THE REMARRIAGE BELIEF INVENTORY: TESTING THE FACTORIAL STRUCTURE AND VALIDITY WITH A REMARRIED SAMPLE Except where reference is made to the work of others, the work described in this dissertation is my own or was done in collaboration with my advisory committee. This dissertation does not include proprietary or classified information. ___________________________________________________ Brian Jensen Higginbotham Certificate of Approval: _________________________ ________________________ Jennifer L. Kerpelman Francesca M. Adler-Baeder, Chair Associate Professor Assistant Professor Human Development Human Development and Family Studies and Family Studies _________________________ ________________________ Thomas A. Smith Anthony J. Guarino Associate Professor Associate Professor Human Development Educational Foundations, and Family Studies Leadership and Technology ________________________ Stephen L. McFarland Acting Dean Graduate School THE REMARRIAGE BELIEF INVENTORY: TESTING THE FACTORIAL STRUCTURE AND VALIDITY WITH A REMARRIED SAMPLE Brian Jensen Higginbotham 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 8, 2005 iii DISSERTATION ABSTRACT THE REMARRIAGE BELIEF INVENTORY: TESTING THE FACTORIAL STRUCTURE AND VALIDITY WITH A REMARRIED SAMPLE Brian Jensen Higginbotham Doctor of Philosophy, August 8, 2005 (M.S., Auburn University, August 5, 2002) (B.S. Brigham Young University, 2000) 135 Typed Pages Directed by Francesca M. Adler-Baeder Recent figures have indicated that approximately half of all marriages each year are remarriages for one or both partners and that the majority of these remarriages form stepfamilies. Despite the prevalence of remarriages and stepfamilies, relatively few empirically-validated resources specific to remarriages/stepfamilies are available to researchers, educators, and therapists. The purpose of this study was to refine and test the factorial structure and evaluate the external validity of the Remarriage Belief Inventory (RMBI) ? an instrument developed to assess individual?s beliefs regarding remarriage and stepfamily. iv The Remarriage Belief Inventory was designed to elicit research knowledge about remarital beliefs and assumptions and to enhance programmatic and clinical work with couples preparing for or living in stepfamilies. Using an on-line sample of 344 individuals in remarriages and/or stepfamilies, this study addressed the measurement and validation of specific beliefs and examined how they impact aspects of remarital quality. These beliefs are: (1) Adjustment comes quickly, (2) Stepfamilies are second-class, (3) Children are priority, (4) Past feelings should end, (5) Partner is perfect, (6) Success is slim, and (7) Finances should be pooled. A parsimonious seven-factor, 22-item instrument was confirmed. This structure was cross-validated by an independent sample of another 217 remarried individuals. In tests of external validity, Finances should be pooled and Success is slim had the strongest relationships with remarital quality. The 7 beliefs explained a unique amount of the variance in reports of marital satisfaction above and beyond that explained by a measure of general relationship beliefs. Results from causal invariance tests did not indicate significant differences for the linkages between RMBI factors and remarital satisfaction based on years married or stepfamily type. The Remarriage Belief Inventory represents the first empirically validated measure of seven distinct remarriage beliefs. This instrument can serve to move the study of remarital dynamics forward. Recommended areas of exploration include (a) the etiology of beliefs, (b) stability of beliefs over time and factors that impact their stability/instability, (c) partner congruence on beliefs and the impact of congruence on relational outcomes, and (d) the linkages among behaviors, beliefs, and relational outcomes. Practical applications of the RMBI and considerations for its use are also presented. v ACKNOWLEDGEMENTS The author would like to express sincere appreciation to Dr. Francesca Adler- Baeder for her mentorship. With tremendous gratitude and admiration, I will always remember her example of excellence coupled with humility and kindness. Appreciation is also extended to Tommy, Tony, and Jennifer for their continual encouragement over the years. Finally, a very special thanks to my wife for her patience, love, and commitment. vi Style manual used: Publication Manual of the American Psychological Association, Fifth Edition. Computer software used: Microsoft Word, SPSS 11.5, AMOS 5.0 vii TABLE OF CONTENTS LIST OF TABLES AND FIGURES ???????... viii LIST OF APPENDICES ????????????.. ix I. INTRODUCTION ??????????????... 1 II. REVIEW OF LITERATURE ??????????.. 9 Theoretical Background for the Development of the Remarriage Belief Inventory ?????????? 23 Research Objectives and Questions ???????. 31 III. METHODS ?????????????????.. 33 Participants ????????????????... 33 Procedure ?????????????????.. 38 Measures ?????????????????.. 40 IV. RESULTS ?????????????????? 46 Testing for the Factorial Validity of the Seven Individual Constructs ???????????....... 47 Testing for the Factorial Validity of the Seven Factor Model ??????????????????... 52 Cross-Validating the Factorial Structure with an Independent Sample ?????????????. 53 Testing for the Validity of the Causal Structure ??.. 55 Testing for Causal Invariance ?????????.. 57 Testing for Unique Contributions of the Remarriage Belief Inventory ??????????????... 61 V. DISCUSSION ????????????????.. 63 VI. REFERENCES ???????????????? 81 VII. APPENDICES ????????????????. 94 viii LIST OF TABLES AND FIGURES Table 1: Family Composition of the Your Stepfamily Magazine Sample ??????????????????????. 34 Table 2: Family Composition of the Smartmarriages.com Sample ??.. 36 Table 3: Family Composition of the Utah Sample ????????? 38 Table 4: Descriptive Statistics of the 22 Items Retained after Confirmatory Factor Analysis of Each Subscale ?????? 49 Table 5: Goodness-of-Fit Indices for 4-item Factors ???????? 51 Table 6: Group Response Characteristics on Outcome Variables ???. 54 Table 7: Summary of SEM Estimates for RMBI Factors Theorized to Predict Satisfaction ?????????????????.. 56 Table 8: Summary of SEM Estimates for RMBI Factors Theorized to Predict Adjustment ?????????????????.. 57 Table 9: Frequencies of Stepfamily Type ????????????. 58 Table 10: Critical Ratios for Differences between Parameters for Individuals in Complex and Simple Stepfamilies and Individuals Married More or Less than three Years ????... 60 Table 11: Summary of Hierarchical Regression Analysis for the RBI and RMBI ????????????????????? 62 Table 12: Cost Estimate for Postage for On-line Sample ??????? 73 Figure 1: Full Standardized Model of the Remarriage Belief Inventory ? 53 ix LIST OF APPENDICES Appendix A: Text of the Invitation Email to the Smartmarriages.com and Your Stepfamily Magazine Samples ???????.. 95 Appendix B: Informed Consent Letter ?????????...???? 97 Appendix C: Text of the E-reminder to the Smartmarriages.com and Your Stepfamily Magazine participants ??????. 99 Appendix D: Text of the Invitation Letter to Utah Newlywed Sample ?.. 101 Appendix E: Text of Reminder Postcard to the Utah Newlywed Sample... 103 Appendix F: 48 Item Remarriage Belief Inventory (RMBI) ?????. 105 Appendix G: Kansas Marital Satisfaction Scale (KMS) ???????. 108 Appendix H: Modified Version of the Marital Adjustment Test (MAT) ... 110 Appendix I: Relationship Belief Inventory (RBI) ?????????. 112 Appendix J: Demographic Questions ???????????...??. 114 Appendix K: RMBI Empirical Indicators Grouped According to Content Areas ?????????????????????.. 117 Appendix L: Refined RMBI Empirical Indicators Grouped According to Content Areas ?????????????????? 120 Appendix M: Results from Validity Tests of the Causal Structure using the Utah Sample ????????????????? 123 Appendix N: Summary of Hierarchical Regression Analysis for the RBI and RMBI using the Utah Sample ?????????? 125 1 INTRODUCTION The need for a valid and reliable instrument of remarriage cognitions was identified in print over a decade ago (Fine & Kurdek, 1994). It has long been purported by clinicians that the endorsement of dysfunctional beliefs (cognitions about what should be [Baucom & Epstein, 1990]), and unrealistic assumptions (cognitions about the way people and relationships are [Baucom & Epstein]) regarding remarriage and stepfamily relationships may negatively impact the quality of these unions (see Coleman & Ganong, 1985; 1990; Walsh, 1992). These clinical assertions have not abated over the years; on the contrary, they appear to be even more pronounced today (i.e., Ganong & Coleman, 2004; Visher, Visher, & Pasley, 2003). Yet, despite the increasing number of individuals entering remarriages and/or stepfamilies (see Bramlett & Mosher, 2002) and the existence of unique remarriages beliefs (Ganong & Coleman, 2004) there are no validated instruments of remarriage cognitions to guide empirical research on the effects of remarriage and stepfamily beliefs on remarriage quality. The validation and accurate measurement of unique remarriage beliefs as well as their documented impact on individual, couple, and family functioning would help fill this gap in the research literature. Furthermore, it could lead to an instrument that would aid preventive and interventive services for this growing population. Remarriage is a phenomenon that is becoming more and more common in the United States (Bramlett & Mosher, 2002; Bumpass & Lu, 2000). Government statistics 2 indicate that seventy-five percent of divorced people remarry within ten years, and serial remarriages are increasingly common (Bramlett & Mosher, 2001; National Center for Health Statistics, 1993). Higher-order marriages account for nearly half of all unions performed in the United States each year (Wilson & Clark, 1992). Typically, these new unions are formed quickly with estimates indicating that the average divorced individual will remarry within 3-5 years of legal dissolution (Bramlett & Mosher, 2002; Kreider & Fields, 2001). These remarriages are, however, at a higher risk for dissolution than first marriages 1 (Bramlett & Mosher, 2001; see also Kreider & Fields, 2001). Furthermore, remarriages where children from a prior relationship are present are at even greater risk of dissolution than those without children 2 (Bramlett & Mosher, 2002). In light of the demographic reports that document the prevalence and dissolution of remarriages and the resulting stepfamilies, many in the field of family studies have called for more research on factors associated with remarital satisfaction (e.g., Coleman, Ganong, & Fine, 2000). However, as the empirical research base on remarriages and stepfamilies continues to grow, research on the cognitions of remarried spouses remains understudied (Allen, Baucom, Burnett, Epstein, & Rankin-Esquer, 2001). Conversely, the clinical literature continues to highlight unique issues and stressors that are thought to influence remarriage satisfaction and stepfamily functioning (see Ganong & Coleman, 2004 for a review; Visher, Visher, & Pasley, 2003). Perhaps because children are thought to be the most salient stressor to remarried persons with children, it is not 1 According to data from the 5 th cycle of the National Survey of Family Growth, the cumulative probability of first marriage dissolution after 10 years of marriage was 0.33 in 1995. The probability of second marriage dissolution after 10 years of marriage was 0.39 in 1995 (Bramlett & Mosher, 2001). 2 Drawing again from the 5 th cycle of the National Survey of Family Growth, Bramlett and Mosher (2002) found that after 10 years of remarriage, the probability of disruption is 32 percent for women with no children at remarriage. For women with children, but none of whom were reported as unwanted, the probability is 40 percent, and for women with ?unwanted? children, the probability is 44 percent. 3 surprising that the majority of remarriage beliefs identified in the clinical literature relate to some aspect of stepparenting or stepfamily functioning (Coleman, Ganong, & Fine, 2000). However, other general content areas also tend to emerge in the literature. These content areas relate to the rate of adjustment, the new partner and relationship, past partners, structure, priorities, and finances. Although throughout the clinical literature the endorsement of beliefs in these areas is thought to influence remarriage satisfaction, empirical research has never documented this relationship across clinical and nonclinical samples. Coleman and Ganong (1990) observed this gap in the literature in their decade review and noted that cognitive variables in remarriages ?have been only superficially investigated in children and ignored in adults? (p. 936). In their 2000 decade review, Coleman and colleagues once again call for more research on cognitive variables. One notable effort in this area is the preliminary work by Fine and Kurdek (1994) on a multidimensional cognitive-developmental model. This model was proposed in an attempt to ?understand the complex processes by which members of stepfamilies adapt to the unique stresses and challenges of this type of family? (p. 15). After they propose their model, the authors conclude their article by highlighting the need for instruments that measure remarriage cognitions. They note that the testing of the validity and utility of their model would have to wait until ?adequate assessment instruments? were developed (p. 31). Only one empirical study has utilized an instrument specifically designed to assess cognitions around unique issues in remarriages with children. In this lone study, Kurdek and Fine (1991) looked at a limited number of myths (which were defined as 4 false beliefs) held by adults in stepfamilies and their reports of relationship satisfaction. Twenty-seven couples in stepfamilies completed various measures of satisfaction (i.e., family, personal, marital, and parenting) and ranked the degree to which they agreed or disagreed with seven myths regarding life in stepfamilies: 1. A stepfamily can never be as good as a family in which children live with both natural parents. 2. Problems that occur in a stepfamily probably would not occur in a family in which children live with both natural parents. 3. A stepfamily will always face problems that will never get solved. 4. Children should immediately feel affection toward their stepparent. 5. A stepfamily cannot offer children the kind of things that a natural family can. 6. Overall, a stepfamily is a poor substitute for a natural family. 7. A stepfamily that has problems early will always have problems. The results of the study indicated that for mothers and stepfathers endorsement of myths was associated with measures of satisfaction (a discussion of the specific associations is included in the review of literature section). However, in terms of instrument development, other than indicating that clinicians had identified common myths held by members of stepfamilies no other discussion or information was provided on their measure of myths/false beliefs (i.e., on what basis are these statements considered to be ?myths?; what was the criteria for a myth to be included or excluded). The authors themselves note in a subsequent publication (Fine & Kurdek, 1994) that the psychometric properties and construct validity of this measure needed to be established. 5 In contrast to this lone empirical study on remarriage beliefs, there are numerous studies that have evaluated relationship beliefs in general. The majority of literature on general relationship beliefs utilizes the Relationship Belief Inventory (RBI; Eidelson & Epstein, 1982). The RBI consists of five subscales that measure what the authors conceptualize as unrealistic and dysfunctional beliefs. They are deemed unrealistic because they are considered ungrounded or unfeasible; dysfunctional, because they lead to unpleasant emotional reactions or consequences such as frustration and disappointment (Epstein & Eidelson, 1981; Rorer, 1989). The first is that disagreement is destructive? the belief that disagreements are threats to a secure relationship, represent a lack of love, or indicate an impending divorce. The second, mind reading is expected, is the belief that a partner who truly cares should be able to sense their needs and preferences without overt communication. The third is partners cannot change?the belief that a partner can neither change him/herself nor the quality of the relationship. Fourth, sexual perfectionism is the belief that one or one?s spouse should be the perfect sexual partner. The fifth subscale comprises the belief that sexes are different?the belief that men and women are inherently different and cannot fully understand each other. In theory, endorsing these relationship beliefs results in less open, honest communication, lower commitment to the relationship, heightened distress and diminished satisfaction (Baucom & Epstein, 1990). Research utilizing the complete RBI, individual subscales of the RBI, or adaptations of the RBI has documented empirical associations between relationship beliefs and various measures of relationship quality. However, two characterizations of this literature should be noted. First, much of the research using the RBI has been cross-sectional and does not validate causal paths. 6 Many of the studies reviewed merely correlate the RBI with measures of relationship quality. Those that do use regression analysis to predict relationship quality with the RBI report standardized coefficients ranging anywhere from close to zero to around -.30 (for examples see Bradbury & Fincham, 1988, Cramer, 2001). A second observation pertains to the type of subjects used. Research on relationship beliefs and relationship quality has used married (Moller & Van Zyl, 1991), homosexual (e.g., Kurdek, 1992), premarital (e.g., Metts & Cupach, 1990), and clinical samples (e.g., Epstein & Eidelson, 1981)?but remarried samples are rarely the target population in empirical studies on relationship beliefs. Rather, in some cases, remarried couples are purposefully left out to ?simplify the interpretation of results? (see Amato & Rogers, 1999, p. 74). Even in the cases where remarried individuals are included the beliefs that have been measured are not inclusive of the unique situations, issues, and structures that remarriages and stepfamilies face. Without studies incorporating remarried samples and instruments that include specific remarriage beliefs it is difficult to know whether the documented associations between relationship beliefs and relationship quality have similar effects on remarriage quality. It is also impossible to know whether specific remarriage beliefs account for more variance in reports of remarital quality than do general relationship beliefs. Related questions also remain unanswered. For example, which remarriage beliefs are the most influential? Which of the unique remarriage beliefs are dysfunctional or unrealistic? By what process do remarriage beliefs impact satisfaction? Is it simply endorsing a remarriage belief or is it disparate beliefs between partners that has the greatest impact on 7 remarital satisfaction (i.e., issues of congruence vs. incongruence)? Do individuals? assumptions change over time or through particular experiences? The next step to answer these questions and to further advance the knowledge base of the role of beliefs in remarriages would be to develop a reliable and valid instrument of remarriage beliefs (see Fine and Kurdek, 1994). To assess their impact on relationship quality, beliefs need to be operationalized, discriminately validated, and reliably measured. Once a valid instrument specific to remarriage beliefs is established, the beliefs (discretely and collectively) can be investigated for their effect on remarital satisfaction. Interrelationships with behaviors, external factors (e.g., cultural stereotypes, social and family, interaction with former spouses, etc.), and internal factors (e.g., parent- child relations, spousal relations, etc.) associated with remarital satisfaction could also be analyzed. Similarly, other cognitive variables (i.e., attributions) could also be researched in conjunction with remarriage beliefs to assess whether the attention given to cognitions in the clinical literature is justified in terms of their relative significance in predicting remarital satisfaction. The purpose of this project was to refine and test the factorial structure and validity of the Remarriage Belief Inventory (RMBI) ? an instrument developed to assess individual?s beliefs and assumptions regarding remarriage and stepfamily (Higginbotham & Adler-Baeder, in press). Using a sample of individuals in remarriages and/or stepfamilies, the study addressed the measurement and validation of specific beliefs and examined how they impact remarital quality. The validated Remarriage Belief Inventory, practical considerations, and suggestions for future studies are offered to researchers and practitioners who may be interested in using the RMBI. It is anticipated that the RMBI 8 will be utilized (a) in future research to elicit knowledge on processes in remarriages and stepfamily development and (b) in applied work with couples preparing for or living in remarriages. 9 REVIEW OF THE LITERATURE The purpose of this chapter is to present (a) a review of relevant literature addressing issues and corresponding beliefs that are specific to remarriages and stepfamilies, (b) the theoretical frameworks that provided the foundation and assumptions for the proposed study, and (c) the guiding research questions. Approximately 40% of all marriages in the United States are remarriages for one or both partners, with the most frequent pattern that of a previously married man marrying a previously married woman (61%) (Wilson & Clarke, 1992; Ganong & Coleman, 1994). The average age of remarriage for men is 38.6 and 34.5 for women. The majority (approximately 65%) of remarrying adults have children from a previous relationship, thus forming stepfamilies (e.g. Chadwick & Heaton, 1999; U.S. Bureau of the Census, 1998). Clinicians report that remarrying couples (particularly those with children) may enter remarriage with distinct beliefs and assumptions that may influence remarriage quality and stepfamilies? adjustment (Berger, 1998; Visher, Visher, & Pasley, 2003; Ganong & Coleman, 2004). For example, Papernow (1987) writes: Remarried couples are impacted by the particular wishes and yearnings generated by their unique history: the wish that the members of the new family will love each other in the way that members of biological families do; the conviction that this new spouse will be a better mother or father to these children than the ex-spouse; the wish that the 10 new family will heal the hurts of the previous divorce or death; the fantasy that the couple?s caring for each other will be experienced between stepparents and their stepchildren. (p. 632) The remarriage literature calls attention to the existence of unique issues and stressors in remarriages ? particularly those with children (Adler-Baeder & Higginbotham, 2004). These unique issues include areas such as co-parenting across households, stepparenting, and life-histories which include at least one marriage or the death of a spouse. Combined with our culture?s nuclear family bias (see Coontz, 1997), a context is created for the development of unique beliefs and assumptions pertaining to remarriages and stepfamilies. Clinicians have provided numerous examples of these unique remarriage beliefs which can be organized into seven content areas. Many remarriage beliefs relate to some aspect of stepparenting or area of stepfamily functioning (Coleman, Ganong, & Fine, 2000); although, other general content areas also tend to emerge in the literature. These content areas address adjustment, the new partner, the new relationship, past partners, structure, priorities, and finances. The following section summarizes these seven content areas with examples of representative beliefs gleaned from the clinical literature. For some of the beliefs the direction of the possible effect on satisfaction seems clear and will be hypothesized, for others it remains unknown if the belief should be considered functional or dysfunctional. Knowledge related to the actual effects of remarriage beliefs will become available only after beliefs are operationalized, tested, and utilized to predict remarital outcomes. 11 Adjustment Quick adjustment in stepfamilies is commonly identified as a prevalent remarriage belief. Beliefs related to adjustment can take many forms including the belief that marriage should result in parental entitlement insomuch that new stepparents expect or assume an intimacy and authority they have yet to earn (Hetherington & Kelly, 2002) or assuming new relationships will be good with little effort (Visher & Visher, 1996). Coleman and Ganong (1985) warn against rigidly endorsing such beliefs. They note that assumptions of, ?instant love,? or ?if you love me you will love my children,? often operates at the time of remarriage and can cause a great deal of grief and misunderstanding. The couple, caught up in the bliss of a new romantic love relationship, may at first be oblivious to the fact that other family members (e.g., children) are less enthralled. It takes extraordinary effort to love a stepchild who blatantly ignores your existence or who is cleverly rude. (p. 118). When instant love and adjustment does not occur family members may look for someone to blame and commonly it is the new spouse (Walsh, 1992). Other adjustment beliefs can also be found in the remarriage literature and media portrayals of stepfamilies (e.g., Coleman, Ganong, & Gingrich, 1985; Coleman, Ganong, & Goodwin, 1994). These include the beliefs that stepchildren resent stepparents and that stepparent and stepchildren can never learn to love each other. In contrast to the instant love assumption which can create guilt and frustration when efforts don?t result in expected outcomes, these negative adjustment beliefs are thought to hinder efforts to work sincerely on fostering love (Ganong & Coleman, 2004). 12 The literature provides support for classifying Adjustment comes quickly as an unrealistic assumption (Hetherington & Kelly, 2002; Visher et al., 2003). There is consistent evidence from both clinical and nonclinical samples that the first several years can be turbulent for stepfamilies (e.g. Bray & Kelly, 1998; Hetherington, 1989) and remarriages are at greatest risk for divorce in the earliest years of marriage (Clarke & Wilson, 1994). One larger sample of stepfamilies has indicated a pattern of 1-2 years of disorganization and turbulence and 1-3 years of stabilization (Hetherington & Kelly, 2002). New Partner Clinicians also note the tendency for remarrying couples to assume their new partner will not have the same negative qualities as previous partners. This is often called the compensation myth because the new spouse is expected to be everything the problematic old mate was not (Hetherington & Kelly, 2002). An example of a belief in this content area is that a new spouse should fill the emotional holes that the previous spouse left empty. Beliefs in this content area may lead to notions of partner perfectionism. Assumptions of Partner perfectionism are inherently unrealistic because no one is perfect. In addition, the literature does indicate that in some cases patterns of unhealthy behavior may reemerge in subsequent unions. For example, Kalmuss and Seltzer (1986) found that spousal violence is likely to continue in ensuing marriages. It has also been suggested that other behaviors, including interactional styles developed in partners? previous relationships, may be repeated in subsequent marriages as a result of behavioral learning and conditioning (see Aguirre & Parr, 1982; Furstenberg & Spanier, 1984). If 13 those behaviors were negative, as was probably the case in a relationship that dissolved, the repetition may undermine subsequent marriages as well (Brody, Neubaum, & Forehand 1988). It is understandable why someone would not want their new marriage partner to have the same negative qualities as their previous spouse(s). However, for this study, it is assumed that believing the new partner will be ?everything and more?, when compared to previous spouses, is considered an unrealistic assumption. Potential for New Relationship For many remarried individuals there is a belief that their new marriage will be better and more successful than their previous one (Sager, Brown, Crohn, Engel, Rodstein, & Walker, 1983). Additionally, some couples endorse the assumption that experience makes things easier the second time around (Prado & Markman (1999). To the degree that these become self-fulfilling prophecies these beliefs in and of themselves are not inherently dysfunctional. However, the literature indicates that assuming that remarriages are inherently easier or more likely to succeed may be unrealistic and/or dysfunctional. For one thing, the proportion of divorces is still slightly higher for remarried couples (Bramlett & Mosher, 2001. Coleman and Ganong (1985) also warn that ?couples who convince themselves that everything was negative in their previous marriages and that everything is going to be perfect in their new marriage are building a relationship based on denial. The first sign of any pattern resembling that of the previous marriage may cause panic? (p. 117). The authors go on to suggest that this belief restricts the open communication necessary to maintain the relationship. Interestingly, longitudinal research has indicated that while levels of positive communication is higher in remarriages the level of negative communication does not change from first marriage 14 to remarriage (Prado & Markman, 1999). In summary, it remains an empirical question as to whether endorsing the belief that Remarriages are more likely to succeed than first marriages, or its? converse Success is slim, is positively or negatively related to spousal reports of remarital quality. Feelings towards Past Partners Cognitions regarding the implications and influence of past partners have been noted in the clinical literature. Examples include the belief that that previous relationships should not interfere with the current relationship (e.g., Sager et al., 1983). Bray & Kelly (1998) indicate that once remarried, some couples develop a just-us assumption where one partner unconsciously assumes their partner?s past (e.g., ex-wives, biological children, ex-in-laws, etc.) just disappears. For some couples this cognition may take shape in the form of beliefs that emotional ties to and traditions from previously relationships should be discontinued. Empirical evidence supports the belief that the Past feelings should end. Research has indicated that both a highly negative and a highly involved relationship with a former spouse negatively affect the new couple?s relationship quality (e.g. Buunk & Matsaers, 1999; Knox & Zusman, 2001). Emotionally divorcing and establishing appropriate boundaries with a former spouse appear to be essential components of remarriage quality (Weston & Macklin, 1990). Structure The cultural belief that remarriages are second-class or that stepfamilies are in some way inferior to first families has been documented through reviews of both the popular media and marriage and family textbooks (e.g., Claxton-Oldfield, 2000; 15 Coleman, Ganong, & Gingrich, 1985; Coleman, Ganong, & Goodwin, 1994). Consistently, it has been observed that stepfamilies and those individuals in stepfamilies are perceived more negatively by society than their first family counterparts (see Ganong & Coleman, 1997; Ganong, Coleman, & Mapes, 1990). This cultural message of inferiority may lead to efforts and/or beliefs that stepfamilies should be just like a nuclear family. Although the structure of a stepfamily includes non-biological relationships, clinicians report that many remarried couples themselves tend to believe that the new family unit will be similar in structure to the former nuclear family (Visher & Visher, 1985). Consequently, remarrying individuals may assume that the stepparent replaces, functionally and emotionally, the nonresidential parent (Ganong & Coleman, 2004). This may include the assumption that the stepmother will take over primary responsibility for the care and nurturing of a stepchild and that the stepchild will respond as he/she did to the biological counterpart (Hetherington & Kelly, 2002). The concern voiced by clinicians is that people with assumptions regarding inferior family structure may fail to acknowledge the strengths of their remarriage/stepfamily (Walsh, 2003) and those with beliefs regarding equivalency too often fail to acknowledge the unique complexities associated with their new family form (Leslie & Epstein, 1988). Priorities Priority and loyalty conflicts are common issues identified in the remarriage/stepfamily literature. For example Coleman and Ganong (1985) write about several mutually exclusive beliefs that may all be operating at one time due to an individuals? desire to make everybody happy: 16 ?always consider yourself first,? ?always consider the other person first,? ?always consider your marriage first,? ?always consider yourself and your children first? (as compared to your spouse and his or her children), and finally, ?always consider everybody first.? Attempting to fulfill these conflicting myths may leave people feeling schizophrenic at worst and frustrated at best. (p. 117). In light of a time period without an intimate partner around, those remarried individuals who raised children as a single divorced parent may have become accustomed to prioritizing their children. As could be the case with any parental loss, effects of the divorce on the children may foster the belief that children need extra time or special attention. During the period of single-parenthood it is not unusual for a child to become a parent?s confidante (Visher & Visher, 1996). In these situations the new spouse may feel unsatisfied for not being considered a priority and the biological parent may feel guilt for not being able to make both children and new spouse feel as if they come first (Coleman & Ganong, 1985). Bray and Kelly (1998) indicate that some remarriage partners, particularly stepfathers, expect their opinions, feelings, desires, and needs to be valued as much as everyone else and expect to feel just as at home as those with whom he is now living. Notwithstanding the potential trade-offs that come from prioritizing, it is commonly accepted that prioritizing the couple bond is vital to remarital satisfaction (Ganong, Coleman, & Weaver, 2001; Visher & Visher, 1996). Consequently, beliefs that place children?s wants, wishes, and time before that of a spouse may likely negatively impact remarital quality. 17 Finances Coleman and Ganong (1985, p. 117) write that assumptions regarding financial resources tend to move developmentally through family stages from marriage (?what is mine is yours?), to divorce (?what is yours is mine?), to single parenthood (?what is mine is mine?), to remarriage (?what is mine is mine, what is yours is yours?). The problem noted with this orientation is a lack of an ?ours? perspective. Coleman and Ganong (1985) note that while it is understandable that remarried individuals may want individual control of their financial assets they caution that ?establishing intimacy in an atmosphere of a business corporation may be difficult, if not impossible? (p. 118). Research, however, has indicated that neither the practice of pooling resources nor retaining separate accounts is significantly more predictive of satisfaction in remarriages (Coleman & Ganong, 1989; Pasley, Sandras, & Edmondson, 1994). In this content area, it appears that mutual agreement of money management practices is more important for satisfaction than any one method of financial management (Lown, McFadden, & Crossman, 1989). Consequently, it is plausible that disparate beliefs amongst partners regarding the pooling of resources may have greater predictive ability for remarital quality than individual beliefs. Unarguably beliefs and decisions about pooling finances are not unique to couples in remarriages. All couples must decide how to manage their resources. However, there is evidence that financial issues are a primary source of conflict in remarriages, particularly when children are involved (Coleman, Ganong, & Weaver, 2001; Engel, 1999). The presence of children often requires additional interaction with and/or dependence on ex-partners (e.g., child support). Highlighting another financial 18 complexity for many remarriages, Ganong and Coleman (2004) describe a scenario in which a remarried couple has to postpone buying a home because the ex-spouse wants to send a mutual child to an expensive private college. These authors note, It is easy to imagine the frustration that remarried partners experience when faced with the dilemma of having ?outsiders? (e.g., former spouses, children from earlier unions) making decisions about how their household income or how a partner?s income will be spent (p. 86) In light of these financial complexities and potential frustrations it is conceivable that remarried individuals, even those who are staunchly committed to the relationship, may have beliefs against pooling their household income. The present study tests for a relationship between such beliefs and remarital quality. Future studies, with couple data, will be needed to compare the present findings with those generated from couple discrepancy scores. Etiology Remarriage beliefs in these seven content areas are thought to originate from a number of sources. One source of remarriage beliefs is from the culture in which we live. Couples do not remarry in a cultural vacuum and the general remarriage beliefs propagated by the broader culture are thought to influence the way stepfamily members think about their relationships and interact together (Berger, 2000; Bray, 1999). In other words, the prevailing ideologies found in the cultural context determine, at least in part, the cognitive context in which individuals evaluate their situation, conduct themselves, and expect to be regarded by others (Ganong & Coleman, 2004; for a detailed discussion on this topic see Dallos, 1991). Social norms and expectations for families still largely 19 adhere to a nuclear family ideology (Coontz, 1997). By some, individuals who divorce are thought to have failed and those that form stepfamilies are thought to be entering a deficient family form (Giles-Sims & Crosbie-Burnett, 1989). Media (from fairy tales to college textbooks to motion pictures) reinforce negative portrayals of stepfamilies (e.g., Claxton-Oldfield, 2000; Coleman, Ganong, & Gingrich, 1985; Coleman, Ganong, & Goodwin, 1994). Even the language in our culture implicitly conveys the message that remarrying individuals who form stepfamilies create a family structure that is abnormal or irregular (Ganong & Coleman, 2004). The term stepfather is frequently associated with images of an abuser or sexual predator (Claxton-Oldfield, Goodyear, Parsons, & Claxton-Oldfield, 2002); stepmother, with the adjectives mean and wicked (Ganong & Coleman, 1995). The term stepchild continues to be used in professional and popular literature and the media to refer to someone that is abused, neglected, or unwanted (Coleman & Ganong, 1997; Ganong & Coleman, 2004). Ganong and Coleman (2004) make the related point that ?Language helps shape thinking, and the lack of language about relationships in stepfamilies may make it more difficult for family members to develop positive identities and satisfying relationships? (p. 31). For example, there is no term for the relationship between a former spouse and current spouse, father and stepfather, etc. (Ganong & Coleman, 1994). Coleman and Ganong (1997) note that ?ironically clinicians suggest that stepfamilies in which biological parents cooperate with each other and with stepparents are those that are creating positive family environments for their children, and yet potentially key relationships are left nameless by societies that do not expect them to exist? (p. 90). Where terms do exist, they implicitly convey the message that stepparents and 20 stepfamilies are ?unreal? or ?unnatural? when compared to biological parents and nuclear families (Ganong & Coleman, 2004, p. 30). A second source of remarriage beliefs may come from the individual?s family of origin. Epstein, Schlesinger, and Dryden (1988) have suggested that beliefs regarding how families should function may be transmitted from one?s family of origin. A parent?s teaching on the morality or appropriateness of traditional families may instill beliefs that divorce constitutes a failure and remarriage is undesirable. Furthermore it has been theorized and documented that experiencing parental divorce is associated with more favorable attitudes toward divorce (Amato, 1996, Wolfinger, 2000) Relationship beliefs are also thought to stem from experiences in a previous relationship (Ganong & Coleman, 2004). For a marriage to be a remarriage at least one partner must have had a previous marriage that ended either by divorce or death. For some, a remarriage may be seen as a chance to become a real family again, another opportunity to have a successful marriage, a chance to get out of poverty, or simply a way to secure more help rearing a child (Crosbie-Burnett, 1989; Kheshgi-Genovese & Genovese, 1997). A potential fourth source of remarriage beliefs are the issues that are unique to stepfamilies. Said another way, some beliefs may develop in the context of living in particular situations and facing unique challenges and/or roles. For example, in addition to the normative issues of sex, in-laws, etc., remarried couples with children face issues related to former partners (e.g., child support and visitation), stepparent-stepchild relationships, negotiating parenting roles among biological and stepparents, and nonexistent or ambiguous laws and social policies (Adler-Baeder & Higginbotham, 2004; 21 Ganong & Coleman, 2004; Visher, Visher, & Pasley, 2003). It could be assumed that those remarrying have some expectations, consciously or unconsciously, regarding these issues. Furthermore, how these cognitions match up with reality may influence remarital satisfaction (Ganong & Coleman, 2004). It is also reasonable to assume that over time remarriage beliefs around stepfamily issues may change. For example, one might enter a remarriage expecting children to quickly adjust to the new stepparent. However, after a few months or years of experiencing disrespect from a stepchild, a stepparent may come to the opinion that his/her original assumption was unrealistic. Notwithstanding the etiology of individual?s remarriage cognitions, it is has been suggested that dissatisfaction in remarriage may be most influenced by incompatible beliefs or incongruent assumptions (e.g., Kaplan & Hennon, 1992). Clashing beliefs or assumptions may be due to different experiences and family histories within the marriage itself and present another potential challenge to remarrying couples (Ahrons & Rodgers 1987). As stated by Leslie and Epstein (1988), ?Regardless of whether each family member holds unrealistic beliefs about remarried family life, conflict may occur when there is incompatibility among members? beliefs? (p. 159). In summary, remarriage beliefs are thought to stem from individuals? history, stereotypes that are held and reinforced in our culture, and family of origin experiences. However, the actual etiology of remarriage beliefs and their effects on satisfaction remains theoretical rather than empirically informed. Empirical Research on Remarriage Beliefs In terms of empirical research, only one study specifically targeting remarriage beliefs was identified. Kurdek and Fine (1991) looked at common myths held by 22 members of stepfamilies and reports of relationship satisfaction. Twenty-seven couples in stepfamilies completed various measures of satisfaction (i.e., family, personal, marital, and parenting) and ranked the degree to which they agreed or disagreed with seven beliefs and assumptions regarding life in stepfamilies. The results of the study indicated that for mothers, endorsement of myths was negatively correlated with personal, family, marital, and parental life satisfaction scores. For the stepfathers, there was a significant negative correlation with parenting satisfaction but not with the other three. To account for these findings the authors suggested that ?because stepfathers are likely to experience some ambiguity and difficulty interacting with their stepchildren, concerns regarding the stepchildren may be more vulnerable or important areas than those regarding marital, family, and personal life? (Kurdek & Fine, 1991, p. 570). In a subsequent article Fine and Kurdek (1994) acknowledge that the psychometric properties and construct validity of their 1991 measure were not established and they encourage more work in the area of instrument development. To date, no work has been done to either validate or enhance this seven-item measure (L. Kurdek, personal communication, April 8, 2005). Summary A summary of the literature leads to seven content areas of remarriage beliefs. While there may exist other remarriage cognitions, those identified in the extant literature could be placed in one of these seven content areas. There exist various assertions and possibilities as to the etiology and impact of remarriage beliefs. However, empirical research is still needed to confirm and clarify how the beliefs represented in these content areas are related to remarital satisfaction and stepfamily functioning. The Kurdek and 23 Fine (1991) seven-item measure identified an association between beliefs and satisfaction; however, there remain questions regarding the psychometric properties and composition of the instrument. The development and validation of a measure of remarriage cognitions, based on a thorough review of the literature, has yet to be undertaken. As previously indicated a valid measure would facilitate empirically documenting the impact (if any) of remarriage beliefs and assumptions on remarriage quality. Theoretical Background for the Development of the Remarriage Belief Inventory Theory can inform and organize research on the impact of remarriage beliefs and consequently are important to review in the process of instrument development (see Boss, Doherty, LaRossa, Schumm, & Steinmetz, 1993; Gable & Wolfe, 1993). For the proposed study, cognitive-behavioral theory and the life-course perspective provided a number of foundational concepts and assumptions for the proposed project. The following section reviews these theoretical perspectives and then turns to a multi- dimensional model that incorporates the two and specifies future directions for the study of remarriage cognitions. Cognitive-Behavioral Theory The cognitive-behavioral literature has maintained that (a) beliefs about how relationships should function and (b) assumptions about why relationships function in a particular way influence individuals? behavior and emotions in relationships (see Baucom & Epstein, 1990). To the extent that the beliefs positively affect relationship outcomes (e.g., satisfaction), the beliefs are considered functional; to the extent they negatively affect relationship outcomes, the beliefs are considered dysfunctional (Epstein & 24 Eidelson, 1981). Cognitive-behavioral theory allows for various processes in which beliefs can influence relationship quality including direct and indirect pathways (Baucom & Epstein, 1990). This study tests specifically for a direct relationship with satisfaction and adjustment. However, it should be noted that remarriage beliefs may also exert their influence (if any) on behaviors, which in turn could affect reports of remarital quality. General beliefs about relationships have consistently shown associations with relationship quality in first marriages, pre-marital relationships, and same-sex relationships (e.g., Bradbury & Fincham, 1988; Fitzpatrick & Sollie, 1999; Kurdek, 1992; Metts & Cupach, 1990; Moller & Van Zyl, 1991). Whereas the existence and impact of general relationship beliefs has been well documented, research on distinct remarriage beliefs is comparatively sparse in the empirical literature (see Allen et al., 2001). Of the five main cognitive dimensions in the cognitive-behavioral literature (i.e., perceptions, attributions, expectancies, assumptions, and beliefs/standards; for details and definitions of each dimension see: Baucom & Epstein, 1990) that could be explored in remarriages it has been ?beliefs? and ?assumptions? that have received the most clinical attention and what little empirical attention exists to date (e.g., Claxton-Oldfield, 2000; Claxton-Oldfield, Goodyear, Parsons, & Claxton-Oldfield, 2002; Kurdek & Fine, 1991; Visher, Visher, & Pasley, 2003). Although the above referenced studies do not consistently use the same terminology for the same cognition ? some use ?myths?, ?expectations?, or ?standards? interchangeably ? the articles reviewed for this paper focused on two specific types of cognition: beliefs and assumptions. In adopting the following definitions of beliefs and assumptions it should be noted that care was taken to review and incorporate into this dissertation project articles that may have used other 25 terms but which were used in such a way that was consistent with the cognitive- behavioral definitions. Assumptions refer to cognitions about how certain types of people behave (i.e., new partners, stepparents, etc.), how relationships usually work (i.e., remarriages and stepfamilies), and the way in which one sees oneself in particular roles (i.e., the second wife, the stepparent, etc.) (Baucom & Epstein, 1990). Standards differ from assumptions in that they are beliefs about how people, things, and relationships should be rather than how they are (Baucom & Epstein, 1990). In a study of both heterosexual and homosexual couples, Kurdek (1992) demonstrated that assumptions and standards are related but distinct cognitions which each influence relationship satisfaction. As the literature utilizes the terms ?standards? and ?beliefs? interchangeably (i.e., Baucom & Epstein, 1990) the paper will continue to use the term ?belief? when referring to cognitions about the ways people, things, and relationships should be. This decision was made in light of ?belief? being used more often than ?standard? in the literature reviewed. Because of unique stressors and issues experienced in remarriages, it can be hypothesized that specific remarriage beliefs and assumptions regarding these issues may contribute to marital quality in higher-order unions. This is not to imply that remarrying individuals are more prone to dysfunctional beliefs. While this may appear to be an unstated assumption in the clinical literature, no empirical study has looked at nor confirmed this possibility. Rather, the emphasis of the Remarriage Belief Inventory is on different and unique beliefs. What is known is that remarriages, particularly those that include stepchildren, deal with a distinct set of attributes and issues that are truly unique (Adler-Baeder & Higginbotham, 2004). The clinical literature identifies the existence of 26 various remarriage beliefs related to these unique issues and clinicians have suggested that the endorsement of these remarriage beliefs influences marital satisfaction (see Ganong & Coleman, 2004; Visher, Visher, & Pasley, 2003). The present study tests for this asserted direct influence in a non-clinical sample of individuals in remarriages. Although from a cognitive-behavioral perspective remarriage beliefs may also influence satisfaction in other ways (e.g., via behaviors), additional studies that include measures of behavior will be needed to evaluate this possibility. Life Course Perspective The life course perspective is particularly helpful in understanding the experience of remarried families, which typically do not map onto the traditional family development cycle (Rodgers & White, 1993). Remarriage can bring together individuals who are at very different stages in their respective developmental cycles. Crosbie- Burnett (1989) highlights a few of the possible mismatches: ?the newlyweds need time alone while children demand attention. The couple may want to create family cohesion by doing family activities while adolescents are individuating and desiring more time with peers. A mature stepfather who has reared one set of children and is ready to be finished with parenting may find himself with a new baby. The new couple may wish to purchase a home, but college tuition bills for stepchildren need to be paid. (p. 326) Consistent with the theoretical proposition that a family out of sequence increases the probability of later life disruptions, Crosbie-Burnett (1989) warns that trying to balance multiple life cycle stages may cause stress and conflict. 27 One of the strengths of the life course perspective is its ability to allow for the consideration of competing trajectories and intersecting stages (Bengtson & Allen, 1993). These considerations raise awareness regarding the possible etiologies and effects of remarriage beliefs. For example, the lack of experience, not some deficiency in family structure, personality or ability, may account for the level of endorsement on particular beliefs (e.g., childless adults who marry someone with children may lack basic child development knowledge and information on effective parenting strategies). Additionally, research from this perspective allows for differences to be studied between early and later remarriages. This is important as major stresses associated with divorce are thought to dissipate over time (Bray & Kelly, 1998; Hetherington, 1989; Hetherington & Kelly, 2002), unhappy couples select out of the latter remarriage pool by divorcing, and ?cognitions are expected to begin to stabilize and become more consistent as new patterns of relating are developed and new traditions and routines are established? (Fine & Kurdek, 1994, p. 30). The Multidimensional Cognitive-Developmental Model The multidimensional cognitive-developmental model (Fine & Kurdek, 1994) integrates key concepts from the two theories discussed above into a four dimensional model. The dimensions include (a) units in the family system, (b) types of cognitions, (c) continua of adjustment, and (d) developmental stages of the family system. The first dimension conceptualizes the stepfamily as a tiered system. This allows for one-person units (e.g., stepfather, mother, child, etc.), two-person units (e.g., the mother-stepfather marital subsystem), three-person units (e.g., the mother-stepfather-child residential 28 subsystem), and four-person unit (e.g., the mother-stepfather-child-nonresidential father system). The second dimension includes various types of cognitions used in the cognitive- behavioral literature that can also be examined in remarriages and stepfamilies. The third dimension depicts adjustment along a continuum from maladaptation to adaptation. In the model, cognitions and adjustment are reciprocally related just as adjustment of each unit member is reciprocally related to the other members of the respective unit. The fourth and final dimension portrays each member of the system as an information- processing organism which attempts to make sense of the experiences stemming from a stepfamily life course. Fine and Kurdek (1994) argue that ?cognitions, and the link between cognitions and adjustment, change over the life span of the stepfamily through such stages as dating and courtship with the eventual stepparent, cohabitation, early remarriage, middle remarriage, and late remarriage? (p. 17). Fine and Kurdek (1994) refer to beliefs as ?standards? which are defined as cognitions referring to how things should be. They also explain that, ?standards represent an ideal comparison level against which actual experiences are compared. Whether the resulting comparison meets or exceeds the standard will affect levels of perceived satisfaction? (Fine & Kurdek, 1994, p. 22). Research has indicated that marital satisfaction is linked to individuals? appraisals of whether internal standards of what relationships should be like are met or exceeded (Rusbult, 1983). Standards do not necessarily have to be realistic; however, when standards are derived from beliefs of how relationships typically function in intact nuclear families the realization of these standards may be increasingly difficult to obtain let alone exceed (see Visher & Visher, 1988). 29 Examples of unrealistic standards include beliefs about functional equivalency to first- marriage families, quick adjustment, and instant love (Visher & Visher, 1988). These standards are considered unrealistic because, according to clinicians, adjustment typically does not occur quickly and stepfamilies have complexities not encountered in nuclear families (Papernow, 1984). These are placed at the maladaptive end of the adjustment continuum because, particularly in the areas of intimacy and cohesion in relationships, when standards are not met frustration may result (Leslie & Epstein, 1988). Fine and Kurdek (1994) define assumptions as cognitions about how types of people behave, how relationships work, and how one sees oneself in particular roles. These assumptions are thought to come in part from one?s family of origin (Epstein, Schlesinger, & Dryden, 1988). For example, if diverse family forms were condoned by one?s parents then the probability of having optimistic beliefs about remarriage is increased. In contrast, if parents condemned all but traditional marriage then divorce may be viewed as a failure. Finally, Fine and Kurdek (1994) posit that beliefs may change over the life span. They explain: Particularly in well-functioning stepfamilies, family members will (a) develop increasingly clear role perceptions, (b) adopt more realistic expectations of stepfamily life, (c) acquire more balanced attributions regarding the role of living in a stepfamily as a cause of pleasant and unpleasant family events, and (d) be less likely to hold the assumption or standard that stepfamilies are functionally equivalent to first families. (p. 30) Hetherington?s (1989) work distinguishing between the early and later stages of remarriage supports the Fine and Kurdek (1994) hypothesis. Hetherington (1989) 30 indicates that the beginning years of a remarriage are devoted to adapting to the new situation. Negotiation of family roles, including issues of clarity and fit, occur in the later stages. The degree of consistency in beliefs between those in subsystems is expected to increase over time as well. Dysfunctional stepfamilies are hypothesized to either have changes in the opposite direction (i.e., less congruence) or to not show the developmental changes. To summarize, the multidimensional model was presented by its? authors to pull together assumptions across theories and articulate relationships among variables in the context of stepfamily functioning. Drawing from prominent family and individual theories the authors provide a framework that not only address the unique stresses and challenges of stepfamilies but also offer testable hypotheses to guide future research. The framework?s characteristics make it a useful resource in the future study of remarriage beliefs. First, it acknowledges the interrelated nature of various remarriage cognitions. Second, the impact of remarriage cognitions on an aspect of remarital quality is the focus of the model. Third, the model allows for a bi-directional influence between beliefs and adjustment. Fourth, the model acknowledges that the stepfamily is not a static unit and that changes will occur over the life span, including more role-clarity and more realistic beliefs and expectations. Fifth, the model differentiates between adaptive and maladaptive adjustment and identifies cognitive pathways that influence feelings of stress and satisfaction. It should be noted, however, that the multidimensional cognitive-developmental model has never been evaluated and limitations have not been addressed (i.e., the role/influence of behaviors is not in the model). The authors openly acknowledge one of 31 the unresolved limitations of the model. They state, ?Considerable work needs to be done before the constructs in the model, particularly cognitions, can be reliably and validly assessed? (Fine & Kurdek, 1994, p. 31). No known efforts have been aimed to overcome this limitation (L. Kurdek, personal communication, April 8, 2005). The development of the Remarriage Belief Inventory is, in part, an attempt to remedy this limitation by providing an instrument that can be used to evaluate the propositions advanced by the multidimensional cognitive-developmental model. Research Objectives and Questions The primary objective of this study was to validate the factorial structure of the Remarriage Belief Inventory?s (RMBI) seven latent variables and association with remarital quality. The first step in this validation process required a testing and refining of the factorial structure for each RMBI subscale. It was hypothesized that through confirmatory factor analysis (CFA) various goodness-of-fit indices would support the RMBI as a seven-factor measure. In CFA one tests the validity of factorial structures by determining the ?extent to which items designed to measure a particular factor actually do so? (Byrne, 2000, p. 99). A second objective was to assess if the RMBI was related to remarital quality. In light of the extant literature, it was expected that the beliefs Adjustment comes quickly, Stepfamilies are second-class, and Children are priority would be negatively associated with remarital satisfaction. The associations of the other beliefs and remarital satisfaction were not hypothesized due to the lack of clear direction from the literature. Of particular interest in this study was the theorized influence of individuals? remarriage beliefs on remarital quality. Previously cited research and theory also elicited the question of 32 equivalence of the RMBI?s causal structure across different subgroups of remarried couples such as between: (a) Remarried individuals with children and remarried individual without children, (b) couples remarried less than 3 years and couples remarried more than 3 years, and (c) males and females. With the exception of the gender comparison, the literature suggests there may be differences between subgroups. Compared to those without children, it was expected that those with children would have stronger associations between Adjustment comes quickly, Stepfamilies are second-class, and Children are priority and remarriage satisfaction. Compared to those married more than 3 years, it was expected that those married less than 3 years would have stronger associations between all the RMBI factors and satisfaction as they have had less time to develop realistic beliefs. Finally, it was hypothesized that the Remarriage Belief Inventory would explain a unique portion of the variance in remarital quality beyond that accounted for by the Relationship Belief Inventory. 33 METHOD Participants Participants for this study included subscribers of the Stepfamily Association of America?s e-Newsletter, the Smartmarriages.com listserv, and a subsample of the Utah Newlywed sample. Details of each sample are presented below. Your Stepfamily Magazine e-subscription list. Your Stepfamily Magazine is the official on-line magazine of the Stepfamily Association of America (SAA). SAA is a national non-profit organization dedicated to providing support and guidance to families with children from previous relationships ? stepfamilies. A 12 month subscription costs $14.95. Currently there are 1,449 subscribers. Subscribers are primarily women (82%). Eighty-one percent have children ages 5-18. They are well educated (80% attended college) and the majority of participants are employed full-time (72%). Forty-seven percent have annual household incomes that exceed $75,000. Data on ethnicity were not available from the magazine?s publishers. Six issues of Your Stepfamily Magazine are emailed to subscribers each year as well as periodic ?e-blasts? on miscellaneous topics Two-hundred and forty five individuals completed the on-line questionnaire. The sample was approximately three quarters female (n = 189). Fifty-six men (22% of sample) also participated. The sample had an average age of 40.78 years (SD = 8.82) with the median age at 39 years. The majority of the sample were Caucasian (93%), well-educated (66.3% had a four-year degree and/or postgraduate schooling), and had a 34 household income over $75,000 (68.2%). One-hundred and seventy-six participants (71.8%) remarried after a divorce, six (2.4%) remarried after the death of a spouse, and sixty-three (25.7%) were not previously married but married someone who had been. Table 1 displays family compositions following remarriage. Sixty-six percent of participants (n = 162) had a child(ren) from a previous relationship and 85.3% of participants (n = 209) were married to someone who had a child(ren) from a previous relationship. The sample included one hundred twenty-seven (51.8%) individuals in complex stepfamilies (both partners had children from previous marriages/relationships) one hundred seventeen individuals in simple stepfamilies (47.8%), and one individual in a childless remarriage (0.4%). The average length of remarriage was 4.98 years (SD = 5.30) with a median of three years. Table 1 Family Composition of the Your Stepfamily Magazine Sample Does your partner have children from a previous marriage/relationship? Yes No Total Yes 127 35 162 Do you have children from a previous marriage/relationship? No 82 1 83 Total 209 36 245 Smartmarriages.com listserv. The smartmarriages.com listserv is managed by the Coalition for Marriage, Family and Couples Education. The coalition serves as an ?information exchange and clearinghouse to help couples locate marriage and relationship courses; to help mental health professionals, clergy and lay educators locate training programs and resources; to connect those with an 35 interest in the continuing development of the field; to support community initiatives, legislation and research; and to promote the effectiveness of the courses and increase their availability in the community? (www.smartmarriages.com). The Coalition is also non-denominational and non-partisan. Approximately 9,000 email addresses are subscribed to the smartmarriages.com listserv. The subscription process only requires a valid email address. Consequently, demographic information is not available on individuals who subscribe. Ninety-nine individuals completed the on-line questionnaire. This sample was also three quarters female (n = 75). The sample had an average age of 46.68 (SD = 10.45) years with the median age at 47 years. Although the sample was predominantly Caucasian (82.3%) there was greater representation of African-American?s (7.1%) and Hispanics (3%) as compared to the other on-line sample. Sixty percent of the sample had more education than a bachelor?s degree and 65.7% had a household income over $75,000. Seventy-one participants (71.7%) remarried after a divorce, four (4.0%) remarried after the death of a spouse, and twenty-four (24.2%) were not previously married but married someone who had been. Table 2 displays the family compositions of participants. Seventy-four percent of participants (n = 73) had a child(ren) from a previous relationship and 75% of participants (n = 74) were married to someone who had a child(ren) from a previous relationship. The sample included participants from fifty- five (55.5%) complex stepfamilies, thirty-seven simple stepfamilies (37.4%), and seven remarriages with no children (7.1%). The average length of remarriage was 8.02 years (SD = 8.46) with a median of 4.75 years. 36 Table 2 Family Composition of the Smartmarriages.com Sample Does your partner have children from a previous marriage/relationship? Yes No Total Yes 55 18 73 Do you have children from a previous marriage/relationship? No 19 7 26 Total 74 25 99 The Utah Newlywed Study ? Remarried sample. The Utah Newlywed Study was a joint venture by researchers at Utah State and the Utah?s Governor?s Commission on Marriage. A random sample of newlywed couples was chosen by selecting every fourth marriage license on file at the state?s Department of Health, from January to July, 2002. A 38-item survey was mailed to a random sample of 2,823 newlywed couples (5,646 individuals). The survey consisted of several demographic questions as well as questions pertaining to marital satisfaction, marital adjustment, and problem areas in marriage. Of the 2,823 surveys mailed, 282 were undeliverable, 12 couples refused to participate, 1,519 did not respond, and 1,010 couples completed and returned surveys. Of the 1,010 surveys returned, 19 contained data for only the husband or wife, while 991 contained data for both husbands and wives. The total response rate for this sample was 40 percent. Participants in the Utah Newlywed Study included husbands and wives between the ages of 16-87. Ages of the wives ranged from 16 to 85 years (M = 27.03, SD = 9.96) while ages of the husbands ranged from 17 to 87 years (M = 29.03, SD = 10.49). The majority (69%) of the participants, however, were between the ages of 20-30, with the median age for wives being 23, and the median age for husbands being 25 years. Length 37 of marriage for couples in the study ranged from two to nine months with an average of five months. For the current study on the Remarriage Belief Inventory only those individuals who indicated being in a remarriage and or stepfamily were invited to participate (n = 622). This was 31% of the sample that participated in the Utah Newlywed Study. Of those 622 who were invited to participate in the Remarriage Belief Study, 217 individuals returned questionnaires (response rate = 35%). Forty-four percent of these participants were males. The mean age was 40.91 (SD = 14.07) with the median at 38 years. The sample was predominantly Caucasian (91.2%) with Hispanic (3.7%) as the only other ethnicity to account for more than two percent of the sample. In contrast to the on-line samples, the Utah participants had less education (only 30% had a bachelor?s degree and/or postgraduate education) and only 27.2% of the sample had household incomes exceeding $75,000. One hundred forty-three participants (65.9%) remarried after a divorce, eighteen (8.3%) remarried after the death of a spouse, and fifty-six (25.8%) were not previously married but married someone who had been. Table 3 displays the family compositions of Utah participants. Sixty-five percent of participants (n = 141) had a child(ren) from a previous relationship and 61.3% of participants (n = 133) were married to someone who had a child(ren) from a previous relationship. The sample included participants from ninety-nine (45.6%) complex stepfamilies, seventy-six simple stepfamilies (35%), and forty-two remarriages (19.4%) with no children. The average length of remarriage was 2.60 years (SD = .30) with a median of 2.58 years. 38 Table 3 Family Composition of the Utah Sample Does your partner have children from a previous marriage/relationship? Yes No Total Yes 99 42 141 Do you have children from a previous marriage/relationship? No 34 42 76 Total 133 84 217 Procedure Permission was granted from the Auburn University Office of Human Subjects to test the validity of the Remarriage Belief Inventory (IRB #04-082 EP 0406). Procedures followed best-practices for mail and internet surveys (see Dillman, 2000). On-line sample. The administrators of Your Stepfamily Magazine and the Smartmarriages.com listserv initiated an email to their subscribers. The email provided a link to an on-line questionnaire and invited individuals in remarriages and/or stepfamilies to participate in a study on remarriage beliefs by clicking on the link (Appendix A). The email also indicated that participants would receive an e-gift certificate if they choose to participate. Best practices in survey methodology emphasize the use of modest ($1-$5) cash incentives to increase the response rate. Dillman (2000), who authors the book "Mail and Internet Surveys: The tailored design method", encourages researchers who use internet-based surveys to deliver incentives by e-mail. Consequently, each participant who completed the on-line questionnaire and provided their e-mail address received, via email, a $5 e-gift certificate to Amazon.com. 39 As this portion of the study was completely paperless, the link immediately took participants to a web-page which contained the information typically found in the informed consent form (Appendix B). It explained the (a) purpose of the study, (b) the volunteer nature of the questionnaire, (c) planned use of the data, and (d) contact information if they had questions or concerns. To complete the questionnaire, participants had to click on a button indicating that they had read the informed consent information. The on-line questionnaire was hosted by Surveymonkey.com ? an on-line, professional, survey company. When data are submitted on Surveymonkey.com they are immediately stored in a secure on-line database. The database is password protected. Participants respond by typing/clicking on their desired answer. They are free to quit the questionnaire at anytime or to choose not to click "submit" upon finishing the questionnaire. As follow-up can improve response rates by 20-40% (Dillman, 2000) a reminder email was sent out by the administrators of Your Stepfamily Magazine and the Smartmarriages.com listserv (see Appendix C). The reminder email was sent out approximately two weeks after the initial invitation. The reminder email included the link to the on-line questionnaire and once again mentioned the e-gift certificate to Amazon.com. The Utah Newlywed Study ? Remarried sample. This sample did not have email addresses so hardcopies of the invitation and questionnaire were sent to their mailing addresses instead (Appendix D). In place of the electronic incentive, a cash incentive 40 was given to these participants. A reminder postcard was sent to the couples approximately two weeks after the questionnaire was mailed (see Appendix E). Measures A variety of instruments were included in the survey including: the Remarriage Belief Inventory (RMBI; Appendix F), the Kansas Marital Satisfaction Scale (Schumm, Paff-Bergan, Hatch, Obiorah, Copeland, Meens, & Bugaihis,1986; Appendix G), a modified version of the Marital Adjustment Test (MAT; Locke & Wallace, 1959; Appendix H), and the Relationship Belief Inventory (RBI; Eidelson & Epstein, 1982; Appendix I). Additionally, various demographic questions were included (Appendix J). Before the individual instruments are detailed an overview of the RMBI?s development is first presented. The Remarriage Belief Inventory (RMBI). In constructing a new instrument, a review of the literature and applicable theories is essential for establishing content and construct validity (Gable & Wolfe, 1993). From the domain-referenced approach to instrument development described by Anderson (1981), existing research and theory are the building blocks from which constructs are operationalized and measures are developed. In applying Anderson?s technique to the development of affective instruments, Gable and Wolfe (1993) explain that the review of literature leads to domain or content categories, which the instrument developers build into the instrument on an a priori basis. Then, following data collection using the instrument, these categories become the constructs evaluated by the instrument. A similar process was undertaken with the development of the Remarriage Belief Inventory. First, a review of the clinical and empirical literature on remarriage cognitions was undertaken. Secondly, prominent, 41 applicable, and previously utilized theories germane to remarriages and stepfamilies were then reviewed. Subsequently, content areas for the unique remarriage beliefs were organized and constructs were operationalized. The development of the Remarriage Belief Inventory (RMBI) was undertaken to produce an instrument that assessed unique remarriage beliefs and assumptions. It was designed for use in (a) research of couples in remarriages and/or stepfamilies in order to further elicit knowledge about the increased divorce risk for remarriages and (b) program work with couples preparing for remarriages or living in stepfamilies. The RMBI has been under development for a number of years and has gone through numerous phases. As indicated above, the first phase was a review of the research and clinical writings related to remarriage and stepfamily living. A special emphasis was placed on those beliefs and assumptions that clinicians noted as prevalent and salient in remarriages. Forty-three items representing five content areas (adjustment, the new partner, structure, priorities, and the past) comprised the original instrument. Following accepted practices of instrument development, the preliminary instrument was subjected to peer evaluation - by University faculty in a Human Development and Family Studies department - before a pilot test was undertaken (DeVellis, 1991). The second phase of the RMBI?s development occurred through piloting testing with a sample of undergraduate students attending a public university in the Southeast. Methodology literature portrays pilot testing as an imperative step in instrument development (DeVellis, 1991; Kirchhoff, 1999). Two of the primary purposes of pilot testing are to identify those items that are misunderstood and to prevent spending valuable resources on a study that uses an instrument that is not reliable (Pett, Lackey, 42 Sullivan, 2003). It was determined that young adults were an acceptable sample for initial validity testing since it is assumed that all adults have some belief or notion about how stepfamilies function based on their experiences within our society. Five hundred forty-six questionnaires were returned. The sample was predominately white (91%) and female (85%). The mean age of participants was 20.5 years. The pilot test confirmed five factors through a principle component factor analysis with promax rotation which assumes correlated factors. These factors were labeled: (1) Quick adjustment - assumed stepparent authority and attachment, (2) Partner is better than previous partner, (3) Stepfamilies are second-class, (4) Children are the priority over the couple relationship, (5) Past should stay in the past. Eliminating items that cross-loaded or loaded less than .45 resulted in a 24-item version of the RMBI with an alpha reliability coefficient of .78. Each factor (a) satisfied Kaiser?s (1958) criterion of eigenvalues greater than 1.00, (b) accounted for an appreciable percentage of total score variance, and (c) had items that principally loaded on one factor (Higginbotham & Adler-Baeder, in-press). The preliminary version of the RMBI and results from the pilot study were subsequently presented at two national conferences (NCFR in 2003 and AAMFT in 2004). Consistent with recommendations for instrument development, feedback was solicited from scholars, clinicians, and pilot participants (Pett, Lackey, & Sullivan, 2003). The feedback included a number of recommended modifications in the pilot 43-item version of the RMBI. Many of the suggestions pertained to the readability of individual items. More substantial recommendations included splitting one of the constructs and adding a seventh construct to assess beliefs related to finances. It was recommended that 43 we split #2 in light of the items encompassing beliefs about both the partner and the relationship. After another review of the literature to find support for these suggestions and additional factor analyses the recommendations were incorporated. Finally, it was recommended that the complete RMBI be tested using a sample of individuals who live in either a remarriage or stepfamily. This recommendation was supported by literature which indicates that the internal structure of instruments, like the RMBI, depend on the sample being studied (Streiner, 2003). Consequently, to assess the internal and external validity of the RMBI for research and interventions for remarried couples, the factorial structure of the RMBI would need to be tested with data from remarried couples. The RMBI tested in this study originally included 48 items as the empirical indicators of the seven constructs. Each individual construct represented one of the seven content categories derived from the literature review (i.e., adjustment, the new partner, the new relationship, past partners, structure, priorities, and finances). Each construct was operationalized in the form of a statement and the indicators/items were all the beliefs and assumptions gleaned from the review of literature. The seven constructs are: (1) Adjustment comes quickly, (2) Stepfamilies are second-class, (3) Children are priority, (4) Past feelings should end, (5) Partner is perfect, (6) Success is slim, and (7) Finances should be pooled. Most of the 48 items were also included in the 43-item version that was used in the initial pilot test. However, many of these items were reworded for clarity. Additionally, items pertaining to finances were added. Appendix K groups the individual RMBI items according to their respective content categories as conceptualized by the authors. Confirmatory factor analysis determined the final number and ultimate placement of these items (see Results section). 44 Demographics. Multiple demographic questions were asked including age, gender, educational attainment, income, and ethnicity. The questionnaire also included questions pertaining to respondents? family type (e.g., complex, simple, or remarriage without children), relationship history (e.g., time together), and children. Kansas Marital Satisfaction Scale. The Kansas Marital Satisfaction Scale (KMSS; Schumm et al., 1986) was developed to evaluate an individual?s satisfaction with their spouse, with their marriage, and with their overall relationship. Satisfaction with each item is indicated on a seven-point scale ranging from extremely dissatisfied to extremely satisfied. Taken together, the three items provide an overall evaluation of the relationship. The reliability of the KMSS has been relatively high and consistent over time, with alpha coefficients ranging from .89 to .97 and intercorrelations among items ranged from .93 to .95 (Mitchell, Newell, & Schumm, 1983; Schumm, Bollman, Jurich, & Hatch, 2001). The alpha for this study was .98. Adjustment. To measure agreement on various issues a modified version of the Locke Wallace Marital Adjustment Test (MAT; Locke & Wallace, 1959) was utilized. The reliability coefficient of the nine-item measure is .88. On a five point scale ranging from ?almost always disagree? to ?almost always agree? respondents indicate the approximate extent of agreement in their relationship on nine items (e.g., Handling finances, religious matters, demonstrations of affection, etc.). General Relationship Beliefs. The original Relationship Belief Inventory (RBI; Eidelson & Epstein, 1982) covers five general relationship beliefs: (1) disagreement is destructive, (2) mindreading is expected, (3) partners cannot change, (4) sexes are different, and (5) sexual perfectionism. Each belief/scale is assessed by 8 items. The 45 RBI has been found to have coefficient alphas for the five scales that range from .72 to .81, with intercorrelations among the scales varying from .17 to .44 (Eidelson & Epstein, 1982). The present study will utilize the abbreviated version of the RBI found in Kurdek?s (1992) study on assumptions and standards/beliefs. Alphas for the four subscales ranged from .63 to .78. The Sexes are different subscale was not included due to published concerns regarding its psychometric properties (Kurdek, 1992). 46 RESULTS The evaluation of the Remarriage Belief Inventory?s (RMBI) factorial structural consisted of three phases. Phase one included confirmatory factor analyses for each of the seven theorized multi-indicator latent constructs. In phase two the full seven- construct model was tested. The sample used in phases one and two comprised the 344 individuals in remarriages who participated via the internet. Phase three was a cross- validation assessment of the final factorial structure using the Utah participants as the independent sample (N = 217). The next set of analyses assessed whether the RMBI is related to remarital quality. Structural equation models, with the seven RMBI factors as exogenous variables and satisfaction and adjustment as endogenous variables, were constructed and tested. Analyses were also conducted to test for invariance in the structural equation models across groups of remarried adults. These analyses were performed to evaluate whether the relationships between remarriage beliefs and remarital quality were consistent for individuals in complex versus simple stepfamilies and for those individuals who had been married more than three years versus those with less than three years of remarriage experience. Additional tests compared the RMBI and RBI in terms of their unique and combined ability to predict satisfaction and adjustment in remarriages. Results of each set of analyses are presented below. 47 Testing for the Factorial Validity of the Seven Individual Constructs Measurement models for each construct were created and tested using SPSS 11.5 and AMOS 5.0. For each of the seven remarriage beliefs, the first model tested included all of the indicators that were originally theorized to measure that given belief. Resulting descriptive statistics, goodness-of-fit indicators, and regression weights were then examined for each measurement model. Items with nonnormative distributions were eliminated as items with a significant nonzero skewness and/or kurtosis lead to (a) underestimations of fit indexes and (b) underestimation of standard errors of parameter estimates (Holye, 1995). As a consequence of underestimated standard errors the ?tests of parameter estimates will not be trustworthy? (Holye, 1995, p. 63). The degree to which the hypothesized model fit the data was measured by a number of goodness-of-fit indices. In structural equation modeling there is ?no single statistical test of significance that identifies a correct model given the sample data? (Schumacker & Lomax, 1996, p. 120). Consequently, four widely used and commonly accepted fit indicators were utilized: the chi-square statistic (??), the goodness-of-fit index (GFI), the comparative fit index (CFI), and the root-mean-square-error of approximation (RMSEA) (Byrne, 2001; Schumacker & Lomax, 1996). Non-significant chi-square values, a GFI and a CFI greater than .90, and an RMSEA less than .08 each indicate a well fitting model (Byrne, 2001). Given findings of an inadequate goodness-of-fit indicators, efforts were taken to detect and eliminate sources of poor fit in the models. These efforts included evaluating the feasibility of the parameter estimates and the statistical significance of the parameter estimates (Byrne, 2001). Items that exhibited unreasonable estimates such as negative 48 variances were excluded as they indicate either an insufficient sample size and/or estimates that are highly correlated. The test statistic of statistical significance for parameter estimates is the critical ratio (which is computed by dividing the parameter estimate by its standard error). The critical ratio needs to be >?1.96 before the hypothesis (that the estimate equals 0.0) can be rejected (Kline, 2005). Byrne (2001) suggests that given adequate sample size and in the interest of scientific parsimony, nonsignificant parameters should be deleted from the model. For each of the seven theorized beliefs the considerations outlined above were utilized to create, test, and refine measurement models that would then be incorporated into the full seven-factor RMBI model. A total of fourteen measurement models were evaluated - two for each factor. For each factor the first model included all the items thought to serve as indicators (see Appendix K). None of the items had significant nonnormal distributions. Consequently, to determine which items to retain for a parsimonious second measurement model I looked for items with coefficients greater than .30 (Hair, 1998) and significant critical values. As CFA is theory driven, care was also taken to ensure that the final indicators were ?sufficiently similar to the judgmentally developed content categories to support the construct validity of the instrument? (Gable & Wolf, 1993, p. 107). Individual items were reviewed before exclusion to ensure that their removal would not compromise or change the theoretical make-up of each factor. This process resulted in the retention of 22 items. Descriptive statistics of the 22 retained items and their respective standardized regression weights are listed in Table 4. 49 Table 4 Descriptive Statistics of the 22 Items Retained after Confirmatory Factor Analysis of each Subscale (N = 344) Range Mean SD Skew Kurtosis ? Adjustment 21. Love should develop quickly between the child and the stepparent. 5.0 2.174 .936 .299 -.514 .81* 26. Adjustment to living in a stepfamily should occur quickly 5.0 1.920 .914 .738 -.068 .76* 28. Stepfamily members should feel close to one another soon after the stepfamily forms. 5.0 2.201 .949 .452 -.292 .87* 29. Stepparents should assume intimacy and authority with the children soon after the stepfamily forms. 5.0 2.325 1.106 .388 -.832 .75* Finances 33. In remarriages, incomes and paychecks should be ?pooled?. 5.0 3.477 1.031 -.344 -.176 .90* 11. Financial resources in a remarriage should be combined. 5.0 3.608 1.027 -.319 -.372 .83* 44. In a remarriage, there should be a distinction between ?mine? and ?your? financial resources.? 5.0 3.637 1.008 -.218 -.598 .72* Partner 17. A new spouse should be a better marriage partner than the one he/she replaces. 5.0 3.541 1.018 -.428 .009 .61* 20. In comparison to ex-spouses, a new spouse should be more ?in-tune? to the quality of the spousal relationship. 5.0 3.277 .922 -.293 .319 .67* 31. A new spouse should be everything the problematic old spouse was not. 5.0 2.280 1.059 .412 -.551 .51* 34. A new spouse should be more understanding than a previous spouse. 5.0 3.201 .899 -.365 .474 .55* 50 Priority 23. Fulfilling the desires of a new spouse should come before fulfilling the desires of biological children.? 5.0 3.304 1.097 -.115 -.563 .56* 47. Wishes of the children should take priority over the wishes of the new spouse. 5.0 2.397 .945 .395 .061 .84* 48. Giving attention to the children is more important than giving attention to the new spouse in a remarriage. 5.0 2.225 .974 .554 .052 .82* Success 36. When a person gets remarried, it is likely that their new spouse will have some of the same flaws they saw in their previous spouse. 5.0 3.262 .955 -.364 -.039 .59* 27. People who have divorced are likely to divorce again. 5.0 3.163 1.032 -.469 -.261 .74* 41. People in remarriages are likely to make the same mistakes they made in previous marriages. 5.0 3.104 .913 -.196 -.128 .78* 35. People in remarriages are likely to repeat the same patterns/behaviors as those in their previous marriage(s). 5.0 3.343 .980 -.519 -.062 .70* Stepfamily 12. A stepfamily cannot offer children everything that a biological family can. 5.0 2.372 1.236 .543 -.814 Unidentified 24. All things considered, a stepfamily is a poor substitute for a biological family. 5.0 1.884 1.089 1.148 .607 Unidentified Past 4. Emotional connection/feelings to an ex-spouse should end with a new marriage. 5.0 3.346 1.312 -.213 -1.187 Unidentified 46. Emotional ties to the previous marriages/relationships should be severed prior to a remarriage. 5.0 3.456 1.227 -.381 -.825 Unidentified ? Item is reverse coded * p < .001 51 All retained items had a ? > .50, which surpasses the value needed (.30) to identify a significant loading based on a sample size of 350 and a power level of 80% (Hair, 1998). Regression weights for the indicators of the Stepfamily is second class and Past feelings should end subscales were unidentified due to insufficient degrees of freedom in the final measurement model. Insufficient degrees of freedom (resulting from an equal or lesser number of sample moments than distinct parameters to be estimated) also accounts for the absence of goodness-of-fit indices for four of the seven final measurement models. For the three models that retained four items, a solution and probability level was determined and accepted (see Table 5). Table 5 Goodness-of-Fit Indices for 4-Item Factors ?? df p GFI CFI RMSEA Adjustment 2.910 2 .233 .996 .999 .036 Partner 6.459 2 .040 .990 .977 .081 Success 1.641 2 .440 .998 1.000 .000 The decision to accept the remaining four models without goodness-of-fit indicators was based on several factors. First, the retained items are representative of their respective construct. Second, in the initial and subsequent model the retained items consistently demonstrated high loadings on their respective factor. Pending the confirmatory factor analyses on the full measurement model and the resulting goodness- of-fit statistics, modification indices, estimated regression weights, and critical ratios the seven measurement models are accepted as constituted in Table 4. 52 Testing for the Factorial Validity of the Seven Factor Model With the same sample of 344 on-line participants the full seven-factor model was evaluated through confirmatory factor analysis. The full model with standardized estimates is depicted in Figure 1. Support for the factorial structure is provided by the estimates and goodness-of-fit indices. Standardized regression weights range from .525 to .899 and critical ratios for all of the regression coefficients were significant (p < .001). The measures of goodness-of-fit indicate a strong fit of the data to the model with a GFI of .932, CFI of .965, and RMSEA of .037. Further support for the model can be inferred from the absence of large modification indices (the largest meaningful MI was 10.504). Finally, support for seven discrete constructs is provided by low correlations among factors (r = .070 to .215). As no additional modification appeared warranted, the presented seven-factor model was accepted as the refined Remarriage Belief Inventory (RMBI; Appendix L). 53 Figure 1 Full Standardized Model of the Remarriage Belief Inventory Past Ties (46) e1 Connenction (4) e4 Stepfamilly Substitute (24) e6 Cannot offer (12) e7 Success Repeat (35) e9 Mistakes (41) e10 Divorce (27) e11 Flaws (36) e12 Priority Attention (48) e13 Wishes (47) e14 Desires (23) e15 Partner Understanding (34) e17 Problematic (31) e18 In-tune (20) e19 Better (17) e20 Finances Distinction (44) e21 Combined (11) e22 Pooled (33) e24 Adjustment Authority (29) e25 Close (28) e26 Quickly (26) e27 Love (21) e28 .07 -.19 -.08 .21 .16 .18 .32 -.05 -.10 .00 -.07 -.07 -.32 -.34 .19 -.06 -.02 .26 .44 -.09 .16 chi-square = 277.326, df =188, p =.000, GFI =.932, CFI =.965, RMSEA =.037 .77 .76 .78 .65 .74 .59 .78 .71 .84 .84 .53 .55 .61 .59 .58 .72 .90 .83 .76 .87 .76 .81 Cross-Validating the Factorial Structure with an Independent Sample The sample of remarried couples from the Utah Newlywed Study was used to cross-validate the factorial structure of the RMBI. The Utah sample (N = 217) differed significantly from the on-line sample on multiple demographic variables. As a group, the Utah sample was less educated (F[1, 558] = 125.87, p < .001) and had a lower household income (F[1, 555] = 125.79, p < .001) than the on-line participants. The Utah sample, as 54 a whole, had also been married for a shorter period of time (F[1, 548] = 5.68, p = .018), had been married more times (F[1, 560] = 130.29, p < .001), and was more likely to have married someone who had children from a previous relationship (F[1, 560] = 32.194, p < .001). The Utah sample was also more satisfied with their remarriage. As illustrated in Table 6 both the on-line and Utah samples reported high marital quality. The maximum score possible for the composite satisfaction score was 21; for adjustment; 45. Utah participants had a mean satisfaction score of 18.36 (SD = 3.67) and the distribution was highly skewed (-2.317). On-line participants were less satisfied (F [1, 560] = 13.025, p < .001) even though their average score still fell in the range between ?somewhat? to ?very? satisfied. The groups did not differ significantly on reports of adjustment. Table 6 Group Response Characteristics on Outcome Variables Although the goodness-of-fit indicators were not quite as strong as those computed for the on-line sample, the Utah data did fit the model (?? [188, N = 217] = 272.72, p < .001; CFI = .938; RMSEA = .046). Betas ranged from .460 to .852 and all of the critical ratios were significant (p < .001). Inter-factor correlations ranged from -.063 N M SD Min. Max Kurtosis Skewness Online Satisfaction 344 17.055 4.478 3.00 21.00 1.771 -1.500 Adjustment 344 35.613 6.351 9.00 45.00 .589 -.842 Utah Satisfaction 217 18.365 3.673 3.00 21.00 5.941 -2.317 Adjustment 217 36.800 5.974 9.00 45.00 2.113 -1.134 55 to .416. Only one covariance modification indices surpassed the 10.0 threshold. However, treating the indicated covariance as a free parameter would have only lowered the discrepancy by 11.66. As there was no clear theoretical basis to make this modification, it was not performed. In summary, the data from this independent and statistically dissimilar sample cross-validates and further supports the reliability of the factorial structure of the Remarriage Belief Inventory. Testing for the Validity of the Causal Structure Two structural equation models were constructed to test the RMBI?s causal structure with two aspects of remarital quality. It was determined that the on-line sample would be used to test these models rather than a combined sample of on-line and Utah participants. This decision was based on significant group differences on various demographic variables (discussed above) as well as on scores of marital satisfaction. As a group the Utah sample rated themselves significantly higher on satisfaction (F[1,560] = 13.025, p < .001). The variability of satisfaction scores in the Utah sample was also approximately one fourth of that of the on-line sample (SD = 3.67 versus SD = 4.48). Outcomes on adjustment were also less variable in the Utah sample and higher, although the difference was not significant (F[1, 560] = 4.861, p = .28). These characteristics raised concerns about committing a type II error (failing to find significant relationships that may exist) due to the restriction of range and ceiling effect in the Utah sample. Analyses were still performed with the Utah sample and are included in Appendix M; however, only results from the on-line sample are reported in the text. The first structural equation model evaluated theoretically causal relations between the seven RMBI beliefs and satisfaction. Satisfaction was modeled as an 56 unobserved endogenous variable with the three items from the KMSS (Schumm et al., 1986) as indicators. Goodness-of-fit indices provided support for a well fitting model. Although the ?? (247, N = 344) = 272.72 is significant (p < .001), a finding which is likely an artifact of the large sample size, the other indices indicate a strong fit (GFI = .929; CFI = .980; RMSEA = .031). Estimates for the causal relations between the RMBI subscales and satisfaction are reported in Table 7 Table 7 Summary of SEM Estimates for RMBI Factors Theorized to Predict Satisfaction B SE B ? Critical Ratio Satisfaction <--- Adjustment -.189 .131 -.098 -1.446 Satisfaction <--- Finances .453 .130 .209 3.494* Satisfaction <--- Partner -.138 .256 -.045 -.539 Satisfaction <--- Priority -.119 .118 -.061 -1.009 Satisfaction <--- Success -.563 .144 -.276 -3.897* Satisfaction <--- Stepfamily -.209 .120 -.119 -1.737 Satisfaction <--- Past .059 .101 .038 .586 * p < .001 The second model positioned adjustment, measured by nine indicators, as the unobserved endogenous variable. This second model also demonstrated acceptable levels of fit (?? [406, N = 344] = 272.72, p < .001; GFI = .898; CFI = .945; RMSEA = .039). Estimates for the causal relations between the RMBI subscales and Satisfaction are reported in Table 8. 57 Table 8 Summary of SEM Estimates for RMBI Factors Theorized to Predict Adjustment B SE B ? Critical Ratio Adjust <--- Adjustment -.048 .061 -.058 -.793 Adjust <--- Finances .194 .062 .208 3.137* Adjust <--- Partner -.091 .123 -.069 -.742 Adjust <--- Priority -.068 .056 -.081 -1.210 Adjust <--- Success -.212 .070 -.240 -3.029* Adjust <--- Stepfamily -.006 .059 .007 -.096 Adjust <--- Past -.027 .047 -.039 -.561 * p < .01 Finances should be pooled and Success is slim are the only two beliefs with statistically significant direct effects on both satisfaction and adjustment. A consistent pattern of negative coefficients is observed across all the beliefs with the exception of Finances should be pooled and Past feelings should end (in the model with satisfaction). However, these coefficients are not significant and consequently the pattern could be spurious. Notwithstanding nonsignificant individual effects, taken as a whole, the RMBI factors explain 14.8 percent of the variance in satisfaction. The seven RMBI factors explain 9.6 percent of the variance in adjustment. Testing for Causal Invariance Three tests of invariance were originally planned. These tests were intended to assess for invariant patterns of causal structures across samples of: (a) Remarried individuals with children and remarried individual without children, (b) remarried individuals married less than three years and those married more than three years, and (c) remarried men and remarried women. These tests essentially act as additional cross- 58 validations of the paths and estimated effects of the Remarriage Belief Inventory as well as indicating possible moderating effects of these categories. Drawing from the on-line sample (N = 344) it was not possible to test for invariance between individuals with children and remarried individual without children due to an insufficient number of childless remarriages (n = 8). However, as indicated in Table 9, there was a large enough sample to compare individuals in complex stepfamilies (both partners had children from previous relationships) versus those in simple stepfamilies (only one partner had children from a previous relationship). Although there where sufficient data to compute solutions and probability levels, because the sample is less than 200 in both groups the solution may be unstable (Kline, 2005). Consequently results should be interpreted with caution. Table 9 Frequencies of Stepfamily Type (N = 344) Frequency Percent Cumulative Percent Complex 182 52.9 52.9 Simple 154 44.8 97.7 No children 8 2.3 100.0 Total 344 100.0 As a first step in testing for invariance, two models (one for complex and one for simple stepfamilies) were computed with the seven RMBI factors as exogenous variables and satisfaction as the endogenous variable. The goodness-of-fit indicators used throughout this study indicate that the data from complex stepfamilies fits the model slightly better (?? [248, n = 182] = 316.907, p = .002; GFI = .879, CFI = .968; RMSEA = .039) than the data from simple stepfamilies (?? [248, n = 154] = 338.353, p < .001, GFI = .859, CFI = .953, RMSEA = .049). In comparison to 16.8% of explained variance in 59 satisfaction in the complex stepfamily model, the RMBI only accounts for 13.1% of variance in satisfaction for individuals in simple stepfamilies. To further explore the comparative fit between the two samples the expected cross-validation index (ECVI) was evaluated. The ECVI assesses, in a single sample, the likelihood that a model cross-validates across similar-sized samples from the same population (Browne & Cudeck, 1993). Models having the smallest ECVI values exhibit the greatest potential for replication. Consistent with the other goodness-of-fit indices, the complex stepfamily sample had a slightly smaller ECVI (2.602) than the simple stepfamily sample (3.218). Furthermore, the ECVI from the simple stepfamily solution fell above the 90% confidence interval of the complex stepfamily ECVI (2.372-2.877) suggesting that the two groups may have nonequivalent structures. To determine the source of noninvariance it is necessary to compare individual parameter pairs across groups. However, in the present model by virtue of the sheer number of parameters (112) there was likely to be some group differences in at least one of the parameter comparisons, if not more. Byrne (2001) indicates that to test all parameters ?represents an overly restrictive test of the data? (p. 175). Consequently, only the parameters which would indicate whether the RMBI affects satisfaction invariantly across groups were compared. These parameters consisted of the seven regression weights predicting satisfaction; one from each of the seven RMBI factors. Table 10 depicts the critical ratios for differences between the seven targeted parameters for individuals in complex stepfamilies versus individuals in simple stepfamilies. The critical ratio is the difference between regression weights for individuals in simple stepfamilies and for those in complex stepfamilies, divided by an 60 estimate of the standard error of the difference. With a correct model this critical ratio (a z statistic) has a standard normal distribution and consequently a value greater than?1.96?would indicate the null hypothesis (the difference between the parameters = 0) could be rejected. None of the critical ratios exceeded ?1.96?which indicates that there is invariance between the two groups on all of the seven coefficients estimating the effect of the RMBI factors on satisfaction. However, caution should be taken in accepting this null finding in light of the sample sizes < 200. Table 10 Critical Ratios for Differences between Parameters for Individuals in Complex (n = 182) and Simple (n = 154) Stepfamilies and Individuals Married More (n = 204) or Less (n = 140) than three Years. Complex and Simple Stepfamilies Married More or Less than Three Years Adjustment 1.166 -1.511 Finances -1.499 .024 Partner -.499 .242 Priority -.077 -.185 Success .192 -.425 Stepfamily .568 .972 Past -.140 .002 Next, remarried individuals married less than three years were compared to those married more than three years. For the newlyweds the RMBI explained 17.8% of the variance in satisfaction and their model had a ??(249, n = 140) = 293.783, p = .027; GFI = .859; CFI = .973; RMSEA = .036. For individuals married more than three years the R? = 13.6 and the ?? (249, n = 204) = 327.394, p < .001; GFI = .889; CFI = .968; RMSEA = .039. With the exception of the GFI which incorporates a penalty for additional parameters and can be influenced by sample size (Byrne, 2001), these fit indices provide support for the data fitting the models. The ECVI for the sample married more than three 61 years was less than the comparison ECVI (2.362 versus 3.207). Consequently, additional tests of the causal paths were evaluated for invariance. As summarized in Table 10 each parameter was statistically invariant in that no single critical ratio exceeded ?1.96?. Collectively, these findings indicate that although the fit and R? do differ in magnitude, the individual paths between the RMBI and satisfaction are statistically invariant across the two groups. However, caution should be taken with these findings as the sample of those married less than three years only consisted of 140 individuals. A test of invariance was also planned for males and females. However, insufficient males (n = 80) made the solution not admissible. The model for the females, however, demonstrated excellent fit (??[249, N = 264] = 322.130, p < .001; GFI = .912; CFI = .975; RMSEA = .033). In the model for females, the RMBI accounted for 17.8% of the variance in satisfaction. Testing for Unique Contributions of the Remarriage Belief Inventory Hierarchical regressions were used to assess the variance accounted for by the RBI and RMBI in the remarital quality of the on-line sample (results from the Utah sample are presented in Appendix N). For both dependent variables (satisfaction and adjustment) the four RBI subscales were entered as step 1 and the RMBI subscales were entered as step 2. A summary of both hierarchical regressions is presented in Table 11. The R? for step 1 of the Satisfaction regression was .256. The strongest predictor was the belief that Disagreement is destructive. Step 2 had a R? = .309. This additional contribution of the RMBI in predicting satisfaction (?R? = .053) was statistically significant (F Change [7, 332] = 3.635, p < .001). For adjustment, the R? equaled .235 for step 1. The addition of the RMBI subscales increased the R? to .260. This additional 62 2.5% of variance explained by the full RMBI was not statistically significant (F change [7, 332] = 1.612, p = .131) although two individual subscales had significant estimates (Finances and Success). Table 11 Summary of Hierarchical Regression Analysis for the RBI and RMBI (N = 344) Satisfaction Adjustment B SE B ? B SE B ? Step 1 Disagreement -.647 .082 -.444 ** -.768 .118 -.371 ** Sex .186 .075 .130 * .267 .107 .131 * Mindreading .200 .091 .121 * .046 .131 .020 Change -.373 .077 -.254 ** -.547 .111 -.263 ** Step 2 Disagreement -.611 .081 -.419 ** -.737 .118 -.356 ** Sex .179 .074 .125 * .264 .108 .130 * Mindreading .135 .091 .082 -.019 .133 -.008 Change -.281 .079 -.192 ** -.450 .115 -.217 ** Adjustment -.057 .077 -.038 -.028 .114 -.013 Stepfamily -.178 .108 -.078 .060 .158 .019 Priority -.047 .089 -.024 -.176 .131 -.065 Success -.238 .076 -.159 * -.249 .112 -.117 * Partner .023 .084 .014 -.012 .123 -.005 Past .070 .093 .035 -.034 .137 -.012 Finances .205 .084 .119 * .242 .123 .099 * * p < .01 ** p < .001 63 DISCUSSION The purpose of this study was to refine and test the factorial structure of the Remarriage Belief Inventory (RMBI) ? an instrument developed to assess individual?s beliefs and assumptions regarding remarriages and stepfamilies. The development of the RMBI began with a review of the literature and theory on remarriage quality. The review resulted in the identification of seven content categories of unique remarriage and stepfamily beliefs which have been suggested to impact remarriage quality. To test this assertion the RMBI was created by building the seven content categories into an instrument on an a priori basis. Then, following data collection with the instrument from 344 on-line participants in remarriages and 217 participants from Utah in remarriages, these categories become the constructs evaluated in the models. Using these two samples of individuals in remarriages, this study examined the measurement and validation of the seven factorial models and then assessed how they affect two aspects of marital quality. Each factor was individually evaluated, refined, and evaluated again using accepted criteria for factor analysis. Multiple goodness-of-fit indicators were utilized to determine how well the models fit the data including the chi- square statistic (??), the goodness-of-fit index (GFI), the comparative fit index (CFI), and the root-mean-square-error of approximation (RMSEA). With the intention to refine the original 48 items into a parsimonious yet valid instrument, items were only retained if their coefficients exceeded the acceptable value 64 needed to identify a significant loading. For a sample size of 350 and a power level of 80% a significant loading should be greater than .30 (Hair, 1998). In this study all retained items had coefficients greater than .50. Items with coefficients below .50 were still carefully reviewed before being excluded from subsequent models. Because confirmatory factor analysis is theory driven, at each step in the refinement process care was taken to ensure that the removal of any item would not compromise the construct validity of the factor. The findings from the study support a seven factor structure. The seven factors are: (1) Adjustment comes quickly, (2) Stepfamilies are second-class, (3) Children are priority, (4) Past emotions should stay in the past, (5) Partner is perfect, (6) Success is slim, and (7) Finances should be pooled. Low inter-factor correlations support the a priori conceptualization of the seven remarriage beliefs as being distinct factors. Two independent samples provide consistent support for a seven factor structure using 22 of the original 48 items. Each factor consists of two to four indicators (see Appendix L). Now validated these constructs may be used individually or collectively in future studies involving remarriage cognitions. When modeled collectively to predict remarital quality, the seven factors account for between 10-18% of the variance in satisfaction, depending on the sample. It also appears that the direct effects of the individual RMBI subscales are invariant for individuals in complex and simple stepfamilies as well as for those married more or less than three years. In other words, the effects of causal pathways do not differ depending on the subgroup. However, due to the sample size of these subgroups the solutions may 65 not be stable (Kline, 2005) and consequently additional studies, with larger samples, will be needed to confirm the structural invariance of the RMBI. The Relationship Belief Inventory (RBI) has a demonstrated utility and validity in various samples (e.g., Epstein & Eidelson, 1981; Kurdek, 1992; Moller & Van Zyl, 1991; Metts & Cupach, 1990). However, the RBI has rarely been used to predict remarital quality. The results presented in this study confirm the predictive ability of the RBI in accounting for variance in the satisfaction and adjustment of a remarried sample. Furthermore, the analyses indicated that the RMBI adds to the RBI in explaining the variance in satisfaction. Together, the seven RMBI beliefs explained an additional 5% of the variance in remarital satisfaction beyond that accounted for by general relationship beliefs alone (as measured by the RBI). Two individual beliefs emerged as the strongest predictors of adjustment or satisfaction. While not exclusively entertained by remarried individuals, the first belief, Finances should be pooled, was incorporated into the RMBI in light of the extant literature which suggests financial issues are a primary source of conflict in remarriages, particularly decisions regarding the combining of separate assets and the support of children. Furthermore, communication and decisions regarding these issues are believed to affect remarital functioning (Coleman, Ganong, & Weaver, 2001; Engel, 1999). Two studies are particularly informative on this topic. The first study (Coleman & Ganong, 1989) reported that while most remarried couples in their sample (76%) managed income in a ?common pot,? couples were generally satisfied whether they pooled their finances or retained separate accounts. Pasley, Sandras, and Edmondson (1994) similarly found no differences in types of financial handling and satisfaction with remarriage. 66 In contrast to what may have been expected from the extant research, the results from this study indicate that endorsing the belief Finances should be pooled is positively associated with remarital quality. In fact, the greater endorsement of this belief the greater satisfaction and adjustment. These findings appear to support the hypothesis that pooling resources may result in improved remarriage relationships because economic futures become intertwined and communication is facilitated out of the necessity to cooperate (Fishman, 1983). However, Fishman?s hypothesis is certainly not confirmed by this study either. As will be discussed in more detail below, actual behaviors were not measured in the current study. Consequently, despite the endorsement of beliefs in support of financial pooling, there was no measure for whether these participants actually pool their finances or if the spouses had congruent beliefs. Future studies will need to test for the extent to which believing that Finances should be pooled actually equates to a reality of shared finances. The second statistically significant predictor of adjustment and satisfaction was the Success is slim belief. This factor was designed to capture the belief about the potential for remarriages to be more successful than previous marriages. The belief was operationalized in such a way that a high score would be indicative of a strong belief that success is slim. Although this assumption is not totally ungrounded (the estimates of divorces are slightly higher for remarried couples (Bramlett & Mosher, 2001), remarriages can be very strong, satisfying and stable (Walsh, 2003). Consequently, the belief could be dysfunctional if a self-fulfilling prophecy ensued. Results from this study indicate a negative association between Success is slim and both adjustment and satisfaction. In other words, remarriage quality goes down 67 when individuals believe their remarriage is unlikely to be any better than their first marriage. As this is cross-sectional data, another interpretation of this association is that as remarriage satisfaction and adjustment declines the less participants may believe in a positive future for their remarriage. In terms of the other five beliefs, results from the structural equation models did not generate statistically significant effects on remarital quality. This was an unexpected finding in light of the theories which suggest that relationships beliefs are associated with remarital quality. Furthermore, the clinical literature has long proposed a link between remarriage beliefs and remarital quality. Three possibilities, which may also be considered limitations to the existing study, may account for the null findings of five RMBI factor?s ability to individually predict remarital quality. They may also account for why the two previously discussed beliefs had small, albeit significant, coefficients. The first possibility is that congruence of beliefs between spouses is more predictive than ones? individual beliefs. Second, consistent with other tenants of cognitive behavioral theory, it may be that beliefs exert their influence on remarriage quality in the context of behaviors and that behaviors then have a direct effect on satisfaction and adjustment. This explanation allows for the effects of remarriage beliefs to occur over time. Thirdly, limited variability in the outcome measures may account for the weak associations. To discuss these possibilities I return to the two theories that guided this study. Congruence. The multidimensional cognitive-developmental model was adopted as a guiding framework for this project because it incorporates various concepts that make it particularly useful in the study of remarriage beliefs. For example, the effect of remarriage cognitions in the context of remarriage quality is the focus of the model. This 68 theorized effect was the focus of this study. However, a key concept of the multidimensional cognitive-developmental model was not controlled for nor tested and may account for the finding that five of the RMBI factors had weak to no direct effect on remarital quality. The concept that was unaccounted for was that of ?congruence?. Fine and Kurdek (1994) explain: In multiperson units, the key issue is the extent to which the cognitions held by each member of the subsystem are consistent and in balance with those held by other members of the subsystem. A balanced subsystem is one in which the cognitions of the relevant members are consistent with each other, while an unbalanced subsystem is one in which members? cognitions are dissimilar and incompatible with each other. (p. 22) Only individual endorsement of the seven remarriage beliefs was identified in this current study. Without couple data it is not possible to assess level of dyadic congruence on the RMBI subscales. Yet, if Fine and Kurdek are correct then congruence may be more predictive of satisfaction and adjustment that one?s individual beliefs. Of this possibility Leslie and Epstein (1988) indicate, ?Regardless of whether each family member holds unrealistic beliefs about remarried family life, conflict may occur when there is incompatibility among members? beliefs? (p. 159). The findings from this study give tacit support to the observation that the endorsement of cognitions measured by the RMBI may ?not be inherently good or bad? (Ganong & Coleman, 2004, p. 203). Although some RMBI beliefs may be less realistic than others, in terms of remarital satisfaction it may be that outcomes are most influenced by incompatible beliefs or incongruent assumptions. Future studies, with data from 69 couples, would help clarify the extent to which congruence of remarriage beliefs predicts remarital quality. Behaviors. A second explanation for the estimated effects of RMBI factors on remarital quality may be found in cognitive behavioral theory. In the cognitive literature there remains an ongoing discussion regarding the directionality of influence among cognitions, behaviors, and marital quality. While the multidimensional cognitive- development model does specify a bi-directional relationship between cognitions and adjustment, the interplay with a behavioral dimension is absent from this model. This is a significant oversight, as the literature consistently points to an interaction between cognitions, behaviors, and satisfaction (i.e., Huston, 2000; Karney & Bradbury, 1995). Cognitive behavioral theories describe cognitions and behavior interacting to exert both direct and indirect effects on satisfaction (Baucom & Epstein, 1990). Relationship cognitions are thought to influence not just satisfaction but also how events and stress are perceived and the availability, use, and effectiveness of coping strategies (Cohen and Edwards, 1989; Lazarus and Folkman, 1984). For cognitive behaviorists, relationship cognitions are also an integral part of the sequential flow of behavioral interactions in that an individual?s cognitions are thought to influence one?s behavior and vice versa (Baucom & Epstein, 1990). As an example, consider a man who believes that in his remarriage he should get more attention from his wife than her children receive from her. If he strongly holds this belief and he perceives himself getting less attention than the children he may behave in ways that are predictive of marital dissatisfaction (e.g., criticizing or withdrawing from his wife). However, if his beliefs are flexible and he evaluates the situation as one that can be tolerated or improved, then his reaction is 70 more likely to include constructive behaviors that are predictive of marital satisfaction (e.g., talking with his wife or scheduling and participating in shared activities) (example adapted from Baucom & Epstein, 1990). This example illustrates a process in which beliefs exert a differential effect on marital satisfaction through behaviors. So instead of being manifest in strong significant regression coefficients on satisfaction or adjustment, the influence of RMBI beliefs on remarital quality may be most potent on behavioral variables which then in turn affect satisfaction and adjustment. Support for this hypothesis can be found in the literature. For example, those individuals who endorse the assumption that Adjustment comes quickly may push too hard for a parental relationship with their stepchildren (Gerlach, 2001). Ganong and Coleman (2004) note that stepparents who move too quickly into parental roles usually encounter resistance from their spouses and stepchildren. This resistance then leaves the stepparent ?confused and anxious, and eventually dissatisfied? (p. 205). The words ?eventually dissatisfied? imply that remarriage beliefs may not have an observable direct or immediate effect on remarriage quality. Rather, relationship beliefs may be most influential in terms of impacting the sequential flow of behavioral interactions and that these interactions are in turn predictive of remarital satisfaction and adjustment. Response characteristics. A third explanation for the small direct estimates of the RMBI factors may be the restricted range associated with high reports of satisfaction and adjustment. As illustrated in Table 6 both the on-line and Utah samples reported high marital quality. The high levels of remarital quality restrict the range in which the predictor variables can explain variability. Consequently, strong direct relationships may 71 exist between RMBI factors and relationship quality but may not be identifiable with the present samples. Utility of Internet Although not a focus of the study, this research contributes to the knowledge about on-line methodologies. This study indicates that on-line methodologies have the capacity to gather large samples in a relatively short period of time. Using email and the internet data was gathered from over 300 individuals in remarriages and stepfamilies ? a group which has historically been difficult to identify and access (Coleman & Ganong, 1990). This was done in a period of approximately six weeks. Various methodological constraints have historically limited researcher?s ability gather large samples of stepfamilies (see Coleman & Ganong, 1990). A primary reason is that recruiting a large, representative sample is too costly for most researchers. Coleman, Ganong, and Fine (2000) noted ?it is prohibitively expensive to recruit as many stepfamilies as are needed to examine or control for all relevant structural variables? (p. 1300). Furthermore, remarried couples are considered a more mobile population than other family types and it takes time and money to locate them (Coleman & Ganong, 1990). A second obstacle in generating large representative samples has been observed by clinicians who note that stepfamilies are an ?invisible? population (Visher & Visher, 1979). Despite the prevalence of couples in stepfamilies, many do not willingly or knowingly self-identify as different than first-marriage couples (e.g., Visher & Visher, 1996) and consequently individuals in remarriages and stepfamilies may omit references 72 to ?step? relations or remarriages to avoid negative stereotypes or stigma (Ganong and Coleman, 1984; 1994). In light of these methodological constraints this study incorporated an on-line methodology. The potential cost effectiveness and efficiency of on-line data collection has been deemed an ?untapped tool for research? (Murray & Fisher, 2002, p. 5). It has been observed that ?online surveys have the potential to improve dramatically the ways in which research participants are recruited and data are collected? (Northey, 2005, p. 99). This present study serves as a case-in-point for these observations in the context of remarriage/stepfamily research. On-line subscriptions tap into large groups of individuals who self-identify with the web-based organization or e-list. In the case of ?Your Stepfamily Magazine? potential participants identified themselves, by virtue of their subscription, as individuals who willingly self-identify as being involved in a stepfamily. Furthermore, data was gathered from on-line participants at a fraction of the cost of traditional pen and paper methods. For example, the cost to implement this online study using comparable ?paper? methods would have been $12, 538. Table 12 itemizes the projected postal costs of this study if all 10,449 email address (9,000 from Smartmarriages.com and 1,449 from Your Stepfamily Magazine) were send hardcopies of the evaluation materials instead of emails. 73 Table 12 Cost Estimate for Postage for On-line Sample # of participants Cost/participant Total Cost Description of cost 10449 $.60 $6,269.40 Postage for consent letter and questionnaire packet. 10449 $.37 $3,866.13 Pre-paid postage for the return of the consent letter and questionnaire. 10449 $.23 $2,403.27 Postage for follow-up postcard. $12,538.8 Total cost for postage Of course the cost projections in Table 12 include the estimated postal costs for all Smartmarriages.com subscribers, regardless of marriage type. Because demographic information is not available on the Smartmarriages.com subscribers it is not possible to know who/how many subscribers are in a first marriage versus a higher order marriage. With traditional methodologies, it would be extremely risky to expend thousands of dollars on postage for a mailing list if the researcher did not know how many of the target population, if any at all, would be reached. Online researchers do not have to worry about this ?gamble? for sending an email to 10,000 e-mail address is just as easy as sending an email to 10 recipients. In the present study, there was no cost to the researcher to solicit potential participants or to place the questionnaire at their fingertips. Rather, all 10,499 e-subscribers received information about the study and those who were remarried were invited to log onto the web-based questionnaire. Notwithstanding the advantages of on-line surveying, concerns and limitations still surround this emerging methodology (Dillman, 2002). For example, there are general concerns around issues such as accessibility, sample characteristics, and response rates. Demographics from this study raise these same concerns for remarried 74 participants (the on-line sample was older, more educated, and had a higher household income than the remarried sample solicited from a random sample of Utah newlyweds). What is becoming clear is that on-line methodologies are not the panacea for low response, non-representative sampling procedures (Dillman, 2002). However, this study does suggest that the internet is a viable modality for data collection and when included in a multi-method experimental design it may improve the quantity and quality of data. By improved data quality I am referring to the criticism that studies often fail to assess for the structural complexity and heterogeneity of remarriages and resulting stepfamily types (see Coleman & Ganong, 1990; Berger, 1998). Remarriage has become a ?catch-all? term that covers multiple types of relationships. For example, a remarriage may consist of two individuals in a second, third, or fourth marriage; or, a remarriage may be the first marriage for an individual who is marrying a widow or divorcee. Large samples can facilitate the comparison of these and other subsamples. As a case in point, in the present study 89 (25.9%) of the on-line participants were in their own first marriage but married to someone who had been previously married, 206 (59.9) were in their second marriage, 36 (10.5%) were in their third marriage, eight (2.3%) were in their fourth marriage, and one participant (.3%) was now in a sixth marriage. Another set of comparisons that may also be of interest is the pathway in which individuals enter remarriage (i.e., divorce, bereavement and never married). With these pathways, there can result at least 18 possible partner combinations depending on whether one or both partners already have children (Ganong & Coleman, 2004). Whatever the path to remarriage, it is commonly believed that the history and composition of the remarriage creates significant within-group differences with important 75 implications for marital dynamics and satisfaction (Berger, 1998; Visher & Visher, & Pasley, 2003). Unfortunately, it is not uncommon for researchers to ignore differences among stepfamily types and to lump all groups together as they compare stepfamilies to other family structures (i.e., nuclear families). There is arguably value (and clearly an interest) in evaluating difference between remarriages with first marriages. However, when stepfamily configurations are combined so as to compare stepfamilies with biological families important questions remained unanswered. Coleman and Ganong (1990) explain: Researchers may find that stepfamilies differ as a group from nuclear families but do not learn if the dependent variables are related to such structural characteristics as different custody patterns, sex of stepparent or sex of child variations, cause of prior marriage dissolutions, or reproduction in the remarriage. No information is gained about such process variables as role performance of stepmothers compared to stepfathers or the levels of marital conflict in different types of households. (p. 934) Another between-stepfamily-group comparison which is commonly glossed over deals with the three types of stepfamilies (Ganong & Coleman, 2004). In remarriages that do not create stepfamilies neither adult brings children into the marriage. Simple stepfamilies include one or more child from one parent. In complex stepfamilies both adults bring children into the relationship. Because of the dynamics associated with stepparenting non-biological children and potential loyalty conflicts, there are potential differences between complex and simple stepfamilies. Table 9 indicated that the online 76 participants provide a fairly equitable distribution of simple and complex stepfamilies. This allowed for specific tests between simple and complex stepfamilies. In summary, this study suggests that online methodologies may offer family science researchers a viable, inexpensive mechanism to overcome two primary limitations in stepfamily research: identification and access to a large enough group of remarried participants that group differences may be evaluated. Although online methodologies are not perfect and limitations regarding generalizability need be acknowledged and addressed; with large enough datasets a smaller representative subsample may be extractable. Future research should continue to explore the characteristics of online participants, keeping in mind the possibility that, if refined, online methodologies may be key to gathering large enough data sets that will allow researchers to ?examine or control for all relevant structural variables? (Coleman, Ganong, & Fine, 2000, p. 1300). Through examining and controlling for the heterogeneity of stepfamilies we, as a field, will be able to expand, strengthen and enhance our understanding of remarriages beliefs and how cognitions, behaviors and other factors influence remarital quality and stepfamily functioning. Future Research Questions Although the internal validity of the RMBI factors was supported, the total effect of remarriage beliefs on remarital quality cannot be concluded based on this study alone. Indeed, some of the direct effects were nonsignificant for the samples under study. However, as previously discussed, it may be that the effects of remarriage beliefs would be more easily detected when behaviors and interactions are considered. It has been noted that ?work exploring the interrelations that exist between behavior and cognition is 77 not well advanced? (Gottman & Notarius, 2000, p. 938). Future studies exploring interrelations between the RMBI and behaviors would help to answer questions of relative importance, directionality of influence, and processes leading to higher marital quality. Assessments of partner concurrence may also identify another process by which remarriage beliefs influence remarital and stepfamily function. With data on partners, researchers would have a better sense as to whether consistency and consensus of beliefs are as important to remarital satisfaction as Fine and Kurdek (1994) have presented them to be. Such data would also shed light on questions that arise if the theorized importance of consistency and consensus is supported. Specifically, what occurs when a couple jointly endorses an unrealistic belief? Would congruence around unrealistic beliefs foster remarital satisfaction? Future studies with couples will be able to assess whether beliefs themselves or spousal agreement on remarriage beliefs is the more potent predictor of satisfaction. The RMBI may also be utilized in other basic research efforts to understand processes involved in the etiology and the impact of remarriage beliefs on remarriage quality, stability, and stepfamily functioning. For example, future research could explore how life-events (i.e., parental divorce, death of a spouse, divorce, etc.) and preparation for remarriage influence the endorsement of remarriage beliefs. Future studies could also analyze the stability of beliefs over time and factors that impact their stability/instability. Future research with the RMBI may be used to explore differences between types of remarriages and stepfamilies. Fortunately, more researchers in the last decade have attempted to account for the complexities of remarriage and associated stepfamily types 78 in their work. There is work distinguishing remarried stepparents from cohabiting and de facto stepparents (e.g., Bulcroft, Carmody, & Bulcroft, 1998) and books have been written to address variations among remarriage types (e.g., Berger, 1998). While these referenced authors do not explicitly discuss how remarriage beliefs may differ across remarriage types, one implication from their work is that the field cannot assume that all stepfamilies think and function in the same way. One contribution of the present study is the indication that the causal structure of the RMBI is invariant between individuals in complex and simple stepfamilies. Unfortunately, due to the small subgroup size, tests of invariance where not conducted with individuals in childless remarriages. It is reasonable to assume that adults in remarriages without children and remarriages with children may exhibit some similarity in terms of beliefs related to their new partners and their previous relationships. However, there is no compelling evidence to suggest they would share similar beliefs regarding stepfamily adjustment or stepparenting issues. In fact on these types of remarriage beliefs, it would be more reasonable to assume that childless remarrying couples would be more similar to those in childless first marriages. After all, in terms of structure and constraints, those remarriages without children could be considered to have more commonalities with childless first marriages than they do with remarriages that form stepfamilies. However, these are merely assumptions that still lack empirical support. Additional studies with a larger subgroup of childless remarriages will be needed to explore whether/how the presence of children impacts the endorsement of remarriage beliefs. Until more research is done to access subgroup invariance an abbreviated version of the RMBI may be more appropriate with individuals without children. The following 79 subscales would be applicable to all remarriages: Past emotions should stay in the past, Partner is perfect, Success is slim, and Finances should be pooled. In samples in which participants have children the other subscales may also be included: Adjustment comes quickly, Stepfamilies are second-class, and Children are priority. Practical Applications Practical applications may include using the RMBI as a starter activity in family- life education classes on stepfamilies or as an activity to explore remarriage beliefs. Couples could complete the RMBI separately and then discuss their similar and dissimilar beliefs. Therapists may also use the RMBI as a discussion starter or as an intervention to promote communication. Future research regarding congruence of beliefs and the interaction of beliefs with behaviors may also have practical implications. For example, the use of the RMBI in clinical samples may prove to illicit support for ?dysfunctional? remarriage beliefs. However, in the non-clinical sample used for this study there was not strong evidence that the remarriage beliefs under study were ?dysfunctional?. As noted, this may be due to the high reported levels of high marital quality. Consequently, at this point in the development of the RMBI clinical cut-off scores can not be devised. Conclusion A parsimonious, 22-item RMBI was developed, refined, and presented for (a) future research to elicit knowledge on processes in remarriages and stepfamily development and (b) in applied work with couples preparing for or living in stepfamilies. The RMBI?s seven factor structure has been confirmed with a sample of 344 individuals in remarriages. This structure was also cross-validated by another sample of 217 80 remarried individuals. Invariance tests were run for individuals in simple versus complex stepfamilies and for individuals married less than three years versus those married more than three years. Tests for invariance did not indicate significant differences for RMBI factors on remarital satisfaction although small subgroup sizes make these results preliminary. Just as measures of general relationship beliefs have aided the researchers and practitioners who work with couples in first marriages, it is hoped that the RMBI will aid researchers and practitioners who work with remarried couples. Up until now and despite the increasing number of individuals entering remarriages and/or stepfamilies (see Bramlett & Mosher, 2002) there has been little effort to empirically define remarriage and stepfamily beliefs and to empirically examine their impact on individual, couple, and family functioning. This study begins to fill that gap. By validating a measure of frequently cited beliefs, this study provides the field with a tool with which to explore clinical assertions, theoretical propositions, and empirical associations pertaining to the development and influence of remarriage beliefs. 81 REFERENCES Adler-Baeder, A., & Higginbotham, B. (2004). Implications of remarriage and stepfamily formation for marriage education. Family Relations, 53, 448-458. Aguirre, B. E. & Parr, W. C. (1982). Husbands' marriage order and the stability of first and second marriages of white and black women. Journal of Marriage and the Family, 44, 605-620. Ahrons, C. R., & Rodgers, R. H. (1987). Divorced families: A multidisciplinary developmental view. New York: Norton. Allen, E. S., Baucom, D. H., Burnett, C. K., Epstein, N., & Rankin-Esquer, L. A. (2001). Decision-making power, autonomy, and communication in remarried spouses compared with first-married spouses. Family Relations, 50, 326-334. Amato, P. R., & Rogers, S. J. (1999). Do attitudes toward divorce affect marital quality? Journal of Family Issues, 20(1), 69-86. Amato, P. R. (1996). Explaining the intergenerational transfer of divorce. Journal of Marriage and the Family, 58, 628-640. Anderson, L. W. (1981). Assessing affective characteristics in the schools. Boston: Allyn and Bacon. Arbuckle, J. L. (1999). Amos 4.0 [Computer software]. Chicago: Smallwaters. Baucom, D. H., & Epstein, N. (1990). Cognitive-behavioral marital therapy. New York: Brunner/Mazel. 82 Bengtson, V. L., & Allen, K. R. (1993). The life course perspective applied to families over time. In P. Boss, W. Doherty, R. LaRossa, W. Schumm, & S. Steinmetz (Eds.), Sourcebook of family theories and methods: A contextual approach (pp. 469-499). New York: Plenum. Berger, R. (1998). Stepfamilies: A multi-dimensional perspective. New York: Haworth. Berger, R. (2000). Stepfamilies in cultural context. Journal of Divorce and Remarriage, 24, 35-50. Boss, P., Doherty, W., LaRossa, R., Schumm, W., & Steinmetz, S. (1993). Sourcebook of family theories and methods: A contextual approach. New York: Plenum. Bradbury, T. N., & Fincham, F. D. (1988). Individual difference variables in close relationships: A contextual model of marriage as an integrative framework. Journal of Personality and Social Psychology, 54, 713-721. Bramlett, M. D., & Mosler, W. D. (2001). First marriage dissolution, divorce, and remarriage: United States: Advanced data from vital and health statistics (No. 323). Bramlett, M. D., & Mosher, W. D. (2002). Cohabitation, Marriage, Divorce, and Remarriage in the United States. Vital Health Stat 23(22). National Center for Health Statistics. Bray, J., & Kelly, J. (1998). Stepfamilies: Love, marriage, and parenting in the first decade. New York: Broadway. Bray, J. H. (1999). From marriage to remarriage and beyond: Findings from the Developmental Issues in Stepfamilies Research Project. In E. M. Hetherington 83 (Ed.), Coping with divorce, single parenting and remarriage: A risk and resiliency approach (pp. 253-271). Mahwah, NJ: Erlbaum. Brody, G., Neubaum, E., & Forehand, R. (1988). Serial marriage: A heuristic analysis of an emerging family form." Psychological Bulletin, 103, 211 -222. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 445-455). Newbury Park, CA: Sage. Bulcroft, R., Carmody, D., & Bulcroft, K. (1998). Family structure and patterns of independence giving to adolescents. Journal of Family Issues, 19, 404-435. Bumpass, L. L., & Lu, H. (2000). Trends in cohabitation and implications for children's family contexts in the United States. Population Studies, 54(1), 29-41. Buunk, B. P., & Mutsaers, W. (1999a) The nature of the relationship between remarried individuals and former spouses and its impact on marital satisfaction. Journal of Family Psychology, 13(2), 165-174. Byrne, B. M. (2001). Structural equation modeling with AMOS0: Basic concepts, applications, and programming. Mahwah, N.J.: Lawrence Erlbaum Associates. Chadwick, B. A., & Heaton, T. B. (1999). Statistical handbook on the American family (2 nd ed.). Phoenix, AZ: Oryx. Clarke, S. C., & Wilson, B. F. (1994). The relative stability of remarriages: A cohort approach using vital statistics. Family Relations, 43, 305-310. Claxton-Oldfield, S, (2000). Deconstructing the myth of the wicked stepparent. Marriage & Family Review, 30(1/2), 51-58. 84 Claxton-Oldfield, S., Goodyear, C., Parsons, T., & Claxton-Oldfield, J. (2002). Some possible implications of negative stepfather stereotypes. Journal of Divorce and Remarriage, 36, 77-88. Coalition for Marriage, Family and Couples Education (n.d.). http://www.smartmarriages.com. Cohen, S., & Edwards, J. R., (1989). Personality characteristics as moderators of the relationship between stress and disorder. In R W. J. Neufeld (Ed.), Advances in the Investigation of Psychological Stress (pp. 235-283), New York: Wiley. Coleman, M., & Ganong, L. H. (1985). Remarriage myths: Implications for the helping professions. Journal of Counseling and Development, 64, 116-120. Coleman, M., & Ganong, L. (1989). Financial management in stepfamilies. Lifestyles: Family and Economic Issues, 10, 217-232. Coleman, M., & Ganong, L. H. (1990). Remarriage and stepfamily research in the 1980s: Increased interest in an old family form. Journal of Marriage and the Family, 52, 925-941. Coleman, M., & Ganong, L. (1997). Stepfamilies from the stepfamily?s perspective. Marriage and Family Review, 26, 107-122. Coleman, M., Ganong, L., & Fine, M. A. (2000). Reinvestigating remarriage: Another decade of progress. Journal of Marriage and the Family, 62, 1288-1307. Coleman, M., Ganong, L. H., & Gingrich, R. (1985). Stepfamily strengths: A review of popular literature. Family Relations, 34(4), 583-590. Coleman, M., Ganong, L. H., & Goodwin, C. (1994). The presentation of stepfamilies in marriage and family textbooks: A reexamination. Family Relations, 43, 289-297. 85 Coleman, M., Ganong, L., & Weaver, S. (2001). Maintenance and enhancement in remarried families. In J. Harvey & A. Wenzel (Eds.), Close romantic relationships: Maintenance and enhancement (pp. 255-276). Hillsdale, NJ: Erlbaum. Coontz, S. (1997). The way we really are: Coming to terms with America?s changing families. New York: Basic. Cramer, D. (2001). Destructive disagreement belief, disagreement, negative conflict, and relationship satisfaction in romantic relationships. Genetic, Social, and General Psychology Monographs, 127(3), 301-317. Crosbie-Burnett, M. (1989). Application of family stress theory to remarriage: A model for assessing and helping stepfamilies. Family Relations, 38, 323-331. Dallos, R. (1991). Family beliefs systems, therapy, and change. Open University Press: Philadelphia. DeVellis, R. F. (1991). Scale development: Theory and applications. Newbury Park, CA: Sage. Dillman, D. A. (2000). Mail and internet survey: The tailored design method. New York, NY: Wiley. Eidelson, R. J. & Epstein, N. (1982). Cognition and relationship maladjustment: Development of a measure of dysfunctional relationship beliefs. Journal of Consulting and Clinical Psychology, 50, 715-720. Engel, M. (1999). Pockets of poverty: The second wives club?examining the financial [in] security of women in remarriage. William and Mary Journal of Women and the Law, 5, 309-381. 86 Epstein, N., & Eidelson, R. J. (1981). Unrealistic beliefs of clinical couples: Their relationship to expectations, goals and satisfaction. American Journal of Fatuity Therapy, 9, 13-22. Epstein, N., Schlesinger, S., & Dryden, W. (1988). Cognitive-behavioral therapy with families. New York: Brunner/Mazel. Fine, M., & Kurdek, L. (1994). A multidimensional cognitive-developmental of stepfamily adjustment. In K. Pasley & M. Ihinger-Tallman (Eds.), Stepparenting: Issues in theory, research, and practice (pp. 1532). Westport, CT: Greenwood. Fishman, B. (1983). The economic behavior of stepfamilies. Family Relations, 32, 359- 366. Fitzpatrick, J., & Sollie, D. L. (1999). Unrealistic gendered and relationship-specific beliefs: Contributions to investments and commitment in dating relationships. Journal of Social & Personal Relationships, 16(6), 852-867. Furstenberg, F. F. Jr., & Spanier, G. B. (1984). The risk of dissolution in remarriage: An examination of Cherlin?s hypothesis of incomplete institutionalization. Family Relations, 33, 433-441. Gable, R. K., & Wolfe, M. B. (1993). Instrument development in the affective domain. Boston: Kluwer. Ganong, L., & Coleman, M. (1984). The effects of remarriage on children: A review of the empirical literature. Family Review, 33, 389-406. Ganong, L. H., & Coleman, M. (1994). Remarried family relationships. Thousand Oaks, CA: Sage. 87 Ganong, L. H., & Coleman, M. (1995). The content of mother stereotypes. Sex Roles: A Journal of Research, 32(7/8), 495-512. Ganong, L., & Coleman, M. (1997). How society views stepfamilies. Marriage and Family Review, 26, 85-106. Ganong, L. H., & Coleman, M. (2004). Stepfamily relationships: Development, dynamics, and interventions. New York: Kluwer. Ganong, L., Coleman, M., & Mapes, D. (1990). A meta-analytic review of family structure stereotypes. Journal of Marriage and the Family, 52(2), 287-297. Ganong, L., & Coleman, M., & Weaver, S. (2001). Maintenance and enhancement in remarried families: Clinical applications. In J. Harvey & A. Wenzel (Eds.), A clinicians guide to maintaining and enhancing close relationships (pp. 105-129). Hillsdale, NJ: Erlbaum. Gerlack, P. (2001). Building a high-nurturance stepfamily. Philadelphia: Hibris Corporation. Giles-Sims, J., & Crosbie-Burnett, M. (1989). Stepfamily research: Implications for policy, clinical interventions, and further research. Family Relations, 38, 19-24. Gottman, J. M., & Notarius, C. I. (2000). Decade review: Observing marital interaction. Journal of Marriage and the Family, 62, 927-947. Hair, J. F. (1998). Multivariate data analysis. Upper Saddle River, New Jersey: Prentice- Hall. Hendrick, S. (1988). A generic measure of relationship satisfaction. Journal of Marriage and the Family, 50, 93-98. 88 Hetherington, E. M. (1989). Coping with family transitions: Winners, losers, and survivors. Child Development, 60, 1-14. Hetherington, E. M., & Kelly, J. (2002). For better or for worse: Divorce reconsidered. New York, NY: W.W. Norton & Co. Higginbotham, B. J., & Adler-Baeder, F. (in-press). Development of the Remarriage Belief Inventory for researchers and educators. Journal of Extension. Holye, R. H. (1995). Structural equation modeling: Concepts, issues, and applications. Thousand Oaks, CA: Sage. Huston, T. L. (2000). The social ecology of marriage and other intimate unions. Journal of Marriage & the Family, 62(2), 298-320. Kalmuss, D. & Seltzer, J. A. (1986). Continuity of marital behavior in remarriage: The case of spouse abuse. Journal of Marriage and the Family, 48, 113-120. Kaplan, M., & Hennon, C. B. (1992). Remarriage education: The Personal Reflections Program. Family Relations, 41(2), 127-134. Karney, B. R., & Bradbury, T. N. (1995). The longitudinal course of marital quality and stability: A review of theory, method, and research. Psychological Bulletin, 118, 3-34. Kheshgi-Genovese, A., & Genovese, T. A. (1997). Developing the spousal relationship within stepfamilies. Families in Society: The Journal of Contemporary Human Services, 78(3), 255-264. Kirchhoff, K. T. (1999). Design of questionnaires and structured interviews. In M. A. Mateo & K. T. Kirchoff (Eds.), Using and conducting nursing research in the clinical setting (2 nd ed., pp. 229-239). Philadelphia: W. B. Sanders. 89 Kline, R. B. (2005). Principles and practice of structural equation modeling. New York: Guilford. Knox, D., & Zusman, M. E. (2001). Marrying a man with 'Baggage': Implications for second wives. Journal of Divorce and Remarriage, 35(3/4), 67-80. Kreider, R. M., & Fields, J. M. (2001). Number, timing and duration of marriages and divorces: Fall 1996. Current Population Reports, P70-80. U.S. Census Bureau, Washington, D.C. Kurdek, L. A. (1992). Assumptions versus standards: The validity of two relationship cognitions in heterosexual and homosexual couples. Journal of Family Psychology, 6(2), 164-170. Kurdek, L., & Fine, M. (1991). Cognitive correlates of satisfaction for mothers and stepfathers in stepfather families. Journal of Marriage and Family, 53, 565-572. Lazarus, R. S. & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer. Lee, T. (2001). Utah Newlywed Study. Utah State University. Leslie, L. A., & Epstein, N. (1988). Cognitive-behavioral treatment of remarried families. In N. Epstein, S. E. Schlesinger, & W. Dryden (Eds.), Cognitive-behavioral therapy with families (pp. 151-182). New York: Brunner/Mazel. Locke, H. J., & Wallace, K. M. (1959). Short marital-adjustment and prediction tests: Their reliability and validity. Marriage and Family Living, 21, 251-255. Lown, J. M., McFadden, J. R., & Crossman, S. M. (1989). Family life education for remarriage focus on financial management. Family Relations, 38, 40-46. 90 Metts, S., & Cupach, W. (1990). The influence of relationship beliefs and problem- solving responses on satisfaction in romantic relationships. Human Communication Research, 17(1), 170-185. Mitchell, S. E., Newell, G. K., & Schumm, W. R. (1983). Test-retest reliability of the Kansas marital satisfaction scale. Psychological Reports, 53, 567-572. Moller, A. T., & Van Zyl, P. D. (1991). Relationship beliefs, interpersonal perception, and marital adjustment. Journal of Clinical Psychology, 47(1), 28-33. Murray, D. M., & Fisher, J. D. (2002). The Internet: A virtually untapped tool for research. Journal of Technology in Human Services, 19(2-3), 5-18. National Center for Health Statistics. (1993). 1988 marriages: Number of the marriage by of bride by groom [Computer program]. Washington, DC: NCHS Computer Center. Northey, W. F. (2005). Studying marriage and family therapists in the 21 st century: Methodological and technological issues. Journal of Marital and Family Therapy, 31(1), 99-105. Papernow, P. L. (1984). The stepfamily cycle: An experiential model of stepfamily development. Family Relations, 33, 355-363. Papernow, P. (1987). Thickening the ?middle ground?: Dilemmas and vulnerabilities of remarried couples. Psychotherapy, 24, 630-639. Pasley, K., Sandras, E., & Edmondson, M. (1994). The effects of financial management strategies on the quality of family life in remarriage. Journal of Family and Economic Issues, 15, 53-70. 91 Pett, M. A., Lackey, N. R., & Sullivan, J. J. (2003). Making sense of factor analysis. Thousand Oaks, CA: Sage. Prada, L. M., & Markman, H. J. (1999). Unearthing the seeds of marital distress: What we have learned from married and remarried couples. In M. J. Cox & J. Brooks- Gunn (Eds.), Conflict and cohesion in families (pp. 51-85). Mahwah, NJ: Erlbaum. Rodgers, R. H., & White, J. M. (1993). Family development theory. In P. Boss, W. Doherty, R. LaRossa, W. Schumm, & S. Steinmetz (Eds.), Sourcebook of family theories and methods: A contextual approach (pp. 225-254). New York: Plenum. Rorer, L. (1989). Rational-emotive therapy: An integrated psychological and philosophical basis. Cognitive Therapy and Research, 13(5), 475-492. Rusbult, C. E. (1983). A longitudinal test of the investment model: The development (and deterioration) of satisfaction and commitment in heterosexual involvements. Journal of Personality and Social Psychology, 45, 101-117. Sager, C. J., Brown, H. S., Crohn, H., Engel, T., Rodstein, E., & Walker, E. (1983). Treating the remarried family. New York: Brunner/Mazel. Schumacker, R. E., & Lomax, R. G. (1996). A beginner?s guide to structural equation modeling. Mahwah, NJ: Lawrence Erlbaum Associates. Schumm, W. R., Bollman, S. R., Jurich, A. P., & Hatch, R. C. (2001). Family strengths and the Kansad Marital Satisfaction Scale: A factor analytic study. Psychological Reports, 88(3). 965-975. Schumm, W. R., Paff-Bergen, L. A., Hatch, R. C., Obiorah, F. C., Copeland, J. M., Meens, L. D., & Bugaihis, M. A. (1986). Concurrent and discriminant validity of 92 the Kansas Marital Satisfaction Scale. Journal of Marriage and the Family, 48, 381-387. Spanier, G. B. (1976). Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. Journal of Marriage and the Family, 38, 15-28. Stanley, S., Markman, H., & Whitton, (2002). Communication, conflict, and commitment: Insights on the foundations of relationship success from a national survey. Family Process, 41, 659-675. Stepfamily Association of America (n.d.). http://www.saafamilies.org Streiner, D. L. (2003). Being inconsistent about consistency: When coefficient alpha does and doesn?t matter. Journal of Personality Assessment, 80(3), 217-222. U.S. Bureau of the Census. (1998). Marital status and living arrangements, March 1998 (update) (P-70-63). Washington, DC: US Government Printing Office. Visher, E. B., & Visher, J. S. (1979). Stepfamilies: A guide to working with stepparents and stepchildren. New York: Brunner/Mazel. Visher, E. B., & Visher, J. S. (1985). Stepfamilies are different. Journal of Family Therapy, 7, 9-18. Visher, E. B., & Visher, J. S. (1988). Old loyalties, new ties: Therapeutic strategies with stepfamilies. New York: Brunner/Mazel. Visher, E. B., & Visher, J. S. (1996). Therapy with stepfamilies. New York: Brunner/Mazel. Visher, E. B., Visher, J. S., & Pasley, K. (2003). Remarriage families and stepparenting. In F.Walsh (Ed.), Normal family processes: growing diversity and complexity (3 rd ed., pp. 153-175). NY: Guilford. 93 Walsh, F. (Ed.) (2003). Normal family processes: growing diversity and complexity. New York: Guilford. Walsh, W. M. (1992). Twenty major issues in remarriage families. Journal of Counseling & Development, 70, 709-715. Weston, C. A., & Macklin, E. D. (1990). The relationship between former-spousal contact and remarital satisfaction in stepfather families. Journal of Divorce & Remarriage, 14(2), 25-47. Wilson, B., & Clarke, S. (1992). Remarriages: A demographic profile. Journal of Family Issues, 13, 123-141. 94 APPENDICES Appendix A: Text of the Invitation Email to the Smartmarriages.com and Your Stepfamily Magazine Samples Appendix B: Informed Consent Letter Appendix C: Text of the E-reminder to the Smartmarriages.com and Your Stepfamily Magazine participants Appendix D: Text of the Invitation Letter to Utah Newlywed Sample Appendix E: Text of Reminder Postcard to the Utah Newlywed Sample Appendix F: 48 Item Remarriage Belief Inventory (RMBI) Appendix G: Kansas Marital Satisfaction Scale (KMS) Appendix H: Modified Version of the Marital Adjustment Test (MAT) Appendix I: Relationship Belief Inventory (RBI) Appendix J: Demographic Questions Appendix K: RMBI Empirical Indicators Grouped According to Content Areas Appendix L: Refined RMBI Empirical Indicators Grouped According to Content Areas Appendix M: Results from Validity Tests of the Causal Structure using the Utah Sample Appendix N: Summary of Hierarchical Regression Analysis for the RBI and RMBI using the Utah Sample 95 Appendix A Text of the Invitation Email to the Smartmarriages.com and Your Stepfamily Magazine Samples 96 Take a 15 minute survey and receive a gift certificate to Amazon.com! If you are remarried or living in a stepfamily, we invite you to participate in the web- based research study on remarriage expectations being conducted by Brian Higginbotham and Dr. Francesca Adler-Baeder from Auburn University. The on-line questionnaire takes approximately 15 minutes to complete. Those who choose to participate will receive an e-gift certificate for Amazon.com. Remarriage has its challenges and rewards. While it has some things in common with a first marriage, there are important ways in which a remarriage is different. Unfortunately we know a lot more about first marriages than we do about remarriages. We would like to change that, and we are asking for your help. From this study we hope to learn about the specific expectations held by adults in remarriages. Your responses will help us to identify and measure the effects of these remarriage expectations on remarital quality. We hope you choose to participate as the responses you provide will assist therapists, researchers, and family life educators in their efforts to strengthen stepfamilies and remarriages. To participate, please use the following link to complete the on-line questionnaire: http://www.surveymonkey.com/s.asp?u=69848722521 97 Appendix B Informed Consent Letter 98 INFORMATION AND CONSENT PAGE For the Remarriage Belief Inventory (RMBI) Evaluation Study. If you are remarried or living in a stepfamily, we invite you to participate in a research study on remarriage beliefs being conducted by Brian Higginbotham and Dr. Francesca Adler-Baeder from the Human Development and Family Studies Department at Auburn University, AL. From this study we hope to learn about the specific beliefs and expectations held by adults in remarriages. Your responses will help us to identify and measure the effects of remarriage expectations on remarital quality and functioning. Participation in this study is voluntary. If you decide to participate, please click on the link below to complete the on-line questionnaire. You do not have to submit your answers if you do not want them recorded. Those who do choose to submit their answers to all the questions will be emailed a $5 E-GIFT CERTIFICATE to Amazon.com as a token of our appreciation. Those who choose to submit their answers will also be emailed in three to six months to complete a brief follow-up questionnaire. All of the information that you provide on the questionnaire will be treated as private and CONFIDENTIAL. You will NOT be asked for your name nor social security number. Your EMAIL address will serve as your code number, which will be used to send your e- gift certificate and to organize the information you provide. If results of the study are published in a journal, no names of individuals will be included in these reports. We hope you choose to participate as the responses you provide will assist therapists, researchers, and family life educators in their efforts to strengthen stepfamilies and remarriages. If you have questions about this study, Brian Higginbotham (higgibr@auburn.edu) will be happy to answer them. For more information regarding your rights as a research participant you may contact the Office of Human Subjects Research by phone or e-mail. The people to contact are Mr. Chip Burson, Executive Director at (334) 844-5966 (bursoen@auburn.edu) or Dr. Peter Grandjean, IRB Chair at (334) 844-1462 (grandpw@auburn.edu). HAVING READ THE INFORMATION PROVIDED, YOU MUST DECIDE WHETHER OR NOT YOU WISH TO PARTICIPATE IN THIS RESEARCH STUDY. CLICKING ON THE BUTTON BELOW INDICATES YOU HAVE READ THIS PAGE AND ARE WILLING TO PARTICIPATE. 99 Appendix C Text of the E-reminder to the Smartmarriages.com and Your Stepfamily Magazine participants 100 There is still time to participate in the ?Remarriage Belief Inventory? Evaluation Study. This is the web-based research study on remarriage expectations and beliefs being conducted by Brian Higginbotham and Dr. Francesca Adler-Baeder from Auburn University. The on-line questionnaire takes approximately 15 minutes to complete and those who choose to participate will receive an e-gift certificate for Amazon.com. Your participation will help the researchers measure the impact of remarriage expectations on remarital quality. To participate, use the following link to complete the on-line questionnaire: http://www.surveymonkey.com/s.asp?u=95614686260 101 Appendix D Text of the Invitation Letter to Utah Newlywed Sample 102 You are invited to participate in a study on remarriage expectations being conducted by researchers from Auburn University. Remarriage has its challenges and rewards. While it has some things in common with a first marriage, there are important ways in which a remarriage is different. However, in terms of the research, we know a lot more about first marriages than we do about remarriages. We would like to change that, and we are asking for your help. From this study we hope to learn about the specific beliefs and assumptions held by adults in remarriages and/or stepfamilies. Your responses will help us to identify and measure the effects of these remarriage expectations on remarital quality. Included in this packet is small token of our appreciation and two sets of questionnaires. The blue questionnaire is for the husband and the gold is for the wife. The questionnaire takes approximately 15 minutes to complete and is divided into four sections: 1) Demographic Information 2) Relationship Quality 3) Relationship and Remarriage Beliefs 4) Family History We have also put the questionnaire on the web. If you choose to participate on-line please go to http://www.surveymonkey.com/s.asp?u=2262780576 On the web, you will be prompted to enter your participant code number which can be found on your questionnaire (Question #1). In the case you prefer to answer the hard- copy version of the questionnaire we have included two self-addressed envelopes so you can return the questionnaires separately. Please answer each question as honestly and accurately as possible. There are no "right" answers. All responses will remain confidential and will not be seen by your partner. We hope you choose to complete and return the survey as the responses you provide will assist therapists, researchers, and family life educators in their efforts to strengthen stepfamilies and remarriages. Thank you! Brian Higginbotham, M.S. 103 Appendix E Text of Reminder Postcard to the Utah Newlywed Sample 104 Two weeks ago a questionnaire designed to study remarriage beliefs and expectations was mailed to you. Your name was drawn from a list of marriage licenses in Utah. If both of you have already completed and returned the questionnaire to us, please accept our sincere thanks. If not, please do so today. We are particularly grateful for your help because it is only by asking those in remarriages and stepfamilies to share their experiences that we can understand adults in these relationships. Your responses will help us to identify and measure the effects of remarriage expectations on remarital quality. If you did not receive a questionnaire, or if it was misplaced, please call us (334- 844-3299) or complete the questionnaire on-line: www.surveymonkey.com/s.asp?u=2262780576 . The participant number for you and your partner can be found directly underneath your name on the address label of this postcard. 105 Appendix F 48-Item Remarriage Belief Inventory (RMBI) 106 Using the following scale, please indicate the extent to which you believe that each of the following statements is true or false. 1 2 3 4 5 Very False More False than True Neither True nor False More True than False Very True 1 A stepfamily should operate like a biological family. 2 People in remarriages should be able to step right into the role of parent to their stepchildren. 3 In a stepfamily, children should feel they come first. 4 Emotional connection/feelings to an ex-spouse should end with a new marriage. 5 Stepfamily members should immediately love one another. 6 A new stepparent should feel just as ?at home? as those with whom he/she has moved in with. 7 A remarriage can be more fulfilling and satisfying than previous marriages. 8 People who have been married before are more choosy about whom they marry. 9 A new spouse should fill the emotional holes that the previous spouse left empty. 10 People in remarriages are likely to marry someone who has the same ?bad? traits as their previous spouse. 11 Financial resources in a remarriage should be combined. 12 A stepfamily cannot offer children everything that a biological family can. 13 A stepparent should have his/her opinions and feelings valued as much as those held by the other members of the family he/she married into. 14 Adjustment to married life should occur quickly in remarriage. 15 A person?s past relationship history has little impact on a remarriage. 16 A stepparent should share in child-discipline responsibilities from the onset of the remarriage. 17 A new spouse should be a better marriage partner than the one he/she replaces. 18 Decisions in a stepfamily should be made based on the best interest of the new spouse rather than the children. 19 People in remarriages are less committed to making their relationship last than people in first marriages. 20 In comparison to ex-spouses, a new spouse should be more ?in-tune? to the quality of the spousal relationship. 21 Love should develop quickly between the child and the stepparent. 22 People in remarriages should maintain individual control of their own financial assets. 23 Fulfilling the desires of a new spouse should come before fulfilling the desires of biological children. 24 All things considered, a stepfamily is a poor substitute for a biological family. 25 Roles in remarriages should be similar to those in previous marriages. 26 Adjustment to living in a stepfamily should occur quickly. 27 People who have divorced are likely to divorce again. 107 28 Stepfamily members should feel close to one another soon after the stepfamily forms. 29 Stepparents should assume intimacy and authority with the children soon after the stepfamily forms. 30 The success/failure of past relationships has little to do with remarriage success. 31 A new spouse should be everything the problematic old spouse was not. 32 Family members should open their hearts to the new stepparent soon after the stepfamily forms. 33 In remarriages, incomes and paychecks should be ?pooled? together. 34 A new spouse should be more understanding than a previous spouse. 35 People in remarriages are likely to repeat the same patterns/behaviors as those in their previous marriage(s). 36 When a person gets remarried, it is likely that their new spouse will have some of the same flaws they saw in their previous spouse. 37 Problems that occur in a stepfamily probably would not occur in a family in which children live with both biological parents. 38 The odds of marital success are better the second time around. 39 The spousal relationship is the most critical to the success of a stepfamily as a whole. 40 People in remarriages are more experienced and thus better prepared for marriage. 41 People in remarriages are likely to make the same mistakes they made in previous marriages. 42 There are no major differences between biological families and stepfamilies. 43 Traditions from previous relationships/families should be integrated into the new family. 44 In a remarriage, there should be a distinction between ?mine? and ?your? financial resources. 45 In a stepfamily, child outcomes are better when the spousal relationship is strong. 46 Emotional ties to the previous marriages/relationships should be severed prior to a remarriage. 47 Wishes of the children should take priority over the wishes of the new spouse. 48 Giving attention to the children is more important than giving attention to the new spouse in a remarriage. 108 Appendix G Kansas Marital Satisfaction Scale (KMS) 109 Please check one box per question. 1 2 3 4 5 6 7 Extremely Dissatisfied Very Dissatisfied Somewhat Dissatisfied Mixed Somewhat Satisfied Very Satisfied Extremely Satisfied 1. How satisfied are you with your marriage (or relationship) in general? 2. How satisfied are you with your partner as a spouse (spouse-to-be)? 3. How satisfied are you with your relationship with your partner? 110 Appendix H Modified Version of the Marital Adjustment Test (MAT) 111 Please indicate the approximate extent of agreement or disagreement between you and your partner for each item on the following list. 1 2 3 4 5 Almost Always Disagree Frequently Disagree Equally Agree/Disagree Frequently Agree Almost Always Agree 1. Handling finances 2. Religious matters 3. Demonstrations of affection 4. Friends 5. Making major decisions 6. Sex relations 7. Child rearing 8. Use of time 9. Extended family 112 Appendix I Relationship Belief Inventory (RBI) 113 Using the following scale, please circle the score that indicates the extent to which you believe that each of the following statements is true or false for you. 1 2 3 4 5 Very False More False than True Neither True nor False More True than False Very True 1. I do not expect my partner to sense all my moods. 2. I get upset if I think I have not completely satisfied my partner sexually. 3. I cannot accept it when my partner disagrees with me. 4. If I have to tell my partner that something is important to me, it does not mean that s/he is insensitive to me. 5. My partner does not seem capable of behaving other than s/he does now. 6. I take it as a personal insult when my partner disagrees with an important idea of mine. 7. I get very upset if my partner does not recognize how I am feeling and I have to tell him/her. 8. Just because my partner has acted in ways that upset me does not mean that s/he will do so in the future. 9. If I cannot perform well sexually whenever my partner is in the mood, I would consider that I have a problem. 10. I get very upset when my partner and I cannot see things the same way. 11. It is important to me for my partner to anticipate my needs by sensing changes in my moods. 12. I can feel okay about my lovemaking even if my partner does not achieve orgasm. 13. If my partner wants to change, I believe that s/he can do it. 14. When my partner and I disagree, I feel like our relationship is falling apart. 15. I do not expect my partner to be able to change. 16. When I do not seem to be performing well sexually, I get upset. 114 Appendix J Demographic Questions 115 Please answer the following questions about yourself as honestly and accurately as possible. All responses will remain confidential and will not be seen by your partner. Please check only one answer per question. 1. Participant #: ______ 2. Age ______ 2. Sex ( 1 1) Male ( 2 2) Female 4. What is the highest level of education you have completed (check only one): ( 1 1) Some high school ( 5 5) Associates degree ( 2 2) High school or GED ( 6 6) Bachelors degree ( 3 3) Technical school ( 7 7) Higher than a bachelors degree ( 4 4) Some college 5. For statistical purposes only, what is your current (combined) household annual income? ( 1 1) Under $9,999 ( 2 2) $10,000 - $19,999 ( 3 3) $20,000 - $29,999 ( 4 4) $30,000 - $39,999 ( 5 5) $40,000 - $49,999 ( 6 6) $50,000 - $74,999 7 7) Over $75,000 6. Which of the following racial groups best describes you? (check only one): ( 1 1) African-American ( 2 2) Asian-American ( 3 3) Caucasian/White 4 4) Hispanic ( 5 5) Mulitracial ( 6 6) Native American ( 7 7) Pacific Islander ( 8 8) Other: ______ 7. Please indicate your present religious affiliation (if any)? ( 1 1) Buddhist ( 2 2) Catholic ( 3 3) Hindu ( 4 4) Islamic ( 5 5) Jewish ( 6 6) Latter-Day Saint ( 7 7) Protestant ( 8 8) No formal religious affiliation ( 9 9) Agnostic ( 1 1 0 0) Atheist ( 1 1 1 1) Other: ______ 8. Please indicate if you and your partner are of the same religious denomination/orientation: ( 1 1) Same ( 2 2) Different 9. Has your partner been previously married to someone else? ( 1 1) Yes ( 2 2) No (skip to question 11) 10. Regarding your partner?s relationship history: Is he/she separated? ( 1 1) Yes ( 2 2) No Is he/she divorced? ( 1 1) Yes ( 2 2) No Is he/she widowed? ( 1 1) Yes ( 2 2) No Regarding your current couple relationship: 11. What is your present relationship status? (check only one) ( 1 1) Never Married ( 5 5) Engaged to be Remarried (Divorced) ( 2 2) First marriage ( 6 6) Engaged to be Remarried (Widowed) ( 3 3) Cohabitating (Divorced) ( 7 7) Remarried (Divorced) ( 4 4) Cohabitating (Widowed) ( 8 8) Remarried (Widowed) 9 9) Other (please specify): ___________ 116 If married please answer questions 12 and 13 then skip to 16. If unmarried please skip to question 14 and 15. If you are not in a couple relationship, please skip to 16. 12. (If married?) How long have you been married to your present spouse? _______ Years _______ Months 13. If you lived with your partner before marriage, how long did you cohabit before marrying? _______ Years _______ Months (If you did not live together please enter 0.) 14. (If in an unmarried couple relationship?) How long have you been with your current partner? _______ Years _______ Months 15 If you are currently living together and are not married, how long have you cohabited together? _______ Years _______ Months (If you do not live together please enter 0.) 16. How many times (including your current marriage) have you been married? _______ If neither you nor your partner has children please skip to question 18. 17a. For each of YOUR children from previous marriages/relationships, please indicate: Age Sex Does he/she live with you? 1 st Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 2 nd Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 3 rd Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 4 th Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 5 th Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 17b. For each of your PARTNER?S children from previous marriages/relationships, please indicate: Age Sex Does he/she live with you? 1 st Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 2 nd Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 3 rd Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 4 th Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 5 th Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 17c. For each of the children you and your partner have had/adopted TOGETHER, please indicate: Age Sex Does he/she live with you? 1 st Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 2 nd Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 3 rd Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 4 th Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 5 th Child ( 1 1) Male ( 2 2) Female ( 1 1) Yes ( 2 2) No 18. Are you expecting a child? ( 1 1) Yes ( 2 2) No 117 Appendix K RMBI Empirical Indicators Grouped According to Content Areas 118 Adjustment comes quickly (Adjustment) 2. People in remarriages should be able to step right into the role of parent to their stepchildren. 5. Stepfamily members should immediately love one another. 6. A new stepparent should feel just as ?at home? as those with whom he/she has moved in with. 14. Adjustment to married life should occur quickly in remarriage. 16. A stepparent should share in child-discipline responsibilities from the onset of the remarriage. 21. Love should develop quickly between the child and the stepparent. 26. Adjustment to living in a stepfamily should occur quickly. 28. Stepfamily members should feel close to one another soon after the stepfamily forms. 29. Stepparents should assume intimacy and authority with the children soon after the stepfamily forms. 32. Family members should open their hearts to the new stepparent soon after the stepfamily forms. Stepfamilies are second-class (Stepfamily) 1. A stepfamily should operate like a biological family. 7. A remarriage can be more fulfilling and satisfying than previous marriages. 12. A stepfamily cannot offer children everything that a biological family can. 24. All things considered, a stepfamily is a poor substitute for a biological family. 37. Problems that occur in a stepfamily probably would not occur in a family in which children live with both biological parents. 42. There are no major differences between biological families and stepfamilies. Children are priority (Priority) 3. In a stepfamily, children should feel they come first. 13. A stepparent should have his/her opinions and feelings valued as much as those held by the other members of the family he/she married into. 18. Decisions in a stepfamily should be made based on the best interest of the new spouse rather than the children. 23. Fulfilling the desires of a new spouse should come before fulfilling the desires of biological children. 47. Wishes of the children should take priority over the wishes of the new spouse. 39. The spousal relationship is the most critical to the success of a stepfamily as a whole. 45. In a stepfamily, child outcomes are better when the spousal relationship is strong. 48. Giving attention to the children is more important than giving attention to the new spouse in a remarriage. 119 Past should stay in the past (Past) 4. Emotional connection/feelings to an ex-spouse should end with a new marriage. 25. Roles in remarriages should be similar to those in previous marriages. 43. Traditions from previous relationships/families should be integrated into the new family 46. Emotional ties to the previous marriages/relationships should be severed prior to a remarriage. Partner is perfect (Partner) 8. People who have been married before are more choosy about whom they marry. 9. A new spouse should fill the emotional holes that the previous spouse left empty. 10. People in remarriages are likely to marry someone who has the same ?bad? traits as their previous spouse. 17. A new spouse should be a better marriage partner than the one he/she replaces. 20. In comparison to ex-spouses, a new spouse should be more ?in-tune? to the quality of the spousal relationship. 31. A new spouse should be everything the problematic old spouse was not. 34. A new spouse should be more understanding than a previous spouse. Success is slim (Success) 15. A person?s past relationship history has little impact on a remarriage. 19. People in remarriages are less committed to making their relationship last than people in first marriages. 27. People who have divorced are likely to divorce again. 30. The success/failure of past relationships has little to do with remarriage success. 36. When a person gets remarried, it is likely that their new spouse will have some of the same flaws they saw in their previous spouse. 38. The odds of marital success are better the second time around. 40. People in remarriages are more experienced and thus better prepared for marriage. 41. People in remarriages are likely to make the same mistakes they made in previous marriages. 35. People in remarriages are likely to repeat the same patterns/behaviors as those in their previous marriage(s). Finances should be pooled (Finances) 11. Financial resources in a remarriage should be combined. 22. People in remarriages should maintain individual control of their own financial assets. 33. In remarriages, incomes and paychecks should be ?pooled? together. 44. In a remarriage, there should be a distinction between ?mine? and ?your? financial resources. 120 Appendix L Refined RMBI Empirical Indicators Grouped According to Content Areas 121 Adjustment comes quickly (Adjustment) 21. Love should develop quickly between the child and the stepparent. 26. Adjustment to living in a stepfamily should occur quickly. 28. Stepfamily members should feel close to one another soon after the stepfamily forms. 29. Stepparents should assume intimacy and authority with the children soon after the stepfamily forms. Stepfamilies are second-class (Stepfamily) 12. A stepfamily cannot offer children everything that a biological family can. 24. All things considered, a stepfamily is a poor substitute for a biological family. Children are priority (Priority) 23. Fulfilling the desires of a new spouse should come before fulfilling the desires of biological children. (Reverse coded) 47. Wishes of the children should take priority over the wishes of the new spouse. 48. Giving attention to the children is more important than giving attention to the new spouse in a remarriage. Past should stay in the past (Past) 4. Emotional connection/feelings to an ex-spouse should end with a new marriage. 46. Emotional ties to the previous marriages/relationships should be severed prior to a remarriage. Partner is perfect (Partner) 17. A new spouse should be a better marriage partner than the one he/she replaces. 20. In comparison to ex-spouses, a new spouse should be more ?in-tune? to the quality of the spousal relationship. 31. A new spouse should be everything the problematic old spouse was not. 34. A new spouse should be more understanding than a previous spouse. Success is slim (Success) 27. People who have divorced are likely to divorce again. 35. People in remarriages are likely to repeat the same patterns/behaviors as those in their previous marriage(s). 36. When a person gets remarried, it is likely that their new spouse will have some of the same flaws they saw in their previous spouse. 41. People in remarriages are likely to make the same mistakes they made in previous marriages. 122 Finances should be pooled (Finances) 11. Financial resources in a remarriage should be combined. 33. In remarriages, incomes and paychecks should be ?pooled? together. 44. In a remarriage, there should be a distinction between ?mine? and ?your? financial resources. (Reverse coded) 123 Appendix M Results from Validity Tests of the Causal Structure using the Utah Sample 124 Summary of SEM Estimates for RMBI Factors Theorized to Predict Satisfaction (N = 217; R? = .085) B SE B ? Critical Ratio Satisfaction <--- Adjustment -.249 .151 -.157 -1.648 Satisfaction <--- Finances .052 .207 .023 .252 Satisfaction <--- Partner .213 .209 .111 1.019 Satisfaction <--- Priority .178 .115 .129 1.545 Satisfaction <--- Success .086 .172 .052 .499 Satisfaction <--- Stepfamily -.264 .195 -.174 -1.357 Satisfaction <--- Past .247 .120 .205 2.063* * p < .001 Summary of SEM Estimates for RMBI Factors Theorized to Predict Adjustment (N = 217; R? = .146) B SE B ? Critical Ratio Adjust <--- Adjustment -.029 .071 -.040 -.408 Adjust <--- Finances .368 .110 .356 3.338* Adjust <--- Partner .070 .098 .080 .716 Adjust <--- Priority -.071 .055 -.111 -1.293 Adjust <--- Success -.034 .083 -.044 -.406 Adjust <--- Stepfamily .016 .097 .021 .163 Adjust <--- Past -.096 .072 -.153 -1.326 * p < .01 125 Appendix N Summary of Hierarchical Regression Analysis for the RBI and RMBI using the Utah Sample 126 Hierarchical Regression Analysis for the RBI and RMBI (N = 217) Satisfaction Adjustment B SE B ? B SE B ? Step 1 Disagreement -.147 .094 -.135 -.495 .148 -.279 ** Sex -.025 .077 -.023 .190 .122 .110 Mindreading -.014 .103 -.011 -.078 .163 -.039 Change -.336 .095 -.245 ** -.490 .150 -.219 ** Step 2 Disagreement -.187 .098 -.171 -.464 .153 -.261 * Sex -.021 .079 -.020 .091 .123 .052 Mindreading -.016 .104 -.013 -.064 .162 -.031 Change -.323 .096 -.235 ** -.415 .149 -.186 * Adjustment -.105 .084 -.089 -.029 .131 -.015 Stepfamily -.098 .128 -.055 .160 .200 .055 Priority .192 .097 .136 * -.157 .151 -.068 Success .041 .086 .034 .012 .135 .006 Partner .184 .094 .146 * .180 .146 .088 Past .171 .116 .103 -.281 .181 -.104 Finances .006 .097 .004 .531 .151 .229 ** * p < .01 ** p < .001 Change Statistics Model R R? R? Change F Change df1 df2 Sig. F Change Satisfaction 1 .328 .107 .107 6.379 4 212 .000 2 .393 .155 .047 1.642 7 205 .125 Adjustment 1 .398 .159 .159 9.993 4 212 .000 2 .473 .224 .065 2.447 7 205 .020