TRUST IN DATING RELATIONSHIPS: THE ROLE OF COMMUNAL ORIENTATION Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. This thesis does not include proprietary or classified information. _________________________________________ Philip Edward Thorsen Certificate of Approval: _______________________ _______________________ Francesca Adler-Baeder Leanne K. Lamke, Chair Associate Professor Professor Human Development and Human Development and Family Studies Family Studies _______________________ _______________________ Donna L. Sollie Joe F. Pittman Profesor Interim Dean Human Development and Graduate School Family Studies TRUST IN DATING RELATIONSHIPS: THE ROLE OF COMMUNAL ORIENTATION Philip Edward Thorsen A Thesis Submitted to the Graduate Faculty of Auburn University in Partial Fulfillment of the Requirements for the Degree of Master of Science Auburn, Alabama December 15, 2006 iii TRUST IN DATING RELATIONSHIPS: THE ROLE OF COMMUNAL ORIENTATION Philip Edward Thorsen Permission is granted to Auburn University to make copies of this thesis at its discretion, upon request of individuals or institutions and at their expense. The author reserves all publication rights. _______________________ Signature of Author _______________________ Date of Graduation iv THESIS ABSTRACT TRUST IN DATING RELATIONSHIPS: THE ROLE OF COMMUNAL ORIENTATION Philip Edward Thorsen Master of Science, December 15, 2006 (B.A., Cedarville University, 2001) 97 Typed Pages Directed by Leanne Lamke The purpose of this study was to examine the relationship between communal orientation and trust in recently established dating relationships. We expected that the degree to which one?s partner was perceived to have a communal orientation would be associated with trust. We also expected that perceptions of one?s partner would be related to one?s own communal orientation as well as the self-reported communal orientation of one?s partner. Perceptions of the communal orientation of one?s ideal partner were anticipated to partially mediate the association between one?s communal orientation and one?s perceptions of his or her partner. The sample included 302 undergraduate students and their relationship partners (151 couples). Of these couples, only 10 reported relationship durations of one month or less, limiting our findings to established relationships. Participants completed the Trust Scale (Rempel, Holmes, & Zanna, 1985) and the Communal Orientation Scale (Clark, Powell, Ouellette, & Milberg, v 1987) pertaining to the one?s own communal orientation as well that of their partner and their ideal partner. We found that women?s perceptions of their partner were related to their levels of trust and these perceptions were partially based on men?s self-reported communal orientation. Women?s perceptions of the communal orientation of their ideal partner accounted for a large portion of the remaining variance in perceptions of their partner. Perceptions of one?s ideal also fully mediated the relationship between women?s self- reported communal orientation and their perceptions of their partner. Men?s perceptions of the communal orientation of their ideal partner also mediated the relationship between their own communal orientation and their perceptions of their partner, but their perceptions of their partner were only marginally associated with trust. Instead, men?s self-reported communal orientation was related to their perception of trust in the relationship. These findings suggest that men?s and women?s communal orientation play an important role in understanding perceptions of trust in dating relationships. vi ACKNOWLEDGEMENTS I would like to thank my major professor, Leanne Lamke, for seeing me through the odyssey that we call, my thesis, and providing me with motivational words when I needed them the most. Thanks to Donna Sollie and Francesca Adler-Baeder who served on my committee and provided helpful insights for the project. I also would like to thank any number of friends and colleagues who have been supportive along the way. vii Style manual or journal used American Psychological Association, Fifth edition Computer software used Microsoft Word viii TABLE OF CONTENTS LIST OF TABLES????????????????????????....... ix LIST OF FIGURES?????????????????????????.. x INTRODUCTION?????????????????????????? 1 REVIEW OF LITERATURE????????????????????...... 10 Trust???????????????????????????...... 10 Communal orientation????????????????????....... 21 METHODS????????????????????????????... 36 RESULTS?????????????????????????????. 43 DISCUSSION???????????????????????????... 58 REFERENCES???????????????????????????. 68 APPENDICES???????????????????????????.. 79 Appendix A?????????????????????????.. 80 Appendix B?????????????????????????.. 81 Appendix C?????????????????????????... 83 Appendix D?????????????????????????.. 86 Appendix E?????????????????????????... 87 ix LIST OF TABLES TABLE 1 Means and standard deviations for the subscales of the Perceived Relationship Quality Components questionnaire??????...... 43 TABLE 2 Zero order correlations, means, and standard deviations for the study variables????????????????????... 45 TABLE 3 Goodness-of-fit indicators for the confirmatory factor analyses?.. 46 TABLE 4 Path coefficients for the hypothesized model????????... 51 x LIST OF FIGURES FIGURE 1. Hypothesized model of associations between trust and communal orientation???????????????????????? 8 FIGURE 2. Final model depicting the associations and path coefficients for each relationship???????????????????????... 56 1 INTRODUCTION Often trust is considered to be a significant part of relationships, especially intimate relationships. Trust provides a foundation upon which close relationships can be built and allows relationships to function effectively and to the satisfaction of the persons involved. Initial investigations of trust focused on individual differences in expectancies for people to behave in honest and reliable ways (Rotter, 1967; Wrightsman, 1964). Other researchers have examined cooperation during the Prisoner?s Dilemma game as a proxy for trust, and found that cooperation varied according to the game partner as well as circumstances under which the game was played (Deutsch, 1958; Gallo & McClintock, 1972; Lave, 1965; Loomis, 1959; Schlenker, Helm, & Tedeshci, 1973). These investigations were followed by research that posited trust was based on the history of two people rather than individual differences, and that different types of trust come to bear in different situations (Johnson-George & Swap, 1982). Other research focused on attributions of benevolence or sincerity made toward one?s marriage partner (Larzelere & Huston, 1980). Most recently, Rempel, Holmes and Zanna (1985) have offered a process view of trust in close relationships. According to this conceptualization, trust is established over a period of time as judgments about predictability, and dependability are made, which in turn lead to faith concerning one?s partner. According to Rempel et al. (1985), partners in a relationship judge each other to be trustworthy based on information attained through continued interaction. They 2 explain that predictability is the first facet of trust which is given attention. Predictability is established as one?s partner behaves in reliable and consistent ways in a number of interactions. Predictability is at its fullest point when a partner?s future behavior can be forecast or anticipated. The second contributor to trust is dependability. Dependability is the process by which trust is attributed to the partner as a character trait. Judgments of dependability are based on whether or not one?s partner chooses to respond to one?s needs in situations where risk and vulnerability are prominent concerns. Choosing to respond to needs in these situations is central to attributions of dependability and a growing level of trust. As partners care for each other?s needs, certainty about the relationship grows and trust enters a third stage. This facet of trust involves making judgments that a partner will continue to respond to one?s needs in the future, even in circumstances that cannot be anticipated. This level of trust is called faith, based on the leap of faith that individuals make when committing to a relationship to the extent that whatever happens, they will trust their partner to care for their needs. Thus, trust develops over the course of multiple interactions and moves from an emphasis on predictability to making character attributions about dependability to the emotional security found in faith. Based on this conceptualization, Rempel et al. (1985) devised the Trust Scale which includes items written to specifically assess predictability, dependability and faith. The development of this relational measure of trust prompted other researchers to use the Rempel et al. conceptualization of trust to guide their research. These investigations have examined the impact that personality traits such as uncertainty orientation (Sorrentino, Holmes, Hanna, & Sharp, 1995), and self-monitoring (Norris & Zweigenhaft, 1999) have 3 on levels of trust. Other researchers have studied the association between trust and different processes such as self-perception (Zak, Gold, Ryckman, & Lenney, 1998), making attributions (Miller & Rempel, 2004; Rempel, Ross, & Holmes, 2001) and the development of commitment (Wieselquist, Rusbult, Foster, & Agnew, 1999). It is interesting to note that the majority of these studies involved samples composed of couples in longer term relationships. Of the 349 couples that participated in these studies, approximately 74% had been married or cohabiting for at least 2 years, 23% had been dating steadily or for at least two months, 2% were engaged or married and 1% were casually dating or in the beginning stages of their dating relationships. One additional study described their sample as consisting of 64 couples who had been dating for a span of 1 to 96 months. These numbers reflect an emphasis on trust and couple interactions that have occurred over a period of time in established relationships. In contrast, a study conducted by Fletcher, Simpson, and Thomas (2000b) suggests that an extended amount of time or interaction is not necessary for high levels of trust to develop. Fletcher et al. investigated the effect that the comparison of beliefs about one?s ideal partner to perceptions of one?s current partner has on relationship quality among college student couples during the first month of dating. In order to examine current perceptions of relationship quality, a measure was developed that assessed levels of seven different relationship constructs including relationship satisfaction, romance, passion, commitment, intimacy, love and trust. Results indicated that levels of trust were consistently higher than the other aspects of relationship quality in the first month of dating. Often the difference between levels of trust and the other relationship quality components was statistically significant. Fletcher et al. concluded from these findings 4 that a somewhat high level of trust is present at the beginning of a dating relationship, or perhaps even before the onset of the relationship. If high levels of trust are found early in dating relationships before partners can interact to the extent which Rempel et al. (1985) propose is necessary for trust to develop, then alternative explanations for trust development must be considered. Perhaps trust does not only lie in the ability of an individual to consistently care for the needs of a partner. It may be that individuals begin relationships ready to respond to the needs of one?s partner, understanding that a dating relationship is communal in nature and part of the communal norm is responding to the needs of one?s partner (Clark & Mills, 1979). Communal relationships are characterized by a motivation to care for the welfare of another person, whether it is by responding to the needs of that person, or incurring cost without being compensated (Clark & Mills, 1979). Exemplars of this type of relationship are friendships, family relationships, and intimate relationships. In contrast, exchange relationships are characterized by a balanced exchange of benefits and a motivation to settle a debt in a relationship as soon as opportunity arises. This type of relationship is best modeled by strangers, business partners, and coworkers who do not consider themselves to be friends. Experimental studies have shown that individuals are likely to behave in communal ways if they believe the possibility of a communal relationship exists (Clark & Mills, 1979). If individuals are motivated to establish a communal relationship, they may act as though such a relationship already exists to indicate their intentions for the relationship to continue and grow closer. Their behavior may include monitoring the needs of the other person so that a need may be addressed as it arises, without regard for 5 the cost of fulfilling the need or concern that the benefit will be reciprocated (Clark, 1984; Clark, Mills, & Corcoran, 1989; Clark, Mills & Powell, 1986). Another aspect of communal relationships is that benefits given and received are often not comparable to one another, because comparable benefits may be a sign of repayment and this may cause the relationship to regress (Clark, 1981). Individuals in a communal relationship may even avoid giving comparable benefits so that their partner in the interaction is not given the false impression that an exchange relationship is desired. Responding to the needs of one?s partner is an expression of the person?s intentions to begin a closer relationship such as a friendship or romantic relationship, and functions to build certainty about the type of relationship that exists. As each individual in the relationship acts in communal ways, each becomes more certain of the relationship and confident that one?s needs will be met. Variability in one?s propensity to act in accord with a communal norm may be rooted in an individual?s communal orientation. A communal orientation is a person?s willingness to respond to others? needs and the expectation that others will respond to his or her needs; simply put, a willingness to give help and a willingness to be helped (Clark, Ouellette, Powell, & Milberg, 1987). In research designed to investigate the concept of communal orientation, Clark et al. specifically focused on the relationship between communal orientation and helping behavior. Individuals high in communal orientation helped more than individuals low in communal orientation. Also, individuals high in communal orientation helped more in response to a confederate?s sad mood than when a confederate seemed to have a neutral mood. These results support the assertion that a communal orientation is not solely determined by who the other is, but may be a source 6 of individual difference which affects how an individual approaches close relationships in terms of their own communal behavior. Research by Ross, Greene, & House (1977) indicates that individuals tend to assume that others share the same attitudes and would make similar decisions as they would in similar situations. If one were to have a high communal orientation then, according to Ross et al., one also would believe that one?s partner would have a similarly high communal orientation, and trust in the partner would result. This assertion is also supported by McCall, Reno, Jalbert, and West (2000) who found that participants reported that task partners, whom they knew very little about, had levels of communal orientation similar to their own. Perceiving one?s partner as having a high level of communal orientation similar to one?s self would then lead to higher levels of perceived trust in one?s dating partner. Support for this link has been provided by Zak et al. (1998) who found a positive correlation between trust and communal orientation. That is, individuals in dating relationships who were high in communal orientation also reported high levels of trust. Perceiving one?s partner as similar to one?s self also may serve to strengthen the relationship by establishing certainty about the relationship (Fletcher, Fincham, Cramer, & Heron, 1987) and security or confidence in the continuation of the relationship (Murray & Holmes, 1997). This perception also would function to supplement shortages of knowledge about one?s partner in the early stages of relationship development. Murray, Holmes, and Griffin (1996) found support for the similarity of partner beliefs and further found that individuals often viewed their partners more positively than partners viewed themselves. 7 Murray et al. (1996) attributed this seemingly unfounded positive view of one?s partner to one?s beliefs about the ideal partner. One?s ideal partner is an image of traits and behaviors which an individual desires and expects of a partner based on present and past experiences (Sternberg & Barnes, 1985). This ideal often functions as a standard by which current partners are judged and therefore often possesses qualities which one?s current partner may not have, or has to a lesser extent (Fletcher, Simpson, Thomas & Giles, 1999). Murray et al. proposed, and found support for, a process of partner perception in which self-perceptions pertaining to individual difference variables (i.e. patient, understanding, intelligent, etc.) led to perceptions of these characteristics in one?s ideal partner which then contributed to perceiving these characteristics in one?s current partner. A similar process also may take place in the association between communal orientation and trust such that one?s own communal orientation may contribute to perceptions of one?s ideal partner leading one to perceive that one?s partner has a high level of communal orientation which would then lead to higher levels of trust. It is important to note, however, that perceptions of one?s current partner also are constrained by some degree of reality. In other words, perceiving one?s partner as having a communal orientation is not based only on the projection of one?s communal orientation, but also on the actual communal orientation of one?s partner as reported by the partner. Consistent with this proposition is the finding of Murray et al. (1996) that beliefs about an ideal partner and self-perceptions reported by a current partner both contributed to views of one?s partner. In summary, it appears likely that levels of communal orientation may influence perceptions of trust in the early stages of dating relationships. This link may occur through the pathways characterized in Figure 1. First, one?s communal orientation may be directly associated with perceptions of trust. Second, this link may be mediated by Figure 1: Possible pathways by which communal orientation may be projected onto one?s partner Male Level of Trust in Partner Predictability Dependability Faith Female Self C. O. Self- perceived help Self Involve -ment Female Ideal Partner C. O. Ideal Partner help Ideal Involve -ment Perceived Partner help Perceived Partner Involve- ment Female Perceived Partner C. O. Female Level of Trust in Partner Dependability Male Self C. O. Self- perceived help Self Invole- ment Male Perceived Partner C. O. Perceived Partner help Perceived Partner Involve- ment Male Ideal Partner C. O. Ideal Partner help Ideal Involve ment Faith Predictability perceptions of one?s partner?s level of communal orientation. The perception that one?s partner will respond to one?s needs also will be a function of the partner?s actual level of communal orientation. Additionally, a partner?s communal orientation is likely to be 8 9 based on views of an ideal partner, which is a function of one?s own communal orientation. It is expected that high levels of communal orientation will be associated with high levels of trust in dating relationships that are in the first month of existence. More specifically, it is hypothesized that: 1. Consistent with the work of Fletcher, et al. (2000b), levels of trust among individuals who have been dating for one month or less will be significantly higher than levels of satisfaction, commitment, intimacy, passion and love. 2. Self-perceived communal orientation will be associated with perceptions of trust. 3. Perceiving one?s partner as having a communal orientation will be related to perceptions of trust in the relationship. 4. Perceptions of one?s partner?s level of communal orientation will be influenced by one?s partner?s self-reported communal orientation as well as one?s self-perceived level of communal orientation. 5. The link between self-perceived levels of communal orientation and perceptions of partner?s communal orientation will be mediated by perception of an ideal partner?s level of communal orientation. 10 LITERATURE REVIEW In order to better understand the link between trust and communal orientation it is important to describe the two concepts in more detail. First, an historical overview of the concept of trust in intimate relationships will be presented. Next, the relationship between communal norm as the basis for the conceptualization of a communal orientation will be discussed. Finally, research that has focused on communal orientation will be reviewed. Trust Past inquiries examining trust reflect a range of different conceptualizations. One early conceptualization of trust suggested that one has trust to the extent that he or she expects an event to occur which has a greater potential to harm the person if the trust is not fulfilled than to benefit the person if the event does occur and trust is fulfilled (Deutsch, 1958). In other words, trust is predicated on the risk of exploitation or the hope of benefit. This notion of trust has been tested using mixed motive games, and more specifically, the Prisoner?s Dilemma Game, in which the interests of one player are pitted against the interests of the other (Brickman, Becker & Castle, 1979; Gallo & McClintock, 1972; Kelley & Stahelski, 1970; Lave, 1965; Schlenker, Helm, & Tedeschi, 1973). Research using the Prisoner?s Dilemma Game has focused primarily on trusting behavior, and the situations in which that behavior occurred. Although some of this research did try to associate attitudes of trust with trusting behavior (Deutsch, 1958; 11 Loomis, 1959) these studies typically dealt with situationally manipulated attitudes. An early conceptualization of trust that did focus on attitudes was formulated by Wrightsman (1964) who included trustworthiness as one subscale in his measure of Philosophies of Human Nature (PHN). Wrightsman asserted that one?s philosophy of human nature is what one believes about how people generally act, and these beliefs are used to make sense of other?s behavior as well as one?s own. He also believed that trustworthiness or ?the extent to which people are seen as moral, honest and reliable? (p. 744), is a fundamental element of how one perceives other individuals. Wrightsman tested his conceptualization of trustworthiness using fourteen items measuring the expectation of how people would act when placed in different situations involving honesty or reliability. Other subscales of the PHN measure included altruism, independence of conviction, strength of will and rationality, complexity of human nature and variability of human nature. Another conceptualization of trust, explicated by Rotter (1967), emphasized the importance of trust for cooperation in society and social stability. Rotter defined interpersonal trust as ?a generalized expectancy held by an individual that the word, promise, oral or written statement of another individual or group can be relied on? (p. 1). It is this ability to believe that others are reliable which allows people in society to cooperate in everyday activities. Rotter couched this conceptualization of trust in social learning theory such that individuals will trust or distrust others based on the expectation that a given behavior will lead to a certain outcome which acts to reinforce the trusting or distrusting behavior. In this way, an individual may come to believe certain social agents 12 are trustworthy, or are able to be relied upon, and this attitude may generalize to other social agents. When considering trustworthiness (Wrightsman, 1964) and interpersonal trust (Rotter, 1967) together, it is clear that one?s ability to believe that what someone says is true, or what is promised will take place, is a key component of trust. Correlations between these conceptualizations of trust and the tendency to cooperate in the Prisoner?s Dilemma Game (Schlenker, Helm, & Tedeschi, 1973; Wrightsman, 1966) show that these three measures probe the same attitude and provide evidence that a trusting attitude does lead to trusting behaviors. It is also clear that trust, as conceptualized by Wrightsman and Rotter, is composed of two dimensions. One dimension concerns trust attitudes toward public figures or bodies with which an individual would have little or no direct contact. The second dimension deals with trust in more interpersonal relationships which occur in everyday situations (Chun & Campbell, 1974; 1975). These conceptualizations of trust do not, however, investigate the nature of trust in intimate relationships. In an effort to move from more general to specific attitudes of trust, Johnson- George and Swap (1982) suggested that trust was a function of specific situations and specific relationships. Consistent with Swinth (1967), trust was expressed in terms of a risk being taken by an individual in the hope of reaching a common positive outcome when the risk is accepted by the other person, without being exploited, but the relational context was much more specific. An investigation of specific trust was performed which operationalized trusting situations as involving material possessions, dependability or reliability, personal confidences and physical safety (Johnson-George and Swap, 1982). 13 Participants were directed to think about a specific same-sex person whom they trusted a great deal as they completed the trust measure. An analysis of the responses showed evidence for a general trust component including a broad array of trust situations, emotional trust dealing with respect for one?s feelings, reliableness dealing with keeping promises and commitments, dependability involving confidence in another person?s help or sense of responsibility and a physical trust component which dealt with one?s physical safety or well being. Emotional trust and reliableness were further investigated by presenting scenarios in which a person of the same-sex as the participant either keeps or fails to keep a confidence (i.e. emotional trust) or arrive on time or late for an appointment (i.e. reliableness). Both of these manipulations were successful in affecting the prospective attitude of trust toward the specific person, supporting these elements of trust for friendships. In contrast to this emphasis, Larzelere and Huston (1980) focused on dyadic trust, or trust in romantic heterosexual relationships, such as those between dating or married partners. Larzelere and Huston proposed that the benevolence of one?s partner, or genuine interest in the welfare of one?s partner, is an important aspect of relationship trust. Benevolence seems to share the same properties as the cooperative orientation as used in investigations of the Prisoner?s Dilemma Game (Deutsch, 1958). The second attribute they thought necessary for trust was honesty, or the extent that the disclosures of one?s partner about future intensions may be believed. This conceptualization of trust led to a definition of trust which states that, ?trust exists to the extent that a person believes another person (or persons) to be benevolent and honest? (p. 596). 14 This conceptualization was tested using an eight item measure which fit this specific definition of dyadic trust (Larzelere and Huston, 1980). Both a dating sample and a married sample were recruited for this investigation. Each sample was divided into smaller groups according to commitment level, operationalized as relationship length. The dating sample included 16 individuals who were casually dating, 90 individuals who were exclusively dating, 54 individuals engaged or living together and 35 individuals not currently in a relationship, who reported about a past dating relationship. There were 40 couples represented by both partners in the dating sample. The married sample included 20 newlywed couples (married less than two months), 20 longer married couples, and 45 divorced individuals. The findings of this investigation revealed that trust was a unidimensional construct which was minimally correlated with Wrightsman?s (1964) Trustworthiness subscale and not correlated with Rotter?s (1967) Interpersonal Trust Scale, thus demonstrating the discriminant validity of relationship trust from these two measures of general trust. Larzelere and Huston also found that dyadic trust was highly correlated with love and self?disclosure. Also, the degree to which love and trust were correlated varied according to commitment level. Longer married couples? scores of trust and love were most highly correlated, followed by exclusively dating couples, newlyweds, and then engaged or cohabiting couples. Levels of trust also varied according to commitment level. Divorced individuals had the lowest levels of trust. Trust levels then progressively increased among the different groups with ex-dating partners having the second lowest level of trust, followed by casual daters, exclusive daters, and engaged or cohabiting couples. Longer married couples and newlyweds had the highest levels of trust. Trust also was significantly reciprocated between partners, 15 whereas love and self-disclosure were not. This finding was replicated by Butler (1986) who also found that marital status had an effect on levels of trust, in a similar pattern with Larzelere and Huston. These findings suggest that relationship partners who are more committed to one another, or have been together longer, have higher levels of trust than partners in relationships with less commitment and of shorter duration. In summary, research investigating conceptualizations of trust concerning specific relationships such as friendships and romantic relationships seem to show that trust is characterized by dependability, reliableness, respect for one?s feelings and keeping confidences (Johnson?George & Swap, 1982) honesty and benevolence (Larzelere & Huston, 1980). These components of trust served as the foundation upon which Rempel, Holmes and Zanna (1985) developed their process conceptualization of trust in romantic heterosexual relationships. According to Rempel et al. (1985), early in a relationship when partners have had limited experiences together as a couple, uncertainty may arise as an individual realizes the degree of dependence he or she has upon the intentions of the partner for the relationship to continue. The uncertainty is the result of the realization that the more the individual cares for the partner the more he or she has to lose. The situation is worsened because the individual does not hold sufficient knowledge about the partner toward making a judgment of these intentions. The prospect that the partner has little interest in continuing the relationship long term would have consequences for the individual because there is a dependence upon the partner to provide the benefits or rewards that are expected of a close relationship. Also, if the relationship ends, any investments made in the relationship at the expense of the individual would be lost. 16 Investment in the relationship is the catalyst that causes the relationship to continue and contributes to a process of trust building that has at its core three concepts: dependence, uncertainty and uncertainty reduction (Holmes & Rempel, 1989). A partner who continues to invest in the relationship does so at the expense of some resource such as time or money. Increasing investment in the relationship results in dependence upon the partner for some judgment of whether the investment is acceptable or not. In this stage of the relationship there is a great amount of risk due to the high level of investment and dependence. Uncertainty about the future of the relationship may rise as this judgment of investments is not offered and assurance of the continuation of the relationship is not demonstrated. Such assurance might be found in behavior during situations that give the partner the option to choose to invest in the relationship and respond to the needs of the partner, or to choose in favor of self interest. Behavior which shows responsiveness to the needs of the partner and the relationship work to reduce feelings of uncertainty which gives relationship partners a sense of security that the relationship will continue. Certainty of the relationship is established in the balance of mutual reward and equal involvement. In order for the level of uncertainty to decrease, and the level of trust to increase, two processes take place. The first process involves interpreting behavior so that a judgment about the partner?s commitment may be made. Behaviors which are most useful for making decisions about the commitment of the partner are behaviors that put the partner in a vulnerable position. Such behaviors involve taking some degree of risk, as with sharing personal feelings and thoughts or believing the claims of a partner to fulfill a commitment. Other behaviors also involve sacrificing self-interest which might mean 17 skipping a sporting event in lieu of attending the high school graduation of a partner?s sibling or helping the partner with a project instead of going out with friends. Perception of such behaviors leads an individual to believe that the partner is committed. The second process occurs as the risky or self-sacrificial behavior is reciprocated by the individual. When the behavior is reciprocated it is a signal that offers reassurance that the partner is also willing to invest in the relationship. This reciprocity may be in the form of agreements that if the partner will do this, then the individual will do that. For example, one might attend an organizational party if the other will attend the wedding of a friend. As partners continue to interact, information about the partner accumulates. Rempel et al. (1985) explain that the information which influences levels of trust the most is in the areas of predictability and dependability. Predictability is the most basic contributor to levels of trust. The degree to which a partner is predictable depends upon the consistency with which the person has performed in past reoccurring behaviors. These behaviors may be any form of interaction such as nightly phone calls to discuss the day?s events, or even the way a person becomes moody after performing poorly on a task. If the conduct of a partner in a relationship has been consistent to the degree that future behavior may be forecast, then a partner would be considered predictable and worthy of trust. Information about dependability is the second contributor to trust and is based less upon specific behaviors and more upon an ?evaluation of the qualities and characteristics attributed to the partner? (Rempel, et al., 1985, p. 96). Attributions of dependability are based upon behaviors performed in specific circumstances involving risk and 18 vulnerability where issues of trust are a prominent concern. Such situations present partners with a choice between a caring response and rejection, such as when a new hair style is attempted or when deeply personal topics are disclosed to one?s partner. When the partner has responded consistently in a caring way, then the individual judges the partner to be dependable and trust is enhanced. Thus, trust develops as individuals behave in predictable and dependable ways, thereby decreasing uncertainty about the individuals? commitment to the relationship. As a result of this decrease in uncertainty, the nature of the exchange between partners changes. As the feelings of trust continue to grow, partners are less motivated to immediately reciprocate caring behavior, or expect reciprocation, because there is confidence that the partner will respond when a need arises (Holmes & Rempel, 1989). This responsiveness to the needs of one?s partner further decreases uncertainty in the relationship because it is evidence that one?s partner is concerned for one?s well-being and is sensitive to providing benefits that will maintain that well-being (Boon & Holmes, 1991). Partners continue to exchange caring behaviors, but the equity of this balance is not scrutinized as closely as it was earlier in the relationship when mutual feelings of trust were not yet established. Trust is further developed as the areas in which partners are involved continue to expand. Interactions in new contexts present further information about the interests and preferences of the partner. In this deeper intimacy it may be discovered that the partner has less in common with the individual than was originally supposed. The problems caused by this conflict of interests are compounded by the realization of heightened dependence upon the partner and the increasing difficulty involved in leaving the 19 relationship. Feelings of trust and commitment are tested as the benefits of continuing to meet partner needs and continuing involvement in the relationship are weighed against costs of sacrificing those values and attitudes which are central to the individual?s personality. As partners encounter and negotiate these situations, they are aware of the impact these behaviors have for the future of the relationship. Accommodation to these preferences or caring response to the vulnerability of these situations affirms the individual?s identity as separate from the relationship and demonstrates commitment to the relationship. Following such accommodations the individual judges that the partner can be trusted to do what is necessary to meet future needs. After relationship partners have encountered and navigated these challenges, and trust continues to grow, a point arises when evidence of past caring responses can no longer mirror the situations that might be found in the future of the relationship. In this instance Rempel et al. (1985) speculate that individuals must make a leap of faith that the partner will be willing to respond to the needs of the individual in those future circumstances. At this point the individual has reached the stage of trust development which Rempel et al. characterize as faith. Faith functions to make judgments about the future without reassuring evidence, whereas judgments of predictability and dependability are made upon sufficient evidence found in past experiences with the partner. At this level, trust reflects an attitude that is not threatened by uncertainty and looks to the future without anxiety. In summary, the development of interpersonal trust involves a process of uncertainty reduction which occurs as partners interact in new situations and contexts. The development of trust depends on the willingness to demonstrate caring for one?s 20 partner by taking risks and responding to needs at some personal cost. As partners perceive each other acting in predictable and dependable ways, they learn about each other?s character and orientation toward the relationship, and trust develops. From this growing body of knowledge individuals judge whether the partner and relationship are worth continued investment. After extended engagement in interactions, individuals move forward basing their future on faith that their partner will continue to respond to their needs whatever situations that will be encountered. This conceptualization of trust was tested by using a 17 item trust scale devised by Rempel et al. (1985) emphasizing predictability, dependability and faith. The development of the measure prompted other researchers to use this conceptualization of trust to guide their research. Some of these investigations examined the impact that personality traits have on levels of trust. One investigation examined the relationship between trust and styles of coping characterized as uncertainty or certainty orientation (Sorrentino, Holmes, Hanna, & Sharp, 1995). An individual with an uncertainty orientation is willing and able to accept and learn from new information and experiences where there is uncertainty about self or the situation. A certainty oriented individual is most comfortable in situations that do not offer uncertainty about self or situation, and avoids circumstances that might present uncertainty, or ambiguity. Seventy?seven couples who had been living together for at least two years participated in this study. They found that those who are able to accept uncertainty in their relationships held moderate levels of trust while those who must have a sense of certainty held either high or low levels of trust. By having a low or high level of trust certainty oriented individuals could avoid the ambiguity which might be a result of a moderate level of trust. 21 Another study performed by Norris and Zweigenhaft (1999) investigated the effect of self-monitoring orientation on levels of trust in romantic heterosexual relationships. High self?monitors and low self?monitors have different approaches to social situations, in that high self-monitors tend to focus on external factors for friendships and do not have a high level of commitment in romantic relationships, while low self-monitors focus on likeability and affection in friendships and hold attitudes of commitment in romantic relationships. Thirty-eight heterosexual couples who had been dating for at least two months participated in the study. This investigation revealed that low self-monitors were more likely to have high levels of trust, while high self-monitors held low levels of trust. These results suggest that individual difference variables such as self-monitoring and uncertainty orientation are related to individual attitudes of trust. As will be seen in the next review of literature section, it is also likely that communal orientation or the tendency for an individual to meet the needs of someone else is associated with trust in a manner similar to other individual difference factors. Communal Orientation In contrast to the proposition that trust in a relationship is based on gathered information about a relationship partner, several studies by Clark and her associates suggest that a person?s orientation toward responding to the needs of another may lead one to behave as though trust already exists in a relationship, even though the length of acquaintance is brief. This propensity to respond to the needs of another person has been characterized as a communal orientation (Clark, Ouellette, Powell, & Milberg, 1987) and 22 is based on research which distinguishes between communal and exchange relationships and the norms by which they function. Exchange relationships involve a norm of reciprocation in which benefits are given and received one for another, or tit-for-tat (Clark & Mills, 1979). In contrast, a communal relationship is based on a norm which reflects a motivation to care for the welfare of another person, whether it is by responding to the needs of that person, or incurring cost without being compensated. Participants in a communal relationship abide by the communal norm by providing for the needs of one another as opportunities arise. The provision of a benefit does not oblige the other to reciprocate, nor does it relieve the requirement to attend to one another?s needs. Because benefits in response to needs may be given and received in close succession, the provision of these benefits may be viewed as repayments as part of an exchange norm, however the motivation remains to meet a need which is consistent with a communal norm. Thus, although the communal norm may appear to be reciprocal at times, the motivation is in response to needs. Research investigating the difference between communal and exchange norms has been based primarily on the use of experimental designs that manipulate expectations among strangers. Clark & Mills (1979) performed one such study which examined the association between attraction, and giving and receiving benefits. The goal was to identify how individuals in exchange and communal conditions would respond to receiving a benefit from a task partner (i.e. enactment of a communal norm). Clark and Mills expected that participants expecting a communal relationship would form a positive impression of one?s partner if a benefit was not followed by a repayment, while participants who expected an exchange relationship would have a more positive 23 impression of one?s partner if a benefit was followed by a repayment. In the joint task method, participants had an opportunity to help their partner with her task and all did so. When the partner gave one of the points she earned from completing the task to the participant in return for his help, participants in the communal condition responded negatively, while participants in the exchange condition responded positively. The opposite also was true, in that when a point was not given, exchange participants responded negatively and communal participants responded positively. These findings suggest that believing that a communal relationship is desirable (i.e. attractive) and possible (i.e. available) lead to communal behavior in responding to needs or receiving a benefit. These findings also suggest that people are aware of the communal and exchange norms at the outset of relationships, and pattern their behavior after these norms, thus attempting to build a closer relationship or friendship in communal ways, or maintaining the distance and formality of an exchange relationship. Another characteristic of the communal norm that describes friendships, family relationships and romantic relationships is the dissimilarity of benefits that members of these relationships provide to one another. Clark (1981) performed a study composed of three experiments to investigate differences in how relationships were viewed based on whether or not benefits were comparable. Reasoning that the motivation for providing benefits in communal relationships is to provide for a need, Clark expected that relationships in which benefits exchanged were dissimilar would be viewed as friendships while relationships which exchanged similar benefits would be viewed as exchange relationships. 24 The results indicated that the kind of benefits given and received in relationships are viewed as cues about the type of relationship which is intended or already exists because interactions which involved comparable benefits were not as consistently rated as friendships as those interactions which involved dissimilar benefits. The more alike the benefits, the more they were perceived as repayment, whereas the more dissimilar the benefits, the more they were viewed as meeting a need or as an effort to please the other person, and thus the relationships which involved dissimilar benefits were perceived as friendships. An additional way to distinguish between communal and exchange relationships is how one monitors the giving and receiving of benefits in the relationship. In a communal relationship the focus is on responding to a current need without regard for when the last benefit was given. In exchange relationships giving and receiving benefits is strictly monitored so that equity might be maintained. Monitoring needs in communal and exchange relationships has been investigated in three different studies (Clark, 1984; Clark, Mills, & Corcoran, 1989; Clark, Mills, & Powell, 1986). Clark (1984) performed three experiments to investigate the hypotheses that people desiring an exchange relationship will monitor investments in a joint task for which rewards will be offered, and people desiring a communal relationship will not monitor such investments. In the first experiment relationship orientation was experimentally manipulated, but in experiments 2 and 3 relationship orientation was manipulated by pairing participants with a friend who agreed to participate in the experiment with them (communal condition), or with another participant?s friend whom they did not know (exchange condition). Partners participated in a joint task and 25 responses to the task were recorded on the same sheet of paper. Whether the participants monitored investments or not was judged by whether the participant used a different colored pen than their partner. Clark reasoned that a different color pen would allow easy recognition of inputs contributed by either partner, which would allow fair dispersal of the reward when the task was completed. The combined findings of these three experiments revealed that people expecting the communal norm or currently in communal relationships did not monitor contributions to a joint task even though rewards were to be given based on performance. On the other hand, people in an exchange relationship or expecting the exchange norm were motivated to monitor contributions so that equity could be maintained when rewards were distributed. Clark, Mills, & Powell (1986) further investigated the monitoring of partners? needs. It was hypothesized that if no opportunity was available for one?s task partner to reciprocate help given, then people desiring a communal relationship would monitor needs more than people desiring exchange relationships. On the other hand, if one?s partner has an opportunity to reciprocate, then people desiring an exchange relationship will monitor needs more than when no opportunity to reciprocate is given, while this opportunity to reciprocate will make no difference to people desiring a communal relationship. Results showed that participants in the communal-no opportunity condition monitored the needs of one?s task partner more often than exchange-no opportunity participants. Also, as expected, the exchange-opportunity participants monitored the needs of one?s task partner more than participants in the exchange-no opportunity 26 condition. There was no difference in monitoring between participants in the communal- opportunity and communal-no opportunity conditions. These findings indicate that people desiring a communal relationship monitor needs even when they were not allowed to help. Monitoring needs was further investigated by Clark, Mills, and Corcoran (1989) in relationships between friends and strangers. The first purpose of this study was to investigate the difference between monitoring for information about needs and monitoring for other types of information such as contributions to a joint task. The second purpose was to investigate if the past finding that people intending to form a communal relationship monitor needs more than people intending to maintain an exchange relationship also applies to established communal relationships. Clark et al. expected to find that members of continuing friendships monitor each other?s needs more than strangers, and that strangers will monitor contributions to a joint task more than friends. Consistent with expectations, friends monitored more than strangers in the needs condition but not in the contributions condition. Also as expected, strangers monitored more than friends in the contributions condition. These findings reveal that the motivation of friends was to monitor needs for help, even though they could not provide help. Friends were less concerned with contributions to the joint task than they were with needs. On the other hand, strangers were more motivated to monitor contributions into the task, rather than needs, so that rewards could be distributed fairly. Feelings of exploitation also differentiate exchange and communal norms. When two people are interacting according to exchange norms, the recipient of a benefit is expected to reciprocate a similar benefit. If this reciprocation does not occur then the 27 benefactor will feel like she or he was exploited. In communal relationships, however, failure to reciprocate a benefit should not create feelings of exploitation because there is no debt accrued in the interaction. Clark and Waddell (1985) performed an experiment which investigated feelings of exploitation in communal and exchange relationships. Researchers anticipated this experiment would show that individuals expecting an exchange relationship would feel more exploited when they were not repaid for a benefit given to a task partner, and that individuals expecting a communal relationship would not feel exploited when a benefit they gave was not repaid. Results revealed that participants who expected an exchange relationship and were led to believe they would not be benefited felt feelings of exploitation significantly more than exchange participants who believed they would be benefited. Participants in the communal condition did not show such a difference because the basis of their help was not on the receipt of repayment. Mills and Clark (1982) discuss another experiment meant to investigate conditions under which individuals in a communal relationship would feel exploited. Again, researchers expected exchange participants to feel exploited at the failure of another person to repay a benefit, but communal participants would not feel exploited by this failure. They also expected that individuals in communal relationships would feel exploited when a need was not provided for and that individuals in exchange relationships would not feel exploited at such a failure. Existing communal relationships were investigated because the communal norm is more certain and established than when a communal relationship is only expected, and not yet certain. Results revealed that participants in exchange relationships, as compared with communal relationships, felt more exploited when a benefit was not repaid, and participants in communal 28 relationships, as compared to exchange relationships, felt exploited when a need was not provided for. The distinction between communal and exchange orientations was further investigated by scrutinizing the association between helping, affect, and types of relationship (Williamson & Clark, 1992). The experimenters specifically investigated situations of helping and not helping as well as situations in which one is required to help or one chooses to help The experimenters anticipated that participants expecting a communal relationship would experience an increase in positive affect when able to help, as compared to not helping. They also expected that positive affect would not increase for participants in the exchange condition who gave help. Positive affect of participants in the communal condition increased, whether or not the individuals were required to help or chose to help. In the exchange condition, choosing to help decreased positive emotions, while being required to help and no help conditions showed little change in affect. These results give further evidence that people behaving according to a communal norm are motivated to meet another person?s needs irrespective of whether they choose to or are required to. Further research was performed by Williamson, Clark, Pegalis, and Behan (1996) who investigated the effect of the type of relationship desired on helping and emotional outcomes using two experiments. In the first experiment, results revealed that participants in the communal condition who rejected the opportunity to help due to the large time commitment required to provide the help experienced more of a decrease in positive affect than participants in the exchange condition. 29 A second experiment was then performed to investigate if similar reports of emotion change would occur when a person recalled a situation in which he or she either succeeded or failed to meet a need of someone with whom a communal relationship already existed, or someone else either failed or succeeded to meet that person?s needs. Williamson et al. (1996) expected that recalling failure to personally meet one?s needs in communal conditions would decrease positive affect and that recalling an event in which someone else failed to meet the needs of one?s communal partner would not result in a decrease in affect unless one was concerned with every instance in which one?s communal partner had a need. Results revealed that positive emotion decreased when one was not able to personally help a communal relationship partner, while positive affect did not decrease among participants who recalled a time when they did help a communal partner or when they either helped or did not help a stranger. When someone else either helped or failed to help, positive emotions did not decrease but showed an increase among participants in a communal relationship. These results reveal that communally motivated people are most concerned with how they personally provide for the needs of their communal partners and might enjoy learning that others are not able to meet the needs of their family member or dating partner as well as they. In summary, this body of research reveals that individuals who intend to form a communal relationship or already have established a communal relationship act in response to the needs of their friend or partner. Individuals behaving according to this communal norm tend to monitor the needs of their partner whether or not their partner has the opportunity to reciprocate the aid given, and avoid monitoring contributions to the 30 relationship so as to prevent the appearance of being motivated by any other purpose than to provide for needs as they arise. Also, these individuals appreciate when others provide benefits to them without having an expectation for reciprocation to occur, interpreting this as evidence that a communal relationship is beginning and one?s needs will be considered. Dissimilar benefits also seem to be an indication of a communal relationship. Individuals expecting communal relationships also enjoy helping others and view themselves more positively when they help whether they choose to help or are required to help, and feel badly when they must pass up an opportunity to help someone, or are unable to help. Finally, individuals in communal relationships are most concerned with their own performance in meeting the needs of a communal relationship partner. These findings are based on research that investigated the communal and exchange norms by manipulating the type of relationship expected or that already existed. Research also has been performed that examines the tendency for individuals to approach relationships in communal ways. The initial study of communal orientation as a disposition examined both the effect of expecting a specific type of relationship and the effect of communal orientation on helping, and mood using two experiments (Clark et al., 1987). In the first study, Clark et al. hypothesized that people high in a communal orientation as indicated by reports on the Communal Orientation Scale (Clark et al., 1987) would help and respond to the sadness of the recipient of the aid more than people low in communal orientation. In the second experiment, Clark et al. anticipated that individuals expecting a communal relationship would help more and respond to the sadness of the recipient more than individuals expecting an exchange relationship. 31 Results for the first experiment revealed that participants who scored high in communal orientation helped more than participants low in communal orientation. Sad mood of the recipient of the help also increased the amount of help given by participants high in communal orientation, as compared to participants low in communal orientation. Results for the second experiment in which relationship orientation was situationally manipulated revealed that participants led to expect a communal relationship helped more than participants expecting an exchange relationship. Sad mood also increased the amount of help given by participants in the communal condition, relative to the exchange condition. Thus, individuals with a high communal orientation are particularly likely to approach relationships according to the communal norm and will respond to the other?s needs, even before a relationship is established. Building upon this initial research, investigations have examined the communal orientation across a range of helping relationships. For example, Bryan, Hammer, and Fisher (2000) investigated what motivates some people to help homeless individuals and found that students with high levels of communal orientation held more positive attitudes about homeless people, reported more empathy and were more likely to help by giving money, as compared to students with a low communal orientation. Williamson and Schulz (1990) also investigated the effect that communal orientation and relationship quality had on family members who become primary caregivers for a related Alzheimer?s disease patient. Results revealed that reports of depressive symptoms were higher among men with low communal orientation scores who rated their prior relationship with the patient as not close, and among women low in communal orientation who rated the relationship as close. Also, feelings of being burdened were highest among men who 32 were low in communal orientation and did not report a close relationship with the patient prior to becoming a caregiver, and among women high in communal orientation who did not have a close relationship Other research has found that high communal orientation among nurses buffers the effects of burnout when inequity is perceived in the relationship with the client (VanYperen, 1996; VanYperen, Buunk, & Schaufeli, 1992). On the other hand, this effect was not found among social workers who experienced lower personal accomplishment when they were high in communal orientation and experienced inequity in their client relationships (Truchot, Keirsebilck, & Meyer, 2000). Truchot and Deregard (2001) sought to replicate this research, and found, as in previous research, that nurses high in communal orientation reporting inequity in their client relationships also reported more feelings of personal accomplishment, while social workers reporting high communal orientation and inequity reported less feelings of personal accomplishment. This research shows that individuals who report high, as compared to low, levels of communal orientation tend to have a better response to helping, especially when the nature of the relationship is such that the person being helped is not expected to reciprocate the benefit. The relationship between communal orientation and perceptions of equity also has been investigated in a sample of Dutch railway employees. Buunk, Doosje, Jans, and Hopstaken (1993) found that workers high in communal orientation reported the highest levels of negative affect when they felt they were under-benefited, reported the lowest levels of negative affect when over-benefited and reported moderate levels of negative affect when they perceived reciprocity. According to these findings, individuals high in 33 communal orientation tend to feel more positive about their work environments when they perceive their needs are being met. Research investigating the relationship between communal orientation and self- serving or other-serving attributions also has been performed. McCall (1995) found that participants claimed more responsibility for success than failure, as consistent with self- serving attributions, but participants high in communal orientation attributed more responsibility for failure to themselves, than they attributed to their partners, and gave more credit for success to their partner than they attributed to themselves. In contrast, people low in communal orientation attributed more responsibility for success to themselves than their partner. This relationship between attributions and communal orientation was further investigated by McCall, Reno, Jalbert, and West (2000) who also examined how perceptions of one?s partner?s communal orientation would affect attributions. They found that participants tended to see their partners as similar to themselves in communal orientation; however, self reports and partner reports of communal orientation did not correspond to the same extent. Also, participants high in communal orientation continued to attribute responsibility to oneself for success, but also attributed responsibility to oneself in failure, especially when they perceived their task partner to be low in communal orientation. Both of these studies give evidence that communally oriented individuals are willing to take blame for failure in an effort to save their task partner from disappointment. Other research focusing more on friendships found that conflict and negativity had a negative impact on satisfaction while communal orientation had a positive impact on satisfaction in the best friendships of senior adults (Jones & Vaughn, 1990). Another 34 study found that communal orientation was a significant predictor of friendship quality for adolescent boys (Jones & Costin, 1995). Thompson and DeHarpport (1998) also found that friends who were similar in communal orientation were able to cooperate more during a negotiation task, such as making decisions about a vacation, when they viewed that task as a problem solving exercise as compared with friends who viewed the task as a bargaining exercise. Also, friends dispersed resources more evenly when both were high in communal orientation than when both were low in communal orientation. Another study investigated concordance, or agreement upon the occurrence of events, between friends in perceiving helping behaviors around a stressful event and found that higher levels of communal orientation were associated with greater agreement about the performance and receipt of socially supportive behaviors (Coriell & Cohen, 1995). The communal orientation disposition also has been investigated as a factor in dating relationships. For example, Williamson and Silverman (2001) found that a communal orientation buffered the relationship between family or peer exposure to violence and perpetrating such violent acts. Results revealed that men high in communal orientation, as compared to men low in communal orientation, were less likely to abuse their partners and were less likely to have abusive friends although such friendships did occur. Another study of college students in dating or more serious relationships found that students who were high in communal orientation were most satisfied with intimate relationships when they were over-benefited but less satisfied when they were equitably treated, as compared to students who reported low communal orientation (Vanyperen & Buunk, 1991). 35 From this research investigating the effects of communal orientation it seems that people high in communal orientation are more willing to help, like helping more, enjoy relationships in which they feel that their needs are being met, are generous to give credit to others for success, and tend to have better quality relationships, as compared to individuals low in communal orientation. These individuals seem to subscribe to the communal norm as an approach to relationships, even with people they don?t know. Thus, it is likely that individuals high in communal orientation who begin dating relationships would be motivated to help and meet the needs of one?s partner. In the process of behaving in communal ways, dating partners also would be building trust because behaviors that meet the needs of one?s partner also have been described as behaviors that contribute to higher levels of trust. As such, it is anticipated that high levels of communal orientation will be associated with high levels of trust in the early stages of dating relationships. 36 METHOD Participants The initial intent of this study was to recruit dating partners who were in their first month of a dating relationship; however, of the 151 couples in this sample, only 10 couples fit this criterion. Women reported being in the relationship for 18.9 months on average ranging from .25 months to 156 months having a standard deviation of 20.7, whereas men reported a mean relationship duration of 18.3 months ranging from .25 months to 91.5 months, with a standard deviation of 17.2. Women averaged an age of 20 years ranging from 19 to 27; 28% were freshman, 23% were sophomores, 33% were juniors and 27% were seniors. Also, 72% reported that they were seriously dating, 21% were casually dating, 2.6% were engaged, 1.3% were living together, and 2% were married. Ninety percent of these women were Caucasian, 7% were African-American, 1% were Hispanic, 1% were Asian. Men were 21 years old, on average, with ages ranging from 19 to 28. Thirteen percent were Freshman, 31% were sophomores, 21% were juniors, 26% were seniors and 3.4% were not undergraduate students. Seventy percent reported that they were seriously dating, 27% reported casually dating, 3% reported being engaged, 1.3 were living together and 2% were married. 37 Procedures Participants for this study were recruited from undergraduate Human Development and Family Studies and Psychology classes. The researcher made several presentations in undergraduate classes for which the instructors would accept participation in this study as an extra credit assignment. Students were provided with an information letter describing the study and directed to look for postings for dates, times and classroom numbers when they could complete the questionnaire. Students also were invited to bring their partners to complete the questionnaire. Upon arriving at the data collection location, each participant received an envelope that contained the questionnaire. The questionnaire included the Trust Scale (Rempel, et al., 1985), the Communal Orientation Scale (Clark, et al., 1987), and the Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1960). Additional measures of relationship quality (Fletcher, et al., 2000a), self-esteem (Rosenberg, 1965) and love and liking (Rubin, 1970) were included as fillers between trust and communal orientation. These measures were ordered in the questionnaire so the participants reported levels of trust for their partner first, then own communal orientation, relationship quality, self-esteem, and social desirability. Participants then reported the perceived level of communal orientation of their current partner, and the communal orientation of their ideal partner with the Love Scale in between these reports. Following these measures, participants completed the Social Desirability Scale, then measures of trust for their ideal partner, then the level of trust that their current partner perceived in them with the liking scale between these two trust measures. In completing items about the ideal partner, participants were told to think about the characteristics that they preferred most in a 38 dating partner. These trust measures were followed by a number of questions pertaining to the nature of the relationship prior to the beginning of their dating relationship and when they began dating. Each envelope had a number code which was assigned to the couple. No names were attached to the number, therefore maintaining the anonymity of the respondents. If the partner of the class member also attended the data collection session, they were given the corresponding partner questionnaire and asked to complete the questionnaire in a different part of the classroom from his or her partner to ensure that neither partner knew how the other responded. When participants completed the questionnaires, they were returned to the envelope, and placed in a box at the front of the room. The participant then filled out an extra credit voucher with his or her name and the class for which extra credit would be received. The voucher was then returned to the researcher who used them to provide lists of participants to the appropriate course instructor so that extra credit could be awarded. If a class member?s partner was not able to attend the data collection session, the individual was asked if they thought his or her partner also would be interested in completing the questionnaire. If they decided that their partner would be willing to complete the questionnaire they were given the envelope marked with the number corresponding with that participant and asked to deliver the envelope to their partner. They also were directed not to be present when their partner filled out the questionnaire. Inside the envelope was a document describing how to return the envelop to the researcher either by campus mail or by returning it to the researcher?s office. This letter 39 also reminded the participant about the anonymity of the responses provided and directed the participant to complete the questionnaire independently of one?s partner. Measures Trust. The Trust Scale was developed by Rempel et al. (1985) and is composed of 17 items designed to measure levels of predictability, dependability and faith (see Appendix B). Items related to predictability (5) were designed to measure the consistency of a partner?s behavior. Dependability items (5) reflect attitudes about the character of one?s partner that allow one to be confident that one?s partner will respond in a caring, rather than a hurtful, way in a situation involving risk. Items designed to assess faith (7) focus on feelings of confidence that one?s partner will respond in caring ways when the outcome of the situation is uncertain. Responses to these items range from 1 (strongly disagree) to 5 (strongly agree). The score for each subscale is the sum of the responses to each item fitting that category. Negative items are reverse scored so that high scores reflect more trust in one?s partner. An examination of the reliability of the scale revealed an overall Cronbach?s alpha of .85 for women with subscale reliabilities of .68, .67, and .82 for predictability, dependability and faith, respectively. For men, the Cronbach?s alpha for the overall Trust scale was .88 with subscale alphas of .73, .76, and .84 for predictability, dependability and faith, respectively. The subscales also were moderately correlated with one another as coefficients ranged from r = .41 to r = .57 for women and r = .47 to r = .58 for men. The correlation between partner scores on the subscales ranged from r = .13 to r = .41. The correlation between women?s dependability and men?s subscale scores for dependability, predictability and faith were r = .38, r = .22, and r = .32, respectively (all 40 statistically significant at the p<.01 level). The correlations between women?s predictability and men?s dependability, predictability and faith were r = .30, r = .29, and r = .33, respectively (all statistically significant at the p<.01 level). The correlations between women?s faith and men?s dependability, predictability and faith were r = .32 (p < .01), r =.13, and r = .41 (p < .01), respectively. Evidence in support of the validity of the Trust Scale was provided by Rempel et al. (1985) through comparisons of scores on the Trust Scale with scores on a measure of motivation and the Loving and Liking Scale (Rubin, 1973). They found that love and faith had a stronger correlation than love and dependability while the correlation between love and predictability was very small. They also found that feelings of love were strongly correlated with intrinsic motivation as they expected but instrumental motivation also was strongly correlated with love, which was not expected. The strongest correlation between trust and motivation was between intrinsic motivation and faith. Intrinsic motivation and dependability also were related but the relationship between intrinsic motivation and predictability was very weak. All three factors of trust showed a negative relationship with extrinsic motivation. These results provide evidence that the Trust Scale shows good discriminant validity in measuring trust and lend support to the conceptualization of trust proposed by Rempel et al. (1985). Communal orientation. The Communal Orientation Scale (Clark, et al., 1987) was developed to measure individuals? communal orientation (see Appendix C). The 14 items were designed to assess one?s tendency to act in communal ways towards others and one?s expectation for other?s to direct communal behavior towards them. Responses range from 1 (extremely uncharacteristic) to 5 (extremely characteristic) with higher 41 scores indicating higher levels of communal orientation. Participants completed the measure reporting their own communal orientation as well as that of their current partner and ideal partner. For women, the Cronbach?s alpha reliability values were .74, .79 and .82 for self-perceived, perceptions of partner and ideal partner communal orientation, respectively. Cronbach?s reliability values for the men were .76, .80 and .80 for self- perceived, perceptions of partner and ideal partner communal orientation, respectively. Information about validity of the communal orientation scale provided by Clark et al. (1987) showed the communal orientation scale to be significantly correlated with similar constructs such as social responsibility and emotional empathy (Clark, et al., 1987). Clark et al. (1987) performed a principle components analysis (PCA) to explore the properties of the measure. In this analysis, three factors emerged, the first of which was a general communal factor on which all fourteen items loaded positively. The second factor that emerged was called ?desire for others? help? and was composed of four items (e.g. It bothers me when other people neglect my needs). The third factor was called ?locus of initiation? because the items that loaded well, positively or negatively, assessed attitudes about how others express emotion or need (e.g. People should keep their troubles to themselves). Four items composed this factor also. These last two factors seem to deal more with the communal orientation of a generalized other, not the specific other, such as a relationship partner, to which we refer in this study. As a result, we excluded the items that loaded on these two factors from further analyses and used the remaining six items, which refer to personal communal behavior. Specifically, three of these items address helping behavior (e.g. I don?t especially enjoy giving others aid) and three items address involvement (e.g. I often go 42 out of my way to help another person). This decision is consistent with our argument that trust and communal orientation are related because a person with a communal orientation is more likely to behave in trustworthy ways. We also performed exploratory factor analyses (EFA) with the intention of replicating the principle component analysis results of Clark et al. (1987) in this sample. In the EFA we found evidence for a general communal orientation factor and ?desire for others? help? factor. The results of the EFA were consistent, for the most part, with the PCA findings of Clark et al. (1987), resulting in six items when the items that concern the feelings of others were excluded. One change to the original six items was made as a result of the EFA, excluding item eight (I often go out of my way to help another person), and including item twelve (When people get emotionally upset, I tend to avoid them). This decision was made based on the results of the EFA. The overall Cronbach?s alpha coefficients for the remaining six items were, for women, .69, .80, and .76, for perceptions of one?s own communal orientation, perceptions of one?s partner?s communal orientation and perceptions of the communal orientation of one?s ideal partner, respectively. For men the alpha coefficients were .73, .76, and .74 for self-perceived communal orientation, perceptions of one?s partner?s communal orientation and perceptions of the communal orientation of one?s ideal partner, respectively. Structural equation modeling uses latent variables in the analyses, and latent variables must have at least two observed or indicator variables (Meyers, Gamst, & Guarino, 2006). Therefore, the six communal orientation items were divided into two three-item parcels. The first parcel included item 3 (?I?m not especially sensitive to other people?s feelings.?), item 4 (?I don?t consider myself to be a particularly helpful person.?) 43 and item 6 (?I don?t especially enjoy giving others aid.?). The second parcel included item 9 (?I believe it is best not to get involved in taking care of other people?s personal needs.?), item 10 (?I?m not the sort of person who often comes to the aid of others.?) and item 12 (?When people get emotionally upset, I tend to avoid them.?). The alpha coefficient for women?s self-perceived communal orientation for the first parcel was .61 and .59 for the second parcel. The alpha coefficient for the first parcel of perceptions of one?s partner was .72 and .77 for the second parcel, and for perceptions of one?s ideal partner the first parcel had an alpha coefficient of .70 and .60 for the second parcel. For men, the first and second parcel of self-perceived communal orientation yielded alpha coefficients of .70 and .60 respectively. The alpha coefficient for the first parcel of perceptions of one?s partner was .69 and .61 for the second parcel, while the first and second parcel alpha coefficients for men?s perceptions of their ideal partner were .72 and .69 respectively. 44 RESULTS Levels of trust The first goal of this study was to investigate the perceived level of trust as compared with other relationship quality factors as measured by the Perceived Relationship Quality Components questionnaire (Fletcher et al., 1999; see Appendix D). The means and standard deviations can be found in Table 1. For men, trust was rated more highly than the other variables, on average, whereas, for women commitment was rated more highly followed by love and then trust. A series of paired samples t tests were performed to ascertain whether or not these differences were significant. For men, the difference Table 1. Means and standard deviations for the subscales of the Perceived Relationship Quality Components Questionnaire (Fletcher et al., 2000a). Male Female subscale M SD M SD Satisfaction 12.85 2.36 12.84 2.15 Commitment 13.23 2.38 13.52 2.1 Trust 13.27 2.18 13.0 2.23 Intimacy 12.77 2.23 12.79 2.21 Passion 11.18 2.79 11.03 2.70 Love 13.24 2.47 13.43 2.35 45 between trust and satisfaction was significant (p < .01), as well as for intimacy (p < .01) and passion (p < .001). The differences between trust and commitment and love were not significant. For women, commitment was rated significantly higher than trust (p < .01), as was love (p < .05). Trust was significantly higher than passion (p < .001) while the differences between levels of trust and satisfaction and intimacy were not significant. Confirmatory Factor Analyses The means, standard deviations and zero order correlations among the eight variables are presented in Table 2. The hypothesized model (see Figure 1) was composed of the eight latent variables, as indicated by the ovals. Arrows indicate hypothesized direct effects. In the hypothesized model, the trust variables each had three measured variables, indicated by rectangles, which correspond to the three subscales of the Trust Scale. The communal orientation variables each had two measured variables. These two indicator variables were the helping and involvement factors that were identified on the basis of the factor analyses. Structural equation modeling (SEM) was used to first assess the measurement model components of the hypothesized model through a series of confirmatory factor analyses, and then in assessing both the entire measurement model and the structural model. SEM allows for the examination of relationships in the model simultaneously. The two-step model fitting analyses were performed using Amos 4.0 (Arbuckle & Wothke, 1999) with the maximum likelihood method of estimation. Amos 4.0 provides a number of model fit indices. The ? 2 is usually the first index that is consulted to assess model fit. It is an estimate of the discrepancy between the hypothesized model and the perfect model, or a simultaneous test of the extent to which all of the residuals are equal 46 to zero (Byrne, 1998). One limitation of the ? 2 statistic is that it is sensitive to sample size so that as sample size increases the likelihood of detecting a statistically significant Table 2. Zero order correlations, means, and standard deviations for the study variables. 1 2 3 4 5 6 7 8 1. Female self- perceived communal orientation - 2. Female perceptions of partner communal orientation .34** -- 3. Female perceptions of ideal partner communal orientation .37** .58** -- 4. Female trust in male partner .18* .27** .09 -- 5. Male self-perceived communal orientation .11 .32** .33** .14 -- 6. Male perceptions of partner communal orientation .17* .18* .21* .16 .31** -- 7. Male perceptions of ideal partner communal orientation .21* .27** .31** .07 .54** .60** -- 8. Male trust in female partner .04 .002 .10 .45** .31** .33** .33** -- M 25.1 22.8 24.8 68.0 22.7 23.7 23.4 66.8 SD 3.5 4.5 3.7 8.7 4.0 4.2 4.2 9.4 * p < .05. **p < .01. ***p < .001. 47 difference between the hypothesized and observed models also increases. It is recommended that the ? 2 be used as a goodness of fit index instead of a test statistic, so that a ? 2 value that is twice as large as the number of degrees of freedom indicates a poor fit and a ? 2 value of smaller than double the degrees of freedom indicates a good fit (Carmines & McIver, 1981). In addition to the ? 2 , other fit indices may be used including Table 3. Goodness-of-fit indicators for the confirmatory factor analyses. Variables df ? 2 CFI GFI RMSEA Female Self perceived communal orientation 8 13.7 .96 .97 .069 Female Perceptions of partner communal orientation 8 5.6 1.0 .99 .000 Female perceptions of ideal partner communal orientation 8 12.7 .98 .97 .063 Female trust in partner 116 152.3* .95 .90 .046 Male Self perceived communal orientation 8 8.6 1.0 .98 .022 Male Perceptions of partner communal orientation 8 22.6** .94 .96 .110 Male perceptions of ideal partner communal orientation 8 17.8* .96 .96 .090 Male trust in partner 116 174.8*** .93 .89 .058 Note. GFI = Goodness of Fit Index; CFI = Comparative Fit Index; RMSEA = Root Mean Square Error of Approximation. * p < .05. **p < .01. ***p < .001. 48 the Goodness of Fit Index (GFI), Comparative Fit Index (CFI) and the Root Mean Square Error of Approximation (RMSEA) because each one presents a different perspective on the quality of model fit. The GFI provides an estimate of the amount of variance in the sample correlation matrix accounted for by the hypothesized model (Guarino, Shannon & Ross, 2001). The CFI is a relative fit measure that compares the hypothesized model with models having perfect fit and no fit, and indicates the placement of the hypothesized model on the continuum between these two extremes (Byrne, 1998). Values between .90 and 1.0 indicate acceptable fit of the model to the data for the GFI and CFI. The RMSEA is the square root of the difference between the hypothesized model and the perfect model. Values of less than .08 indicate acceptable fit for the RMSEA (Tabachnick & Fidell, 2001). Each of these indices will be presented for the confirmatory factor analyses and the test of the full hypothesized model. The confirmatory factor analyses performed for the communal orientation variables applied a two factor structure with helping behavior and involvement as the two latent constructs, each having three indicators. The two latent variables were allowed to covary. This procedure was repeated for each communal orientation variable, for men and women. Each of these analyses resulted in excellent fit indices (see Table 3). A similar procedure was performed for the confirmatory factor analyses for the trust variables for men and women. The structure of the models used in these confirmatory factor analyses was based on the three subscales of the Trust Scale: predictability (five items), dependability (five items) and faith (seven items). The three factors were allowed to covary. Both models yielded fit indices suggesting a good fit between the model and the data. The fit indices for the trust variables and communal 49 orientation variables provide evidence that these measured variables reliably reflect their respective latent constructs (Bentler & Newcomb, 1986). Structural Model The test of the full hypothesized model, including the measurement and structural model revealed a significant ? 2 value of 215.13 (123, N =151, p <.001). This value taken together with the GFI (.87), the CFI (.90) and the RMSEA (.071) indicates a moderate fit between the model and the data. This result is not necessarily unexpected because the initial fit of a hypothesized model is often moderate, especially when several variables are in the model (Newcomb & Bentler, 1988). The moderate fit of this model is probably caused by associations between variables that were not included or considered to be a part of the hypothesized model (Newcomb & Bentler, 1988). Amos 4.0 provides a listing of modification indices which suggest changes that might be made to the model in order to improve the fit of the model to the data. The decision of whether or not to use these fit indices should be informed by theoretical reasoning (Sorbom, 1975), and the anticipated magnitude that the change would make in the fit of the model. Based on these criteria, the correlation between the residuals of male trust and female trust was added to the model. The addition of this correlation makes theoretical sense because it is likely that perceptions of trust in a relationship are affected by factors that are not included in the model. For example, it could be that relationship satisfaction influences perceptions of trust for both males and females. Also, based on the modification index (31.1), this change would have a greater effect on the fit of the model than any of the other changes (14.3 is the second highest modification index value). 50 The modification indices suggest that several other changes to the model could be made to improve model fit. The decision not to utilize the modification indices any further was made because this strategy is more consistent with exploratory data analysis rather than confirming the model (Meyers, Gamst, & Guarino, 2006). Also, the remaining modifications were for correlations between the error terms of indicators across latent variables, rather than correlations between error terms of the indicators of the same latent variable. The resulting ? 2 of this respecification was 176.8 (122, N = 151, p = .001) which is a statistically significant change from the original hypothesized model based on the critical chi-square test (Tabachnick & Fidell, 2001). Also, the ratio of the ? 2 to the degrees of freedom decreased to a value of 1.45. Other fit indices improved as well, with a GFI of .89, a CFI of .94 and an RMSEA of .055. These findings provide evidence that an improvement to the model was made, resulting in a good fit of the model to the data. Two problems were found with this model, however. First, there was a negative path coefficient between male self-perceived communal orientation and perceptions of partner communal orientation (? = -.53, p = .06). The zero order correlation, however, indicates a positive association between these two variables (r = .31, p < .01). According to Tabachnick and Fidell (2001), a negative beta coefficient and a positive zero order correlation is a sign of suppression. A second problem was that the path between male perceptions of ideal partner communal orientation and male perceptions of partner communal orientation had a standardized beta coefficient of 1.2 which could indicate colinearity. The two problems are probably related because the suppressor variable increases the ability of one variable to predict the other and is assigned a negative weight 51 in the process (Meyers, Gamst, & Guarino, 2006). It is likely that male perceptions of their communal orientation is acting as a suppressor variable by reducing its own association with male perceptions of partner communal orientation and enhancing the association between perceptions of the communal orientation of one?s ideal partner and male perceptions of partner communal orientation. In order to address this problem, the model was respecified a second time by removing the direct path from male self-perceived communal orientation to male perceptions of one?s partner?s communal orientation. This model resulted in a ? 2 of 183.2 (123, N = 151, p < .001) and a ? 2 to degrees of freedom ratio of 1.49. Other model fit indices were a GFI of .88, a CFI of .93 and an RMSEA of .057. Though this respecification resulted in a statistically significant change in the ? 2 value (?df = 1, ? ? 2 = 6.4, critical ? 2 = 3.84, p < .05), the other fit indices seemed to change very little. These results suggest a good fit of the hypothesized model to the data. Hypotheses Based on evidence that the structural and measurement models represent our data well, we can address the hypotheses that are reflected in the model (see Table 4). Hypothesis two, which was the first hypothesis addressed by the model, was that self- reported communal orientation would be associated with high levels of trust. The zero order correlation for women was r = .18 (p <.05) and r = .31 (p <.01) for men. For men, the path coefficient was .32 (p = .013) meaning that a 1 standard deviation change in male perceptions of their own communal orientation was associated with a change of .32 standard deviations in male perceptions of trust. For women, the path coefficient between communal orientation and perceptions of trust was .02 (ns) which indicates 52 Table 4. Path coefficients for the hypothesized model. Hypothesized Paths ? Measurement Model Female self-perceived help to self-perceived communal orientation .53*** Female self-perceived involvement to self-perceived communal orientation 1.05*** Female perceptions of partner help to perceptions of partner communal orientation .74*** Female perceptions of partner involvement to perceptions of partner communal orientation .88*** Female perceptions of ideal partner help to perceptions of ideal partner communal orientation .65*** Female perceptions of ideal partner involvement to perceptions of ideal partner communal orientation .75*** Female perceived dependability to trust in partner .71*** Female perceived predictability to trust in partner .65*** Female perceived faith to trust in partner .80*** Male self-perceived help to self-perceived communal orientation .69*** Male self-perceived involvement to self-perceived communal orientation .73*** Male perceptions of partner help to perceptions of partner communal orientation .73*** Male perceptions of partner involvement to perceptions of partner communal orientation .87*** ? p = .10. * p < .05. **p < .01. ***p < .001. 53 Table 4 continued. Hypothesized Path ? Male perceptions of ideal partner help to perceptions of ideal partner communal orientation .74*** Male perceptions of ideal partner involvement to perceptions of ideal partner communal orientation .80*** Male perceived dependability to trust in partner .75*** Male perceived predictability to trust in partner .64*** Male perceived faith to trust in partner .77*** Structural Model Hypothesis 1 Female Self-perceived communal orientation to female levels of trust .02 Male self-perceived communal orientation to male levels of trust .32** Hypothesis 2 Female perceptions of partner communal orientation to female level of trust .33*** Male perceptions of partner communal orientation to male level of trust .19 ? Hypothesis 3 Female self perceptions of communal orientation to perceptions of partner communal orientation .06 Male self-perceptions of communal orientation to female perceptions of partner communal orientation .17* Female self-perceived communal orientation to male perceptions of partner communal orientation .06 ? p = .10. * p < .05. **p < .01. ***p < .001. 54 Table 4. Continued Hypothesis 4 Hypothesized Path ? Female self-perceived communal orientation to perceptions of ideal partner communal orientation .49*** Female perceptions of ideal partner communal orientation to perceptions of partner communal orientation .72*** Male self-perceived communal orientation to perceptions of ideal partner communal orientation .71*** Male perceptions of ideal partner communal orientation to perceptions of partner communal orientation .75*** ? p = .10. * p < .05. **p < .01. ***p < .001. that self perceptions of communal orientation were not associated with perceptions of trust when controlling for the other variables in the model. The third hypothesis proposed that perceiving one?s partner as having a communal orientation would be related to perceptions of trust. Zero order correlations as presented in Table 1 show that there was a significant relationship between women?s perceptions of their partner?s communal orientation and perceptions of trust in their partners (r = .27, p <.01). The relationship between men?s perceptions of their partner?s communal orientation and trust also was significant (r = .36, p < .01). In the hypothesized model, men?s perceptions of their partner?s communal orientation was not significantly associated with their perceptions of trust (? = .19, ns) in the presence of other predictors. For women, the path coefficient for this relationship was .33 (p = .001) 55 indicating that perceptions of their partner?s communal orientation were associated with their perceptions of trust. Hypothesis four made two assertions. First, we suggested that self-perceived communal orientation would contribute to perceptions of one?s partner?s communal orientation. Also, we expected that the actual partner reports of communal orientation would contribute to one?s perceptions of his or her partner?s communal orientation. The zero order correlation shows a significant relationship between perceptions of one?s own communal orientation and perceptions of their partner?s communal orientation (r = .34, p < .01) for women. This relationship was not significant when controlling for other variables in the model (? = .06, ns). The zero order correlation reflecting the association between women?s perceptions of their partner?s communal orientation and the self- reported communal orientation of their partner showed a significant relationship (r = .32, p < .01), and a significant path coefficient (? = .17, p = .05). For men, the zero order correlation between perceptions of their own communal orientation and perceptions of their partner?s communal orientation revealed a significant relationship (r = .31, p < .01), but as discussed, this path was removed from the hypothesized model under the suspicion of suppression. The zero order correlation between men?s perception of their partner?s communal orientation and their partner?s self-reported communal orientation showed a significant relationship (r = .17, p < .05) but this path was not significant when controlling for other variables in the model (? = .06, ns). The final hypothesis was that perceptions of the communal orientation of one?s ideal partner would have an indirect effect on the relationship between perceptions of one?s own communal orientation and perceptions of the communal orientation of one?s 56 ideal partner communal orientation. Support for this hypothesis was found for both men and women. For men, the zero order correlation between self-perceived communal orientation and perceptions of ideal partner communal orientation was .57 (p <.01) and .60 (p <.01) for the relationship between ideal and partner communal orientation. The path coefficient between men?s perceptions of their own communal orientation and perceptions of their ideal partner?s communal orientation was .71 (p <.001). The path coefficient between perceptions of their ideal partner?s communal orientation and perceptions of their partner?s communal orientation was .75 (p < .001). The zero order correlation for women was .37 (p < .01) between perceptions of one?s own communal orientation and perceptions of the communal orientation of one?s ideal partner, and .58 (p < .01) between perceptions of the communal orientation of one?s ideal partner and perceptions of one?s partner?s communal orientation. The path coefficient between perceptions of their own communal orientation and perceptions of the communal orientation of their ideal partner was .50, (p <.001) and the path coefficient between perceptions of the communal orientation of their ideal partner and perceptions of their current partner?s communal orientation was .72 (p <.001). The indirect effect can be found by multiplying the two path coefficients together. For men the indirect effect was .53 and for women the indirect effect was .36. In summary, it seems that communal orientation functions somewhat differently for men than it does for women. Perceptions of one?s own communal orientation were more closely associated with perceptions of trust for men than women, whereas perceptions of one?s partner?s communal orientation were more related to trust for women than for men. Also, the relationship between perceptions of partner communal Figure 2: The associations and path coefficients included in the respecified model. 57 ? p = .10. * p < .05. **p < .01. ***p < .001. orientation and the self-perceptions of partners was not significant for men, but was significant for women. There was no support for a direct effect of perceptions of one?s own communal orientation on perceptions of the communal orientation of one?s partner because this path was removed for men and was non-significant for women. Finally, there was evidence supporting the indirect effect of perceptions of one?s own communal 1.05*** .74*** .71*** .06 .72*** .33*** .19? .53*** .65*** Ideal Ideal .75*** .49*** .06 .17* .02 .32** .65*** .80*** .73*** .69*** .71*** .75*** .74*** .80*** .73*** .87*** .64*** .75*** .77*** .88*** Male Level of Trust in Partner Predictability Dependability Faith Female Self C. O. Self- perceived help Self Involve -ment Female Ideal Partner C. O. Partner help Involve -ment Perceived Partner help Perceived Partner Involve- ment Female Perceived Partner C. O. Female Level of Trust in Partner Dependability Predictability Faith Male Self C. O. Self- perceived help Self Male Perceived Partner Male Ideal Partner C. O. Invole- ment Perceived Partner help C. O. Perceived Partner Ideal Partner help Ideal Involve- mentInvolve ment 58 orientation on perceptions of one?s partner?s communal orientation through beliefs about the communal orientation of one?s ideal partner. Taken together, the amount of variance captured by these associations can be ascertained by examining the squared multiple correlations or R 2 values. The R 2 for male trust was .20 and .12 for female trust, indicating that about 20% of the variance in trust for men and 12% of the variance in trust for women was accounted for by perceptions of one?s own communal orientation, perceptions of one?s partner?s communal orientation, partner?s perceptions of their own communal orientation and beliefs about the communal orientation of one?s ideal partner. Men?s perceptions of their own communal orientation and beliefs about the communal orientation of their ideal partner accounted for 57% of the variance in perceptions of their partner?s communal orientation and for women these variables accounted for 58% of the variance in perceptions of their partner?s communal orientation. Finally, the R 2 for men?s beliefs about the communal orientation of their ideal partner was .50 and this R 2 for women was .24, indicating that perceptions of one?s own communal orientation accounted for twice as much variance in beliefs about one?s ideal partner for men than for women. 59 DISCUSSION Based on existing research that revealed high levels of trust in the first month of dating relationships (Fletcher, et al., 2000b), the first aim of this study was to investigate levels of trust relative to other relationship quality contributors in the first month of dating. Unfortunately, this goal was not met because we were unable to recruit an adequate number of dating couples who were in the first month of dating (only 10 couples in our sample had been dating for one month or less). There are several possible reasons for our difficulty in recruiting these couples, such as the implications that making such a request of one?s partner might have. Perhaps students who have just begun dating were afraid that asking their partners to participate in this study would be viewed as a sign that the relationship was more substantial than it actually was. Also, students may not have been certain of their partner?s response and were unwilling to test the relationship by making this request. Students who had friends who were in the first weeks of a dating relationship also may have been uncomfortable inviting these friends to participate in the study because of this uncertainty. Another possible explanation for this shortcoming in our sample is that many of the participants were in relationships that had been initiated during high school. For example, 66% of the freshman women reported being in their relationships longer than five months, and had been enrolled in college for about four months. This figure reflects about 12% of the women in this sample. 60 In order to address these issues in the future, a number of strategies could be used to increase the likelihood of recruiting couples in recently established relationships. In the recruitment process a more concerted effort to recruit members of dating couples who are in the early stages of their relationship could be made by using language during the class visits that includes terms that reference the early stages of the relationship before students consider the relationship specifically dating. Also, recruiting students over the span of one semester or two rather than only during a limited amount of time such as one month might allow more individuals to participate as well. This strategy might involve making frequent visits to the classrooms, perhaps once every two or three weeks to keep this study fresh in their minds. Another idea would be to provide students with a card with contact information that could be turned in if they began a new relationship. An additional possibility would be to allow students to receive extra credit for their own relationship and another newly-formed relationship which may encourage higher recruitment when students in the recruited classes are in predominantly established relationships. Exploring the possibility of recruiting from other departments and from high schools might also help as well as using alternative incentives such an opportunity to win a gift card or some other more tangible compensation. Because we were not able to gather a sufficient amount of data from students in the beginning of their relationships, our findings regarding levels of trust compared with other relationship factors replicate previous work that indicates that trust, commitment, love, intimacy and passion are relationship components that arise in satisfying intimate relationships (Fehr, 1988; Rempel, et al., 1985; Weiselquist, et al., 1999; Larzelere & Huston, 1980; Fletcher, et al., 2000b). Our findings are consistent with this research and 61 revealed levels of trust that were the same as those for commitment and love, and higher than satisfaction intimacy and passion for men. Levels of trust for women were rated as lower than satisfaction and love, but higher than passion, and similar with satisfaction and intimacy. This study also contributes to our understanding of relationship dynamics in that it is only the second to reveal an empirical link between individual levels of communal orientation and variability in trust. Perhaps more importantly, it examines two potential mediators of this relationship. We anticipated that both self-perceptions of communal orientation and perceptions of one?s partner?s communal orientation would be related directly to trust. We also expected an indirect relationship whereby one?s self-perceived communal orientation would relate to trust through perceptions of partner. In addition, we anticipated an indirect relationship between self-perceived communal orientation and perceptions of partner communal orientation by way of perceptions of one?s ideal partner. We found that these variables, taken together, accounted for about 12% of the variance in trust for women and 20% of the variance for men, thus providing evidence that communal orientation plays a role in the development of trust in romantic relationships. Within our model, the first association we anticipated was between perceptions of one?s partner?s communal orientation and perceptions of trust in one?s partner. Support for this relationship was found for women, indicating that women who perceived their partners as being helpful, sensitive to others? feelings and willing to care for others? needs, were likely to report high levels of trust in their dating relationships. The notion that characterizing one?s partner as someone who is helpful, or communal is consistent with the process view of how trust develops (Holmes & Rempel, 1989). As women 62 perceive their partners meeting their needs in dependable and predictable ways, they trust their partners more. Because perceptions of partner appear to influence levels of trust, it is important to identify factors that contribute to women?s perceptions of their partner?s communal orientation. The first possibility explored in this study was that perceptions of partner would be a reflection of one?s own level of communal orientation (Ross, Greene, & House, 1977; McCall, Reno, Jalbert, & West, 2000). Our results, however, revealed no direct association between women?s self-perceived communal orientation and perceptions of their partner?s communal orientation. This finding is somewhat surprising because previous research has found that men and women perceive their partners as having characteristics similar to their own across a range of attitudes including caring feelings, equity and enjoyment of sex (Kenny & Acitelli, 2001), as well as the performance of positive and negative conflict behaviors (Acitelli, Douvan, & Veroff, 1993). This non- significant finding, however, needs to be considered in light of the fact that the zero-order correlation was significant (r = .31, p< .01), and that self-perceived communal orientation explained perceptions of one?s ideal partner which, in turn, predicted perceptions of partner. This pattern of associations suggests the presence of full mediation, as described by Baron and Kenny (1986). Perceptions of one?s ideal partner appear to be the mechanism through which self-perceptions of communal orientation influence trust in dating relationships. Women perceive themselves to be helpful and sensitive to the needs of others and value these traits in an ideal relationship partner. They then appear to use their characterization of the ideal partner as a lens through which they view their current partner, so their partner is perceived as having these ideal traits. 63 Having a view of one?s current relationship partner that is a function of perceptions of one?s ideal partner may enhance the relationship by filling in the gaps of perceptions about one?s partner. The perceptions that partially reflect one?s ideal partner may help women anticipate how their partner will behave in communal ways in future situations involving further interdependence and reliance on the positive response of their partner. Perceiving one?s partner in this fashion likely builds trust and confidence in the strength of the relationship. Our finding that perceptions of one?s ideal partner act as a mediator is consistent with the work of Murray et al. (1996) who found that perceptions of one?s ideal partner fully mediated the relationship between self-perceived communal orientation and perceptions of partner communal orientation for married couples, and partially mediated the relationship for dating couples. Another contributing factor to women?s perceptions of their partner?s communal orientation was the self-perceptions reported by their partners. This means that men?s level of communal orientation seems to be a shared reality in these relationships and women?s perceptions are based, presumably, on observations of instances in which their partners acted in communal ways. Past research also has used partner self-perceptions as an approximation of reality by which to compare partner perceptions (Murray, et al., 1996) and found both men?s and women?s perceptions of their partners also reflected partner self-perceptions. Thus, women?s perceptions of their partners appear to be based both in reality and in an idealized view of their partner. These two sources of information account for more than half of the variance in women?s perceptions of their partners (R 2 = .59). Though perceptions of ideal partner claimed the lion?s share of influence on women?s perceptions 64 of their partners (? = .72), compared with partner?s self-perceptions (? = .17), this combination has a sizable association with trust in the relationship. These findings indicate that communal orientation is an important individual difference variable when investigating women?s relationship trust. For men, the pattern of relationships that emerged was somewhat different. Similar to women, men?s perceptions of their ideal partner mediated the relationship between self-perceived communal orientation and perceptions of their partner?s communal orientation. However, there was evidence of a suppressor variable (Meyers, et al., 2006), as indicated by the large negative path coefficient between self-perceived communal orientation and perceptions of partner and the enhancement of the association between perceptions of ideal partner and perceptions of partner. It appears that inclusion of men?s self-perceived communal orientation as a predictor of perceptions of one?s partner suppresses the part of variance in perceptions of their ideal partner that is not relevant to perceptions of partner, thereby increasing the ability of perceptions of ideal to predict perceptions of partner (Tzelgove & Henik, 1991). In order to have more interpretable results (Wiggins, 1973), we removed the direct path between men?s self- perceived communal orientation and perceptions of their partner. This strategy clarified the associations and revealed that men who viewed themselves as helpful and sensitive to the needs of their partners, also perceived their ideal partner to have high levels of communal orientation. These ideal perceptions were then used to characterize their partner. However, men?s perceptions of their partner were only marginally related to perceptions of trust (p = .10). 65 We had anticipated that men?s perceptions of their partner would partially mediate the relationship between self-perceived communal orientation and trust; however, our findings indicate that men?s level of trust is significantly related to one?s own communal orientation. It is not clear, however, what mechanism might facilitate the association between men?s communal orientation and perceptions of trust in their dating relationships. One possibility is that men?s actual communal behavior may serve as a mediator. This proposition is consistent with the work of Zak et al. (1998) who found that relationship partners who were experimentally induced to perform a trusting behavior later rated their level of trust higher than participants who did not perform trusting behavior. For men then, levels of trust may be based more on their own actions than on their perceptions of their partner. Conclusion Our findings suggest that men?s communal orientation plays a role in the development of trust in dating relationships for both men and women. Although specific behaviors were not examined in this study, it is likely that communal orientation is a proxy for behavior because this orientation or propensity for being sensitive to others? needs has been associated with pro-relationship behaviors (Clark et al., 1987; Coriell & Cohen, 1995; Williamson and Silverman, 2001). It is speculated that a higher level of communal orientation for men increases the likelihood that women observe men performing communal behaviors in the relationship, leading to higher levels of trust. On the other hand, men?s performance of caring and need fulfilling behaviors not only reveal their positive intentions for the relationship to their partners, but also gives them an awareness of their own intentions for the relationship leading them to higher levels of 66 trust. Women?s communal orientation also is important for trust in relationships because perceptions of partner, which appear to influence women?s trust, are a partial reflection of women?s own communal orientation, as partially mediated by perceptions of the communal orientation of one?s ideal partner. Our study highlights the importance of further examining the role of communal orientation in intimate relationships. Specifically, understanding the extent to which self-reported communal orientation is an indicator of actual communal behaviors is of central importance. Our findings also contribute to past relationship research in that they are consistent with studies that indicate women attend to relationship issues and that concordance between women?s perceptions of their partner and partner self-perceptions is related to more stable relationships (Neff & Karney, 2005; Acitelli, Douvan, and Veroff, 1993). It is likely that women in stable relationships have gathered information about their partner throughout the span of their relationship, and then are able to compile that data into a profile which is corroborated by their partners? perceptions of themselves. These findings provide support for the process of trust development as proposed by Rempel et al. (1985; also Holmes & Rempel, 1989). Communal orientation appears to be similar to dependability in that both are personal dispositions attributed to the person based on the fulfillment of the needs of another person, or in this case, the needs of one?s partner. As one?s partner acts communally (i.e. according to one?s communal orientation) by being responsive to one?s needs, the partner will be perceived as predictable and dependable. Over the span of multiple occasions in which the communal orientation of one?s partner has been exercised, contributing to the dependability 67 attributed to that person, one has faith that one?s partner will continue to respond to one?s needs in unforeseeable future scenarios. Limitations and Future Directions Although this study provides insights about how communal orientation contributes to trust in close relationships, there are several limitations. These findings are based on cross-sectional data, which only allows us to reveal associations, not causality. Having reports of communal orientation and trust, in addition to other factors relevant to relationship development and maintenance, at different time points would allow us to assess how the levels of these constructs change over time. Also, longitudinal data would allow us to better examine the nature of the relationship between these two relationship factors. It may be that a person will perform communal behaviors and these behaviors will contribute to the level of trust in the relationship, or a certain level of trust in the relationship might be necessary for a person to be confident enough in the relationship to consistently perform communal behaviors. Another alternative is the relationship between communal orientation and trust is reciprocal whereby each informs the other. While longitudinal data would assess change or stability over time, participant reports of communal behaviors and how those behaviors relate to reports of communal orientation and trust would provide further insights into relationship dynamics. Whereas this study found that men and women differ in their accuracy in perceiving and summarizing the communal orientation of their partners, having daily reports of communal behavior to compare with summary measures would provide a richer picture of the relationship. Collecting the perspectives of family members or friends for the communal orientation of the relationship partners also would provide us with a way to 68 judge whether one?s communal orientation is the same across several close relationships, or operates in different ways in different relationships. Our findings also are limited by the use of a college student sample. Though many close relationships are formed during this time in students? lives, it is might be difficult to generalize our findings to other populations. For example, among newlywed couples the nature of the relationship between trust and communal orientation might change. Once committed in marriage, the pro-relationship interactions between relationship partners might change, which would also change the association between trust and communal orientation could grow weaker or stronger Though this research contributes to our understanding of the association between communal orientation and trust in relationships, the limitations of this study necessitate further research. Further investigations should examine the association between communal orientation and communal behaviors and the impact of these behaviors on trust as well as other relationship factors. Also, it would be helpful to investigate factors that contribute to the performance of communal behaviors so these behaviors might increase in frequency. These further insights might find utility in other close relationships such as friend, sibling, or parent-child relationships. 69 REFERENCES Acitelli, L. K., Douvan, E., & Veroff, J. (1993). Perceptions of conflict in the first years of marriage: How important are similarity and understanding? Journal of Social and Personal Relationships, 10, 5?19. Arbuckle, J. L., & Wothke, W. (1999). Amos 4.0 user?s guide. Chicago: SmallWaters. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173?1182. Bentler, P., & Newcomb, M. (1986). Personality, sexual behavior, and drug use revealed through latent variable methods. Clinical Psychology Review, 6, 363?385. Boon, S. D., & Holmes, J. G. (1991). The dynamics of interpersonal trust: Resolving uncertainty in the face of risk. In R. A. Hinde, & J. Groebel (Eds.), Cooperation and prosocial behavior (pp. 190?211). Cambridge: Cambridge University Press. Brickman, P., Becker, L. J., & Castle, S. (1979). Making trust easier and harder through two forms of sequential interaction. Journal of Personality and Social Psychology, 37, 515?521. Bryan, A. D., Hammer, J. C., & Fisher, J. D. (2000). Whose hands reach out to the homeless? Patterns of helping among high and low communally oriented individuals. Journal of Applied Social Psychology, 30, 887?905. 70 Butler, J. K. (1986). Reciprocity of dyadic trust in close male-female relationships. The Journal of Social Psychology, 126, 579?591. Buunk, B. P., Doosje, B. J., Jans, L. G. J. M., & Hopstaken, L. E. M. (1993). Perceived reciprocity, social support, and stress at work: The role of exchange and communal orientation. Journal of Personality and Social Psychology, 65, 801?811. Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS and SIMPLIS. Hillsdale, NJ: Lawrence Erlbaum. Carmines, E., & McIver, J. (1981). Analyzing models with unobserved variables: Analysis of covariance structures. In G. Bornestedt, & E. Gorgatta (Eds.), Social Measurement: Current Issues. (pp. 65?115). Beverly Hills, CA: Sage. Chun, K., & Campbell, J. B. (1974). Dimensionality of the Rotter interpersonal trust scale. Psychological Reports, 35, 1059?1070. Chun, K., & Campbell, J. B. (1975). Notes on the internal structure of Wrightsman?s measure of trustworthiness. Psychological Reports, 37, 323?330. Clark, M. S. (1981). Noncomparability of benefits given and received: A cue to the existence of friendship. Social Psychology Quarterly, 44, 375?381. Clark, M. S. (1984). Record keeping in two types of relationships. Journal of Personality and Social Psychology, 47, 549?557. Clark, M. S., & Mills, J. (1979). Interpersonal attraction in exchange and communal relationships. Journal of Personality and Social Psychology, 37, 12?24. 71 Clark, M. S., Mills, J. R., & Corcoran, D. M. (1989). Keeping track of needs and inputs of friends and strangers. Personality and Social Psychology Bulletin, 15, 533?542. Clark, M. S., Mills, J., & Powell, M. C. (1986). Keeping track of needs in communal and exchange relationships. Journal of Personality and Social Psychology, 51, 333? 338. Clark, M. S., Ouellette, R., Powell, M. C., & Milberg, S. (1987). Recipient?s mood, relationship type, and helping. Journal of Personality and Social Psychology,53, 94?103. Clark, M. S., & Waddell, B. (1985). Perceptions of exploitation in communal and exchange relationships. Journal of Social and Personal Relationships, 2, 403? 418. Coriell, M., & Cohen, S. (1995). Concordance in the face of a stressful event: When do members of a dyad agree that one person supported the other? Journal of Personality and Social Psychology, 69, 289?299. Crowne, D. P., & Marlowe, D. A. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24, 349?381. Deutsch, M. (1958). Trust and suspicion. The Journal of Conflict Resolution, 2, 265?279. Fehr, B. (1988). Prototype analysis of the concepts of love and commitment. Journal of Personality and Social Psychology, 55, 557?579. 72 Fletcher, G. J. O., Fincham, F. D., Cramer, L., & Heron, N. (1987). The role of attributions in the development of dating relationships. Journal of Personality and Social Psychology, 53, 481?489. Fletcher, G. J. O., Simpson, J. A., & Thomas, G. (2000a). The measurement of perceived relationship quality components: A confirmatory factor analysis approach. Personality and Social Psychology Bulletin, 26, 340?354. Fletcher, G. J. O., Simpson, J. A., & Thomas, G. (2000b). Ideals, perceptions, and evaluations in early relationship development. Journal of Personality and Social Psychology, 79, 933?940. Fletcher, G. J. O., Simpson, J. A., Thomas, G., & Giles, L. (1999). Ideals in intimate relationships. Journal of Personality and Social Psychology, 76, 72?89. Gallo, P. S., & McClintock, C. G. (1972). Cooperative and competitive behavior in mixed-motive games. In L.S. Wrightsman, J. O?Connor, & N.J. Baker (Eds.), Cooperation and competition: Readings on mixed?motive games (pp. 8?18). Belmont, CA: Brooks/Cole. Guarino, A. J., Shannon, D.M., & Ross, M. (2001). Making sense of fit indices in structural equation modeling (SEM). Journal of Research in Education, 11(1), 130-134. Holmes, J. G., & Rempel, J. K. (1989). Trust in close relationships. In C. Hendrick, (Ed.), Close relationships: Review of personality and social psychology (pp. 187?220). Thousand Oaks, CA: Sage Publications. 73 Johnson-George, C., & Swap, W. C. (1982). Measurement of specific interpersonal trust: Construction and validation of a scale to assess trust in a specific other. Journal of Personality and Social Psychology, 43, 1306?1317. Jones, D. C., & Costin, S. E. (1995). Friendship quality during preadolescence and adolescence: The contributions of relationship orientations, instrumentality, and expressivity. Merrill-Palmer Quarterly, 41, 517?535. Jones, D. C., & Vaughn, K. (1990). Close friendships among senior adults. Psychology and Aging, 5, 451?457. Kelley, H. H., & Stahelski, A. J. (1970). The inference of intentions from moves in the prisoner?s dilemma game. Journal of Experimental Social Psychology, 6, 401?419. Kenny, D. A., & Acitelli, L. K. (2001). Accuracy and bias in the perception of the partner in a close relationship. Journal of Personality and Social Psychology, 80, 439?448. Larzelere, R. E., & Huston, T. L. (1980). The dyadic trust scale: Toward understanding interpersonal trust in close relationships. Journal of Marriage and the Family, 42, 595?604. Lave, L. B. (1965). Factors affecting cooperation in the prisoner?s dilemma. Behavioral Science, 10, p. 26?38. Loomis, J. L. (1959). Communication, the development of trust and cooperative behavior. Human Relations, 12, 305?315. McCall, M. (1995). Orientation, outcome, and other?serving attributions. Basic and Applied Social Psychology, 17, 49?64. 74 McCall, M., Reno, R. R., Jalbert, N., & West, S. G. (2000). Communal orientation and attributions between self and other. Basic and Applied Social Psychology, 22, 301?308. Meyers, L. S., Gamst, G., & Guarino, A. J. (2006). Applied multivariate research: Design and interpretation. Thousand Oaks, CA: Sage Publications. Miller, P. J. E., & Rempel, J. K. (2004). Trust and partner-enhancing attributions in close relationships. Personality and Social Psychology Bulletin, 695?705. Mills, J., & Clark, M. S. (1982). Exchange and communal relationships. In L. Wheeler (Ed.) Review of personality and social psychology (pp. 121?144). Beverly Hills, CA: Sage Publications. Murray, S. L., & Holmes, J. G. (1997). A leap of faith? Positive illusions in romantic relationships. Personality and Social Psychology Bulletin, 23, 586?604. Murray, S. L., Holmes, J. G., & Griffin, D. W. (1996). The benefits of positive illusions: Idealization and the construction of satisfaction in close relationships. Journal of Personality and Social Psychology, 70(1), 79?98. Neff, L. A., & Karney, B. R. (2005). To know you is to love you: The implications of global adoration and specific accuracy for marital relationships. Journal of Personality and Social Psychology, 88, 480?497. Newcomb, M. & Bentler, P. (1988). Impact of adolescent drug use and social support on problems of young adults: A longitudinal study. Journal of Abnormal Psychology, 97, 64?75. Norris, S. L., & Zweigenhaft, R. L. (1999). Self-monitoring, trust, and commitment in romantic relationships. The Journal of Social Psychology, 139, 215?220. 75 Rempel, J. K., Holmes, J, G., & Zanna, M. P. (1985). Trust in close relationships. Journal of Personality and Social Psychology, 49, 95?112. Rempel, J. K., Ross, M., & Holmes, J. G. (2001). Trust and communicated attributions in close relationships. Journal of Personality and Social Psychology, 81, 57-64. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press. Ross, L., Greene, D, & House, P. (1977). The ?false consensus effect?: An egocentric bias in social perception and attribution processes. Journal of Experimental Social Psychology, 13, 279?301. Rotter, J. B. (1967). A new scale for the measurement of interpersonal trust. Journal of Personality, 35, 651?665. Rubin, Z. (1970). Measurement of romantic love. Journal of Personality and Social Psychology, 16, 265?273. Rubin, Z. (1973). Liking and loving: An invitation to social psychology. New York: Holt, Rinehart & Winston. Schlenker, B. R., Helm, B., & Tedeschi, J. T. (1973). The effects of personality and situational variables on behavioral trust. Journal of Personality and Social Psychology, 25, 419?427. Sorbom, D. (1975). Detection of correlated errors in longitudinal data. British Journal of Mathematical and Statistical Psychology, 28, 138?151. 76 Sorrentino, R. M., Holmes, J. G., Hanna, S. E., & Sharp. A. (1995). Uncertainty orientation and trust in close relationships: Individual differences in cognitive styles. Journal of Personality and Social Psychology, 68, 314?327. Sternberg, R. J., & Barnes, M. L. (1985). Real and ideal others in romantic relationships: Is four a crowd? Journal of Personality and Social Psychology, 49, 1586-1608. Swinth, R. L. (1967). The establishment of the trust relationship. Journal of Conflict Resolution, 11, 335?344. Tabachnick, B. G. and Fidel, L. S. (2001). Using multivariate statistics. Boston: Allyn and Bacon. Thompson, L., & DeHarpport, T. (1998). Relationships, goal incompatibility, and communal orientation in negotiations. Basic and Applied Social Psychology, 20, 33?44. Truchot, D., & Deregard, M. (2001). Perceived inequity, communal orientation and burnout: The role of helping models. Work & Stress, 15, 347?356. Truchot, D., Keirsebilck, L., & Meyer, S. (2000). Communal orientation may not buffer burnout. Psychological Reports, 86, 872?878. Tzelgov, J., & Henik, A. (1991). Suppression situations in psychological research: Definitions, implications, and applications. Psychological Bulletin, 109, 524?536. VanYperen, N. W. (1996). Communal orientation and the burnout syndrome among nurses: A replication and extension. Journal of Applied Social Psychology, 26, 338?354. 77 VanYperen, N. W., & Buunk, B. P. (1991). Equity theory and exchange and communal orientation from a cross?national perspective. The Journal of Social Psychology, 131, 5?20. VanYperen, N. W., Buunk, B. P., & Schaufeli, W. B. (1992). Communal orientation and the burnout syndrome among nurses. Journal of Applied Social Psychology, 22, 173?189. Weiselquist, J., Rusbult, C. E., Foster, C. A., & Agnew, C. R. (1999). Commitment, pro-relationship behavior, and trust in close relationships. Journal of Personality and Social Psychology, 77, 942-966. Wiggins, J. S. (1973). Personality and prediction: Principles of personality assessment. Reading, MA: Addison-Wesley. Williamson, G. M., & Clark, M. S. (1992). Impact of desired relationship type on affective reactions to choosing and being required to help. Personality and Social Psychology Bulletin, 18, 10?18. Williamson, G. M., Clark, M. S., Pegalis, L. J., & Behan, A. (1996). Affective consequences of refusing to help in communal and exchange relationships. Personality and Social Psychology Bulletin, 22, 34?47. Williamson, G. M., & Schulz, R. (1990). Relationship orientation, quality of prior relationship, and distress among caregivers of Alzheimer?s patients. Psychology and Aging, 5, 502?509. Williamson, G. M., & Silverman, J. G. (2001). Violence against female partners: Direct and interactive effects of family history, communal orientation, and peer?related variables. Journal of Social and Personal Relationships, 18, 535?549. 78 Wrightsman, L. S. (1964). Measurement of philosophies of human nature. Psychological Reports, 14, 743?751. Wrightsman, L.S. (1966). Personality and attitudinal correlates of trusting and trustworthy behaviors in a two-person game. Journal of Personality and Social Psychology,4, 328?332. Zak, A. M., Gold, J. A., Ryckman, R. M, & Lenney, E. (1998). Assessments of trust in intimate relationships and the self-perception process. The Journal of Social Psychology, 138, 217-228. 79 APPENDICES Appendix A Demographic items 80 Appendix B The Trust Scale (Rempel et al., 1985) 81 Appendix B continued The Trust Scale (Rempel et al., 1985) 82 Appendix C The Communal Orientation Scale (Clark et al., 1987) 83 Appendix C continued The Communal Orientation Scale (Clark et al., 1987) 84 Appendix C continued The Communal Orientation Scale (Clark et al., 1987) Note. Items marked by an asterisk (*) were used in our measure of communal orientation (Clark, et al., 1987). 85 Appendix D The Perceived Relationship Quality Components Questionnaire (Fletcher, et al., 2000a) 86 Appendix F Information Letter 87