The fourth hypothesis posited that the experiences related to marital status transitions would be related to greater gender role malleability, such that those who experience divorce and remarriage demonstrate greater changes across time in their gender role orientation compared to those who do not experience divorce and remarriage. To test for significant between group differences in those individuals who remained in their first marital relationships and those who experienced marital status change, the researcher conducted a mixed between-within subjects repeated measures 36 analysis of variance (RMANOVA), wherein the gender role attitude mean score in 1980, 1988, and 2000 again served as the repeated measure, and the between-subjects variable groups depended on marital status (i.e. continuously married or non-continuously married). These finding lead to further examination of the data set in which mixed between-within subjects repeated measures analysis of variance (RMANOVA) were utilized to compare participants who were continuously married, those in remarital relationships in 2000, and those who experienced divorce and were still single in 2000. All of the statistical analyses were performed in SPSS for Windows. 37 IV. RESULTS Preliminary Analysis Prior to testing specific hypotheses, initial descriptive statistics for the Gender Role Attitude Scale were computed across each of the three time points. The statistics are included in Table 3. Preliminary assumption testing was also conducted to check for independence, normality, and sphericity/equality in variance, with no serious violations noted. Table 3. Means, Standard Deviations, and Reliabilities for Survey Scales (n = 590) Gender Role Attitude Scale Mean Standard Deviation Reliability 1980 16.02 9.71 .70 1988 15.11 9.46 .72 2000 14.99 6.97 .60 Hypothesis 1: There have been significant period effects on gender role attitudes between 1980 and 2000. A paired-samples t-test was conducted to evaluate the change in gender role attitudes for the entire sample. On average, there was a statistically significant difference in gender role attitude scores from 1980 (M = 2.29, SD = .45) to 2000 (M = 38 2.14, SD = .38), t (589) = 8.47, p < .001 (two tailed). An inspection of the mean scores at each time point indicated that, on average, the entire sample reported more egalitarian attitudes in 2000 compared to 1980. The magnitude of change can be indicated by comparing the mean level difference to the standard deviation. From 1980 to 2000 the mean level gender role attitude score changed approximately .35 standard deviations, indicating a modest shift. Research Question 1: Does age group and/or sex affect gender role attitudes across time? A mixed between-within subjects repeated measures analysis of variance (RMANOVA) was conducted to assess the impact of the different age-related cohorts on participants? gender role attitude scores over time. While there was a significant main effect for time, Wilks? Lamdba =.89, F (1, 588) = 72.26, p < .001, partial eta squared = .109, with both groups reporting more egalitarian gender role attitudes between 1980 and 2000, there was no significant interaction effect for age cohort X time, Wilks? Lamdba =.99, F (1, 588) = 0.66, p = .418, partial eta squared = .001. A mixed between-within subjects repeated measures analysis of variance (RMANOVA) was also conducted to assess the impact of sex on participants? gender role attitude scores between 1980 and 2000. There was significant interaction effect for sex X time, Wilks? Lamdba =.99, F (1, 588) = 4.15, p < .05, partial eta squared = .001, such that women?s increase slope in egalitarian attitudes across time was significantly greater than men?s. The main effect for time was also significant, Wilks? Lamdba =.89, F (1, 588) = 76.33, p < .001, partial eta squared = .115, but this finding should be 39 interpreted cautiously as the change across time may be driven by the group that experienced significantly greater change. Hypothesis 2: Individuals who divorce and remarry will, on average, report more egalitarian attitudes in their remarried relationship when compared to their attitudes in their first marital relationship. A paired-samples t-test was conducted to assess if individuals in remarital relationships (n = 54) in 2000 report statistically significantly different gender role attitudes in their remarital relationship compared to their reported gender role attitudes when they were in first marital relationships (i.e., 1980). Results revealed that the difference in mean level gender role attitude scores between 1980 when the participants where in first marital relationships (M = 2.24, SD = .50) and 2000 when the participants were in remarital relationships (M = 2.11, SD = .39), did not reach a level of statistical significance [t(53) = 1.90, p = .06] when utilizing a two-tailed method of analysis. Because this was a directional hypothesis predicting that attitudes would shift towards a more egalitarian direction, one-tailed analysis revealed a statistically significant difference in reported gender role attitudes between first marriages and remarriages [t(53) = 1.90, p = .03]. The magnitude of change can be indicated by comparing the mean level difference to the standard deviation. From 1980 to 2000 the mean level gender role attitude score changed approximately .25 standard deviations, indicating somewhat modest shift. This hypothesis was supported. 40 Hypothesis 3: Individuals who divorce will, on average, report more egalitarian attitudes when they are divorced compared to their gender role attitudes in their marriage. A paired-samples t-test was conducted to assess if divorced individuals (n = 48) in 2000 would report statistically significantly mean level gender role attitude score during singlehood compared to their reported gender role attitudes when they were in first marital relationships (i.e., 1980). Again this was a directional hypothesis predicting that attitudes would shift towards a more egalitarian direction. Accordingly, one-tailed analysis revealed that the difference in mean level gender role attitude scores between 1980 when the participants where in first marital relationships (M = 2.21, SD = .43) and 2000 when the participants were single (M = 2.00, SD = .43) reached statistically significant [t(47) = 3.86, p < .0005 (one tailed)]. As expected, inspection of the mean scores at each time point indicated that, on average, individuals who had undergone marital status change into divorce/singlehood reported more egalitarian attitudes. The magnitude of change can be indicated by comparing the mean level difference to the standard deviation. From 1980 to 2000 the mean level gender role attitude score changed more than a half of a standard deviation (i.e., .56), indicating a moderately large shift. This hypothesis was supported. Hypothesis 4: The experience of marital status change will be related to greater gender role change, such that those who experience marital status change will demonstrate significantly greater malleability in gender role attitudes across time 41 compared to individuals who remain in first marital relationships. That is, there will be a time X marital status change interaction effect on gender role attitudes. A mixed between-within subjects repeated measures analysis of variance (RMANOVA) was conducted to assess the impact of the different marital status experiences on participants gender role attitude scores across three time periods. There was no significant time X marital status experience interaction effect for gender role attitudes when comparing participants who had remained in first marital relationships and those who experienced any type of marital status change, Wilks? Lamdba =.99, F (2, 587) = 1.78, p = .17. There was a substantial main effect for time, Wilks? Lamdba =.90, F (2, 587) = 32.59, p < .001. This hypothesis was not supported when participants were dichotomously split (i.e., those who experienced marital continuity and those who experienced some type of marital status change). Due to this finding, and given that other analyses comparing gender role attitudes in 1980 and 2000 revealed a greater shift for the group that remained divorced, a fourth research question was examined; Do different patterns of marital status change affect the level of change in gender role attitudes across time? That is, are there differences between those who are continuously married and those who are in remarital relationships? Furthermore, are there differences between those who are continuously married, those who divorce and remarry, and those who divorce and do not remarry? A mixed between-within subjects repeated measures analysis of variance (RMANOVA) was conducted between marital change groups; first, eliminating those who experienced divorce only. When comparing those who were continuously married and those who were remarried in 2000, a significant time X marital status experience interaction effect for gender role attitudes was found, Wilks? Lamdba =.99, F (2, 539) = 3.98, p > .05, such that those who experienced divorce and remarriage demonstrated a pattern of change across time that was significantly different than those who remained married across time. While reaching statistical significance, the effect size of the change is considered small to modest based on the partial eta squared, .02 (Cohen, 1973). Analyses reveal a significant difference in gender role attitudes between groups in 1988 (see Figure 1 and Table 4), with those who experienced divorce and remarriage reporting significantly more egalitarian gender role attitudes in 1988 than those continuously married. Between group contrasts showed no significant differences in 1980 and 2000. Figure 1. RMANOVA of Gender Role Attitude Scores 42 Importantly, this time X marital status experience interaction effect remained significant when controlling for the age, race and sex of the participants (See Figure 2), Wilks? Lamdba =.99, F (2, 536) = 3.51, p = .03. Again, the effect size of this change is considered small, partial eta squared = .013. Figure 2. RMANOVA of Gender Role Attitude Scores Controlling for Age, Sex, and Race A mixed between-within subjects repeated measures analysis of variance was then conducted to assess the differences in changes in gender role attitudes across time among three categories of participants? marital experiences: those who experienced marital continuity; those who divorced and were remarried in 2000; and those who divorced and were not remarried in 2000. When comparing these three groups between 1980 and 2000, the time X marital status experience interaction effect for gender role attitudes was statistically significantly different, Wilks? Lamdba =.98, F (4, 1,172) = 2.33, p < .05, partial eta squared = .01, such that the pattern of change across time was 43 significantly different based on one?s marital status experiences when comparing those who remained in first marital relationships, those who experienced divorce only, and those that experienced divorce and remarriage (see Table 4 and Figure 3). Post hoc comparisons using the Least Significant Difference (LSD) test indicated that the mean score for the continuously married group was significantly different in 1988 from the remarried group and in 2000 from the group that was divorced and single. Table 4. Between-Groups Analysis of Variance of Gender Role Attitudes Time Continuously Married (n = 488) Remarried (n = 54) Divorced (n = 48) M SD M SD M SD F Sig 1980 2.30 .44 2.24 .50 2.21 .43 1.13 .325 1988 2.18 .43 2.01 .47 2.07 .48 4.91 .008 2000 2.16 .37 2.11 .39 2.00 .43 3.89 .021 Figure 3. RMANOVA of Gender Role Attitude Scores 44 This time X marital status experience interaction effect remained significant when controlling for the age, sex, and race of the participants (See Figure 4), Wilks? Lamdba =.98, F (4, 1164) = 2.45, p < .05. Again, the effect size of this change is considered small, partial eta squared = .01. Post hoc comparisons at time two (1988) using the Least Significant Difference (LSD) test indicated that the mean gender role attitude score of the continuously married group was significantly different from those in the remarital group. Post hoc comparisons at time three (2000) using the Least Significant Difference (LSD) test indicated that the mean score of the group who remained divorced was significantly different from those who were continuously married and those who remarried. Figure 4. RMANOVA of Gender Role Attitude Scores Controlling for Age, Sex, and Race 45 46 V. DISCUSSION Based on previous findings in the gender role literature and in accordance with the assumptions of symbolic interactionism and identity theory, the purpose of this study was to explore gender role attitude malleability within the context of marital status transitions. As earlier noted, the majority of findings related to gender role attitude malleability occur at the sociological level of study. The current study attempted to examine life course experiences that theoretically could alter gender role attitudes within an individual?s lifecourse. This analysis was able to examine profiles of individuals who experienced similar marital status transitions to explore gender role attitude change. Sociological Effects on Gender Role Attitudes Examination of gender role attitude change between 1980 and 2000 revealed significant period effects. On average, participants reported more egalitarian gender role attitudes over the course of this twenty year period. This finding is consistent with previous findings of macro-level influences on individuals? gender role attitudes (Ciabattari, 2001; Rogers & Amato, 2000). It is possible that significant cultural events contributed to this distal effect on gender role attitudes including advances for women in the workforce and in politics with the appointment of the first woman to the Supreme Court in 1981, reaffirmation of a woman?s right to decided on abortion (i.e. Planned Parenthood v. Casey, 1992), and even the birth of the National Fatherhood Initiative in 47 1994 that promotes the need for responsible fathers in the home and society to name a few. Demographic Effects for Gender Role Attitudes: Age and Sex Demographic variables have previously been found to influence gender role attitudes. In the current study, the researcher first examined age-related cohort differences. There were no significant between group differences in the rate of change in gender role attitudes over time based on age groups. The younger and older age group did not appear to experience the time between 1980 and 2000 differently in terms of their gender role attitudinal change. Conversely, significant between group differences in reported gender role attitudes over time were found when examining men and women. Although all participants, on average, reported more egalitarian beliefs over the course of 1980 to 2000, women were found to have a significantly different rate of change than men. This finding is consistent with the majority of studies examining sex differences in gender role attitudes across a variety of contexts including adolescent development, marital relationships, and politics (Cassidy & Warren, 1992; Fan & Marini, 2000; Roper & Labeff, 1977). The current study suggests that while men appear to alter their gender role attitudes towards a more egalitarian direction over time, the rate of change is significantly different for women such that women reported a sharper decrease in traditional gender role attitudes. This pattern of increasingly less congruence over time may have implications for marital instability should this difference in the rate of change in gender role attitudes occur within dyads. 48 Gender Role Attitude Malleability in the Context of Marital Status Transitions Examination of gender role attitude change over time in the context of marital status transitions revealed statistically significant differences in reported attitudes for individuals between their first marital and their remarital relationships when utilizing a directional analysis, such that individuals reported more egalitarian gender role attitudes in their remarriages. This finding is consistent with the few studies that have focused on changes within individuals who experience a first and a subsequent marriage. Smith et al. (1991) found that individuals retrospectively comparing their first marital relationship to their remarriage reported more egalitarian gender role orientations in their remarriages. Qualitative studies indicate that divorced individuals frequently report ?redoing? or reevaluating gender-specific roles (Sakraida, 2005; Walzer, 2008). Importantly, the current study is the first to utilize longitudinal data and prospectively examine gender role attitude change among those who experience marital transitions. While the change in gender role attitudes was different when comparing individuals in their first marriage and in their second marriage, it was clear that the magnitude of change was much more modest when compared to the change in gender role attitudes for those who were in their first marriage in 1980 and divorced in 2000. This lead to the uncovering of differential patterns of change across time between those who remarried and those who remained divorced by examining a mid-time-point in the 20-year span. When 3 timepoints were examined and subjects were grouped by those who experienced any type of marital status transition and those who remained in first marital 49 relationships, there were no between group interaction effects for time and marital status, only a main effect for time. Given the difference in degree of change found for the simple paired-sample t-tests with those who remarried and those who remained divorced, we suspected that these two groups? experiences across time differed significantly and that this information is lost if the groups are lumped together as ?experienced marital transition.? To explore this possibility, the group who experienced marital status transitions was divided into two sub-groups: those who experienced divorce only and those who experienced both divorce and remarriage. Subsequently, a significant between group interaction effect for time and marital status for gender role attitudes was found between individuals who remained in first marital relationships and those who experienced both divorce and remarriage. As seen in Figures 1 and 2, this examination indicated that in 1988, those who remarried had significantly more egalitarian gender role attitudes than the continuously married group, but in 2000, the two groups did not differ from each other. This interaction effect was significant even when controlling for age, race, and sex. Thus, a pattern across time of significantly greater egalitarian gender role attitudes, followed by significantly greater traditional gender role attitudes was evident for those who were remarried in 2000. Those who remained continuously married showed a steady decline in traditional gender role attitudes across time. Interestingly, in 1988, a demographic profile of the remarried group revealed that the participants report diverse marital statuses at that time point of the study wherein 29.6% were already in remarital relationships, 25.9% were divorced/single, yet 44.4% 50 were still in their first marital relationship. Nearly half of individuals were still in their first marriage at the time when this group ? who would all eventually remarry by the conclusion of the study ? reported the most egalitarian attitudes. Thus, it can be speculated that gender role attitudes begin to shift in a more egalitarian direction prior to divorce. This supports Hetherington and Kelly?s (2002) theory that ?postdivorce life begins within the marriage? (p. 42); meaning that, snap decisions to divorce rarely occur. Individuals, particularly women, experience a breakdown of the marriage and often have some time to prepare for life after marriage, and this seems to be a context that would promote attitudinal change related to self, the relationship, and even roles within the relationship. Hetherington and Kelly (2002, p. 33) additionally note that ?for many couples, the fighting is long over and a period of distancing and sadness occurs as affection, respect, and the marriage unravel.? This paints a picture of the couple unraveling to become two distinct individuals with separate ? and potentially novel ? attitudes about the relationship as well as discrepancies between actual and desired roles. It also suggests that the event of divorce does not necessarily dictate or begin the process of attitudinal change surrounding gender roles, but rather attitude change often occurs in the context of the divorce process, and may even contribute to it. In this case, it is possible that the event of divorce or remarriage may alter gender role attitudes for some, but it also seems to indicate that some individuals began to perceive gender role attitudes differently before undergoing any marital status transition. 51 Hetherington and Kelly (2002) observed the stability/instability of traditional marriages when one or both partners alter their gender role behaviors and attitudes and noted "the Achilles heel of traditional marriage is change...When one or the other partner begins to behave untraditionally, trouble follows" (p. 34). Equally interesting is the unexpected pattern that emerged, showing those in the remarital group becoming more traditional in 2000, such that there were no significant differences between those in first marital relationships and those in remarriages. While those who remarry do not return to their previous level of traditional gender role attitudes, there is a significant change in their gender role attitudes around the time of divorce and again, around the time of their remarriage. This pattern of slight reversal to more traditional gender role attitudes in remarriage has not been previously revealed in research. To further examine patterns across time, additional analyses examined gender role attitude malleability comparing the reported gender role attitude scores of those who were continuously married, those who divorced and remarried, and those who divorced and remained single. Again, a significant between group interaction effect for time and marital status for gender role attitudes was found between these three groups, indicating the significant difference in change patterns between those who remarry and those who do not. Specifically, an added dimension of this test showed that those who were divorced at time three (2000) became increasingly egalitarian over time and by 2000, had significantly more egalitarian gender role attitudes than either those in first marriages or those remarried. 52 Unlike the divorced/remarried group, the divorced/single group does not at any point in the study have an increase in traditional gender role attitudes. Instead, they become more egalitarian over time. Hetherington and Kelly (2002) documented that men and women experience events and difficulties that were not expected in the years following their divorce if a remarriage does not occur. They continue to operate in less gendered roles. It may be that the necessity of performing both ?male? and ?female? roles is associated with more egalitarian gender role attitudes and then those who remarry may resume some gender-specific roles and behaviors, leading to an increase in traditional gender role attitudes. This interplay of behaviors and attitudes across time, however, can only be speculated in the current study. Findings here identify a distinct pattern of change for those who remarry that is distinct from those who do not and opens the way for further exploration of shifts in gender role attitudes for those who experience marital transitions. Both predictors and outcomes of different patterns of change will inform our empirical knowledge base, as well as inform practice. Limitations The current study offers some indications that gender role attitudes are malleable at the individual-level of development particularly in the context of marital status transitions. Limitations of the study, however, are acknowledged and some cautions in the interpretations of the findings are suggested. Although the overall sample size is substantial, the sub-samples of those who experiences marital status transitions are relatively small. In addition, the data collected in 1980 were representative of the United States population in terms of age, sex, and race; however, attrition rates differed by race 53 and socio-economic status. Data for these analyses were not weighted and therefore, the results of the current study pertain predominantly to European American couples. An additional limitation of this study is that the data on gender role attitudes was collected independently. There is a growing literature on gender role congruence that examines spousal gender role attitude congruence in relation to marital satisfaction, distress, and even instability (Bollman, Schumm, Jurich, & Yoon, 1997; Lye & Biblarz, 1993). Assessing an individual?s position on gender roles independently does not provide information on the context of the marriage during marital transitions. That is, we have no information on the spouses? trajectories of change in their gender role attitudes and how this dyadic assessment affects and is affected by marital status change. Also, because these data were collected at prescribed intervals and patterns of change were assessed only at three of those intervals, participants in each group varied by length of marriage, divorce, and/or remarriage, as well as by marital status. Therefore, only general patterns of change between groups could be assessed. Suggestions for Future Research Given the prevalence of the experiences of divorce and remarriage in the United States (Fields, 2003) and the influential nature of gender role attitudes (Fan & Marini, 2000; Softas-Nall & Sukhodolsky, 2006; Stone & Lovejoy, 2004), further study of gender role attitude change in the context of marital status transitions will serve to promote understanding of the processes for altering and creating more functional and/or desirable gender role beliefs. Comparison studies based on marital status groups provide some useful information in describing the group profiles of individuals who experience 54 divorce and remarriage and those that experience divorce and continued singlehood. Continued work using more sophisticated methods for modeling change at the individual level can reveal more of the nuances regarding timing and the context of change, as well as the implications of change for marital quality and stability. In future research, examination of additional demographic/individual-level variables such as the presence of children, socio-economic status, and work force participation will be valuable in the process of understanding how these factors influence gender role attitude stability and change. Additionally, individual and relational well- being indicators can be used to investigate how certain gender role attitudes are related to individual satisfaction as well as relational quality and stability. To date, there is little to no information available addressing the functionality of certain gender role attitudes in the context of marital status transitions. That is, are certain gender based attitudes more useful or practical to endure and flourish during the transition to divorce as well as the transition to remarriage? Congruence studies addressing gender-based attitudes and behaviors would also make a major contribution to the literature on marital status transitions. Examining the timing and context of attitudinal and behavioral change can be valuable information for researchers, therapists, and marriage educators. That is, do attitudes and behaviors change somewhat concurrently over the course of divorce and remarriage or do patterns of change in attitudes and behaviors operate more independently based on the context of the divorce and/or remarriage (i.e. sex of the individual, whose decision it was to divorce, becoming a single parent, length of time divorced before remarrying, etc.)? 55 How do gender role behaviors and gender role attitudes influence each other across time? Finally, the field of human development and family studies should strive to continually enhance diversity in research. Thus, employing an ethnically diverse sample of continuously married and noncontinuously married individuals would serve researchers and clinicians as a means to better understand the process of attitudinal change across and within racial groups. Practical Implications Gender is a predominant organizing principle among cultures and societies as well as within marital relationships (Coltrane, 1998). Whether or not couples talk explicitly about gender based roles and their attitudes pertaining to those roles, research continues to demonstrate that gender role attitudes have considerable bearing on marital quality and relational distress. The current study contributes to this body of research by first demonstrating that gender role attitudes are not fixed belief systems across individual life course development. Instead, it appears that both sociological influences and life course factors influence one?s perception of appropriate or desirable gender roles. Demographic factors and early childhood experiences are important influences that initially shape and form gender role attitudes, but this study found that life course experiences ? particularly marital status transitions and the contexts that surround these transitions ? either change the way people think about appropriate or desired gender roles or result from changes in 56 gender role attitudes. This has important implications for therapists, marriage educators, and couples. Therapists and marriage educators can utilize this information with clients and consumers by first creating an awareness of gender role attitudes within marital relationships. As indicated previously, the theoretical process of role making maintains that these attitudes and beliefs become second nature and ingrained within the individual and/or the relationship. Bringing these implicit beliefs and assumptions into the realm of awareness may help the couple to create shared meaning about gender based roles within the marriage. Therapists and family life educators may also benefit from using this information in premarital settings to help couples begin a dialogue about roles that are traditionally masculine or feminine. Because gender based roles are a predominant feature within marital relationships, it is important for couples at any stage ? premarital, remarital, or long-term marital relationships ? to share and potentially rework some of their conceptualizations for appropriate or desired gender roles. It is important for the individual to have a level of awareness of their own process of ?doing gender? within their relationship or ?re-doing gender? (Walzer, 2008) during marital status transitions to potentially recognize their desires and expectations for future relationships. Because the state of the research does not allow for a ?prescription? of patterns of functional gender role attitudes (i.e., we have no information from this study or consensus from others on what patterns of change over time predict marital quality and marital stability), we can only emphasize the evidence that gender role attitudes can be altered in adulthood. Evidence suggests that dyadic 57 congruence in gender role attitude is associated with greater marital quality and marital stability. Thus, the recognition that gender role attitudes can change can lead to greater couple congruence. 58 REFERENCES Aldous, J. (1990). Family development and the life course: Two perspectives on family change. Journal of Marriage and the Family, 52, 571-583. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC: Author. Bernard, J. (1956). Remarriage: A study of marriage. American Sociological Review, 21, 652-653. Bernstein, A. C. (2000). Remarriage: Redesigning couplehood. In P. Papp (Ed.), Couples on the fault line: New directions for therapists (pp. 284-311). New York: Guilford Press. Blair, S., & Lichter, D. (1991). Measuring the division of household labor: Gender segregation of housework among American couples. Journal of Family Issues, 12, 91-113. Blee K. M., & Tickamyer, A. R. (1995). Racial differences in men?s attitudes about women?s gender roles. Journal of Marriage and the Family, 57, 21-30. Bollman, S., Schumm, W., Jurich, A., & Yoon, G. (1997). Predicting marital satisfaction from ideal and actual maternal roles. Psychological Reports, 80, 99-106. Booth, A., David J., Paul A., & Stacy R. Marital instability over the life course (United States): A five-wave panel study, 1980, 1983, 1988, 1992-1994, 1997, 2000 59 [Computer file]. 1st ICPSR version. University Park, PA: Alan Booth et al., Pennsylvania State University [producers], 1998. Ann Arbor, MI: Inter- university Consortium for Political and Social Research [distributor], 2003. Booth, A., Amato, P., Johnson, D., & Edwards, J. Marital instability over the life course (United States): A five-wave panel study, 1980, 1983, 1988, 1992-1994, 1997 [Computer file]. 2nd ICPSR version. University Park, PA: Alan Booth et al., Pennsylvania State University [producers], 1998. Ann Arbor, MI: Inter- university Consortium for Political and Social Research [distributor], 2001. Booth, A., Johnson, D., Amato, P. & Rogers, S. Marital instability over the life course (United States): A six-wave panel study, 1980, 1983, 1988, 1992-1994, 1997, 2000 [Computer file]. ICPSR03812-v1. University Park, PA: Alan Booth et al., Pennsylvania State University [producers], 1998. Ann Arbor, MI: Inter- university Consortium for Political and Social Research [distributor], 2003. Brooks, C., & Bolzendahl, C. (2004). The transformation of US gender role attitudes: Cohort replacement, social-structural change, and ideological learning. Social Science Research, 33, 106-133. Bryant, A. (2003). Changes in attitudes toward women's roles: Predicting gender-role traditionalism among college students. Sex Roles, 48, 131-142. Burke, P.J. (1991). Identity processes and social stress. American Sociological Review, 56, 836 849. Burke, P. J, & Reitzes, D. C. (1991). An identity theory approach to commitment. Social Psychology Quarterly, 54, 239-251. 60 Cassidy, M., & Warren, B. (1996). Family employment status and gender role attitudes: A comparison of women and men college graduates. Gender & Society, 10, 312- 329. Chisholm, J., & Greene, B. (2008). Women of color: Perspectives on 'multiple identities' in psychological theory, research, and practice. In F.L. Denmark & M.A. Paludi (Eds.), Psychology of women: A handbook of issues and theories (2nd ed.) (pp. 40-69). New York: Praeger Publishers. Ciabattari, T. (2001). Changes in men's conservative gender ideologies: Cohort and period influences. Gender and Society, 15, 574-591. Clarke, L. (2005). Remarriage in later life: Older women's negotiation of power, resources, and domestic labor. Journal of Women and Aging, 17, 21-41. Cohen, J. (1973). Eta-squared and partial eta-squared in fixed factor anova designs. Educational and Psychological Measurement, 33, 107-12. Coltrane, S. (1998). Gender, power, and emotional expression: Social and historical contexts for a process model of men in marriages and families. Mahwah, NJ: Lawrence Erlbaum Associates Publishers. Cooley, C. H. (1902). Human nature and the social order. New York: Scribner. Desrochers, S., Andreassi, J. & Thompson, C. (2002). Identity Theory. Retrieved June 10, 2008, from Boston College, Sloan Work and Family Research Network Web site: http://wfnetwork.bc.edu/encyclopedia_entry.php?id=242/ Diedrick, P. (1991). Gender differences in divorce adjustment. Journal of Divorce and Remarriage, 14, 35-44. 61 Diekman, A., Goodfriend, W., & Goodwin, S. (2004). Dynamic stereotypes of power: Perceived change and stability in gender hierarchies. Sex Roles, 50, 201-215. Fan, P., & Marini, M. (2000). Influences on gender-role attitudes during the transition to adulthood. Social Science Research, 29, 258-283. Fields, J. (2003). America?s families and living arrangements: 2003. Current Population Reports, P20-553. U.S. Census Bureau, Washington, DC. Fowers, B. (1991). His and her marriage: A multivariate study of gender and marital satisfaction. Sex Roles, 24, 209-221. Guilbert, D., Vacc, N., & Pasley, K. (2000). The relationship of gender role beliefs, negativity, distancing, and marital instability. The Family Journal, 8, 124-132. Hetherington, E. M. & Kelly, J. (2002). For better or for worse: Divorce reconsidered. New York: Norton. Hill, L., & Hilton, J. (1999). Changes in roles following divorce: Comparison of factors contributing to depression in custodial single mothers and custodial single fathers. Journal of Divorce and Remarriage, 31, 91-114. Kaufman, G. (2000). Do gender role attitudes matter? Family formation and dissolution among traditional and egalitarian men and women. Journal of Family Issues, 21, 128-144. Kelley, P. (1995). Developing healthy stepfamilies: Twenty families tell their stories. New York: The Haworth Press. 62 King, D. W., Beere, C. A., King, L. A., & Beere, D. B. (1981). A new measure of sex- role attitudes. Paper presented at the annual conference of the Midwestern Psychological Association, Detroit, MI. LaRossa, R. & Reitzes, D. C. (1993). Symbolic interactionism and family studies. In P. G. Boss, W. J. Doherty, R. LaRossa, W. R. Schumm, & S. K. Steinmetz (Eds.), Sourcebook of family theories and methods: A contextual approach. (pp.135- 160). New York: Plenum Press. Larson, L. E. (1974). System and subsystem perception of family roles. Journal of Marriage and the Family, 36, 123-138. Lye, D., & Biblarz, T. (1993). The effects of attitudes toward family life and gender roles on marital satisfaction. Journal of Family Issues, 14, 157-188. Myers, S. M. (2004). Religion and intergenerational assistance: Distinct differences by adult children's gender and parent's marital status. Sociological Quarterly, 45, 67 - 89. Osborne, J. (2008). Best practices in quantitative methods. Los Angeles: Sage. Pasley, K., Kerpelman, J., & Guilbert, D. E. (2001). Gendered conflict, identity disruption, and marital instability: Expanding Gottman's model. Journal of Social and Personal Relationships, 18, 5-27. Rogers, S., & Amato, P. (2000). Have changes in gender relations affected marital quality? Social Forces, 79, 731-753. 63 Roper, B. S., & Labeff, E. (1977). Sex roles and feminism revisited: An intergenerational attitude comparison. Journal of Marriage and the Family, 39, 113-119. Sakraida, T. (2005). Common themes in the divorce transition experience of midlife women. Journal of Divorce and Remarriage, 43, 69-88. Schuchts, R., & Witkin, S. (1989). Assessing marital change during the transition to parenthood. Social Casework, 70, 67-75. Seitz, S., Lord, C., & Taylor, C. (2007). Beyond pleasure: Emotion activity affects the relationship between attitudes and behavior. Personality and Social Psychology Bulletin, 33, 933-947. Smith, R., Goslen, M., Byrd, A., & Reece, L. (1991). Self-other orientation and sex-role orientation of men and women who remarry. Journal of Divorce and Remarriage, 14, 3-32. Stryker, S., & Burke, P. (2000). The past, present, and future of an identity theory. Social Psychology Quarterly, 63, 284-297. Softas-Nall, B., & Sukhodolsky, D. (2006). Family in the United States: Social context, structure, and roles. In J. Georgas, J. Berry, F. van de Vijver, ?. Ka??t??ba??, and Y. Poortinga (Eds.), Families across cultures: A 30-nation psychological study (pp. 483-490). New York: Cambridge University Press. Stone, P., & Lovejoy, M. (2004). Fast-track women and the 'choice' to stay home. The ANNALS of the American Academy of Political and Social Science, 596, 62-83. 64 Stryker, S. (1968). Identity salience and role performance. Journal of Marriage and the Family, 4, 558-564. Stryker, S. (1980). Symbolic interactionism: A social structural version. Menla Park, CA: Benjamin Cummings. Stryker, S. (1987). The vitalization of symbolic interactionism. Social Psychology Quarterly, 50, 83-94. Stryker, S., & Burke, P. J. (2000). The past, present and future of identity theory. Social Psychology Quarterly, 63, 284-297. Stryker, S., & Serpe, R. (1982). Commitment, identity salience and role behavior: Theory and research example. In W. Ickes & E. S. Knowles (Eds.), Personality, Roles, and Social Behavior. New York: Springer-Verlag. Sullivan, C., & Lewis, S. (2006). Work at home and the work-family interface. In F. Jones, R. Burke, & M. Westman (Eds.), Work-life balance: A psychological perspective (pp. 143-162). New York: Psychology Press. Thornton, A. (1989). Changing attitudes toward family issues in the United States. Journal of Marriage and the Family, 51, 873-893. Thornton, R. & Nardi, P. M. (1975). The dynamics of role acquisition. The American Journal of Sociology, 80, 870-885. Wallace, D., Paulson, R., Lord, C., & Bond, C. (2005). Which behaviors do attitudes predict? Meta-analyzing the effects of social pressure and perceived difficulty. Review of General Psychology, 9, 214-227. 65 Walzer, S. (2008). Redoing gender through divorce. Journal of Social and Personal Relationships, 25, 5-21. 66 APPENDIX A MEASURES Gender Role Attitudes Scale (Booth et al., 1980) 67 Strongly Agree Agree Disagree Strongly Disagree Don?t Know Refuse 1. A woman's most important task in life should be taking care of her children. 1 2 3 4 8 9 2. A husband should earn a larger salary than his wife. 1 2 3 4 8 9 3. It should not bother the husband if a wife's job sometimes requires her to be away from home overnight. 1 2 3 4 8 9 4. If his wife works full-time, a husband should share equally in household chores such as cooking, cleaning, and washing. 1 2 3 4 8 9 5. If jobs are scarce, a woman whose husband can support her ought not to have a job. 1 2 3 4 8 9 6. A working mother can establish just as good a relationship with her children as a mother who does not work. 1 2 3 4 8 9 7. Even though a wife works outside the home, the husband should be the main breadwinner and the wife should have the responsibility for the home and children. 1 2 3 4 8 9 68 APPENDIX B CHARTING MARITAL STATUS Possible Marital Status Trajectories 69 1980 1983 1988 1992-1994 1997 2000 Divorced (1 st Marriage) Divorced (1 st Marriage) Divorced (2 nd Marriage) First Marriage 3 rd Marriage 2 nd Marriage Divorced (1 st Marriage) Divorced (2 nd Marriage) First Marriage 3 rd Marriage 2 nd Marriage Divorced (1 st Marriage) Divorced (2 nd Marriage) First Marriage 3 rd Marriage 2 nd Marriage Divorced (1 st Marriage) Divorced (2 nd Marriage) First Marriage 2 nd Marriage 2 nd Marriage Divorced (3 rd Marriage) Divorced (3 rd Marriage) First Marriage First Marriage