Gender role attitudes: An examination of within-individual malleability and the value of dyadic congruence by Mallory Joy Lucier-Greer A dissertation submitted to the Graduate Faculty of Auburn University in partial fulfillment of the requirements for the Degree of Doctor of Philosophy Auburn, Alabama August 4, 2012 Keywords: attitudinal congruence, divorce, gender role attitude malleability, gender role attitudes, relationship education, remarriage Copyright 2012 by Mallory Joy Lucier-Greer Approved by Francesca Adler-Baeder, Chair, Professor of Human Development and Family Studies Jennifer Kerpelman, Professor of Human Development and Family Studies Joe Pittman, Professor of Human Development and Family Studies Amy Rauer, Assistant Professor of Human Development and Family Studies ii Abstract Suggestions from structural symbolic interactionism and recent empirical studies are that gender role attitudes are malleable ? able to change and be redefined ? at the individual-level in relation to significant experiences. This dissertation explored the process of gender role attitude malleability and the implications of attitudinal malleability in the following studies. The first study examined how divorce and remarriage affect within-individual gender role attitude growth across a 20-year period. Longitudinal analyses (N = 1,731) found that marital status transitions uniquely affect gender role attitude change patterns. Remaining in a first marital relationship was related to a slight, but significant decline in traditional gender role attitudes. The experience of divorce was associated with a slight acceleration in the decline in traditional attitudes, while the experience of remarriage was associated with a slower rate of decline in traditional attitudes. As such, transitions out of marriage were related to more egalitarian attitudes, but transitions into remarriage were associated with attitudinal shifts similar to being in a first marriage. Preliminary evidence is also presented related to the possibility that attitudinal shifts towards more egalitarian attitudes may precipitate divorce for some. Implications for theoretical refinement are given. The second study sought to simultaneously evaluate three theories that hypothesize about the link between gender role attitudes and couple quality ? the companionate theory of marriage, the gender theory of marriage, and a dyadic attitudinal congruence model. Data from couples (N = 383 couples) engaged in couple and relationship education (CRE) provided a context to iii evaluate how gender role attitudes and changes in couple-level attitudinal congruence influence relational outcomes. Using structural equation modeling and the actor-partner interdependence model (APIM), findings indicate no ?best? type of gender role attitude (either traditional or egalitarian) for predicting couple quality for men and women. Rather, greater gender role attitude congruence between couples predicts increased intimacy for the husband/male partner. In turn, higher levels of male intimacy predict increased perceptions of couple quality for both partners. Tests of moderation reveal that findings were similar, regardless of race. Theoretical and practical implications can be drawn as this study empirically evaluates several theoretical perspectives and provides insight for practitioners and the content of CRE. iv Acknowledgements I have learned over the past year that it takes a village?to write a dissertation. The completion of this dissertation would not have been possible without the help, support, encouragement, and push of those around me. First, I want to thank my family. To my parents, Laura Luecke and Geoffrey Lucier: thank you for encouraging me to follow my dreams and reminding me take breaks from work and enjoy life. To my sister, Jessica Smith: thank you for being my biggest cheerleader and my very best friend. To my husband, Adam Greer: words cannot adequately thank you for walking with me during this journey. I truly consider this ?our? product, because it would not have happened without you; we did it! I also want to thank my ?professional family? who has continually provided me with methodological assistance, writing suggestions, and all around professional development support. To my committee, Dr. Jennifer Kerpelman, Dr. Joe Pittman, and Dr. Amy Rauer: thank you for your thoughtful feedback and insights and for pushing me to critically evaluate my own work. Finally, to my major professor, Dr. Francesca Adler-Baeder: I hope that you know that you have made a forever impact on my life personally and professionally. Thank you for taking a chance on me six years ago and investing endless energy and resources into my development. Launching from Auburn University comes with mixed emotions as I feel like you have prepared me to enter the workforce, but it is hard to imagine you not being part of my everyday working interactions. Your mentorship inspires me, and I hope to make you proud now and in the future. v Table of Contents Abstract ......................................................................................................................................... ii Acknowledgements ..................................................................................................................... iv List of Table ............................................................................................................................... vii List of Figures ........................................................................................................................... viii Chapter 1 ? General Introduction ................................................................................................. 1 Chapter 2 ? Gender role attitude malleability in the context of marital status change: A longitudinal growth curve perspective ............................................................................ 12 Introduction ................................................................................................................... 13 Method ........................................................................................................................... 24 Results ............................................................................................................................ 28 Discussion ...................................................................................................................... 35 Chapter 3 ? Dyadic gender role attitude congruence as a predictor of couple quality in the context of coupe and relationship education .................................................................. 46 Introduction ................................................................................................................... 47 Method ........................................................................................................................... 58 Results ............................................................................................................................ 63 Discussion ...................................................................................................................... 69 Chapter 4 ? General Discussion .................................................................................................. 81 References ................................................................................................................................. 85 Appendix A ............................................................................................................................... 88 Appendix B ............................................................................................................................... 89 Appendix C ............................................................................................................................... 90 vi Appendix D ............................................................................................................................... 91 vii List of Tables Study 1 Table 1 ....................................................................................................................................... 26 Table 2 ....................................................................................................................................... 33 Table 3 ....................................................................................................................................... 34 Table 4 ....................................................................................................................................... 35 Study 2 Table 1 ....................................................................................................................................... 65 Table 2 ....................................................................................................................................... 68 Table 3 ....................................................................................................................................... 69 viii List of Figures Study 1 Figure 1 ...................................................................................................................................... 31 Study 2 Figure 1 ...................................................................................................................................... 58 Figure 2 ...................................................................................................................................... 64 1 I. General Introduction ?Gender is not merely something that happens in the nooks and crannies of interaction, fitted in here and there and not interfering with the serious business of life? (West & Zimmerman, 1987, p. 130). Twenty five years ago West and Zimmerman challenged us to recognize the complex influence that gender has on our lives and in our relationships. Since that time, myriad studies have been conducted to examine the ways in which gender influences personal and interpersonal functioning. One way studies operationalize gender is by measuring personal interpretations of the appropriate roles of men and women ? also known as gender role attitudes. These attitudes are most often conceptualized as ranging on a continuum from traditional to egalitarian (e.g., Beere, King, Beere, & King, 1984; Davis & Greenstein, 2009). Those with traditional attitudes tend to endorse beliefs related to the unique and separate roles of men and women and focus on the differences between the sexes, while those with egalitarian attitudes support notions of shared roles and responsibilities and emphasize common understanding and functioning between the sexes. As hypothesized by West and Zimmerman (1987), research has found that gender role attitudes influence a variety of individual and interpersonal contexts. For example, women with traditional gender role attitudes report poorer mental health, namely anxiety and depression (Read & Grundy, 2011), slightly lower workplace earnings (Judge & Livingston, 2008), lower age at first motherhood (Stewart, 2003), but also higher levels of couple quality (Amato & Booth, 1995); egalitarian attitudes are associated with higher wages and working more hours 2 (Corrigall & Konrad, 2007), lower intentions of having a child (Kaufman, 2000), and a decreased likelihood of actually having children (Kaufman, 2000). For men, traditional gender role attitudes are associated with higher earnings (Judge & Livingston, 2008); egalitarian attitudes are related to higher levels of couple quality (Bowen & Orthner, 1983), more frequent participation in household tasks (Cunningham, 2005), greater intentions to have children (Kaufman, 2000), and higher levels of parental involvement (Gaunt, 2006). From these studies, it seems that gender role attitudes may have different effects depending on the context (i.e., inside or outside the home) and the sex of the individual. In terms of the couple relationship, competing hypotheses and mixed findings exist related to the influence of gender role attitudes on relational quality and stability (Davis & Greenstein, 2004). Some support has been found for the gender theory of marriage which proposes that more traditional gender role attitudes are related to better relationship functioning (e.g., Amato & Booth, 1995; Wilcox & Nock, 2006), and, conversely, support also exists for the companionate theory of marriage specifying that egalitarian attitudes are related to better relational outcomes (e.g., Amato, Johnson, Booth, & Rogers, 2003; Juni & Grim, 1994). While these theories have different premises, both seem to suggest that who you are married to matters. In other words, the combination of attitudes across partners appears to influence couple outcomes. It should be noted that couple outcomes, such as relational quality and stability, imply a couple-level analysis, yet to date, most studies examine independent spousal reports of the relationship between gender role attitudes and relational outcomes (e.g., Amato & Booth, 1995; Kaufman & Taniguchi, 2006). These studies are informative, but do not speak to the interactional nature that gender role attitudes are thought to have between partners in a couple 3 relationship (West & Zimmerman, 1987) nor do they respond to the call for researchers to move beyond independent reports to understand the dyadic interplay of gender role attitudes on couple functioning (Ferree, 1990; Thompson, 1993). Researchers within the last two decades have begun to theorize and evaluate the ways in which gender role attitudes influence personal and interpersonal outcomes as well as the ways in which changes in one?s social context may alter gender role attitudes. Roles are conceptualized by symbolic interactionism as transactional or negotiated concepts between individuals and their social structures; accordingly, changes in roles and role-related attitudes are predicted to occur across the life course as a function of new experiences and interactions (Kelley, 1995; Stryker, 2008). A small, but growing literature substantiates this assumption. Specifically, gender role attitudes have been found to be generally stable constructs across time; yet, role changes and significant life experiences appear to provide individuals the opportunity to re-conceptualize their gender role attitudes (Fan & Marini, 2000). Life experiences, such as attending college (Bryant, 2003), entry into and exit from the workforce (Fan & Marini, 2000; Vespa, 2009), entry into marriage and subsequent marital status transitions (Fan & Marini, 2000; Lucier-Greer & Adler-Baeder, 2011; Vespa, 2009), the transition to parenthood (Fan & Marini, 2000; Katz-Wise, Priess, & Hyde, 2010), and participating in a couple and relationship education class (Lucier- Greer, Adler-Baeder, Ketring, & Smith, 2011), have all been linked to subsequent attitudinal change. Studies that examine gender role attitude malleability as a function of significant life experiences serve to promote theoretical refinement related to gender socialization and inform the literature on gender role attitude malleability in adulthood (Moen, Kelly, & Magennis, 2008). This work has important implications in challenging the notion that gender role attitudes are 4 stable constructs following childhood, and research of this nature would extend the cultural explanation of gender development (Maccoby & Jacklin, 1974) by providing evidence that gender is learned throughout the life course. Because gender role attitudes influence individual and interpersonal outcomes, research on gender role development opens the door to examining how changes in attitudes may influence changes in individual and relational functioning. The studies in this dissertation examined the continued socialization process of gender role attitudes within individuals across adulthood and the dyadic interplay of how changes in gender role attitudes influence perceptions of couple quality ? two relatively uncharted areas of study. Gender Socialization throughout the Lifespan Gender role attitudes represent complex individual-level interpretations of societal messages and interpersonal experiences about the appropriate roles of men and women (Renk, Donnelly, McKinney, & Agliata, 2006). The process of synthesizing external cues related to gender into internal, useable frameworks is known as gender socialization (Moen et al., 2008). Gender socialization is thought to begin early in childhood as children are readily exposed to gendered information and rules about appropriate behaviors according to one?s sex (Renk et al., 2006). In fact, the majority of theories on gender role socialization focus on the formative years in childhood. Social cognitive theory emphasizes the ways in which parents directly and indirectly teach their children about appropriate gender role behavior (Bussey & Bandura, 1999); cognitive developmental theory illustrates how cognitive milestones such as gender labeling and gender constancy serve as critical organizing factors in social development and lay the foundation for value-based assessments, attitudes, and behaviors pertaining to gender roles (Kohlberg, 1966); and finally, gender schema theory describes how gender schemas permit the individual to anticipate and make judgments about their environment based on gender-associated 5 information gathered from prior experiences (Bem, 1981). These theories each posit that experiences teach children about gender roles, and experiences both create an understanding of and reinforce culturally acceptable messages related to gender roles. Messages about appropriate sex-typed roles and behaviors, then, contribute to one?s gender identity or personal conceptualization of self as either a male or female (Bussey, 2011). Gender identity is expressed interpersonally via gender role attitudes, which summarize one?s beliefs about appropriate behaviors, responsibilities, and interactions of men and women (Renk et al., 2006). Scholars conceptually frame and measure gender role attitudes on a continuum from a traditional perspective which endorses the unique roles of and differences between men and women to an egalitarian perspective which endorses similarities and shared roles across the sexes (e.g., Larsen & Long, 1988). In other words, ?an attitude that causes one to respond to another individual independently of the other individual?s sex? is thought to be egalitarian in nature, while responses based on a person?s sex would be more traditionally based (Beere et al., 1984, p. 564). This continuum has been used since the 1970?s (e.g., Kalin & Tilby, 1978; Parelius, 1975), although the term egalitarian was not regularly applied until the 1980?s (e.g., Beere et al., 1984). Before then, the term feminist perspective was used to describe the endorsement of shared roles and equal responsibilities. Current Research on Gender Role Attitudes Many studies that examine gender role attitudes utilize a sociological perspective to understand shifts in gender role attitudes across socio-historical time and across cohorts (e.g., Ciabattari, 2001; Rogers & Amato, 2000). Research has found evidence of both period effects and cohort effects in examining gender role attitudes across time; period effects are thought to be influential times that influence all of society simultaneously, and cohort effects emphasize 6 distinct characteristics of individual groups based on their specified life stage (Ciabattari). For example, the feminist movement of the 1970?s was found to have a significant influence on all people that experienced it (period effect), but age-based cohorts appeared to vary in their attitudinal responses to this movement (cohort effect). Rogers and Amato (2000) also report cohort shifts as gender role attitudes differed significantly for two cohorts of married couples, such that the younger group of married persons reported more egalitarian attitudes, more equality in relationship functioning and the division of labor inside the home, and higher rates of both sexes working outside the home. Fewer studies have examined gender role attitude change from a microsociological perspective which emphasizes how dyadic and relational interactions create unique experiences at the individual-level of development (Desrochers, Andreassi & Cynthia, 2002). In fact, research on gender role attitude development at the individual-level is sparse beyond the childhood years. It would seem that gender socialization persists into adulthood as individuals continue to be inundated by external cues related to ?doing gender? via interpersonal interactions and societal messages. Yet, only within the last few years have theoretical developments begun to hypothesize about gender socialization at the individual-level in adulthood (e.g., Moen et al., 2009). Moen and colleagues call researchers to examine the ecology of adult development, which can also be thought of as the relationship between the person and their environment. Kalmijn (2005) used this framework of continued gendered socialization to demonstrate that over the course of marriage, couples tend to align in their gender role attitudes. Thus, the marital environment seemed to have an influence on personal perceptions of gender roles. Theoretical and empirical advancements, such as those discussed, suggest that adult development is rooted in 7 child and adolescent development, but that experiences in adulthood continue to create meaning and shape socialization ? specifically, gender socialization ? throughout the life course. Symbolic Interactionist Perspective on Gender Role Attitude Development The theoretical foundation for this dissertation was rooted in symbolic interactionism and the notion that meaning is a social construction (Blumer, 1969). This theory proposes that the meaning ascribed to a stimulus directs an individual?s interactions with regards to that stimulus. It also assumes that meaning may be altered to adapt to or in light of new experiences (Serpe, 1987). Thus, the meaning ascribed to a particular stimulus is not static. Instead, it can be modified through a process of interacting with one?s environment. These tenets of symbolic interactionism led to the development of theories that detail how aspects of one?s self are socially created. First, structural symbolic interactionism posits that there is an interconnected relationship between self and environment (Serpe, 1987). Similarly, identity theory suggests that it is not possible to understand self apart from one?s social structure (Wells & Stryker, 1988). Serpe (1987) describes this relationship between self and social structure by noting that there is ?both relative constancy in the structure of the self, given absence of movement within the social structure, and relative change in the structure of the self, given such movement? (p. 44). Social structure is conceptualized as all of the tangible and intangible factors that make up one?s environment (e.g., interpersonal and romantic relationships, the cultural macrosystem, etc.). Therefore, the self ? which is defined by identity theory as encompassing affective, conative, and cognitive aspects of an individual (Stryker, 1968; Stryker, 2008) ? may be modified through the process of interacting with others and society. As such, one?s gendered self is also thought to be developed through relevant social interactions (Bussey, 2011) and is expressed interpersonally via gender role attitudes. Structural 8 symbolic interactionism and identity theory serve to inform the study of gender role attitudes by first recognizing that development can and does occur across the life course and is not limited to childhood. Although the majority of theories and empirical studies related to gender role attitude development emphasize developmental stages that occur in childhood (e.g., Bussey & Bandura, 1999; Kohlberg, 1966), symbolic interactionism and its derivatives posit that development and socialization are continual, interpretative processes that occur throughout adult development (Serpe, 1987; Wells & Stryker, 1988). More specifically, suggestions are that gender role attitudes may be altered in response to changes in one?s social structure which may include novel social interactions and/or significant life events. Additionally, structural symbolic interactionism and identity theory suggest that new meaning may lead to changes in interpersonal functioning (Serpe, 1987). This change is thought to be related to changes in personal identities (Stryker & Wells, 1988). As life events alter our identities and our relative identity salience (Stryker, 1980), our roles and positions within organized social relationships also change (Stryker, 2008). Specifically, if one?s gender role attitudes were to be altered due to changes in one?s social structure, changes in functioning would also be expected. Thus, it may be that gender role attitude development is cyclical in nature, i.e., with social interactions influencing gender role attitudes and gender role attitudes influencing subsequent social interaction. Furthermore, tenets of symbolic interactionism provide insight into the means by which gender role attitudes may influence the couple relationship. Several studies have hypothesized about the correct ?type? of gender role attitude for each individual ? either egalitarian (e.g., Amato et al., 2003) or traditional (e.g., Dizard & Gadlin, 1990) ? as a predictor of higher levels of couple quality, but suggestions are that the link between gender role attitudes and couple 9 quality may take place at the dyadic level as individual and partner effects influence perceptions of couple quality (Thompson, 1993; West & Zimmerman, 1987). This theory posits that meaning is not created in a vacuum void of influence from interpersonal experiences, but rather that meaning is co-created and shared with significant others. Likewise, gender role attitudes can be thought of as co-created and renegotiated in relation to interpersonal interactions including the couple?s unique history and experiences (Gottman & Gottman, 2008). Rather than a single, prescribed gender role attitude type predicting couple quality, it may be that meaning lies in the interaction or level of similarity across the individuals in the couple relationship. Higher levels of shared meaning/similarity ? which can be conceptualized as dyadic gender role attitude congruence ? may be highly relevant in understanding the relationship between gender role attitudes and couple quality. To date, less than a handful of studies could be found that examined couple functioning in relation to dyadic gender role attitude congruence (e.g., Bowen & Orthner, 1983; Juni & Grim, 1994; Lye & Biblarz, 1993) The Current Studies This dissertation extends beyond the research assessing the effects of individual gender role attitudes to explore (1) gender role attitude development in adulthood and (2) dyadic assessments of gender role attitudes within couples and elements of the process by which attitudes influence the couple relationship. The first study examined prospectively how gender role attitudes changed in relation to significant life events, specifically marital status transitions. Based on assumptions from structural symbolic interactionism and identity theory, hypotheses were that gender role attitudes would be malleable ? able to change and be refined ? across the course of adult development (Bussey, 2011). These theories posit that novel social interactions and experiences (e.g., divorce and remarriage) provide individuals the opportunity to reexamine 10 and alter their gender role attitudes (Serpe, 1987; Wells & Stryker, 1988). This hypothesis is supported by a handful of recent empirical studies that have found links between significant life events and gender role attitude malleability (e.g., Bryant, 2003; Fan & Marini, 2000; Lucier- Greer & Adler-Baeder, 2011; Vespa, 2009). Specifically, this study inquired about the ways in which gender role attitudes change in relational to marital status transitions. The expectation was that transitions into and out of marriage would be related to subtle, but noteworthy refinements in personal gender role attitudes as one?s social structure and perceived identity is altered. Transitions out of marriage and away from marital roles was hypothesized to be related to decreases in traditional attitudes, while reentry into marriage was hypothesized to increase traditional gender role attitudes as individuals choose to re-partner (Lucier-Greer & Adler- Baeder, 2011). Individual growth models were fitted using data from participants initially in first marital relationships using six waves of data gathered across 20 years to explore the influence of divorce and remarriage on gender role attitude malleability. Building on early evidence and the notion that gender role attitudes are malleable, the second study sought to understand how changes in gender role attitudes influence changes in couple quality. Competing models ? the companionate theory of marriage, the gender theory of marriage, and a dyadic gender role attitude congruence model ? were examined to understand the relationship between gender role attitude change and perceptions of couple quality. The companionate and gender theories of marriage posit a direct relationship between gender role attitudes and couple quality. The companionate theory of marriage suggests that egalitarian beliefs are related to higher levels of marital quality (Wilcox & Nock, 2006), thus it was hypothesized that increases in egalitarian attitudes would be related to increases in marital quality. Conversely, the gender theory of marriage posits that traditional beliefs are related to 11 higher levels of couple quality (Dizard & Gadlin, 1990; Wilcox & Nock, 2006). According to this theory, it was hypothesized that increases in traditional attitudes would be predictive of increases in marital quality. Alternatively, a dyadic attitudinal congruence model based on tenets of symbolic interactionism suggests that the relationship between gender role attitudes and marital quality is related to how similar or congruent gender role attitudes are between partners (e.g., Bowen & Orthner, 1983). It may be that there is a direct relationship between increases in congruence and increases in marital quality as proposed by the literature on relational similarity (e.g., Fitzsimons & Anderson, 2011), but the central hypothesis of this study was that the relationship was expected to be mediated by increases in shared experience, operationalized as intimacy and closeness. Thus, as attitudes become more congruent between couples, a general feeling of connection was expected to increase, which, in turn, was hypothesized to lead to increases in couple quality. Each of these models were simultaneously fitted using structural equation modeling to assess which pathways best predict how gender role attitudes influence couple quality. 12 II. Paper 1 ? Within-individual gender role attitude malleability in the context of marital status change: A longitudinal growth curve perspective Suggestions from structural symbolic interactionism and recent empirical studies are that gender role attitudes are malleable ? able to change and be redefined ? at the individual-level in relation to significant experiences. This study examined how divorce and remarriage affect within- individual gender role attitude growth across a 20-year period. Longitudinal analyses (N = 1,731) found that marital status transitions uniquely affect gender role attitude change patterns. Remaining in a first marital relationship was related to a slight, but significant decline in traditional gender role attitudes. The experience of divorce was associated with a slight acceleration in the decline in traditional attitudes, while the experience of remarriage was associated with a slower rate of decline in traditional attitudes. As such, transitions out of marriage were related to more egalitarian attitudes, but transitions into remarriage were associated with attitudinal shifts similar to being in a first marriage. Preliminary evidence is also presented related to the possibility that attitudinal shifts towards more egalitarian attitudes may precipitate divorce for some. Implications for theoretical refinement are given. Keywords: divorce, gender role attitude malleability, remarriage, symbolic interactionism 13 Within-individual gender role attitude malleability in the context of marital status change: A longitudinal growth curve perspective Gender role attitudes are personal conceptualizations about the roles of men and women in the context of society and interpersonal relationships (Larsen & Long, 1988). These attitudes are measured on a continuum from a traditional perspective that emphasizes the unique roles of men and women to an egalitarian perspective that underscores shared roles and equality between the sexes (Beere, King, Beere, & King, 1984; Larsen & Long, 1988). Researchers most often conceptualize and measure gender role attitudes as a stable construct thought to be consistent across adulthood, yet suggestions from symbolic interactionism as well as a few recent empirical studies (e.g., Fan & Marini, 2001; Lucier-Greer & Adler-Baeder, 2011; Vespa, 2009) are that gender role attitudes are malleable ? able to change and be refined ? in the context of significant adult experiences. Gender role attitude malleability has significant implications for theoretical refinement of current gender socialization theories and opens the door for empirical research on the potential consequences of attitudinal change on aspects of intra- and interpersonal functioning. This study serves to further current research on gender role attitude malleability by longitudinally examining individual patterns of gender role attitude change following significant life events, specifically divorce and remarriage. Growth curve models were fitted using six observations obtained across a 20-year period to explore if and how gender role attitude change occurs in relation to these transitions out of and back into the institution of marriage. Background on Gender Role Attitude Malleability To date, the majority of research addressing gender role attitude malleability has utilized a sociological framework in which the researchers gauge attitudinal change across age-related 14 cohorts (e.g., Ciabattari, 2001). Findings indicate that gender role attitudes are becoming more egalitarian across younger cohorts as these individuals tend to endorse higher levels of equality in terms of decision making, household responsibilities, and labor force participation when compared to past generational cohorts (Brooks & Bolzendahl, 2004; Rogers & Amato, 2000). These macro-level changes in gender role attitudes are often attributed to sociological influences including historical and cultural events (e.g., the feminist movement; Ciabattari) While there appears to be strong evidence to suggest sociological shifts in gender role attitudes, far less evidence exists that substantiates gender role attitude malleability at the micro- or individual-level of development, particularly development that occurs in adulthood. Theories that describe the socialization process of gender roles and studies that examine the development of gender role attitudes at the individual-level tend to focus on the formative years. For example, social cognitive theory suggests that gender development is shaped by parental figures who send messages about appropriate gender roles to their children indirectly via modeling acceptable interactions and directly when they sanction certain gendered behaviors (Bussey & Bandura, 1999). Cognitive developmental theory notes that gender-based cognitive milestones in childhood (e.g., gender labeling and gender constancy) serve as critical organizing factors in social development and lay the foundation for value-based assessments, attitudes, and behaviors pertaining to gender roles (Kohlberg, 1966). Finally, gender schema theory posits that gender- based information is rooted in enduring schemas developed from learned societal and interpersonal expectations beginning in childhood; these schemas allow the individual to anticipate and make judgments about their environment based on gender-associated information gathered from prior experiences (Bem, 1981). 15 Because the study of gender role attitude change has occurred predominantly at the macro-level and the theories that do consider individual-level development focus on the formative years of childhood, gender role attitude development has most often been operationalized as static in adulthood (e.g., Kaufman & Taniguchi, 2003). It has only been since the early 2000?s that researchers began to explore and test theories of continued socialization beyond the childhood years and within the last five years these theories have been applied to gender socialization. For example, Arnett (2007) and Fingerman and Pitzer (2007) examined tasks of socialization and influences on socialization in emerging adulthood and later life, respectively. Their suggestions were that individuals continue to be confronted with novel interpersonal circumstances and continue to experience societal changes throughout the life course, thus personal adaptation was seen as essential. More specifically, Moen, Kelly, and Magennis (2008) applied this concept of adult socialization to examine gender role attitude development across the life course suggesting that individuals lead inextricably linked lives to their environment ? both the physical environment and to a web of interpersonal relationships. They propose that individuals do not follow linear, passive paths of development. Rather, unique experiences and one?s personal agency to intentionally select into and out of roles contributes to development in general and gender development, specifically, as one?s context continues to evolve over time. Thus, it seems that developmental theories of childhood address initial processes of gender socialization, but socialization appears to continue in the context of adult development. This premise is supported by principles of structural symbolic interactionism. This theory suggests that development is a continuous process of assessing and redefining self and others based on role changes, influential life events, and/or interactions with others (Blumer, 1969; 16 Serpe, 1987). Social interactions are thought to create personal and shared meaning. Accordingly, individuals, then, apply and act in a way that is congruent with the meaning ascribed to a particular event or interaction. Furthermore, meaning can be modified through a process of adapting to situations the individual encounters. It becomes a continuous and iterative cycle as meaning influences one?s social interactions, and new social interactions alter relevant meaning (Serpe, 1987; Wells & Stryker, 1988). Accordingly, context and personal development are, thus, inextricably related as one influences the other. Utilizing an interactionist perspective, it is expected that gender role attitudes do not remain constant in adulthood. Rather, gender role attitudes are theorized to be malleable ? able to change and be refined ? as socialization processes continue (e.g., Arnett, 2007; Fingerman & Pitzer, 2007; Moen et al., 2008) and individuals reevaluate their roles and perceptions of self throughout the life course and specifically, following meaningful life events. Emerging Evidence of Gender Role Attitude Malleability Although frameworks of gender role attitude development suggest that attitudes are malleable, only a handful of studies to date have examined attitudinal change in adulthood. Fan and Marini (2000) apply these frameworks of attitudinal malleability in an empirical study of young adults using the National Longitudinal Survey of Youth. Findings suggest evidence of both stability and change in gender role attitudes. Attitudes remained considerably stable when measured three times over an eight year timeframe, yet change in gender role attitudes did occur as a function of ?specific socializing experiences? (Fan & Marini, p. 277). In general, social experiences related to transitions or events ?inside the home,? which can also be thought of as intrafamilial experiences (e.g., marriage, the transition to parenthood, and exit from the workforce) tended to prompt more traditional attitudes especially for women. Yet, events related 17 to experiences ?outside the home? or extrafamilial experiences (e.g., entry into the workforce) were related to shifts toward more egalitarian attitudes for women. These socializing experiences appear to prompt previously adopted attitudes to be altered in the context of new information and novel roles. Bryant (2003) explored the process of gender role attitude change related to attending college using Astin?s (1993) Input-Environment-Output model. This model proposes that the ?environment? represents some aspect of contextual change, and contextual change is hypothesized to generate attitudinal malleability (the ?output?) accounting for initial levels (the ?input?). She found that the ?environment? of college (an extrafamilial context) led to changes towards more egalitarian gender role attitudes (the ?output?) for both men and women. While gender role attitudes were only measured at two time points during their first and fourth years of college, this does provide some additional evidence of gender role attitude malleability within individuals after a significant life experience. Corrigall & Konrad (2007) also examined gender role attitude change in relation to workforce participation in a sample of high school seniors who were followed longitudinally for 14 years via the Monitoring the Future Study. Using four data collection points, they found that engaging in more work hours was predictive of changes toward more egalitarian attitudes, while having children predicted later shifts toward more traditional attitudes particularly for women. Similar to the findings of Fan and Marini (2000), extrafamilial experiences (e.g., increased work hours and time outside the home) were related to more egalitarian attitudes, while those related to intrafamilial (e.g., having children) experiences promoted more traditional attitudes particularly for women. 18 Finally, Vespa (2009) found that new life experiences such as marriage, parenthood, and workforce participation altered gender role attitudes uniquely for men and women from different racial backgrounds. Using data collected at two time points in 1979 and 2004 from the National Longitudinal Survey of Youth, he found that attitudes did not remain stable in relation to these life events. For African American participants, entry into marriage was associated with changes toward more egalitarian attitudes, while it was associated with more traditional attitudes for European American participants. The addition of children to a family was subsequently associated with more traditional attitudes for African American men and European American women, but more egalitarian attitudes for African American women. In terms of workforce participation, increases in full time work were associated with shifts towards more traditional attitudes for men of both racial backgrounds, but no changes for women. Vespa concluded that gender role attitudes undergo an ?organic construction process? (p. 365) which allow for refinement of attitudes when one?s social interactions change and/or novel life events occur. Interestingly, these findings did not seem to follow the pattern of gender role attitude change found in previous studies (e.g., transitions related to intrafamilial experiences were predictive of more traditional attitudes and transitions related to extrafamilial experiences were predictive of more egalitarian attitudes) when racial differences where teased out. From these studies, it seems essential to take into account the dynamic nature of gender role attitudes as life experiences appear to influence gender-based attitudes. Furthermore, clarifying patterns of attitudinal change relative to the gendered nature of the experience may provide more understanding the gender socialization process. Thus, this study examined gender role attitude malleability related to marital status transitions. In other words, what effect does the 19 transition to being ?outside? the marriage (i.e., divorce) have on gender role attitudes and what effect does a transition back ?into? marriage (i.e., remarriage) have on attitudes? Marital status transitions. In general, Americans from various racial and economic backgrounds report to highly value marriage (Cherlin, 2004; Edin & Reed, 2005) and approximately 85% will marry at some point in their lives (Krieder & Ellis, 2011). Yet, marital dissolution rates in the United States remain higher than many other industrialized nations (Krieder & Ellis, 2011). Census data suggests that 43% of marriages end in divorce or separation within 15 years (Bramlett & Mosher, 2001). Research suggests that this period of adjustment is a challenging time for some and has been found to be related to decreases in well- being, higher levels of negative affect, more instances of psychopathology, and increases in physiological symptoms (Krumrei, Coit, Martin, Fogo, & Mahoney, 2007). Of those who divorce, approximately 75% will eventually remarry. Remarriage can be associated with economic benefits (Ozawa & Yoon, 2002) and improvements in physical and emotional functioning (Sweeney, 2010) compared to those who remain single, but it is also associated with complexities related to a lack of cultural norms, faulty expectations, and challenges related to stepfamily dynamics (Dupuis, 2007). As such, there is reason to believe that the experiences of divorce and remarriage could be considered significant socializing experiences as these transitional periods influence intra- and interpersonal outcomes. Research can inform our understanding of the ways in which marital status transitions influence personal conceptualizations of gender roles as attitudinal changes have important implications for other intra- and interpersonal outcomes. The majority of the work in this area ? gender role attitude malleability following marital status change ? is exploratory in nature and predominantly consists of retrospective and 20 qualitative analyses. Sakraida (2005) and Walzer (2008) conducted in-depth interviews with divorced individuals about the transition from marriage to divorce/singlehood. Suggestions from these studies were that divorce was associated with the development of new roles and novel responsibilities (e.g., solo homeowner and independent decision maker; Sakraida), and it seems that these new experiences were accompanied by a period of actively reevaluating gender roles and moving towards less gendered differentiation (Walzer). Walzer concludes that ?gendered processes, relatively unrecognized as married people enact them day to day, are identified and revised in retrospective accounts of divorced people? giving individuals an opportunity to ?redo? their conceptualizations of gendered expectations following divorce (p. 5). Suggestions are that gender role attitudes shift toward more egalitarian as divorced individuals critique the traditional roles of marriage and experience life ?outside? of the couple relationship. Additionally, retrospective qualitative work with individuals in remarriages also details a process of reevaluating gendered attitudes (Clarke, 2005; Smith, Goslen, Byrd, & Reece, 1991). Smith and colleagues interviewed individuals to explore their remarital experiences. They found that those in remarriages described their relationship as comparatively more egalitarian than their first marriage and most reported that their decision to remarry was based more on egalitarian principles ? taking into account both individual and partner needs (i.e., they were intentionally seeking a relationship that was more egalitarian). Similarly, Clarke found that older women in remarital relationships described their first marriages as more traditionally gendered with regard to their attitudes and to the actual division of labor. For those who divorced versus those who were widowed, greater equality is cited by participants as something they sought in their remarriage and a primary reason for being more satisfied in their remarital relationship. 21 One longitudinal, quantitative study exists that supports the notion that gender role attitudes are redefined following divorce and remarriage, yet the direction of change revealed patterns not previously captured in qualitative work (Lucier-Greer & Adler-Baeder, 2011). In this study, three time points were examined over a 20-year period. Participants were grouped based on their marital status at the conclusion of the study, so the timing of the marital status change could not be specified. Instead, the middle time point served as a proxy for when marital status change occurred. Findings were that individuals who remained continuously married over 20 years reported slight, but significant shifts toward more egalitarian attitudes similar to sociological studies of period effects in gender role attitudes over time (Ciabattari, 2001). Those who divorced and remained single reported even greater shifts toward more egalitarian gender role attitudes. Finally, individuals who divorced and, subsequently, remarried were found to have a curvilinear pattern of change. These individuals reported a shift toward more egalitarian gender role attitudes at the mid-point of the study, then a change towards more traditional attitudes. At the conclusion of the study, individuals in remarital relationships reported similar gender role attitudes compared to those who remained in first marital relationships. Lucier-Greer and Adler-Baeder (2011) posit that qualitative reports from remarried individuals suggesting that remarital relationships were more egalitarian than first marital relationships (Clarke, 2005; Smith et al., 1991) may represent accurate perceptions of shifts toward more egalitarian attitudes, but longitudinal analyses revealed that continuously married individuals would likely have reported a similar shift over that 20-year period. The longitudinal findings suggest that the beginning and ending ?snapshots? of gender role attitudes appear similar for those in first marriages and remarriages; yet, the difference between those in first marital and remarital relationships lies in the change pattern over time. Interestingly, these 22 patterns of change were similar for men and women. This is noteworthy, because studies fairly consistently find attitudinal differences between men and women, such that women tend to report more egalitarian beliefs than men (e.g., Fan & Marini, 2000). Similar attitudinal change patterns between men and women suggest that proximal life events (e.g., both experience a marital status transition) may be a more salient influence on gender role attitudes than one?s sex, but as previously noted this did not hold true for other studies. From this, it appears that retrospective measurements cannot fully capture attitudinal change patterns. Yet, even current longitudinal assessments of change lack assessment frequency and measurement sophistication to detail the timing and more nuanced patterns of gender role attitude malleability. To date, the only longitudinal assessment of attitudinal malleability following marital status change used broad group profiles of married individuals that forced participants into pre-assigned groups based on their marital status at the end of the study. This type of examination provides only a crude estimate of change patterns. A study that utilizes more frequent assessments over time and allows for individual differences to emerge based on one?s unique timing of the marital status transition can serve to more accurately capture gender role attitude change patterns of individuals with differing marital histories. The Current Study This study serves to substantiate current theoretical conceptualizations of gender role attitude malleability in adulthood and further the empirical literature on attitudinal change in relation to marital status transitions. Hypotheses drawn from symbolic interactionism are that significant life events influence gender role attitude change as individuals reevaluate and refine their gendered attitudes to fit current experiences, yet only a handful of studies have tested this hypothesis. Even fewer studies have examined the ways in which marital status transitions are 23 related to gender role attitude malleability as individuals transition out of and back into a couple relationship, but these studies lack methodological sophistication to thoroughly test this assumption longitudinally. Therefore, this study seeks to explore within-individual gender role attitude malleability following marital status transitions using individual growth modeling. Rationale for the current study is twofold. First, marital status transitions are prevalent in American culture as almost half of first marriages are expected to end in divorce or separation within the first 15 years of marriage and about 3/4 of those who divorce will eventually remarry (Krieder & Ellis, 2011). It is important for researchers and practitioners to have an understanding of how marital status transitions affect individual-level attitudes and personal conceptualizations of gender roles. Second, gender role attitudes have been identified as key predictor variables in several empirical studies related to intra- and interpersonal functioning including mental health (Read & Grundy, 2011), help-seeking behaviors (Shepherd & Rickard, 2011), career choices and work place earnings (Corrigall & Konrad, 2007; Judge & Livingston, 2008), the division of household labor (Cummingham, 2005), familial patterns of interaction (Marks, Lam, & McHale, 2009), and aspects of marital quality (Amato & Booth, 1995; Mickelson, Claffey, & Williams, 2006). Examining malleability of gender role attitudes in relation to divorce and remarriage contributes to our understanding of gender socialization in relation to experiences in adulthood and lays the foundation for future work to explore the bidirectional nature of how experiences influence attitudes and how attitudes influence or alter subsequent experiences. Using data collected at six time points over the course of 20 years, this study utilized a prospective approach in examining growth trajectories of gender role attitude change. The use of several assessment points and a more sophisticated methodology allows for advancements in 24 documenting within-individual gender role attitude malleability in relation to marital status transitions, because longitudinal growth modeling provides researchers the ability to predict variance in the rate of change of a given outcome (gender role attitude growth) based on specified predictors (divorce and remarriage). This study also sought to compare the growth trajectories of men and women to see if marital status transitions influence gender role attitude malleability similarly for men and women. As such, this study examined the following research questions: Research Question 1 ? What are the patterns of growth in reported gender role attitude scores among adults who are in first marital relationships between 1980 and 2000? Research question 2 ? Do individuals who experience marital status transitions ? divorce and remarriage ? differ in their rate of change in gender role attitudes compared to individuals who remain in first marital relationships across a 20-year period? Research question 2a ? Does sex influence gender role attitude change over time? Method Participants A secondary analysis was conducted with a sample of married individuals from the Marital Instability Over the Life Course dataset (Booth, Johnson, Amato, & Silver, 2000). This 20-year longitudinal study began with 2,033 participants recruited from across the continental United States. The original investigators reported that the initial sample was representative of married individuals in the United States with regard to age, race, household size, and presence of children. The sample does represent biases in terms of sex and education as more women and those who were better educated were more likely to participate. Data collection began in 1980 and participants were contacted again in 1983, 1988, 1992, 1997, and 2000 for follow up phone 25 interviews. If an individual did not participate in a given wave of data collection, he/she was not contacted again for subsequent data collection points. Only one spouse from each household was chosen by random selection procedures to participate in the study. Participants did not receive compensation for responses. Full information regarding this sample can be obtained from Booth and colleagues. Procedure The dataset had previously been rendered anonymous as participants are unidentifiable and labeled only by codes. Criteria for the current study were that participants be in first marital relationships at time one and answer the survey questions pertaining to gender role attitudes (N = 1,731). Individuals contributed data for each wave until they self-selected out of the data collection process or when they experienced widowhood or a second divorce as these marital statuses represent qualitatively different contexts and are outside the scope of the current research questions. Retention rates by assessment year range from 79 to 90% and can be found in Table 1. For the analytic sample in 1980, ages ranged from 18 and 55 with a mean of 35.10 years (SD = 9.52 years). The majority of the sample was female (59.3%). Participants had been marriage an average of 13.67 years (SD = 9.88). Eighty eight percent of the participants were European American; 5% were Hispanic; 4% were African American; and 3% reported ?other? as their racial group. The average years of schooling was 13.57 years (SD = 2.6 years), such that on average participants had about 1.5 years of education following high school. 26 Table 1 Retention Rates, Mean Gender Role Attitude Scores, and Scale Reliability Across all Six Waves of Data Collection Assessment Year n Retention Rate from Previous Wave Gender Role Attitude Traditionalism Scale Reliability ? M SD 1980 1,731 2.3775 .45475 .704 1983 1,371 79% 2.2633 .41731 .718 1988 1,232 90% 2.2106 .43759 .716 1992 1,044 85% 2.1988 .42463 .720 1997 891 85% 2.1816 .41398 .704 2000 782 88% 2.1776 .40254 .695 Measures Gender role attitude traditionalism. The outcome of interest was gender role attitude growth. Attitudes were measured by a seven-item scale created by the survey designers (see Booth et al., 2000). Participants rated statements about appropriate and desired roles in the home and workplace for husbands and wives using a 4-point scale (1 = Strongly Agree to 4 = Strongly Disagree) on items such as ?If his wife works full-time, a husband should share equally in household chores such as cooking, cleaning, and washing.? Four of the questions were reverse- coded. Higher scores indicated more traditional attitudes. Mean scores at each measurement interval were used for the analyses. The reliability for the scale in 1980 was adequate (? = .704) and remained consistent across the study. The full survey scale can be found in Appendix A, and scale reliability information across the study can be found in Table 1. Marital status transitions. Two separate variables were used to identify the participants? marital status at each wave of data collection (Divorce: 0 = No, 1 = Yes; Remarriage: 0 = No, 1 = Yes). All participants began the study in a first marital relationship 27 (Divorce = 0; Remarriage = 0). These time varying variables were utilized as within-individual predictors. Time. As noted, data was collected at six assessment points across a twenty year period. For longitudinal growth modeling, time is an essential variable in predicting growth. For this study, time is centered on wave one of data collection and measured by number of years between the first and subsequent assessment points, such that 1980 = 0, 1983 = 3, 1988 = 8, and so on. Therefore, gender role attitude growth can be explained in the measure of units per year. Moderator variable. The sex of the each participant was dummy coded (1 = Male) and used as a moderator variable in research question 2a to examine if gender role attitude change patterns differ between men and women. For the other research questions, sex is considered a control variable on the intercept and slope. Control Variables. Education, race (1 = European American), age, and years married were included in the model as control variables on the intercept and slope based on previous literature indicating their influence of gender role attitudes (Fan & Marini, 2000; Lawrence- Webb, Littlefield, & Okundaye, 2004). All controls variables were time invariant and measured in 1980. Analytic Strategy To address each of the research questions, a longitudinal growth model was fitted in SPSS using the mixed command. Preliminary analyses were conducted to examine the potential functional form of the trajectory. Then, control variables ? sex, education, race, age and years married ? were included in the model to control for their effect on the initial status and rate of change of gender role attitudes, and the marital status variables were added to examine gender role attitude malleability as related to marital status change. Finally, the interaction terms 28 (divorce X male; remarriage X male) were included to conduct the moderation analyses (i.e., do gender role attitudes change similarly for men and women who experience marital status change?). Below is the composite specification used for this study: Yij = [ 00 + 01DIVij + 02REM ij + 03MALEj + 04EDU + 05RACE + 06AGE + 07YRSMAR + 08MALEj*DIVij + 09MALEj*REMij] + [ 10TIMEij + 11TIMEij*DIVij + 12TIMEij*REMij + 13TIMEij*MALEj + 14TIMEij*EDU + 15TIMEij*RACE + 16TIMEij*AGE + 17TIMEij*YRSMAR + 18TIMEij* MALEj*DIVij + 19TIMEij* MALEj*REMij] + [ ?0i + ?1i TIMEij + ?ij ] Note that the interaction terms between male and divorce or remarriage on both the intercept and slope were only included to answer RQ2a. Results Empirical growth trajectories were plotted for a random subsample of the dataset and visual inspection of the trajectories indicated variation in the rate of change and the shape of the trajectory. Some trajectories had a curvilinear pattern, but most had a distinct linear form as attitudes appeared to decrease slightly over time (see Table 1 for mean scores across the study). Both the linear model and the quadratic model were statistically significant, but further inspection of the models provided more detail in selecting the appropriate functional form. Specifically, the curvature of the quadratic model appeared almost imperceptible (.0008). While this may represent actual curvature, it is likely a function of the large sample size. Therefore, the linear functional form appeared to best fit the data in measuring gender role attitude change over time and was fitted to address each of the research questions. Research Question 1: Developing a Model and Examining Growth To answer research question one, an unconditional means model was fitted first (Table 2 29 ? Model A). The intercept (2.2887) indicates that married individuals in this study reported somewhat more egalitarian attitudes as higher scores on this 1 to 4 scale indicate more traditionalism. The model also suggests that there is significant variation in the outcome (Table 3 ? Model A), gender role attitudes, and provides the information needed to calculate the intraclass correlation coefficient (?). This correlation describes where the variation in gender role attitudes exists ? within or between individuals. The interclass correlation for the current study is: ? = 20 / ( 20 + 2 ) = .1217 / (.1217 + .0733) = .62 This value indicates that about two thirds of the variance (62%) is attributable to between individual differences, and over a third (38%) of the variation in gender role attitudes lies within the individual. This study attempted to account for the within-individual variance by examining marital status change over time. Next, the unconditional growth model was fitted (Table 2 ? Model B). This model adds the predictor variable, time, to the equation. The initial status remains significant and similar (2.3374), and we find that the population average rate of change is negative indicating that gender role attitudes shift slightly towards more egalitarian at the rate of -.0080 units per year. This unconditional growth model also provides a baseline with which to compare subsequent models. The variance statistics from the unconditional means and the unconditional growth models were, then, used to compute a pseudo R2 statistic, which quantifies the amount of within- person variation in gender role attitudes explained by linear time (Singer & Willett, 2003). Pseudo R2 = [ 2 (unconditional means) - 2 (unconditional growth)] / 2 (unconditional means) = (.0733 - .0613) / .0733 30 = .012 / .0733 = .16 A pseudo R2 of 0.16 indicates that about 16% of the within-individual variability in gender role attitudes is explained by linear time. This leaves 22% of the within-individual variance left to be explained. Research Question 2: Predicting Growth in Gender Role Attitudes To address research question two, a taxonomy of statistical models were fitted to examine the patterns of growth in reported gender role attitude scores among adults with differing marital status histories between 1980 and 2000 (see Tables 2 and 3). Models were fitted using full maximum likelihood. This allows for estimates of the unknown population parameters (Singer & Willett, 2003). Model C (Table 2) depicts the effects of the time-varying predictor variables (divorce and remarriage) on gender role attitude change controlling for sex, education, race, age, and years married. Estimated initial gender role attitudes were 2.8541 controlling for all else in the model. The time-varying predictors, divorce and remarriage, had a statistically significant, negative effect on this average intercept, such that those who were divorced had initial gender role attitude scores .1015 units lower than those in first marriages (i.e., more egalitarian) and those who were remarried had initial gender role attitude scores of .1326 units lower than those in first marriages (i.e., more egalitarian). Yet, for this study all participants take on the same initial value as everyone began the study in a first marital relationship. On average, the true annual rate of change was estimated to be -.0304. This represents a small, but statistically significant decline towards more egalitarian attitudes over time controlling for all else in the model. For divorced individuals, their annual rate of change was estimated to be -.0372 (-.0304 + -.0068). While still quite small, those who were divorced decrease at a 31 significantly faster rate of change toward egalitarianism. In other words, they experienced acceleration in the decline of their gender role attitude traditionalism compared to those who remained in first marital relationships. For remarried individuals, their annual rate of change was estimated to be -.0206 (-.0304 +.0098), a less steep slope, than those in first marriages. In sum, those who do not experience a marital status transition report a decline in gender role attitudes over time towards egalitarianism at a rate of -.0304 units per year. The experience of divorce was associated with a somewhat steeper slope towards egalitarianism at a rate of - .0372 units per year, and the experience of remarriage lead to a deceleration in the decline of traditional gender role attitudes at a rate of -.0206 units per year. From this, it appears that marital status transitions uniquely predict the rate of change in gender role attitudes. Figure 1. Trajectories of gender role attitude change between 1980 and 2000 for individuals who are not divorced or remarried, individuals who are continuously divorced, and individuals who are continuously remarried controlling sex, education, race, age, and years married. 1.9 2.1 2.3 2.5 2.7 2.9 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 Ge nde r R ole Atti tude Sc ore Time Not divorced, Not Remarried Divorced Remarried 32 Trajectories by marital status can be found in Figure 1. This plot illustrates the average rate of change in gender role attitudes as predicted by one?s marital status. While it was not possible to be divorced or remarried at time 1 in the current study, this figure provides a visual display of the three trajectories to compare the rates of change in gender role attitudes as a function of one?s marital status. Goodness of fit statistics provide indication that this model (Table 3 ? Model C) is a better model in predicted gender role attitude growth than time alone (Table 3 ? Model B). In comparing the deviance statistics [(-22LLFull) ? (-22LLReduced)], the delta deviance was compared to the critical value in the ?2 distribution for 14 degrees of freedom as the predictor variables (divorce and remarriage) and controls (sex, education, race, age, and years married) were included in the model at the intercept and slope. The deviance statistic declines by 334.34, which exceeds the .001 critical value of the ?2 distribution for 14 degrees of freedom (36.12). Finally, Model D was fitted to see if men and women differed in their gender role attitude change patterns (Tables 2 and 3). As expected from previous literature, men and women generally differed in their initial status of gender role attitudes, such that men reported significantly more traditional gender role attitudes than women (Male). Yet, no differences were seen in their initial status or rate of change by marital status (Male X Divorce and Male X Remarriage). Men and women with similar marital status histories appear to change similarly controlling for all else in the model. Model C was selected as the final model as goodness of fit statistics indicate that Model D offers no significant improvement in model fit [ D = 4.14, df = 4; ? Crit (p = .05, df = 4) = 9.49; 4.14 < 9.49]. 33 Table 2 Fixed Effects with Time Centered on Wave One when All Participants (N = 1,731) were in First Marital Relationships Model A Model B Model C Model D Parameter Unconditional Means Unconditional Growth Hypothesized Model Differences by Sex Initial Status ( 0i) 00 2.2887*** (.0093) 2.3374*** (.0102) 2.8541*** (.0728) 2.8530*** (.0728) Divorce 01 -.1045* (.0432) -.0902 (.0564) Remarriage 02 -.1313* (.0617) -.0885 (.0723) Male 03 .1380*** (.0197) .1379*** (.0198) Education 04 -.0547*** (.0038) -.0547*** (.0038) Race 05 -.0866** (.0298) -.0867** (.0298) Age 06 .0067* (.0027) .0067* (.0027) Years Married 07 .0011 (.0027) .0011 (.0027) Male X Divorce 08 -.0338 (.0878) Male X Remarriage 09 -.1627 (.1381) Rate of Change ( 1i) 10 -.0080*** (.0006) -.0304*** (.0054) -.0306*** (.0054) Divorce 11 -.0068** (.0031) -.0074~ (.0039) Remarriage 12 .0098* (.0040) .0092~ (.0049) Male 13 -.0032*** (.0013) -.0025~ (.0011) Education 14 .0014*** (.0003) .0014*** (.0003) Race 15 .0022 (.0023) .0022 (.0023) Age 16 5.82E-5 (.0002) 4.73E-5 (.0002) Years Married 17 3.37E-6 (.0002) 1.14E-5 (.0002) Male X Divorce 18 -.0017 (.0064) Male X Remarriage 19 -.0049 (.0087) Note. ~ p < .10, * p < .05, ** p < .01, *** p < .001 34 Table 3 Variance Components and Goodness of Fit Statistics with Time Centered on Wave One when All Participants (N = 1,731) were in First Marital Relationships Variance Components Level 1 Within-Person 2 .0733*** (.0014) .0613*** (.0013) .0612*** (.0013) .0612*** (.0013) Level 2 Initial Status 20 .1217*** (.0051) .1426*** (.0062) .1133*** (.0052) .1133*** (.0052) Rate of Change 21 .0002*** (1.89E-5) .0002*** (1.84E-5) .0002*** (1.83E-5) Covariance 10 -.0024*** (.0003) -.0018*** (.0003) -.0018*** (.0003) Goodness of Fit Deviance Statistic 4453.18 4118.84 4114.70 ? df 14*** 4 ? D 334.34 4.14 AIC 4465.18 4158.84 4162.70 BIC 4506.23 4295.65 4326.87 Note. *** p < .001 Findings from the current study are such that marital status transitions predicted the rate of change in gender role attitudes. No differences were found for men and women. Because of the interactional nature of life experiences and gender role attitudes (i.e., life experiences influence gender role attitudes as seen in the current study and, alternatively, changes in gender role attitudes are theorized to influence life experiences), a series of logistic regression post hoc analyses were conducted to see if there was any evidence to suggest that changes in gender role attitudes at one time point predicted divorce and remarriage at the subsequent time point. Logistic regression was used to descriptively examine the relationship between attitudes and marital status outcomes as this regression procedure is intended to model dichotomous outcome variables (Divorce: 0 = No; 1 = Yes; Remarriage: 0 = No; 1 = Yes). 35 From these analyses, no evidence was found linking gender role attitude change to remarriage, but data from 1988, 1992, 1997, and 2000 suggest that decreases in gender role attitude scores (i.e., becoming more egalitarian) were related to significant increases in the odds of divorce at the next assessment period (see Table 4). Previously, it was noted that divorce significantly predicts the rate of change in gender role attitudes, such that, on average, gender role attitudes become more egalitarian following divorce. Yet, these findings indicate that it may also be that attitudes begin to change prior to the divorce as individuals prepare to leave the marriage. Implications for these significant post hoc findings are discussed in the next section. Table 4 Summary of Logistic Regression Analyses for Gender Role Attitude Scores Predicting Divorce in the Subsequent Measurement Interval Year B S.E. p 1983 -.621 .364 .087 1988 -.982 .337 .004 1992 -.808 .269 .003 1997 -.971 .278 < .001 2000 -.694 .294 .018 Discussion The current study examined gender role attitude change in relation to marital status transitions. Similar to the findings of Lucier-Greer and Adler-Baeder (2011), results suggest that divorce and remarriage have a unique, albeit small, influence on the rate of change in gender role attitudes. More specifically, those who did not experience a marital status change (i.e., stayed in a first marital relationship) reported a small, but significant decline in gender role attitudes across this 20 year period. The experience of divorce (an extrafamilial experience) 36 was related to a small acceleration in the decline of traditional attitudes, and, conversely, the experience of remarriage (an intrafamilial experience) was related to a small deceleration in the decline of traditional attitudes. In other words, for remarrieds, attitudes continued to move toward more egalitarian over time, but the rate of decline was significantly slower after the experience of remarriage. During the process of divorce and singlehood, attitude change accelerated toward more egalitarian attitudes, but remarriage was associated with a slowing in this rate of change, such that over time remarried individuals tended to endorse similar attitudes as those in first marital relationships (see Figure 1). Interestingly, patterns of gender role attitude change were comparable across the sexes, such that marital status transitions predicted gender role attitude change similarly for men and women as found by Lucier-Greer and Adler-Baeder (2011). While men and women in first marital relationships differed initially in their gender role attitudes (men with more traditional attitudes), divorce and remarriage appeared to be salient life events that similarly affected men and women. This study provides further evidence of gender role attitude malleability and has implications for theoretical refinement and practice. Although growth over time was, overall, quite modest, and there was not a large amount of variability to be predicted within individuals (22% after accounting for time), these findings still offer contributions to the literatures related to divorce and remarriage and gender role attitudes. To date, much of the research on gender role attitudes has demonstrated how attitudes influence personal and interpersonal experiences, and, typically, gender role attitudes are treated as static variables (e.g., Gaunt, 2006; Kaufman, 2000). The research questions for the current study were rooted in theoretical assumptions from structural symbolic interactionism that propose that development continues across adulthood as 37 individuals assess and refine their perspective of self and others based on influential experiences (Blumer, 1969; Serpe, 1987), such as divorce and remarriage. Evidence from the current study is that experiences do, in fact, alter attitudes as within-individual gender role attitude malleability, such that changes in marital status were related to unique rates of change in gender role attitudes as found by Lucier-Greer and Adler-Baeder (2011). Unique from that previous publication, the current study was able to prospectively examine gender role attitude growth based on the specific timing of one?s marital status change. This study also applied the tenets of structural symbolic interactionism as a means to preliminarily examine the timing of gender role attitude change in post hoc analyses. Since attitudes can influence life events, it might also be that attitude changes precede marital status change. Findings were that gender role attitude change may precipitate as well as follow divorce, but no evidence was found to suggest that attitude change occurred prior to remarriage. Overall, conclusions were that experience and personal meaning are inextricably and cyclically related, such that meaning (i.e., attitudes) can affect experiences and experiences can, in turn, affect attitudes as proposed by structural symbolic interactionism. Within-Individual Gender Role Attitude Change Related to Marital Status Transitions It has been fairly well established that gender role attitudes differ across individuals as a function of personal characteristics, such as education (Etaugh & Spandikow, 1981), race (Lawrence-Webb et al., 2004), age (Ciabattari, 2001), and various other between-group variables. The focus of this study was to further the extremely limited literature exploring evidence of within-individual malleability as a means to understand a piece ? arguably a very important piece ? of the gender socialization process in adulthood. Similar to the handful of previous findings, evidence suggested that gender role attitudes can be altered in the context of 38 new roles and significant life experiences (e.g., Fan & Marini, 2000; Katz-Wise et al., 2010). Specifically, this study expanded recent efforts to study gender role attitude malleability in the context of divorce and remarriage. Overall findings were similar to that of Lucier-Greer and Adler-Baeder (2011), such that marital status transitions were found to have a unique influence on gender role attitude change. Yet, the current study was able to move beyond general trend analyses to establish within-individual malleability as a function of marital status change. Previous work was only able to provide rough estimates of attitudinal change based on group profiles, because individuals were grouped by their marital status at the conclusion of the study. This type of grouping did not allow for the analyses to include the timing of marital status change for each individual, thus the researchers were only able to speak generally about between group differences across three time points. While establishing group differences was an important first step, a few additional methodological concerns of that initial examination necessitated the current study. Specifically, conclusions about group differences were made by using a middle time point in the study to serve as a proxy for when marital status change was estimated to have occurred and the grouping procedures in that study led to the omission of over 1,000 participants who did not meet the inclusion criteria (i.e., contributing complete data at every time point over a twenty year period). The current study utilized all available assessments as well as individualized marital status change variables to more precisely document the influence of divorce and remarriage on within-individual gender role attitude malleability. Detailed evidence is provided to establish that different socializing experiences ? specifically, transitions out of and into the institution of marriage ? uniquely predicted gender role attitude change. First, baseline levels of gender role attitude change depicted change patterns for those 39 who did not experience a marital status transition (i.e., they stayed in a first marital relationship). Findings were that small, but significant changes did occur. Over the course of the twenty year study, gender role attitudes shifted approximately one and a half standard deviations. In general, gender role attitude traditionalism tended to decline over time for those that experienced minimal role change. In other words, individuals who remained in first marriages became slightly more egalitarian over time. Because these individuals did not experience a marital status transition, this decline in traditionalism may represent the influence of macro-level or sociological changes that occurred between 1980 and 2000. Similar to the findings of Ciabattari (2001) and Rogers and Amato (2000), a pattern of more egalitarian attitudes across time has been established for individuals in the United States as a result of cultural changes in gender relations (e.g., female employment and the division of household tasks). Thus, consistent with an ecological framework, macro-level changes, also conceptualized as socio-historical changes, in gender relations have an influence on an individual?s personal perceptions of gender roles. Next, it was found that marital role changes uniquely altered individual-level rates of change in gender role attitudes. The transition ?out? of marriage was related to an acceleration in the decline of traditional attitudes compared to staying in a first marital relationship. Thus, this extrafamilial experience of divorce was related to the adoption of more egalitarian attitudes over time. Again, applying an ecological framework, this slight, but significant acceleration in egalitarian attitudes was related to a more proximal, micro-level experience. Additional insight into this altered trajectory of gender role attitude change may be gleaned from the divorce-stress- adjustment perspective (Amato, 2000). This perspective takes into account the varied stressors that accompany divorce, including new roles and responsibilities, independent decision making, changes in social support and declines in one?s economic situation. These stressors are thought 40 to lead to adjustments in intrapersonal functioning and personal identity. As new opportunities for independence arise, it may be that one?s gendered identity is altered and gender role attitudes more rapidly shift toward egalitarian as seen in the current study. Conversely, the current study found that the transition back ?into? marriage was related to a slower rate of change or deceleration in the decline of gender role attitude traditionalism. This intrafamilial experience of remarriage presents a more complex change pattern of gender role attitude growth as individuals recouple and experience another micro-level role change. Attitudes continued to move toward more egalitarian, but the rate of decline was significantly slower than previously seen. Interestingly, over time the slower rate of decline in traditionalism would lead to attitudes that resemble those found by individuals in first marital relationships (see Figure 1). It may be that the transition into remarriage promotes similar attitudinal changes as seen upon entry into first marriages (i.e., toward more traditional; Fan & Marini, 2000). Thus, the context of marriage ? whether it be a first marriage or remarriage ? may promote culturally normative-expectations regarding gendered (i.e., distinct) family roles. These findings do not necessarily contradict past qualitative work which found that decisions to remarry were based on egalitarian principles and that remarriages were often considered more egalitarian than first marital relationships (Clarke, 2005; Smith et al., 1991). However, they may not be as different in their second marriages as they perceive. The slowed rate in traditionalism decline has meaning; distinguishes them from the group who do not remarry. Thus, they become more similar to those who remained married ? and may in actuality, manifest as a slight reversion. The current study depicts that single individuals preparing to remarry endorse more egalitarian attitudes, and those individuals were likely to select a mate who was compatible with their gendered perspective. Furthermore, general trends were that gender role attitudes became more 41 egalitarian over time for the entire sample, including those who remained in first marital relationships, those who divorced and remained single, and those who divorced and remarried. Thus, perceptions of a more egalitarian remarriage are accurate, but it is likely that those who stayed in a first marital relationship would also report more egalitarian attitudes and experiences over time. The greater distinction becomes between those married/remarried and those not married. Timing of Attitudinal Change The broad purpose of this study was to apply the tenets of structural symbolic interactionism to examine whether experiences ? divorce and remarriage ? affect gender role attitudes as a next step in establishing the cyclical nature of influence between experience and personal meaning. Previous work established that gender role attitudes affect experiences (e.g., Cunningham, 2005; Judge & Livingston, 2008), and this work as well as the work of a few other studies (e.g., Fan & Marini, 2000; Vespa, 2009) established that, conversely, experiences also affect attitudes. Because evidence has established the cyclical nature of influence between experience and personal meaning, post hoc analyses were conducted to see if there was any evidence to suggest that attitudes may have changed prior to the marital status transition. Notably, results suggest that, in fact, gender role attitude change may precede as well as follow divorce as higher levels of egalitarian attitudes were related to increased odds of experiencing a divorce (i.e., attitudes affect experiences). This finding may indicate that personal changes in gender role attitudes can precipitate, prompt, or even ?cause? a divorce. Structural symbolic interactionism proposes that individuals act in a way that is congruent with the meaning ascribed to a particular interaction (Stryker, 2008), thus if an individual were to embrace new, more egalitarian meaning surrounding gender 42 roles, this could trigger them to reconsider a relationship that is incongruent with their gendered beliefs. In other words, discrepancies between actual and desired roles may lead to divorce if the discrepancy continues. This hypothesis aligns with observations from a longitudinal study of divorced families which suggests that "the Achilles heel of traditional marriage is change...When one or the other partner begins to behave untraditionally, trouble follows" (Hetherington & Kelly, 2002, p. 34). Attitudinal change prior to divorce could also reflect preparation for the experience of divorce similar to Hetherington and Kelly?s theory that ?postdivorce life begins within the marriage? (p. 42). They posit that individuals experience a breakdown of the marriage and begin to prepare for life after marriage before officially entering singlehood. This context of marital breakdown may prompt changes in gendered expectations related to self, the relationship, and even roles within the relationship as the marriage unravels and individuals mentally prepare for singlehood and life outside the marriage. Further work is needed to move forward these speculations about attitudinal change prior to divorce. Overall conclusions are that some individuals began to perceive gender role attitudes differently before the experience of divorce (i.e., attitudes change predicts the experience/ interaction), while other attitude change follows divorce (i.e., experiences predict attitudes). Unlike divorce, no evidence was found to suggest that attitude change occurred prior to remarriage. This may be because intrafamilial experiences, movement into new relational roles, may prompt change after the experience occurred, while extrafamilial experiences, movement out of relational roles, may be associated with change prior to and after the experience. Implications and Future Directions The dynamic nature of gender role attitudes has now been established in short-term (e.g., Lucier-Greer, Adler-Baeder, Ketring, & Smith, 2011) as well as long-term (e.g., Bryant, 2003) 43 settings. Findings from the current study, in conjunction with other studies that have examined gender role attitude change, should begin to move the field away from asking if gender role attitudes are malleable at the individual-level in adults to questions such as how do gender role attitudes change as a function of an intrafamilial or extrafamilial experiences and for whom do attitudes change in various contexts? In other words, the ability of recent work to document gender role attitude change opens the door for future work to investigate how change occurs and the meaningfulness of this malleability in adulthood. Because gender role attitudes are thought to provide a framework by which we view life (Bem, 1981), attitudinal changes have implications for personal and relational development which may be of interest to both researchers and practitioners. This study lays the ground work for future studies to examine how gender role attitude change influences relevant outcomes and behaviors (e.g., couple quality when one or both partners alter their gendered attitudes, co- parenting with former partners, mate selection after a divorce, and relational interactions in remarriages, etc.). The use of dyadic data in future work will promote understanding about the bidirectional influence of partner attitudes on pertinent outcomes including relationship quality and stability. Understanding, more specifically, the process by which gender role attitudes change and how this change impacts the couple relationship may be useful for practitioners working with individuals and couples. Knowledge about the interactional nature of gender role attitudes may be applied to help couples see how their attitudes affect their personal perceptions of the relationship as well as their partner?s perception of the relationship. Finally, this study has important implications for theoretical refinement and our understanding of gender role development in adulthood. Gender role attitudes are not a fixed belief system, but rather an ever-evolving framework by which we view ourselves and our 44 relationships. This process can be viewed as a cycle by which experiences influence attitudes and attitudes influence experiences. Thus, our gendered belief system impacts our interactions, and significant experiences alter our gendered belief system. Applying this information to the gender socialization process is essential for fully understanding adult development. Limitations As described, this study has several important implications, yet limitations are noted. Because this study was a secondary data analysis, measurement intervals could not be altered. It may have been more meaningful to collect data about gender role attitude change immediately before and after marital status change as some nuances of attitudinal change may have been missed due to the distance between assessment intervals (between three and five years). While this study is an improvement upon studies that make inferences about attitudinal change using two data points, even more frequent and targeted data collection times could have potentially captured subtle changes that may exist preceding and following marital status change. Additionally, a more heterogeneous sample with individuals from various racial backgrounds would have allowed us to examine if the change patterns found in the current study were similar for individuals across racial groups. Because the sample was predominantly European American (88%), we could not meaningfully examine the interaction of race and marital status on gender role attitude change patterns. Finally, this study was only able to capture one dimension of gender; future studies may want to explore other aspects, such as masculinity and femininity or attitudes towards daily family tasks, to examine malleability in these measures. Conclusions Several advances were made in the current study. It stands alone as a longitudinal prospective study of within-individual gender role attitude malleability across marital status 45 transitions, marking improvements over recent retrospective studies and group profile analyses. From this work, suggestions are that divorce and remarriage uniquely influence gender role attitude change and do so similarly for men and women. Implications for theory as well as practice are that gender role attitudes must be considered dynamic variables, able to influence personal and relational outcomes and able to be refined over the course of life experiences. 46 III. Paper 2 ? Dyadic gender role attitude congruence as a predictor of couple quality in the context of coupe and relationship education Abstract This study sought to simultaneously evaluate three theories that hypothesize about the link between gender role attitudes and couple quality ? the companionate theory of marriage, the gender theory of marriage, and a dyadic attitudinal congruence model. Data from couples (N = 383 couples) engaged in couple and relationship education (CRE) provided a context to evaluate how gender role attitudes and changes in couple-level attitudinal congruence influence relational outcomes. Using structural equation modeling and the actor-partner interdependence model (APIM), findings indicate no ?best? type of gender role attitude (either traditional or egalitarian) for predicting couple quality for men and women. Rather, greater gender role attitude congruence between couples predicts increased intimacy for the husband/male partner. In turn, higher levels of male intimacy predict increased perceptions of couple quality for both partners. Findings were similar, regardless of race. Theoretical and practical implications can be drawn as this study empirically evaluates elements of several theories and provides insight for practitioners and the content of CRE. Keywords: attitudinal congruence, couple and relationship education, gender role attitudes 47 Dyadic gender role attitude congruence as a predictor of couple quality in the context of coupe and relationship education A longstanding debate exists in the literature as to if and how gender role attitudes influence romantic relationships. Most studies find some relationship ? either directly or indirectly ? between gender role attitudes and couple functioning (e.g., Amato and Booth, 1995; Mickelson, Claffey, & Williams, 2006), but others have found no relationship between these variables (e.g., Blair, 1993; Xu & Lai, 2004). Hypotheses about optimal or prescriptive gender role attitudes continue to surface, yet the research in this area lacks consensus. This study has important implications for researchers and practitioners as competing models were examined to clarify how gender role attitudes influence couple quality. This study examined whether there is a ?best? type of gender role attitude in terms of predicting couple quality or whether it is more important that couples are congruent in their gendered attitudes. Three models were simultaneously examined. The first two models ? the companionate theory of marriage (Wilcox & Nock, 2006) and the gender theory of marriage (Dizard & Gadlin, 1990; West & Zimmerman, 1987; Wilcox & Nock, 2006) ? have opposing views, but both posit a direct link between gender role attitudes and couple quality. The companionate theory of marriage suggests that egalitarian beliefs are related to higher levels marital quality, while, the gender theory of marriage posits that traditional beliefs are related to higher levels of couple quality. The third model examined was a dyadic gender role attitude congruence model (Bowen & Orthner, 1983; Lye & Biblarz, 1993). This model suggests that more congruent attitudes between partners are associated with higher levels of couple quality. This relationship between congruent attitudes and couple quality was hypothesized to be mediated by an increase in feelings of connectedness. Justification for this mediational model is rooted in symbolic 48 interactionist thought which suggests that the meaning ascribed to a situation (e.g., increases in shared attitudes) influences the ways in which individuals tend to interact (e.g., more shared experiences and feelings of connection; Blumer, 1969; Serpe, 1987). This connectedness is related to assessments of relational quality. Surprisingly, dyadic attitudinal congruence between partners has received little empirical attention in the gender roles literature, yet having shared meaning or a congruent perspective with one?s partner appears to be essential in successful romantic relationships (Gottman & Gottman, 2008). This study will aid in the understanding of the importance of gender role attitudes in romantic relationships by examining both direct and indirect effects of gender role attitudes on couple quality. Specifically, this study examined these linkages in the context of change- initiating events. We sought to examine how changes in gender role attitudes at the individual and dyadic levels influence changes in perceived couple quality directly and indirectly through assessments of intimacy and closeness using the actor-partner interdependence model (APIM; Cook & Kenny, 2005). This type of model measures the dyad as the unit of analysis and examines mutual influence in the relationship. Personal outcomes are hypothesized to be affected by one?s own predictor variables (i.e., actor effect) as well as one?s partner?s predictor variables (i.e., partner effect). Gender Socialization: The Development of Gender Role Attitudes Gender socialization begins at an early age. Children are taught directly and indirectly about the appropriate roles of men and women via gendered messages (Renk, Donnelly, McKinney, & Agliata, 2006). These gendered messages come from one?s interpersonal and cultural context and, ultimately, contribute to one?s gender identity or personal conceptualization of self as either a male or female (Bussey, 2011). Gender identity is expressed publicly via 49 gender role attitudes, which summarize one?s beliefs about appropriate behaviors, responsibilities, and interactions of men and women (Renk et al., 2006). Gender role attitudes differ by individual and range on a continuum from traditional to egalitarian (Beere, King, Beere, & King, 1984; Larsen & Long, 1988). Those with traditional attitudes respond to others based on the sex of the other individual and endorse beliefs related to the unique and separate roles of men and women, while those with egalitarian attitudes respond independent of the other person?s sex and emphasize shared roles and responsibilities between the sexes (Beere et al., 1984). Gender role attitudes are generally thought to be stable, yet suggestions from symbolic interactionism and identity theory are that attitudes may change as individuals encounter significant interactions and/or novel life experiences that alter previously held perceptions (Serpe, 1987; Wells & Stryker, 1988). Aspects of the self are thought to be interconnected to one?s social structure and, thus, the self ? which includes affective, conative, and cognitive components ? may be altered through the process of interacting with others and experiencing novel life events (Serpe, 1987; Stryker, 2008). Recent research supports this notion of gender role attitude malleability, such that significant experiences have been found to be related to changes in gender role attitudes, including marriage and subsequent marital status transitions (Lucier-Greer & Adler-Baeder, 2011; Vespa, 2009), entry into/exit from the workforce (Fan & Marini, 2000), attending college (Bryant, 2003), becoming a parent (Fan & Marini, 2000), and even participation in couple and relationship education (CRE) classes (Lucier-Greer, Ketring, Adler-Baeder, & Smith, 2011). It is suggested that these specific socializing experiences allow the individual to reconceptualize and refine their previously held gender role attitudes. As a result of theoretical understanding and empirical evidence, it appears that gender role attitudes change across the life course. 50 Gender role attitude malleability has ramifications for several intra- and interpersonal variables. Specifically, within the context of romantic relationships, researchers have found that gender role attitudes influence relational outcomes such as the division of labor within the home (Cummingham, 2005), family formation and the intent to have a child (Kaufman, 2000), union stability (Davis & Greenstein, 2004), and aspects of couple quality (Amato and Booth, 1995; Mickelson et al., 2006). The Influence of Gender Role Attitudes on Couple Quality: Competing Perspectives Mixed evidence of the relationship between gender role attitudes and couple quality has resulted in little clarity for researchers. Understanding this relationship has important implications for the literature on gender role attitudes as well as for practitioners who work with couples in educational and therapeutic settings. In fact, recent research has found that curricula focused on strengthening the couple relationship present divergent messages about gender roles in marriage either promoting egalitarian attitudes or endorsing traditional attitudes (Lucier-Greer et al., 2011), but little evidence exists to support teaching a particular type of gender role attitude. This paper sought to provide an overview and examination of three competing perspectives ? the companionate theory of marriage, the gender theory of marriage, and a dyadic gender role attitude congruence model ? as a means to empirically evaluate each perspective and provide needed clarity as to if and how gender role attitudes influence couple quality. The companionate and gender theories of marriage. The companionate and gender theories of marriage propose a direct link between gender role attitudes and couple quality. First, the companionate theory of marriage suggests that when both spouses endorse an egalitarian perspective, marriages tend to be higher in quality and more stable (Wilcox & Nock, 2006). That is, husbands and wives do not see themselves as inherently different nor do they make 51 distinctions about roles and responsibilities based on sex. Instead, a person with an egalitarian perspective would view their partner as an equal and emphasize interpersonal similarities. This notion of equality is hypothesized to lead to higher levels of couple quality. Conversely, the gender theory of marriage hypothesizes that when both spouses endorse more traditional attitudes, relational quality tends to be higher (Dizard & Gadlin, 1990; West & Zimmerman, 1987; Wilcox & Nock, 2006). In general, cultural messages in the United States encourage individuals to recognize and embrace the inherently gendered nature of marriage and the differences between men and women (Wilcox & Nock, 2006). This theory posits that individuals face strong internal and external pressures to ?do gender? in their marriage according to learned gendered norms, thus married individuals are invested in gender roles that map on to societal expectations, i.e., unique roles for husbands and wives. This theory proposes that because traditional gender role attitudes are considered culturally normative, those with traditional attitudes will report higher levels of couple quality (Dizard & Gadlin, 1990). A handful of studies have examined the companionate and gender theories of marriage. Mixed findings tend to surface for both theories. Some research suggests that egalitarian attitudes are related to higher levels of couple quality supporting the companionate theory (e.g., Amato, Johnson, Booth, & Rogers, 2003; Juni & Grim, 1994). Other works support the gender theory in their findings, i.e., egalitarian attitudes were related to poorer couple quality especially for women (e.g., Amato & Booth, 1995; Wilcox & Nock, 2006). Still other studies have found that gender role attitudes do not seem to directly influence couple quality (Blair, 1993; Xu & Lai, 2004). These empirical studies suggest that there is not a prescriptive or ideal gender role attitude that uniformly relates to higher levels of couple quality (Davis & Greenstein, 2009). In other words, because of the diverse findings and lack of consensus when examining the 52 companionate and gender theories of marriage, researchers and practitioners cannot universally prescribe which type of gender role attitude is best for healthy relational outcomes and, ultimately, couple quality. Dyadic gender role attitude congruence. Rather than prescribing the one ?best? type of gender role attitude for healthy couple outcomes, it may be that the ?best? type varies by couple as a function of their co-created meaning. Gottman and Gottman (2008) suggest that strong relationship functioning is related in part to shared meaning and expectations between partners. The creation and maintenance of shared goals, values, and symbols lends itself to developing a sense of connectedness with one?s partner, which may lead to higher levels of couple quality. These ideas are based on premises found in symbolic interactionism, which indicate that individuals are invested in creating shared meaning with others (Blumer, 1969). Shared meaning is created through the process of interacting with others as individuals jointly interpret their experiences. As a result of these interactions, individuals tend to act in a way that is congruent with the co-created meaning they have developed with significant others. In terms of gender role attitudes, it may be that when partners develop shared meaning (conceptualized as attitudinal congruence), their overall level of shared experience and closeness increases. While, there may be a direct link between congruent gender role attitudes and couple quality as hypothesized by the empirical literature related to dyadic similarities (Luo & Klohnen, 2005), the central hypothesis of this study is that more congruent gender role attitudes between spouses are related to increases in the broader sense of shared experiences and closeness, and these increases are predictive of higher levels of couple quality. This indirect link between gender role attitude change and couple quality appears to more adequately address gender within marriage as an interactional process as suggested by West and Zimmerman (1987) and Thompson (1993). 53 To date, only a few early studies have examined the relationship between dyadic gender role attitude congruence and couple quality and none could be found that examine the dynamic nature of gender role attitudes, i.e., how changes in congruence are related to changes in couple quality. Bowen and Orthner (1983) examined the relationship between attitudinal congruence and couple quality with a sample of 331 predominantly European American, military couples. Couple quality was measured with items that tapped into multiple areas of couple functioning including satisfaction with behavioral aspects of the relationship (e.g., ability to resolve arguments, communication, etc.) and the level of agreement/disagreement the couple reported regarding a variety of issues (e.g., leisure time, friends, etc.). Couples were categorized into one of four groups based on their level of attitudinal congruence ? both egalitarian, both traditional, husband egalitarian/wife traditional, and wife egalitarian/husband traditional. Findings were that couples with congruent gender role attitudes ? either both traditional or both egalitarian ? reported higher relational quality than couples categorized as husband traditional/wife egalitarian. Contrary to expectations, non-congruent couples in which the husband reported more egalitarian attitudes and the wife reported more traditional attitudes were found to have similar levels of couple quality compared to the congruent couples. Additionally, Lye and Biblarz (1993) inquired about gender role attitude congruence using data from the National Survey of Families and Households. Controlling for gender role behavior (i.e., labor force participation and domestic work involvement), the researchers found a direct relationship between dyadic gender role attitude congruence and couple satisfaction. Couples who reported greater congruence perceived alternatives to their marriage less favorably and reported higher levels of relational stability and lower levels of relationship disagreement. Results suggested that incongruent gender role attitudes within the couple dyad, specifically 54 husband traditional/wife egalitarian, were related to less satisfying marriages. While congruency appears to be a primary factor related to couple outcomes, similar to Bowen and Orthner (1983), this study found that incongruent couples where the husband endorsed egalitarian attitudes and the wife endorsed traditional attitudes also reported high levels of satisfaction in their marriages. In the most recent study found, Juni and Grim (1994) categorized a sample of 48 married couples into groups based on their level of gender role attitude congruence to understand how congruence was related to different aspects of couple functioning. They found that incongruent couples reported greater dissatisfaction with affective communication, time spent together, and child rearing issues. Conversely, egalitarian attitudinal congruence was related to satisfaction with behavioral aspects of couple functioning including communication within the marriage and the amount of time spent together. No conclusions can be made about couples who shared congruent traditional attitudes from this study, because these couples were not examined. The Current Study The current study sought to examine and expand upon competing perspectives that describe how gender role attitudes influence couple quality. Thus far, these perspectives ? the companionate theory of marriage, the gender theory of marriage, and a dyadic congruence model ? have been conceptualized as static models measured at one time point, but recent suggestions are that gender role attitudes are dynamic variables ? able to change and be refined ? in the context of adult development in relation to both long-term and short-term events (Fan & Marini, 2000; Lucier-Greer & Adler-Baeder, 2011; Moen, Kelly, & Magennis, 2009). Thus, this study sought to build on these recent findings and examine how gender role attitude change is related to changes in couple quality. For the companionate theory, suggestions are that more egalitarian attitudes should be related to increases in couple quality, while gender theory would suggest that 55 more traditional attitudes should be related to increases in couple quality. Similarly, the third perspective, a dyadic attitudinal congruence model, may also be applied to gender role attitude change. The expectation is such that greater congruence should be predictive of increases in martial quality mediated by increases in shared experience and connectedness. For this study, shared experience and connectedness are operationalized as perceptions of intimacy. Greater intimacy, in turn, is hypothesized to lead to enhancement of couple quality. Contributions to the literature. While seeking to better understand the relationship between gender role attitudes and couple quality, this study also attempted to address past methodological issues related to this area of research. First, most studies draw conclusions about the impact of gender role attitudes on the couple dyad using analyses with only one member of the dyad (e.g., Amato et al., 2003; Amato & Booth, 1995; Xu & Lai, 2004). To best understand how gender role attitudes influence couple outcomes, cross partner influences were examined in the current study (Cook & Kenny, 2005). Second, categorizing couples into groups based on their level of congruence may be informative to answer some research questions (e.g., Bowen & Orthner, 1983; Juni & Grim, 1994), but drawing cutoff points to create categories of couples may lose some of the subtle variations that exist in the data. Finally, applying recent theoretical advancements and empirical findings related to gender role attitude malleability was seen as the next logical step in broadening our understanding of the influential nature of gender role attitudes on intra- and interpersonal outcomes. Because gender role attitudes color our understanding of the world (West & Zimmerman, 1987), changes in gender role attitudes have the potential to alter our perception of our situations and/or alter our experiences with others, particularly our romantic partner. 56 In an effort to capture gender role attitude malleability, participants were drawn from a context in which malleability had previously been established ? couple and relationship education (CRE) classes (Lucier-Greer et al., 2011). The purpose of CRE is to promote relational skill-building and provide applicable information on healthy relationship functioning in a shared, educational learning environment. Notably, Lucier-Greer and colleagues (2011) found that individuals leave CRE classes with altered gender role attitudes. It may be that the content of CRE programs intentionally or unintentionally conveys assumptions about the roles of men and women and sparks participants to reconsider previously held gender role attitudes, even though this is typically not explicated as an outcome of focus for CRE (Lucier-Greer, et al., 2011). This interesting recent finding prescribes the next step, which is to understand whether and how these attitudinal changes influence changes in couple quality either directly or indirectly through increases in a general sense of closeness. The anticipated gender role attitude shifts in CRE classes allows us to examine the effect of these changes. Examining these effects of gender role attitude changes in the CRE context also permits implications to be more specifically drawn back to CRE practices. In addition to the examination of patterns of change for men and women, this study examines whether and how change patterns differ by race. To date, studies on the relationship between gender role attitudes and couple quality have generally been conducted with predominantly European American participants (e.g., Amato and Booth, 1995; Bowen & Orthner, 1983). There is fairly consistent evidence to suggest that one?s racial background has an influence on individuals? gender role attitudes particularly within the context of romantic relationships (Hill, 2002; Lawrence-Webb, Littlefield, & Okundaye, 2004), such that African Americans tend to espouse similar-levels of egalitarian gender role attitudes in relation to their 57 romantic partner. Yet, no studies could be found that examined the influence of race on the relationship between gender role attitudes and couple quality. Taking this more nuanced approach will serve to avoid a Eurocentric analysis of the influence of gender role attitudes on couple quality (Lawrence-Webb et al., 2004) and provide practitioners information when working with consumers from diverse backgrounds. Figure 1 illustrates the proposed empirical model of this study, which is based on theory and findings from recent empirical studies (e.g., Lucier-Greer et al., 2011). Change scores for each measure are modeled as post-test scores in which the effects of the pre-test scores had been controlled for (Singer & Willett, 2003). As previously stated, this model allows for the simultaneously examination and direct comparison of three competing theories. Comparisons can be made based on the significance and magnitude of depicted pathways in the comprehensive model. According to this model, evidence for the companionate theory would be depicted by significant, negative paths between changes in gender role attitudes and changes in couple quality (i.e., as traditionalism decreases ? marital quality increases) for both men and women. Conversely, support for the gender theory would be depicted by significant, positive associations between changes in gender role attitudes and couple quality (i.e., as traditionalism increases ? martial quality increases) for men and women. Finally, support for the dyadic congruence model would be displayed by a significant path from dyadic attitudinal congruence to changes in couple quality mediated by intimacy. Multi-group analyses, then, would be able to test whether and how the operation of this model differs by race. 58 Figure 1. Proposed empirical model to test competing perspectives describing the relationship between changes in gender role attitudes and changes in couple quality. Note. GRA = Gender role attitude scores, and higher scores reflect more traditional beliefs. Smaller scores for the GRA Congruence measure represent more congruent attitudes. Error terms and correlations not depicted. Method Participants Secondary data analyses were conducted on a sample of participants who voluntarily self- selected into CRE classes offered in their communities. This sample of couples was drawn from the same participant sample used by Lucier-Greer and colleagues (2011). Participants were given the option to complete pre-test and post-test surveys based on approved Institutional Review Board (IRB) protocol. A total of 643 couples began the CRE class with their partner and 59 submitted pre-test materials. Of those, 383 couples completed the CRE course and submitted useable post-test materials. No significant differences were found between those who completed the post-test survey and those who did not, in terms of the race, age, and sex of the male, race, age, and sex of the female, number of years married, or presence of children. For a model with 121 degrees of freedom, such as the one being tested, a sample size of at least 75 participants is needed to obtain a power level of .90 (Loehlin, 2004). This means that for the current study there was over a 90% probably of observing a significant effect when it occurs. Of the sample, 80.4% were married (M = 10.05 years, SD = 11.11); 8.6% were engaged; and 11% were dating. The sample was economically and ethnically diverse and closely divided between European American (50.5%) and African American participants (49.5%). In terms of education, 32.8% of the sample had a high school diploma or some high school education; 38.9% attended some college or had a 2-year degree; 17.5% had a 4-year degree; and 10.8% had a post- graduate degree. The average age for women was 37.67 years (SD = 11.99) and 40.11 years (SD = 12.65) for men. Procedure Couple and relationship education (CRE) classes were offered by trained relationship educators free of charge for adults throughout a moderately-sized, Southeastern state. Relationship education curricula used in the classes varied slightly relative to the needs of specific audiences (e.g., premarital couples, stepcouples, etc.), but all of the curricula contained the core relationship skills denoted by the National Extension Relationship and Marriage Education Network (NERMEN; Adler-Baeder & Futris, 2005). These principles include choosing to be intentional about your relationship, knowing your partner, caring for your partner and the relationship, connecting to social support, sharing and creating a sense of ?we-ness?, 60 managing differences, and caring for one?s self. Evidence exists to suggest that different gendered messages are presented in different curricula (Lucier-Greer et al., 2011), thus this study controlled for which curriculum the couple received. Classes were taught by two trained relationship educators, typically male-female teams, to a class of 10-16 participants. In general, classes were offered on a weekly basis and lasted approximately two hours each. A minimum of six sessions were offered to each group in an effort to adequately cover the content of the curriculum. Participants were given informed consents prior to research participation to explain the nature of the study and their rights as human subjects. When consent letters were signed, pre-tests were administered prior to program participation at an intake session or immediately before the first session. Post-tests were administered after program participation at the end of the last session. To ensure confidentiality, participants were assigned a code in order to match pre-test and post-test questionnaires. Those who chose not complete the surveys were still able to participate in the CRE class. Measures Predictor variables. Gender role attitude traditionalism. This study used three items from the Traditional- Egalitarian Sex Role Attitude scale (Larsen & Long, 1988) to assess gendered attitudes. Participants used a 7-point Likert scale from 1 (Strongly Disagree) to 7 (Strongly Agree) to report personal attitudes towards gender role traditionalism (e.g., ?Ultimately a woman should submit to her husband?s decision?). See Appendix B for the full measure. Higher scores indicate more traditional gender role attitudes and specifically, a greater tendency to view men as the leaders. The internal consistency of this scale appeared to be relatively high for men (pre ? = .855; post ? = .849) and women (pre ? = .850; post ? = .849). 61 Dyadic gender role attitude congruence. Attitudinal congruence was calculated by examining the absolute value of women?s gender role attitude traditionalism score minus their partner?s score. Values closer to zero indicate less attitudinal difference between partners (i.e., more congruent attitudes), while larger values indicate more difference between the couple in terms of their gender role attitudes. Outcome variables. Intimacy. Three items were used to measure intimacy. Items, such as ?I am able to understand my partner?s thoughts and feelings,? tap into the emotional closeness and feelings of connection between partners (see Appendix C for the full measure). These items are measured on a 5-point Likert scale (1 = Never to 5 = All of the Time) with higher scores indicating more intimacy in the relationship. Reliability for the scale was adequate for men (pre-test ? = 623; post-test ? = .696) and women (pre-test ? = .675; post-test ? = .695). Couple quality. Couple quality was assessed using the Quality Marriage Index (Norton, 1983). This measure provides five statements about one?s couple quality (e.g., ?We have a good marriage?). See Appendix D for the full measure. Participants used a 7-point Likert scale to respond (1 = Very Strongly Disagree to 7 = Very Strongly Agree). Higher scores indicate higher perceptions of couple quality. Reliability for the scale was high for men (pre-test ? = .959; post-test ? = .960) and women (pre-test ? = .960; post-test ? = .970). Control variables. Control variables ? years married, parental status, education, race, and curricula ? were included based on previous literature indicating their influence on gender role attitudes (Fan & Marini, 2000; Lawrence-Webb et al., 2004; Lucier-Greer et al., 2011). Years married, parental status, and curricula were measured as dyad variables, while individual reports of education and race were used. 62 Analytic Strategy A preliminary model was first fitted using data from one time point (pre-test data) to explore the relationships between men?s and women?s gender role attitudes, dyadic gender role attitude congruence, men?s and women?s intimacy, and men?s and women?s couple quality in AMOS 17.0. This structural equation model allowed for an initial comparison of the three theories which posit concurrent relationships between gender role attitudes and couple quality at a given time point ? the companionate theory of marriage, the gender theory of marriage, and a dyadic congruence model. Next, preliminary analyses were conducted to examine whether significant shifts occurred in the predictor and outcome measures from pre-test to post-test in order to be able to test a change model for these three theories. Based on preliminary analyses, an adjusted final structural equation model was, then, fitted. We simultaneously estimated regression coefficients for paths from predictor variables to the specified outcome variables controlling for years married, parental status, education, race, and curricula. This provided a stringent test that individually and concurrently considered each of the competing models and allowed us to compare the potency of primary predictive paths. Furthermore, the actor-partner interdependence model was applied to examine self and partner effects on the outcome of interest, couple quality (Cook & Kenny, 2005). More specifically, the wife/female partner?s score was used to predict changes in her perception of couple quality and her husband/partner?s perception of couple quality. Similarly, his score was used to predict his own and his partner?s couple quality change. After assessing and confirming model fit, specific paths were examined to explore which of the three perspectives best explained the relationship between changes in gender role attitudes 63 and changes in couple quality. Finally, multi-group structural equation analyses were conducted to see whether differences in the model existed by race. Results Preliminary Analyses Analyses were first conducted to examine the relationship between gender role attitudes, congruence, intimacy, and couple quality at Time 1. Pre-test data were applied to the model to explore the basic assumptions of the three competing perspectives before applying the change model (see Figure 2). Overall, model fit indices suggest decent model fit, comparative fit index (CFI) = .90, root mean square error of approximation (RMSEA) = .06, p = .10, and ?2/df ratio = 3.66. A significant path from dyadic attitudinal congruence to husband/male partner?s intimacy (? = -.103) and significant paths from men?s intimacy to both personal perceptions of (? = .497) and partner perceptions of couple quality (? = .233) indicate some evidence for the dyadic congruence model. In other words, greater congruence in gender role attitudes (operationalized as smaller difference scores) between partners were significantly related to higher levels of perceived intimacy for the male partner, and higher levels of men?s intimacy was significantly related to higher levels of both personal and partner perceptions of couple quality. One additional significant path between the female partner?s gender roles attitude and male partner?s perception of couple quality was found (? = .068), such that higher levels of traditionalism reported by the wife/female partner was predictive of higher levels of perceived couple quality for the husband/male partner. The strongest path in predicting perceptions of couple quality for both members of the couple is from personal and partner perceptions of intimacy. 64 Figure 2. Structural equation model with pre-test data. Standardized coefficients for significant paths are listed. Note. GRA = Gender role attitude scores with higher scores reflecting more traditional beliefs. Smaller scores for the GRA Congruence measure represent more congruent attitudes. * p < .05. * p < .01. *** p < .001 In order to examine the change model these theories imply, analyses were conducted to test whether significant shifts occurred over time for the predictor and outcome variables. Table 1 provides change scores for the measures at pre-test and post-test. Significant changes toward higher intimacy and couple quality were found for both men and women following participation in a CRE class, and a significant shift toward more congruent gender role attitudes was found for couples. Although previous reports indicated that CRE was related to a small, but significant individual-level gender role attitude change (Lucier-Greer et al., 2011), no significant shift was 65 found in the current study sample. Thus, these preliminary findings drove subsequent model modification, such that only post-test scores for men?s and women?s gender role attitudes were used in the final model. The use of an auto-regressive model (Time 2 score, accounting for Time 1) results in residual change, unless there is no significant change in a variable. In this case, the resulting measure is mostly error variance, rather than residual change. For the other variables (i.e., intimacy, couple quality, and dyadic gender role attitude congruence) change scores were used. Again, change scores were modeled as post-test scores in which the effects of the pre-test scores had been controlled for (i.e., post-test scores were regressed on pre-program scores; Singer & Willett, 2003). Table 1 Change Scores for Men and Women Who Participated in Couple and Relationship Education (CRE) with Their Partner Pre-Test M (SD) Post-test M (SD) t Cohen?s d Men Gender role attitude traditionalism 3.92 (1.61) 3.80 (1.61)~ 1.85 .09 Intimacy 3.78 (0.73) 3.88 (0.74)*** -3.20 -.14 Couple Quality 5.21 (1.39) 5.66 (1.19)*** -8.51 -.35 Women Gender role attitude traditionalism 3.55 (1.66) 3.62 (1.69) -.727 -.05 Intimacy 3.78 (0.79) 3.92 (0.71)*** -4.60 -.19 Couple Quality 5.05 (1.43) 5.53 (1.31)*** -9.49 -.35 Dyadic Gender Role Attitude Congruence ? |(Female GRA ?Male GRA)| 1.72 (1.15) 1.52 (1.20)** 3.064 .17 ? Smaller scores indicate more congruent attitudes. Note. ~ p < .10. ** p < .01. *** p < .001 66 Final Empirical Model For the final model, overall model fit indices suggest good model fit: comparative fit index (CFI) = .94, Tucker-Lewis index (TLI) = .90, root mean square error of approximation (RMSEA) = .06, p = .13, and ?2/df ratio = 2.99. Regression coefficients controlling for years married, parental status, education, race, and curricula can be found in Table 2 and Figure 3. Evidence or the lack of evidence for each of the theories is examined. Figure 3. Structural equation model with the standardized coefficients of significant paths listed. Note. ? denotes a post-test measure controlling for pre-test levels of that measure. GRA = Gender role attitude scores with higher scores reflecting more traditional beliefs. * p < .05. ** p < .01. *** p < .001 67 In the final model, more traditional attitudes for women at post-test were related to increases in personal perceptions of couple quality as hypothesized by the gender theory (? = .099), but the woman?s gender role attitude did not affect changes in their partner?s couple quality (? = .030) controlling for all else in the model. For men, more egalitarian gender role attitudes at post-test were related to increases in their wife/partner?s perception of couple quality (? = -.069) as hypothesized by the companionate theory. Men?s gender role attitudes did not directly influence their own perceptions of couple quality (? = -.003). Support was found for the dyadic congruence perspective in this change model. While, no direct link was found between attitudinal congruence and perceptions of couple quality for either the husband/male partner (? = .0001) or the wife/female partner (? = -.026), evidence for a mediated relationship was found. Specifically, increases in attitudinal congruence (indicated by lower difference scores) were related to increases in the husband/male partner?s perception of intimacy (? = -.081). In turn, higher levels of men?s intimacy predicted increases in the husband/male partner?s own perception of couple quality (? = .402) and his partner?s perception of couple quality (? = .134). Increases in gender role attitude congruence were not related to increases in the wife/female partner?s intimacy. However, the strongest predictor of increases in women?s couple quality were increases in her own intimacy (? = .298), but gender role attitude congruence was not predictive of increases in the woman?s perceptions of intimacy (? = .043). 68 Table 2 Regression Coefficients (with Standard Errors in Parentheses) Controlling for Years Married, Parental status, Education, and Race Parameter Unstandardized (SE) Standardized Companionate and Gender Theories of Marriage T2 Female Gender Role Attitudes ? ? Female Couple Quality .074 (.026) ** .099 T2 Female Gender Role Attitudes ? ? Male Couple Quality .021 (.025) .030 T2 Male Gender Role Attitudes ? ? Male Couple Quality -.003 (.025) -.004 T2 Male Gender Role Attitudes ? ? Female Couple Quality -.054 (.028) * -.069 Dyadic Gender Role Attitude Congruence Model Direct path ? Gender Role Attitude Congruence ? ? Female Couple Quality -.028 (.037) -.026 ? Gender Role Attitude Congruence ? ? Male Couple Quality .001 (.034) .001 Mediated path ? Gender Role Attitude Congruence ? ? Female Intimacy .025 (.024) .043 ? Female Intimacy ? Female Couple Quality .543 (.066)*** .298 ? Female Intimacy ? Male Couple Quality .224 (.059)*** .133 ? Gender Role Attitude Congruence ? ? Male Intimacy -.051 (.026)* -.081 ? Male Intimacy ? ? Female Couple Quality .226 (.060)*** .134 ? Male Intimacy ? ? Male Couple Quality .631 (.056)*** .402 Note: * p < .05. *** p < .001 69 Multi-Group Analyses Multi-group analyses were conducted to compare the model fit for European and African American couples. Because of the relative lack of interracial couples (7%), individuals in these relationships were removed from this set of analyses. Two models were fitted, a constrained model in which the parameters for the two groups were set equal to each other (H0 = the paths are equal for European and African American couples) and an unconstrained model in which the parameters for the two groups were allowed to vary freely (H1 = the paths differ by race). To test the null hypothesis that the paths are equal for European and African American couples, a ? test was conducted (see Table 3). From the results of the ? test, we cannot reject the null hypothesis that the paths are equal [ ? = 12.81, df = 12; ? Crit (p = .05, df = 12) = 21.03; 12.81 < 21.03]. This indicates that the model fit similarly for couples from both racial backgrounds. Table 3 Delta Chi-Squared Test between the Constrained and Unconstrained Model Model ? 2 df Constrained Model Constrained 1 319.31 162 H0 = The paths are equal for European and African American couples, in the population controlling for all else in the model Unconstrained 2 306.50 150 ??2= 12.81 ?df = 12 Discussion This study contributes to the growing body of literature aimed at understanding how gender role attitudes influence the couple relationship, specifically couple quality. In an attempt 70 to evaluate current theories on the relationship between gender role attitudes and couple quality, a structural equation model was fitted to simultaneously examine the companionate theory of marriage, the gender theory of marriage, and a dyadic congruence model. Several conclusions can be drawn from this study, including some support for the interdependent influence of personal gender role attitudes on couple outcomes and the impact congruent attitudes can have on romantic relationships. The examination among variables was conducted in the context of an event (i.e., CRE), within which small, but significant changes in gender role attitudes have been documented (Lucier-Greer et al., 2011). This study had hoped to provide a next step in the exploration of gender role attitude shifts by examining the ways in which individual changes in gender role attitudes matter for perceptions of personal and partner couple quality. This would have allowed for the testing of concurrent changes in theoretically related areas. Yet, findings in the current study were such that individual-level shifts in gender role attitudes were not significant at the p < .05 level. Significant shifts, however, were found in gender role attitude congruence between romantic partners in the context of a brief six-week, educational intervention, such that couples became more congruent in their attitudes following participation in a CRE class (Cohen?s d = .17). Because shifts in individual-level gender role attitudes were not found, post-test reports of gender role attitudes were used, but the other variables in the model reflected change scores. While not optimal for answering the current research questions related to links among changes, comparative influence of predictors could still be assessed using the concurrent model and the model that included the links among changes in congruence, intimacy, and couple quality. This study provides information on the interdependent influence of gender role attitudes as well as the influence of changes in attitudinal congruence on relational outcomes. Using the 71 APIM (Cook & Kenny, 2005), examinations of how one?s gender role attitude and changes in dyadic attitudinal congruence influence personal and partner perceptions of couple quality were conducted. The findings of this study have implications for basic research related to gender role attitude development and for practitioners in educational and therapeutic settings. ?Best? Type Evidence Results from the current study suggest that there does not appear to be a universal ?best? type of gender role attitude that consistently predicts higher levels of relationship quality for both partners. As previously noted, the companionate and gender theories of marriage hypothesize a direct link between gender role attitudes and marital quality. In the current study, little support was found for these theories in the initial examination of theories with pre-test only data, and due to the nature of the current sample, the change model could not be fully examined as no individual-level, significant shifts were found in gender role attitudes from pre-test to post-test. In the final model, partial support was found for the gender theory of marriage for women (i.e., higher levels of traditional attitudes predicted increases in personal perceptions of couple quality) and the companionate theory for men (i.e., higher levels of egalitarianism for men predicted increases in perceptions of couple quality for their partner). However, these findings, if taken alone, may actually misrepresent the bigger picture of gender role attitudes? influence, which is evident in the current study due to the simultaneous examination of possible pathways to personal and partner assessments of couple quality. Comparison of the magnitude of paths is possible. Within the model we find that higher levels of the wife/woman?s traditionalism were modestly related to increases in her own perception of couple quality. These findings are similar to patterns found in a longitudinal study of individuals (Amato & Booth, 1995). In that study, 72 Amato and Booth were only able to speculate as to why traditional gender role attitudes were related to increases in relational quality (e.g., perhaps, it is because society endorses a more traditional role for women), because the data were individual-level accounts of gender role attitude change and no information was provided on the partners? trajectory of attitudinal shifts or the partners? perception of relationship quality. The current study is able to examine this link between traditional attitudes and increased marital quality for women in the larger context of the dyadic change process between romantic partners. By using couple data, we see that higher levels of traditionalism for women coupled with higher levels of egalitarian attitudes by their partners predicted increased couple quality for women. Wives attitudes were a stronger predictor of increases in her own perceptions of couple quality compared to the influence of her husband?s attitudes. Because of the interdependent nature of the data (and, of course, the couple relationship), the individual-level gender role attitude of each partner appears to influence the entire marital system as suggested by family systems theory, but the influence does not appear to be as straightforward as suggested by the gender and companionate theories of marriage (Bowen, 1978). To further understand the influence of personal and partner attitudes on couple-level outcomes, it is important to place and draw meaning from the group means in the context of the full scale. For this study, gender role attitudes were measured on a one to seven scale with higher scores reflecting more traditional attitudes. In general, both men and women report gender role attitude scores near the middle of the scale (post-test M = 3.62 for women; post-test M = 3.80 for men) with men, on average, reporting more traditional attitudes. Therefore, the finding that more traditional attitudes are related to increased marital quality for women should be examined with a few qualifiers. First, the wife/woman?s gender role attitude score must be 73 examined in relation to her partner?s. Although men?s gender role attitudes are comparatively more traditional than women?s, the findings do not suggest that men and women should be or currently are on opposite ends of the gender role attitude continuum. In general, reports of gender role attitudes for men and women are within a half of a standard deviation of each other. Notably, changes in gender role attitudes for the individual groups of men and women were non-significant, unlike the previous study of participants in CRE who did demonstrate significant change in gender role attitudes, dependent upon the messages in two different curricula (Lucier-Greer et al., 2011). The significant shift in the current study was in the level of congruence within dyads. CRE had a consistent and significant influence on couple congruence, such that, in general, couples left the class with more similar gender role attitudes. Thus, while individual-level shifts were not significant, together the relative shift of each partner created a significant change in the couple?s congruence. In other words, the small, non-significant shifts of each partner lead to a modest, significant movement of the couple toward one another. And it is this factor that is comparatively the stronger predictor of positive changes in couple quality for both men and women, mediated by enhanced feelings of intimacy for men. A Case for Congruence and Feelings of Closeness Previous evidence on the relationship between gender role attitudes and couple quality produced mixed results ? some studies finding evidence for egalitarian attitudes as predictors of marital happiness and marital interaction (e.g., Amato et al., 2003), other results supporting traditional attitudes (e.g., Amato & Booth, 1995), and still others finding no apparent relationship between gender role attitudes and couple quality (Blair, 1993; Xu & Lai, 2004). This is likely because researchers only examined a piece of the interactional process between partners. It is 74 unknown in most previous research how respondents? gender role attitudes aligned with their partner?s. We have some evidence in the current study that the dyadic gender role attitude congruence model ? a model that considers the interactional nature of gender role attitudes and related feelings of connectedness within the couple context ? may better explain the relationship between gender role attitudes and couple quality. Support for this model was such that more congruent attitudes (i.e., less difference between partners) were related to higher levels of perceived intimacy using only pre-test data, and increases in attitudinal congruence were related to increases in intimacy for the male partner. Higher levels of male intimacy were, then, related to higher levels of couple quality for both partners. Similarly for the change model, increases in male intimacy were related to increases in couple quality for both partners. For this model, it may be that his perceptions of intimacy were related to feeling closer to his partner or it might even be a reflection of behavioral changes in the couple relationship (e.g., more interaction or more supportive behaviors exhibited), but it is clear that when the husband/male partner felt closer to his wife/partner, then both experienced an increase in perceptions of couple quality. More congruent gender role attitudes did not lead to increases in intimacy for women. It may be that the development of more congruent gender role attitudes was more salient for men in creating awareness of gendered processes in the marriage, because men are generally less aware of gendered processes compared to women who are more likely to experience gendered inequality and recognize gendered influences in relationships (Dryden, 1999). Although previous research has demonstrated that aspects of relational similarity directly influence couple outcomes and the handful of studies on gender role attitude congruence found direct relationships between congruence and couple outcomes (e.g., similar goals directly 75 influence couple conflict; Fitzsimons & Anderson, 2011), no direct relationship was found between attitudinal congruence and couple quality. Yet, suggestions are that the relationship between similarities and couple quality is not as straightforward as it may appear. Kenny and Acitelli (1994) used the actor-partner interdependence model and also found no direct path between similarity and marital quality. They suggested that researchers should take into account the specificity of the domain of interest when considering similarity. Because there are a multitude of domains in which couples may be more or less similar to one another, it is likely important to consider how salient the measured domain is to the interpersonal relationship (e.g., communication styles versus political perspective). While gender clearly influences interpersonal relationships (e.g., West & Zimmerman, 1987), perhaps, gender role attitudes as measured in this study have more of a general or peripheral influence on the couple relationship since relationships were found between gender role attitudes and feelings of intimacy; however, these were modest at best. Enhanced feelings of intimacy; however, were the strongest predictors of enhanced couple quality for both men and women, verifying the value of strengthening this overarching ?connectedness? element in couple relationships. Notably, the final model fit similarly for European and African American participants, providing more support for a congruence perspective. Although race has been shown to influence baseline levels of gender role attitudes, such that African American individuals often report less traditional attitudes and tend to enact more egalitarian behaviors in couple relationships (Lawrence-Webb et al., 2004), one?s racial background did not appear to influence the relationship between gender role attitudes and changes in couple functioning in the context of CRE. The influence of CRE appears to be a salient and proximal influence that similarly impacts individuals from different racial backgrounds. 76 Implications for Theory and Practitioners The current study has several implications. Theoretically, this study evaluates assumptions of symbolic interactionism in the context of gender role attitudes. Symbolic interactionism suggests that the meaning one has about a particular topic influences one?s social interactions related to that topic (i.e., meaning influences outcomes; Serpe, 1987; Wells & Stryker, 1988). As development continues across the life course and new meaning is developed from new interactions; cyclically, this new meaning is then hypothesized to alter social interactions and interpersonal functioning (Serpe, 1987). In studies of gender roles, a relationship between gender role attitudes and a wide range of personal choices and social interactions have been established (i.e., meaning influences outcomes). This study sought to evaluate and found evidence to support the premise that new meaning (e.g., changes in dyadic gender role attitude congruence), then, alters social interactions (e.g., feelings of intimacy and couple functioning). The application and validation of this theory in a newly examined content area serves to strengthen the applicability of symbolic interactionism and, specifically, it furthers the literature on gender roles in validating the dynamic nature of gender role attitudes in the couple relationship and suggests implications for attitudinal refinements ? even small ones ? on personal and interpersonal functioning. Practically, this work builds on the literature related to CRE. It should be noted that in this study and others similar to it, couples tend to report significant increases in couple quality after participating in a relationship education class (Hawkins, Blanchard, Baldwin, & Fawcett, 2008). Yet, the challenge for researchers and practitioners is to understand the mechanisms of change that lead to increases in couple quality and relational functioning (Wadsworth & Markman, 2012). Hypothesized mechanisms of change in CRE are improved communication 77 skills, better self-regulation, reduced stress, and a more positive connection (Wadsworth & Markman), but few of these links have been empirically validated. This study proposed that another potential mechanism of change in altering couple quality is via increases in shared meaning (i.e., more congruent attitudes) and connection. As noted, the purpose of CRE is to promote relational health, and less so to intentionally alter gender role attitudes, yet a recent study found that gender role attitudes were altered in the context of these short-term, educational programs depending on the (explicit and arguably, implicit) messages they received (Lucier-Greer et al., 2011). The next logical step was to understand whether and how these attitudinal changes impact the couple relationship. Practitioners should note that there is no definitive evidence that there is a universal ?best? type of gender role attitude. Thus, cautions are offered for presenting material that endorses a particular gendered perspective. While teaching implicitly or explicitly gendered messages that promote a particular gendered perspective is not in and of itself inherently invalid, it should be clear that positive benefits to the couple relationship that may result are most likely a function of attitudinal similarly rather than a result of the actual attitude that was endorsed. In other words, couples who hear the same message in a CRE class ? whether it be related to an egalitarian view or are more traditional perspective of gender roles ? are likely to increase attitudinal congruence. This congruence may promote general feelings of connectedness, and thus, may contribute to increased couple quality. Findings here suggest that in discussions of gender roles, relationship educators may be most effective by focusing on the promotion of shared meaning and mutual understanding as a way to enhance connection/intimacy and influence couple quality. Limitations 78 While there are several meaningful findings from the current study, we acknowledge some limitations. To most thoroughly understand the interactional nature of gender role attitudes, it would have been ideal to examine how changes in personal gender role attitudes influenced changes in couple functioning. Thus, one limitation of the study was that individual- level shifts in gender role attitudes were not significant, and modification of the model was necessary. Capturing individual-level gender role attitude change will be necessary to fully examine the interdependent nature of change in gender role attitudes and change in couple outcomes. Second, this is a convenience sample of couples who voluntarily engaged in a CRE class. The sample was diverse in terms of participants? racial and educational backgrounds, but couples who intentionally seek educational resources to build and sustain their relationship may not be representative of couples who do not seek these services. Also, as in other survey research, the measures represent the participants? perceptions. Behavioral assessments of marital gender roles (e.g., who does what?) and observational methods related to assessing couple functioning may serve to enhance the validity of the measures in the current study. Finally, when considering the implications of the study, it is important to remember that these changes in gender role attitude congruence are subtle and reflect change only across two time periods six weeks apart. We maintain that small changes still have practical influences on couple functioning, yet it should be clear that, in general, participants are not adopting drastically different gender role attitudes, but rather they are making slight adaptions to their current perspective. Follow-up analyses would provide more information related to the sustained impact of CRE on gender role attitude malleability and the relationship between attitudes and couple outcomes. This approach may also capture delayed effects particularly in understanding how 79 changes in attitudinal congruence and couple quality influence the wife/female partner?s perception of intimacy. Future Studies and Conclusion This study provides suggestions for future research in both the CRE literature and the literature related to gender role attitudes. First, research is only beginning to uncover the mechanisms that lead to changes in couple functioning in the context of CRE. More work is needed to develop our understanding of why and how changes occur and for whom. This information will provide empirical grounding for developing future iterations of relationship education materials. To further develop the gender roles literature, future work may want to focus on the measurement of gendered attitudes. The addition of assessments that measure multiple dimensions of gender and role related attitudes may provide richer discussion on how CRE influences our perceptions of gender, but more importantly the measurement tools we, as a field, use to evaluate gender role attitudes, perhaps, need to be reconsidered. From theory and empirical evidence, it seems apparent that gender role attitudes influence intra- and interpersonal functioning, but could it be that our perceptions of the roles of men and women differ by context? For some, perceptions of appropriate roles may differ when considering their romantic relationship compared to an evaluation of other interpersonal relationships and contexts (e.g., the workplace). Advancements are needed in the measurement of gender role attitudes to assess how context influences our evaluation of gendered roles. In relation to the current study, it could be that if we had utilized a measure to specifically assess gender role attitudes related to one?s couple relationship, the construct would have been a more salient or even direct predictor of couple outcomes. 80 Overall, this study provided a needed next step in the literature by increasing understanding of how personal gender role attitudes and changes in dyadic gender role attitude congruence influence couple connectedness and quality. Using the context of CRE, this study proposed a potential mechanism of change in predicting couple quality in educational settings based on increases in shared meaning and connection. These findings have implications for empirically validating the tenets of symbolic interactionism as well as serving to promote the well-being of the couple relationship by providing practitioners with insight into how gendered messages alter couple functioning. Evidence for a dyadic gender role attitude congruence model also has implications for moving the field beyond debates about the ?best? type of gender role attitude to creating a dialogue about the role of attitudinal congruence and feelings of connectedness in predicting positive couple outcomes. If individual group shifts in gender role attitudes had been tested separately, as has been done in previous research, we may have missed the bigger picture that points to the value of assessing dyadic congruence in gender role attitudes. 81 I. General Discussion It has been 25 years since West and Zimmerman (1987) challenged researchers to recognize the influence that gender has on personal and interpersonal functioning. Since that time, several empirical studies have validated this claim by examining the ways in which gender role attitudes influence various outcomes (e.g., Judge & Livingston, 2008; Kaufman, 2000). Additionally, a handful of studies have begun to examine how personal and interpersonal experiences, in turn, influence gender role attitudes (e.g., Bryant, 2003; Lucier-Greer & Adler- Baeder, 2011). The notion of gender role attitude malleability ? the concept that attitudes change and can be refined across the life course ? and the exploration of gender socialization in adulthood have only recently surfaced as important ?next steps? in the literature (e.g., Moen, Kelly, & Magennis, 2009). This dissertation sought to establish additional evidence of gender role attitude malleability in the context of significant life experiences, specifically divorce and remarriage, and subsequently, explore the impact of gender role attitude change on couple functioning. The findings have important implications for theory refinement as it relates to gender role attitude development in adulthood. The application of symbolic interactionism to the study of gender role attitude development allows researchers and practitioners to explicitly identify the continued socialization process of gender role attitudes in adulthood and speak to the dynamic relationship between gender role attitudes and life experiences. This cyclical relationship implies that gender role attitudes color our understanding of the world and, thus, influence our experiences, and reciprocally, significant life experiences allow individuals the opportunity to reevaluate and 82 reconceptualize their gendered attitudes. Support for these assumptions and evidence to support theory refinement was found across the two studies in this dissertation. The first study examined within-individual gender role attitude malleability in relation to significant experiences, divorce and remarriage. Using a more sophisticated within individual methodology than has been previously used, longitudinal analyses found that marital status transitions uniquely affect gender role attitude growth across a 20-year period. Specifically, remaining in first marital relationship was related to a significant decline in traditional gender role attitudes; divorce was associated with a small acceleration in the decline of traditional attitudes; and remarriage was associated with a small deceleration in the decline of traditional attitudes. In other words, transitions out of marriage were related to more egalitarian attitudes, but transitions into remarriage were associated with attitudinal shifts similar to being in a first marriage. Perhaps, more importantly, findings provide evidence to support the notion of gender socialization in adulthood through these slightly altered and distinct trajectories and confirm the premise of symbolic interactionism, which suggests that new experiences provide a context in which the individual reevaluates and reconceptualizes their perceptions of gender roles. Because gender role attitudes are thought to provide a framework by which we view life (Bem, 1981), attitudinal changes have implications for personal and relational development which may be of interest to both researchers and practitioners. Furthermore, this study laid the ground work for the second study of this dissertation to examine how gender role attitude change influences relevant outcomes (e.g., couple quality when one or both partners alter their gendered attitudes). Study two sought to simultaneously evaluate how gender role attitudes and changes in couple-level attitudinal congruence influence relational outcomes. Three theories that 83 hypothesize about the link between gender role attitudes and couple quality ? the companionate theory of marriage (Wilcox & Nock, 2006), the gender theory of marriage (Dizard & Gadlin, 1990; West & Zimmerman, 1987; Wilcox & Nock, 2006), and a dyadic attitudinal congruence model (Bowen & Orthner, 1983; Lye & Biblarz, 1993) ? were examined. Data from a diverse sample of couples engaged in couple and relationship education (CRE) provided a context to evaluate cross-partner effects of gender role attitudes and the importance of dyadic congruence for couple quality. Using structural equation modeling and the actor-partner interdependence model (APIM; Cook & Kenny, 2005), findings indicate minimal support for a ?best? type of gender role attitude (either traditional or egalitarian) for predicting couple quality for men and women. Rather, shifts toward greater gender role attitude congruence between couples best predicted increased intimacy for the husband/male partner, and higher levels of male intimacy predicted increased perceptions of couple quality for both partners. Findings were similar, regardless of race. This second study sought to evaluate and found evidence to support the premise that new meaning (e.g., changes in dyadic gender role attitude congruence) alters social interactions (e.g., feelings of intimacy and couple functioning). The application and validation of this theory in a newly examined content area serves to strengthen the applicability of symbolic interactionism. Specifically, it furthers the literature on gender roles in validating the dynamic nature of gender role attitudes in the couple relationship and suggests implications for attitudinal refinements ? even small ones ? on personal and interpersonal functioning. This study also points to the value of promoting attitudinal congruence among couples, rather than the prescription of a specific type of gender role attitude for the enhancement of couple quality. Overall Conclusions 84 Findings from this dissertation, in conjunction with other studies that have examined gender role attitude malleability in adulthood, should begin to move the field away from asking if gender role attitudes are malleable at the individual-level to questions pertaining to how gender role attitudes change as a function of intrafamilial or extrafamilial experiences and for whom attitudes change in various contexts? In other words, the results here open the door and provide rationale for future efforts to continue the development of the literature on the process and meaningfulness of gender role malleability across the lifespan. 85 References Adler-Baeder, F., & Futris, T. (2005). Update from the National Extension Relationship and Marriage Education Network. Identifying core concepts related to marital quality: Implications for marriage education programs. Paper presented at the annual meeting of the National Council on Family Relations, Phoenix, AZ. Amato, P. R. (2000). The consequences of divorce for adults and children. Journal of Marriage and The Family, 62, 1269-1287. doi:10.1111/j.1741-3737.2000.01269.x Amato, P. R., & Booth, A. (1995). Changes in gender role attitudes and perceived marital quality. American Sociological Review, 60, 58-66. doi:10.2307/2096345 Amato, P. R., Johnson, D. R., Booth, A., & Rogers, S. J. (2003). Continuity and change in marital quality between 1980 and 2000. Journal of Marriage and Family, 65, 1-22. doi:10.1111/j.1741-3737.2003.00001.x Arnett, J. (2007). Socialization in emerging adulthood: From the family to the wider world, from socialization to self-socialization. In J. E. Grusec, P. D. Hastings, J. E. Grusec, P. D. Hastings (Eds.), Handbook of socialization: Theory and research (pp. 208-231). New York, NY US: Guilford Press. Astin, A.W. (1993). Assessment for excellence: The philosophy and practice of assessment and evaluation in higher education. Phoenix, AZ US: American Council on Education/Oryx Press. 86 Beere, C. A., King, D. W., Beere, D. B., & King, L. A. (1984). The Sex-Role Egalitarianism Scale: A measure of attitudes toward equality between the sexes. Sex Roles, 10, 563-576. doi:10.1007/BF00287265 Bem, S. L. (1981). Gender schema theory: A cognitive account of sex typing. Psychological Review, 88, 354-364. doi: 10.1037/0033-295X.88.4.354 Blair, S. L. (1993). Employment, family, and perceptions of marital quality among husbands and wives. Journal of Family Issues, 14, 189-212. doi:10.1177/019251393014002003 Blumer, H. (1969). Symbolic interactionism: Perspective and method. Englewood Cliffs, NJ US: Prentice-Hall. Booth, A., Johnson, D., Amato, P., & Silver, S. (2000). Marital Instability Over the Life Course [United States]: A six-wave panel study, 1980, 1983, 1988, 1992?1994, 1997, 2000 [Computer database]. University Park, PA US: Pennsylvania State University. Bowen, G. L., & Orthner, D. K. (1983). Sex-role congruency and marital quality. Journal of Marriage & The Family, 45, 223-230. doi:10.2307/351312 Bramlett, M. D., & Mosher, W. D. (2001). First marriage dissolution, divorce, and remarriage: United States. Advanced Data. Vital and Health Statistics. No 323. Hyattsville, MD US: National Center for Health Statistics. 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. doi:10.1016/S0049-089X(03)00041-3 Bryant, A. N. (2003). Changes in attitudes toward women's roles: Predicting gender-role traditionalism among college students. Sex Roles, 48, 131-142. doi:10.1023/A:1022451205292 87 Bussey, K. (2011). Gender identity development. In S. J. Schwartz, K. Luyckx, V. L. Vignoles, S. J. Schwartz, K. Luyckx, V. L. Vignoles (Eds.), Handbook of identity theory and research (pp. 603-628). New York, NY US: Springer Science + Business Media. Bussey, K., & Bandura, A. (1999). Social cognitive theory of gender development and differentiation. Psychological Review, 106, 676-713. doi:10.1037/0033-295X.106.4.676 Carlson, D. L., & Knoester, C. (2011). Family structure and the intergenerational transmission of gender ideology. Journal of Family Issues, 32, 709-734. doi:10.1177/0192513X10396662 Cherlin, A. J. (2004). The deinstitutionalization of American marriage. Journal of Marriage and Family, 66, 848-861. doi:10.1111/j.0022-2445.2004.00058.x Ciabattari, T. (2001). Changes in men's conservative gender ideologies: Cohort and period influences. Gender & Society, 15, 574-591. doi:10.1177/089124301015004005 Clarke, L. (2005). Remarriage in later life: Older women's negotiation of power, resources and domestic labor. Journal of Women & Aging, 17, 21-41. doi:10.1300/J074v17n04_03 Cook, W. L., & Kenny, D. A. (2005). The Actor-Partner Interdependence Model: A model of bidirectional effects in developmental studies. International Journal of Behavioral Development, 29, 101-109. doi:10.1080/01650250444000405 Corrigall, E. A., & Konrad, A. M. (2007). Gender role attitudes and careers: A longitudinal study. Sex Roles, 56, 847-855. doi:10.1007/s11199-007-9242-0 Cunningham, M. (2005). Gender in cohabitation and marriage: The influence of gender ideology on housework allocation over the life course. Journal of Family Issues, 26, 1037-1061. doi:10.1177/0192513X04273592 88 Davis, S. N., & Greenstein, T. N. (2004). Interactive effects of gender ideology and age at first marriage on women's marital disruption. Journal of Family Issues, 25, 658-682. doi:10.1177/0192513X03257795 Davis, S. N., & Greenstein, T. N. (2009). Gender ideology: Components, predictors, and consequences. Annual Review of Sociology, 35, 87-105. doi:10.1146/annurev-soc- 070308-115920 Desrochers, S., Andreassi, J. & Thompson, C. (2002). Identity Theory. Retrieved February 22, 2012, from Boston College, Sloan Work and Family Research Network Web site: http://workfamily.sas.upenn.edu/wfrn-repo/object/rm30nx4ac1nt212c Dizard, J. E., & Gadlin, H. (1990). The minimal family. Amherst, MA US: University of Massachusetts Press. Dupuis, S. B. (2007). Examining remarriage: A look at issues affecting remarried couples and the implications towards therapeutic techniques. Journal of Divorce & Remarriage, 48, 91-104. doi:10.1300/J087v48n01_05 Dryden, C. (1999). Being married, doing gender: A critical analysis of gender relationships in marriage. New York, NY US: Routledge. Edin, K., & Reed, J. M. (2005). Why don't they just get married? Barriers to marriage among the disadvantaged. The Future of Children, 15, 117-137. doi:10.1353/foc.2005.0017 Fan, P., & Marini, M. (2000). Influences on gender-role attitudes during the transition to adulthood. Social Science Research, 29, 258-283. doi:10.1006/ssre.1999.0669 Ferree, M. (1990). Beyond separate spheres: Feminism and family research. Journal of Marriage and the Family, 52, 866-84. 89 Fingerman, K. L., & Pitzer, L. (2007). Socialization in old age. In J. E. Grusec, P. D. Hastings, J. E. Grusec, P. D. Hastings (Eds.), Handbook of socialization: Theory and research (pp. 232-255). New York, NY US: Guilford Press. Fitzsimons, G. M., & Anderson, J. E. (2011). Interdependent goals and relationship conflict. In J. P. Forgas, A. W. Kruglanski, K. D. Williams, J. P. Forgas, A. W. Kruglanski, K. D. Williams (Eds.), The psychology of social conflict and aggression (pp. 185-199). New York, NY US: Psychology Press. Gaunt, R. (2006). Couple similarity and marital satisfaction: Are similar spouses happier?. Journal of Personality, 74, 1401-1420. doi:10.1111/j.1467-6494.2006.00414.x Gottman, J., & Gottman, J. (2008). Gottman method couple therapy. In A. S. Gurman, A. S. Gurman (Eds.), Clinical handbook of couple therapy (4th ed.) (pp. 138-164). New York, NY US: Guilford Press. Hawkins, A. J., Blanchard, V. L., Baldwin, S. A., & Fawcett, E. B. (2008). Does marriage and relationship education work? A meta-analytic study. Journal of Consulting and Clinical Psychology, 76, 723-734. doi:10.1037/a0012584 Hill, S. (2002). Teaching and doing gender in African American families. Sex Roles, 47, 493- 506. doi:10.1023/A:1022026303937 Hetherington, E. M. & Kelly, J. (2002). For better or for worse: Divorce reconsidered. New York NY US: Norton. Judge, T. A., & Livingston, B. A. (2008). Is the gap more than gender? A longitudinal analysis of gender, gender role orientation, and earnings. Journal of Applied Psychology, 93, 994- 1012. doi:10.1037/0021-9010.93.5.994 90 Juni, S., & Grim, D. W. (1994). Marital satisfaction as a function of dyadic gender-role constellations. American Journal of Family Therapy, 22, 106-112. doi:10.1080/01926189408251304 Kalin, R., & Tilby, P. J. (1978). Development and validation of a sex-role ideology scale. Psychological Reports, 42, 731-738. doi: 10.2466/pr0.1978.42.3.731 Kalmijn, M. (2005). Attitude alignment in marriage and cohabitation: The case of sex-role attitudes. Personal Relationships, 12, 521-535. doi:10.1111/j.1475-6811.2005.00129.x Katz-Wise, S. L., Priess, H. A., & Hyde, J. S. (2010). Gender-role attitudes and behavior across the transition to parenthood. Developmental Psychology, 46, 18-28. doi:10.1037/a0017820 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. doi:10.1177/019251300021001006 Kaufman, G., & Taniguchi, H. (2006). Gender and marital happiness in later life. Journal of Family Issues, 27, 735-757. doi:10.1177/0192513X05285293 Kelley, P. (1995). Developing healthy stepfamilies: Twenty families tell their stories. New York, NY US: The Haworth Press. Kenny, D. A., & Acitelli, L. K. (1994). Measuring similarity in couples. Journal of Family Psychology, 8, 417-431. doi:10.1037/0893-3200.8.4.417 Kolhberg, L. (1966). A cognitive-developmental analysis of children?s sex-role concepts and attitudes. In E.E. Maccoby (Ed.), The development of sex differences (pp. 82-173). Stanford, CA US: Stanford University Press. 91 Kreider, R. M., & Ellis, R. (2011). Number, timing, and duration of marriages and divorces: 2009 (Current Population Reports, P70?125). Retrieved from http://www.census.gov/prod/2011pubs/p70-125.pdf Krumrei, E., Coit, C., Martin, S., Fogo, W., & Mahoney, A. (2007). Post-divorce adjustment and social relationships: A meta-analytic review. Journal of Divorce & Remarriage, 46, 145- 166. doi:10.1300/J087v46n03_09 Larsen, K. S., & Long, E. (1988). Attitudes toward sex-roles: Traditional or egalitarian?. Sex Roles, 19, 1-12. doi:10.1007/BF00292459 Lawrence-Webb, C., Littlefield, M., & Okundaye, J. N. (2004). African American intergender relationships: A theoretical exploration of roles, patriarchy, and love. Journal of Black Studies, 34, 623-639. doi:10.1177/0021934703259014 Lucier-Greer, M., & Adler-Baeder, F. (2011). An examination of gender role attitude change patterns among continuously married, divorced, and remarried individuals. Journal of Divorce & Remarriage, 52, 225-243. doi:10.1080/10502556.2011.556977 Lucier-Greer, M., Adler-Baeder, F., Ketring, S., & Smith, T. (2011, November). Malleability of gender role attitudes: Gendered messages in relationship education. Poster presented at the annual meeting of the National Council on Family Relations, Orlando, FL. Luo, S., & Klohnen, E. C. (2005). Assortative mating and marital quality in newlyweds: A couple-centered approach. Journal of Personality and Social Psychology, 88, 304-326. doi:10.1037/0022-3514.88.2.304 Lye, D. N., & Biblarz, T. J. (1993). The effects of attitudes toward family life and gender roles on marital satisfaction. Journal of Family Issues, 14, 157-188. doi:10.1177/019251393014002002 92 Maccoby, E. E., & Jacklin, C. N. (1974). The psychology of sex differences. Stanford University Press. Marks, J. L., Lam, C., & McHale, S. M. (2009). Family patterns of gender role attitudes. Sex Roles, 61, 221-234. doi:10.1007/s11199-009-9619-3 Mickelson, K. D., Claffey, S. T., & Williams, S. L. (2006). The moderating role of gender and gender role attitudes on the link between spousal support and marital quality. Sex Roles, 55, 73-82. doi:10.1007/s11199-006-9061-8 Moen, P., Kelly, E., & Magennis, R. (2009). Gender strategies: Socialization, allocation, and strategic selection processes shaping the gendered adult life course. In M. Smith, N. DeFrates-Densch, M. Smith, N. DeFrates-Densch (Eds.), Handbook of research on adult learning and development (pp. 378-411). New York, NY US: Routledge/Taylor & Francis Group. Morgan, M. Y. (1987). The impact of religion on gender-role attitudes. Psychology of Women Quarterly, 11, 301-309. doi:10.1111/j.1471-6402.1987.tb00905.x Norton, R. (1983). Measuring marital quality: A critical look at the dependent variable. Journal of Marriage & the Family, 45, 141-151. doi:10.2307/351302 Ozawa, M. N., & Yoon, H. (2002). The economic benefit of remarriage: Gender and income class. Journal of Divorce & Remarriage, 36, 21-39. doi:10.1300/J087v36n03_02 Parelius, A. P. (1975). Emerging sex-role attitudes, expectations, and strains among college women. Journal of Marriage & The Family, 37, 146-153. doi:10.2307/351038 Read, S., & Grundy, E. (2011). Mental health among older married couples: The role of gender and family life. Social Psychiatry and Psychiatric Epidemiology, 46, 331-341. doi:10.1007/s00127-010-0205-3 93 Renk, K., Donnelly, R., McKinney, C., & Agliata, A. (2006). The development of gender identity: Timetables and influences. In K. Yip (Ed.), Psychology of gender identity: An international perspective (pp. 49-68). Hauppauge, NY US: Nova Science Publishers. Risman, B. J. (2004). Gender as a social structure: Theory wrestling with activism. Gender & Society, 18, 429-450. doi:10.1177/0891243204265349 Rogers, S. J., & Amato, P. R. (2000). Have changes in gender relations affected marital quality?. Social Forces, 79, 731-753. doi:10.2307/2675515 Sakraida, T. J. (2005). Common themes in the divorce transition experience of midlife women. Journal of Divorce & Remarriage, 43, 69-88. doi:10.1300/J087v43n01_04 Serpe, R. T. (1987). Stability and change in self: A structural symbolic interactionist explanation. Social Psychology Quarterly, 50, 44-55. doi:10.2307/2786889 Shepherd, C. B., & Rickard, K. M. (2011). Drive for muscularity and help-seeking: The mediational role of gender role conflict, self-stigma, and attitudes. Psychology of Men & Masculinity, doi:10.1037/a0025923 Singer, J. D., & Willett, J. B. (2003). Applied longitudinal data analysis: Modeling change and event occurrence. New York, NY US: Oxford University Press. Smith, R. M., Goslen, M. A., Byrd, A. J., & Reece, L. (1991). Self-other orientation and sex-role orientation of men and women who remarry. Journal of Divorce & Remarriage, 14, 3-32. doi:10.1300/J087v14n03_02 Stewart, J. (2003). The mommy track: The consequences of gender ideology and aspirations on age at first motherhood. Journal of Sociology and Social Welfare, 30, 3-30. Stryker, S. (1980). Symbolic Interactionism: A Social Structural Version. Menlo Park, CA US: Benjamin Cummings. 94 Stryker, S. (2008). From Mead to a structural symbolic interactionism and beyond. Annual Review of Sociology, 34, 15-31. doi:10.1146/annurev.soc.34.040507.134649 Sweeney, M. M. (2010). Remarriage and stepfamilies: Strategic sites for family scholarship in the 21st century. Journal of Marriage and Family, 72, 667-684. doi:10.1111/j.1741- 3737.2010.00724.x Thompson, L. (1993). Conceptualizing gender in marriage: The case of marital care. Journal of Marriage & the Family, 55, 557-569. doi:10.2307/353338 Vespa, J. (2009). Gender ideology construction: A life course and intersectional approach. Gender & Society, 23, 363-387. doi:10.1177/0891243209337507 Wadsworth, M. E., & Markman, H. J. (2012). Where's the action? Understanding what works and why in relationship education. Behavior Therapy, 43, 99-112. doi:10.1016/j.beth.2011.01.006 Walzer, S. (2008). Redoing gender through divorce. Journal of Social and Personal Relationships, 25, 5-21. doi:10.1177/0265407507086803 Wells, L., & Stryker, S. (1988). Stability and change in self over the life course. In P. B. Baltes, D. L. Featherman, R. M. Lerner, P. B. Baltes, D. L. Featherman, R. M. Lerner (Eds.), Life-span development and behavior, (pp. 191-229). Hillsdale, NJ US: Lawrence Erlbaum Associates, Inc. West, C., & Zimmerman, D.H. (1987). Doing gender. Gender & Society, 1, 125-151. doi: 10.1177/0891243287001002002 Wilcox, W., & Nock, S. L. (2006). What's love got to do with it? Equality, equity, commitment and women's marital quality. Social Forces, 84, 1321-1345. doi:10.1353/sof.2006.0076 95 Xu, X., & Lai, S. (2004). Gender ideologies, marital roles, and marital quality in Taiwan. Journal of Family Issues, 25, 318-355. doi:10.1177/0192513X03257709 96 Appendix A Study 1 ? Gender Role Attitude Traditionalism Scale After each of the following statements, indicate whether you strongly agree (1), agree (2), disagree (3), or strongly disagree (4) with each statement. 1. A woman's most important task in life should be taking care of her children.* 2. A husband should earn a larger salary than his wife.* 3. It should not bother the husband if a wife?s job sometimes requires her to be away from home overnight. 4. If his wife works full-time, a husband should share equally in household chores such as cooking, cleaning, and washing. 5. If jobs are scarce, a woman whose husband can support her ought not to have a job.* 6. A working mother can establish just as good a relationship with her children as a mother who does not work. 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.* * Reverse coded 97 Appendix B Study 2 ? Gender Role Attitude Traditionalism Scale After each of the following statements, indicate whether you strongly disagree (1), disagree (2), disagree more than agree (3), feel mixed or neutral (4), agree more than disagree (5), agree (6), or strongly agree (7) with each statement. 1. Ultimately a woman should submit to her husband?s decision. 2. As head of the household, the father should have the final authority over the children. 3. Men make better leaders than women. 98 Appendix C Study 2 ? Intimacy Scale After each of the following statements, indicate whether you strongly disagree (1), disagree (2), disagree more than agree (3), feel mixed or neutral (4), agree more than disagree (5), agree (6), or strongly agree (7) with each statement. 1. I keep very personal information to myself and do not share it with my partner/spouse.* 2. I am able to understand my partner?s thoughts and feelings. 3. I feel close to my partner/spouse. * Reverse coded 99 Appendix D Study 2 ? Couple Quality After each of the following statements, indicate whether you strongly disagree (1), disagree (2), disagree more than agree (3), feel mixed or neutral (4), agree more than disagree (5), agree (6), or strongly agree (7) with each statement. 1. We have a good marriage/relationship. 2. My relationship with my spouse/significant other is very stable. 3. Our marriage/relationship is strong. 4. My relationship with my spouse/significant other makes me happy. 5. I really feel like part of a team with my spouse/significant other.