Running head: ABC of social desires The ABC of social desires: Affiliation, being alone, and closeness to partner

We propose a triadic model of social desires directed at appetence/aversion of affiliation with friends (A), being alone (B), and closeness to one’s partner (C) that account for individual differences in subjectively experienced needs for proximity and distance in serious couple relationships. The model assumes that A, B, C can be conceptualized at the individual level as correlated latent factors measured by appetence and aversion indicators with opposite factor loadings and low shared method variance, and at the couple level assuming the same measurement model and identical (co)variances for men and women. The model was confirmed with confirmatory factor analyses in a sex-balanced internet sample of 476 individuals and a longitudinal sample of both partners of 578 heterosexual couples by assessing the ABC desires with brief appetence/aversion scales. In both samples, the desires showed expected unique associations with the Big Five personality traits, loneliness and relationship satisfaction, perceived available support by friends and partner, and attachment style toward the partner, and high 1-year stability in the longitudinal sample. We suggest that the ABC model helps to integrate research on couples’ distance regulation along the lines of communal and agentic motivation. 2 ABC of social desires The ABC of social desires: Affiliation, being alone, and closeness to partner The regulation of dyadic closeness and distance has been acknowledged by many researchers as a central issue in couple research (Feeney, 1999; Kantor & Lehr, 1975; Pistole, 2010). Finding the right balance between closeness and distance seems to be crucial for relationship functioning and stability and presents a major challenge to any couple (Baxter, 1990; Pistole, 1994). A considerable body of research has addressed patterns of distance regulation in specific situations such as attachment behavior in threatening situations (Mikulincer & Shaver, 2003, 2007) and communication styles during conflict (Christensen, 1987; Gottman, 1994). For instance, the demand-withdrawal pattern, where one partner wants more closeness than the other, has been identified as a maladaptive style of distance regulation that is linked to relationship distress and breakup (e.g., Christensen & Shenk, 1991). How do such maladaptive patterns come about? In line with other researchers, we assume that couples regulate their closeness according to the partners’ subjectively experienced needs, and if the two partners’ needs differ greatly, this can lead to quarrels and conflicts (Birtchnell, 1993; Christensen, Eldridge, Catta-Preta, Lim, & Santagata, 2006; Feeney, 1999; Pistole, 1994). But what are the needs relevant for dyadic distance regulation, and how do they relate to characteristics of the relationship and the partners’ personalities? A number of motivational concepts, such as intimacy and identity goals (Sanderson & Cantor, 1995), preferences for closeness over independence (Christensen et al., 2006), and personal versus relational concerns (Kumashiro, Rusbult, & Finkel, 2008) have been proposed to explain differences in couples’ interactions and distance regulation behaviors. However heterogeneous these approaches are, they all map onto the broad distinction of goal contents as agentic or communal. Originally introduced as the two fundamental modalities of human existence (Bakan, 3 ABC of social desires 1966), the terms agency and communion have been widely used to denominate two higherorder motivational dimensions that subsume the contents of human needs and goals (Brunstein, Schultheiss, & Grässmann, 1998; Prager & Buhrmester, 1998; Read et al., 2010). Agency includes needs that focus on the individual self and on forming separations whereas communion comprises needs that focus on the social aspects of the self and on forming relations (Helgeson, 1994; McAdams, Hoffman, Mansfield, & Day, 1996). In other words, agency refers to needs for autonomy, independence, and dominance whereas communion refers to social needs like intimacy, affiliation, and attachment. Many researchers agree that satisfaction in both motivational domains is necessary for optimal psychological functioning (e.g., Cantor & Malley, 1991; Prager & Buhrmester, 1998). However, different notions concerning the significance of couple relationships for the fulfillment of agentic and communal needs have been proposed. On the one hand, wellfunctioning relationships may support communal as well as agentic needs. For instance, Prager and Buhrmester (1998) reported that couples’ intimacy fostered need fulfillment in both domains. On the other hand, communal and agentic needs are viewed as competing and potentially conflicting motivations (Baxter, 1990; VanLear, 1998). Recently, Kumashiro et al. (2008) suggested that personal (i.e., agentic) and relational (i.e., communal) concerns cannot always be fulfilled simultaneously and compete within individuals for the allocation of time and resources (see also Goldsmith, 1990). Thus, fulfillment of agentic and communal needs is supposedly attained by different instrumental actions and often in different situations. Translating this view of competing motivations to distance regulation in couples, we contend that while communal needs directed at one’s partner can be best fulfilled in spatial proximity, agentic needs often imply distancing and pursuing individual interests by oneself. This point of view emphasizes the contrasting propensities to form relations versus 4 ABC of social desires separations, which are at the heart of Bakan’s (1966) original distinction of communion and agency and seem most relevant for couples’ distance regulation. Experienced desires for proximity or distance can thus be viewed as functions of the two partners’ communal and agentic needs. ABC Model of Social Desires The ABC model distinguishes three aspired goal states, which are differentially favored by individuals in couple relationships: closeness to the partner (C), affiliation with friends (A), and being alone (B). The model reflects a relationship-specific approach to the assessment of social motives and defines the person in the relationship as the unit of analysis. To indicate that the ABC dimensions refer to subjectively experienced motivations in a relationship context and thus to an intermediate level of specificity (i.e., more specific than global motive dispositions but less specific than situation-contingent motivational states), we use the term desire. The ABC model captures three directions distance regulating behavior can take. It therefore allows for more differentiated analyses of the motivational underpinnings of dyadic distance regulation than existing unior two-dimensional models of agentic and communal motivation. For instance, unidimensional models (e.g., Christensen et al., 2006) confound low communal motivation with high agentic motivation in explaining dyadic distance. In contrast, the tripartite ABC model assumes that distancing from the partner is not only and not always due to low closeness motivation, but can also result from other desires. Moreover, in contrast to two-dimensional models (e.g., Sanderson & Cantor, 1995), these other desires can be either communal, that is, seeking contact with other people like friends, or agentic, that is, engaging in activities by oneself. The following sections outline our conceptions of the three ABC desires as well as a theory-based approach to their assessment. Desire for Closeness 5 ABC of social desires The desire for closeness (C) reflects the propensity to seek spatial proximity or psychological closeness to one’s relationship partner. Closeness is thus a more broadly defined goal state than intimacy, which focuses on positive interactions, self-disclosure, and sharing of emotions (McAdams, 1992; Prager, 1995; Reis & Shaver, 1988; Sanderson & Cantor, 2001), the security afforded by a reliable attachment figure in threatening situations (Mikulincer & Shaver, 2007), or sexual desire (Impett, Strachman, Finkel, & Gable, 2008). In terms of the ABC model, intimacy, attachment, and sexuality reflect more specific motivations that may underlie subjectively experienced desires for closeness and thus probably show some empirical overlap. Indeed, previous studies have shown that adult attachment styles are substantially related to proximity and distance motivation both in the realm of couple relationships and relationships in general (e.g., Dewitte & De Houwer, 2008; Feeney, 1999). Thus, individual differences in attachment to one’s partner seem to be important determinants of desired closeness. However, other variables pertaining to the partners’ personalities or characteristics of their relationship may also contribute to the explanation of closeness motivation. For instance, high agreeableness and, in particular, high general satisfaction with the partnership are likely to affect the desire for closeness, too. Empirical research on the relative influences of individual and relationship characteristics on desired closeness is however missing. Desire for Affiliation Communal desires do not necessarily aim at closeness with one’s partner. In contrast, we contend that communal motivation is relationship-specific. Different relationship partners (e.g., partner, friends, relatives) are associated with differentially strong desires and are thus not fully interchangeable with each other. This assumption of relationship-specificity was supported by studies that reported substantial within-person variance in attachment security across different relationship types (Asendorpf & Wilpers, 2000; Cook, 2000; La Guardia, 6 ABC of social desires Ryan, Couchman, & Deci, 2000). To investigate commonalities and differences between communal desires directed at different relationship partners, we included the desire for affiliation with friends (A) in the ABC model. Of particular interest is the question, to what extent the different facets of communal motivation converge or compete within individuals. Friendships seem most suitable for this purpose because they share a lot of common features with couple relationships: Friendships are frequently rated as the second closest relationships next to couple relationships (e.g., Berscheid, Snyder, & Omoto, 1989) and provide social support and intimacy, two important resources for well-being also afforded by couple relationships. Desire for Being Alone Being alone is a double-edged sword. On the one hand, spending much time by oneself can reflect a state of social deprivation, which is accompanied by a lack of social support and feelings of loneliness and isolation. Traditionally, research has mainly focused on this negative side of solitude and its consequences for well-being and health (e.g., Ernst & Cacioppo, 1999). For instance, depression (Beck, 1967), bulimic eating disorder (Larson & Johnson, 1985), and even a generally higher risk for mortality (Baumeister & Leary, 1995; House, Landis, & Umberson, 1988) have been related to social isolation. On the other hand, being alone can be a valued and deliberately aspired state that may serve different psychological functions and have beneficial effects on the individual. People seek solitude for recreational purposes as well as to pursue agentic activities that can be better carried out in privacy, such as creative work, self-reflection, making future plans, or seeking spiritual experiences (Burger, 1995; Long, Seburn, Averill, & More, 2003). Long et al. (2003) were able to differentiate such positive episodes of solitude from negative experiences of loneliness and showed that individuals differ in the motivation to strive for positive experiences of being alone. 7 ABC of social desires In the ABC model, being alone is not merely viewed as the opposite of communal desires (see also Lavee & Ben-Ari, 2007). It rather reflects an agentic goal state in its own right giving partners the opportunity to engage in individual activities and/or recover from the confinements and liabilities of their relationships. The desire for being alone (B) is therefore conceived as reflecting agency needs. This proposition was recently supported in an experimental study. Participants who had visualized an autobiographical episode, in which they had had an intense desire for solitude and distance from their romantic partners, employed more themes and expressions related to independence and power, but not fewer themes related to closeness and intimacy than controls in a subsequent TAT-like motive test (Hagemeyer & Neyer, in press). Thus, the experimentally induced desire for being alone yielded an increase in typical agentic imagery, but no decrease in communal imagery. These findings have two implications relevant for the current study. First, the assumed agentic nature of the desire for being alone was supported. Second, the assumed independence of agentic and communal motivations was corroborated, albeit at the level of motivational states rather than stable dispositions. Appetence and Aversion as Indicators of Social Desires In line with dynamic theories of motivation (e.g., the Zurich model of social motivation; Bischof, 1993; Schneider, 2001; see also Atkinson & Birch, 1970), we conceive of the ABC desires as relatively stable set-points or reference values of feedback control systems. These motivational set-points determine how much of a specific class of experiences an individual typically needs to be satisfied (i.e., how much closeness to the partner, how much affiliation with friends, and how much time spent alone). According to feedback control theory, stable set-points are continuously compared with the actual values, and discrepancies between the two will instigate a reaction to restore concurrence, that is, a state of equilibrium (Powers, 1973; Wiener, 1948). Discrepancies between set-points and actual 8 ABC of social desires values can derive from too little or too much of a specific class of experiences. The two kinds of discrepancies determine two different kinds of affective-motivational reactions. If the actual value falls below the set-point, for instance actual closeness to one’s partner is lower than desired, an appetitive reaction will result: Closeness will be experienced as a positively valued approach goal, and the individual will strive to attain closeness thereby restoring equilibrium. If, on the other hand, the actual state exceeds the set-point, for instance actual closeness exceeds desired closeness, the resulting reaction will be aversive: The surfeit of closeness will be affectively experienced as negative and motivate the individual to avoid closeness. Thus, the same goal content can be experienced as appetitive or aversive depending on the individual’s current state of need (Bischof, 1993; Cabanac, 1971). These characteristics of the actual genesis of motivational states, as derived from feedback control theory, can be utilized to assess individual differences in the ABC desires. Motive dispositions in general are defined as recurrent concerns for the attainment of specific incentives or the avoidance of specific disincentives (McClelland, 1985; Schultheiss, 2008). This definition implies that motive dispositions like the ABC desires can be inferred from the frequency of specific motivational states. According to feedback control theory, individuals with a strong desire for a specific goal content (i.e., a high set-point) are more likely to experience appetence and less likely to experience aversion towards this goal content than individuals with a low desire (i.e., a low set-point). Thus, frequent experiences of appetence and infrequent experiences of aversion indicate a strong desire, whereas frequent experiences of aversion and infrequent experiences of appetence indicate a weak desire. Consequently, we assessed individual differences in the ABC desires by the frequencies of the opposing states of appetence and aversion towards the three goal contents, thereby encircling the motivational set-points. The Present Research 9 ABC of social desires The present investigation introduces brief scales for the assessment of the ABC desires that were built on the aforementioned considerations. In the following, we will outline a measurement model and expectations regarding the relations between the three desires and external criteria. We will then test the factorial validity as well as the convergent and discriminant validities of the ABC desires in an individual and a dyadic sample and, finally, discuss the results with regard to implications, limitations, and prospects for future research. Internal Structure The first aim of the present study was to establish the factorial validity of the threedimensional ABC model. Figure 1 presents the measurement model for the ABC desires (a variant of a confirmatory factor analysis). The three desires are considered latent correlated factors with opposite factor loadings of their two manifest indicators (appetence and aversion subscales). The residuals of the three appetence subscales and the residuals of the three aversion subscales are allowed to covary in order to capture possible shared method variance. We had the following expectations. First, appetence and aversion subscales are considered indicators of specific motivational set-points. Shared method variance due to common affectivity across desires should thus be low. Second, the desires for closeness C and affiliation A are expected to form distinguishable factors. Because these two desires are viewed as facets of a higher-order need for communion, the two factors may be positively related. Because of the assumed relationship-specificity, this correlation should be modestly sized. Third, the desire for being alone B is expected to form a third latent factor. Because the communal desires and the agentic desire cannot be satisfied simultaneously and thus compete for the allocation of resources, high scores on both bear the potential for motivational conflicts within individuals. Thus, individuals will probably tend to favour one over the other, which may result in moderate negative correlations between the agentic factor B and the communal factors A and C. 10 ABC of social desires Convergent and Discriminant Validity of ABC Desires The second aim of this study was the identification of variables that relate to and potentially influence individual differences in the ABC desires differentially, thereby establishing their convergent and discriminant validities. We related the ABC desires to four classes of variables: (a) personality traits, (b) general and relationship-specific satisfaction, (c) perceived available support from partner, friends, and parents, and (d) attachment to one’s partner. Previous research has mainly focused on adult attachment styles in explaining the sources of individual differences in closeness and distance motivation (Dewitte & De Houwer, 2008; Feeney, 1999; Pistole, 1994). The present analyses were intended to establish a more comprehensive and differentiated nomological network of social motivation in couples. We expected the ABC desires to be distinguishable with respect to two attributes: (a) agentic versus communal goal contents and (b) high versus low relationship specificity within the communal domain. More specifically, we had the following expectations. The desires for affiliation with friends A and closeness to the partner C are expected to relate differentially to relationship-general versus partnership-specific variables. A, as the more general communal desire, should show stronger associations with communal personality traits (high sociability, high agreeableness), satisfaction with one’s life and relationships in general, and perceived social support by friends than with partnershipspecific criteria. Previous studies have shown that perceived dyadic closeness is a crucial feature of relationship quality and closely linked to relationship satisfaction and attachment security (Aron, Aron, & Smollan, 1992; Ben-Ari & Lavee, 2007; Berscheid et al., 1989; Feeney, 1999). We assume a reciprocal relation between closeness motivation and relationship quality: A strong desire for closeness will promote relationship satisfaction which in turn enhances closeness motivation. Thus, the partnership-specific C should be associated primarily with variables indicating high partnership quality, particularly 11 ABC of social desires relationship satisfaction, perceived support by the partner rather than others, and security of romantic attachment. C may also relate to communal traits, but, because of the assumed relationship-specific influences, to a much lesser extent than A. The agentic desire for being alone B should be clearly distinguishable from the two communal desires. Drawing on previous research that addressed related constructs (e.g., preference for solitude, rated importance of spending time alone, need for distance to partner), we had the following expectations. First, B should show small to moderate associations with agentic personality traits, particularly high openness to experience (Long et al., 2003; Nestler, Back, & Egloff, 2011). Second, previous research suggests negative relations to perceived relationship quality, particularly relationship satisfaction and support by the partner. These negative relations should however be markedly lower than the positive relations expected for C (Craddock, 1997). Third, both theoretical considerations (Pistole, 1994) and empirical findings (Dewitte & De Houwer, 2008; Feeney, 1999) suggest a strong association between B and avoidant attachment toward the partner. Study 1: Individuals For a first cross-sectional test of our expectations, Study 1 was designed as an internet study in order to achieve a higher age variance and proportion of serious relationships than in dating studies with undergraduates. Method Participants. Participants were invited to a study on serious relationships announced by a German press release in newspapers, magazines, radio stations, online services, and online forums. Immediate feedback on participants’ personality and relationship quality was offered as an incentive. Of the 1,639 individuals completing the whole questionnaire, 1,160 were selected who were at least 18 years old, were currently in a serious sexual relationship of at least 1 year duration (excluding long-distance relationships), and had non-missing data 12 ABC of social desires on the three social desires and relationship satisfaction. Because this sample was highly unbalanced regarding sex (238 males, 922 females), we randomly selected females such that a completely balanced sample resulted concerning sex (238 males, 238 females), coresidence with the partner (yes – no) within sex, and marital status within sex. Thereby we made sure that the sample composition concerning sex, living arrangement, and marital status resembled a sample where one partner was randomly picked from a sample of couples. The final sample consisted of 476 participants (96.4% native German speakers) aged 18 to 71 years (M = 34.6, SD = 10.7) with a relationship duration of 1 to 40 years (M = 6.06, SD = 6.75), 0 to 4 own children (M = 0.66, SD = 1.02); 64% lived with the partner in the same household, 26% were married. Measures. Desire for affiliation, desire for being alone and desire for closeness were each assessed with an 8-item scale (see Appendix). The items were systematically varied regarding appetitive and aversive motivation towards the three goal states. The items were answered on a 7-point frequency scale ranging from 1 to 7 (1 = never, 4 = sometimes, 7 = always). Wording of the items was as parallel as possible across the three scales. The Big Five personality traits were assessed by the German 15-item version (3 items for each trait) of the Big Five Inventory (BFI; John et al., 2008) that is used in the German Socio-economic Panel Study (GSOEP) (Gerlitz & Schupp, 2005). Each item was answered on a 7-point Likert-type scale ranging from 1 to 7 (1 = applies to me not at all; 7 = applies to me completely). The internal consistencies α of the five scales were moderate to satisfactory (Openness .66, Conscientiousness .65, Extraversion, .81, Agreeableness, .53, Neuroticism, . 71). They were similar to those in the nationally representative GSOEP (Dyrenforth, Kashy, Donnellan, & Lucas, 2010), and acceptable for brief scales that stress bandwidth rather than fidelity. Intercorrelations of the scales were low (rs from -.15 to .30). In addition, Sociability was assessed with 5 items adopted from Asendorpf and 13 ABC of social desires Wilpers (1998) that were mixed with the 15 BFI items and had the same response format as the BFI items because this subfactor of extraversion seemed conceptually more closely related to desire for affiliation than the broader factor of extraversion. The internal consistency of this 5-item scale was acceptable (α = .72). Satisfaction with one’s life, oneself, one’s social relationships in general, and one’s partnership was assessed by four scales. Life satisfaction was assessed with the single life satisfaction item of the GSOEP “How satisfied are you with your life overall?” which is rated on an 11-point scale ranging from 0 – 10 (0 = not at all satisfied; 10 = completely satisfied). Global self-worth was assessed with a German 6-item version of the global self-esteem scale of the SDQ III by Marsh and O’Neill (1984) (for example, “Overall, I have a lot of respect for myself“; see also Asendorpf & van Aken, 2003) and Loneliness with a 5-item version of the emotional loneliness scale of the German adaptation of the UCLA loneliness scale by Döring and Bortz (1993) (for example, “I feel lonely”). Relationship satisfaction was assessed with a German 7-item version of the Relationship Assessment Scale (RAS; Hendrick, 1988; for example, “How satisfied are you with your relationship in general”). The items of the latter three scales were answered on a 5-point Likert-type scale ranging from 1 – 5 (1 = applies to me not at all; 5 = applies to me completely). The internal consistencies α of these scales were satisfactory (α > .80). Perceived available support from friends, partner, mother, and father was each assessed with an item adapted from Asendorpf and Wilpers (1998) “If I have problems, I would turn to this person to talk about my problems”. It was rated for each type of potentially supporting person on a 5-point frequency scale ranging from 1 – 5 (1 = never, 5 = always). The two parental support measures correlated .41 and were averaged in order to obtain a single index of parental support. Support from partner could thus be contrasted with support from non-kin and kin relationships at comparable levels of specificity. 14 ABC of social desires Attachment to partner was assessed with a German partnership-specific version of the Relationship Questionnaire (RQ; Bartholomew & Horowitz, 1991). The attachment styles secure, fearful, preoccupied and dismissing were described by short paragraphs in a partnership-specific format (replacing “others” in the original version by “my partner”). For example, the dismissing style was described as “I am comfortable without close emotional relationships. It is very important to me to feel independent and self-sufficient, and I prefer not to depend on my partner or have my partner depend on me.” As recommended by Bartholomew, each style was first answered in a forced-choice format to minimize order effects, and subsequently rated on a 7-point Likert scale ranging from 1 – 7 (1= applies to me not at all; 7 = applies to me completely). Results Descriptive statistics. Table 1 displays descriptive statistics of the main variables. The ABC scales showed only weak intercorrelations except for moderately negative correlations between being alone and closeness to partner, both at the level of scales and of each subscale. The appetence and aversion subscales showed relatively high negative correlations such that the internal consistency of each ABC scale was at least as high as the internal consistencies of its two subscales. Sex and age differences. Sex differences in the means of the 9 variables of Table 1 were analyzed with t tests using stepwise Bonferroni correction to prevent alpha inflation. Only one significant sex difference was found; females reported higher affiliation appetence than males although the effect size was small (t(474) = 3.03, p < .003, Cohen’s d = 0.28). Correlations of the 9 variables with age were also small (ranging from -.17 to .14). For the full desire scales, correlations were -.10, p < .05 (affiliation), .13, p < .01 (being alone), -.16, p < .001 (closeness). Confirmatory factor analysis of the subscales. A test of the ABC model presented 15 ABC of social desires in Figure 1 yielded an acceptable fit, χ2(3) = 6.56, p = .09, CFI = .997, RMSEA = .050. Covariances between desire for affiliation with desire for being alone (c1 in Figure 1) and between the residuals of the three appetence scales (a12, a13, a23 in Figure 1) were nonsignificant. Therefore we set these covariances to zero. The resulting more parsimonious model did not show a significantly poorer fit than the baseline model, χ2(4) = 8.77, p = .07, and still showed an acceptable fit, χ2(7) = 15.33, p = .03, CFI = .994, RMSEA = .050. This final model is presented in Figure 2. The correlations between the latent variables were consistent with the correlations between the ABC scales presented in Table 1. The standardized factor loadings were high for each latent factor. The three aversion subscales showed positively correlated residuals that confirmed our assumption of low shared method variance. The appetence scales did not show shared method variance. All in all, the final model fitted the data and confirmed our expectations. Prediction of ABC desires. We predicted the ABC desires by four blocks of predictor variables, both in terms of unique contributions within a block (multiple regressions on all variables of a block) and in terms of zero-order correlations in order to detect possible suppressor effects. The results are presented in Table 2. In a first block we examined the Big Five scales, and separately the sociability scale because we assumed that sociability might be more strongly related to desire for affiliation than the broader trait of extraversion. This was actually true (correlations of .49 versus .36) as confirmed by Steiger’s (1980) test for differences between dependent correlations (t(473) = 3.54, p < .001). Because suppressor effects were not strong, we describe here only the results of the multiple regressions. Desire for affiliation A was most strongly uniquely related to extraversion whereas desire for closeness to one’s partner C was related to agreeableness. As expected, the more relationshipgeneral desire A showed stronger relations with communal traits than the partnership-specific 16 ABC of social desires C. Desire for being alone B was most strongly uniquely related to openness, as we expected, although the effect size was not large. In addition, all desires showed low negative unique associations with neuroticism, and conscientiousness was unrelated to all desires. Regarding the satisfaction measures, A was most strongly uniquely associated with loneliness whereas C was uniquely associated with relationship satisfaction but not at all with loneliness, confirming the relationship specificity of the communal desires. Consistent with its negative correlation with C, B was negatively associated with relationship satisfaction, but only weakly. Also supporting our assumptions about relationship specificity, the global satisfaction measures life satisfaction and global self-worth did not make significant unique contributions to the prediction of the ABC desires, except for a small association between A and life satisfaction. Similarly, A was uniquely associated only with perceived available support from friends, whereas C and B were uniquely associated only with support from the partner, and all three desires were unrelated to support from the parents. Finally, A showed only small associations with attachment to the partner, whereas B was substantially associated with dismissing versus preoccupied attachment and C with secure versus fearful and dismissing attachment. Thus, B predominantly relates to avoidant attachment (Brennan, Clark, & Shaver, 1998), whereas C relates to the general quality of romantic attachment. Whereas this section has focused on unique contributions of variables in a block of predictors, the unique contributions of the blocks themselves require an analysis of hierarchical regression. For efficient presentation, this analysis is presented together with a similar analysis of the data of Study 2 in the results section of Study 2. Discussion Study 1 confirmed all our expectations, with minor exceptions. The ABC desires could be validly and reliably assessed by the frequencies of appetitive and aversive 17 ABC of social desires experiences. The proposed measurement model fitted the data well, and the amount of shared method variance between subscales was low for aversion and non-significant for appetence. The latent variables of the two communal desires showed a small positive correlation, and the agentic desire for being alone B showed a modest negative correlation with desire for closeness to one’s partner C. However, counter to our predictions, B was unrelated to affiliation with friends A. The desire for being alone thus seems to be better reconcilable with the desire for affiliation with friends than with the desire for closeness to one’s partner. This may be due to the fact that closeness to one’s partner is usually more time-consuming and affords a higher investment than affiliation with friends. The competition for resources is thus stronger between C and B than between A and B. Although unexpected, this finding further supports the relationship-specificity of communal desires. The differentiation of the ABC desires by their relations to external criteria was also successful. Communal and agentic desires were differentially related to corresponding communal and agentic personality traits, although the relation between B and openness for experience was somewhat smaller than expected. In addition, C was most strongly associated with partnership satisfaction, perceived available support from partner, and secure attachment to partner, whereas B showed markedly smaller negative relations to variables of partnership quality and was best predicted by dismissing attachment. Finally, we found support for the relationship specificity of the two communal desires. A was best predicted by relationshipgeneral variables of personality, satisfaction, and available support and showed negligible relations to partnership-specific variables. For C, on the other hand, the reverse pattern was observed. All in all, Study 1 rendered good support for the convergent and discriminant validities of the ABC desires. Implications of these results will be outlined in the general discussion section since they were largely replicated in Study 2. Study 2: Couples 18 ABC of social desires Major limitations of Study 1 were (a) the cross-sectional approach such that the stability of the ABC scales could not be studied, (b) assessments of individuals rather than both partners of a couple such that the between-partner consistency of the ABC scales could not be determined, and (c) the sampling through the internet which resulted in a sample biased toward younger individuals and those regularly using the internet. In order to replicate and complement the validation results of Study 1, we designed a second study that accounted for these limitations by (a) being longitudinal thereby allowing for stability analyses, (b) including both partners of a couple to afford analyses of between-partner associations, and (c) using a different sampling procedure based on households and telephone prescreening to assure a less biased sample in terms of age and internet usage. Method Participants. A sample of 714 German heterosexual couples was recruited by a mass mailing to all households in a non-traditional metropolitan area of Berlin and in a rural catholic area of Germany (this sampling procedure was chosen to study effects of modernization on the couple relationships, a question outside the scope of the present study). A letter from a well-known local university invited couples to participate in a study on relationships of serious couples in different living arrangements excluding long-distance relationships. A brief telephone pre-screening and under-sampling of younger couples made sure that the couples really met the requirements, spoke German fluently, and consisted of similar numbers of couples from each age decade between 18 and 68 years (as defined by women’s age). Participants could opt for a paper-and-pencil version or an online questionnaire and were asked not to involve the partner in answering the questions. In the online version, each participant received by email an individualized link to the study such that no one else could access the data during data acquisition; after completion of the questionnaire, this link became immediately dysfunctional. Couples were offered an 19 ABC of social desires honorarium of 20 € and an individual, confidential feedback for each partner about his or her own ratings of personality and relationship quality. In the present analysis we included only couples with at least 1 year of relationship duration and non-missing data of both partners for all main variables at the first assessment, resulting in a final sample of 578 couples, with women’s age ranging from 18 to 68 years (M = 40.6, SD = 13.3), and men’s age ranging from 19 to 73 years (M = 43.0, SD = 13.8). Relationship duration correlated .99 between men and women and was thus averaged, ranging from 1 to 53 years (M = 12.44, SD = 12.39). Participants reported 0 to 4 children (M = 1.14, SD = 1.20), 67% of the couples lived in the same household, 50% were married. As expected, men were older than women (t(577) = 11.21, p < .001, d = 0.93); no significant sex differences were found for the number of children (t < 1). Exactly one year after participation, each participant was contacted again by email or mail and was asked to participate in a follow-up study. For their participation, we offered participants to take part in lotteries, with prices amounting to a total of 3,700 €. Response rate was 68% for couples (at least 1 partner of each couple responded), 57% for women and 49% for men, a significant sex difference according to a McNemar test at the couple level (χ2(1) = 13.23, p < .001). All participants whose t1-relationships were still intact and who provided valid information about their ABC desires were included in the analyses on temporal stability of the ABC desires reported below. Measures. At Time 1, desire for affiliation, desire for being alone, desire for closeness, personality, life satisfaction, loneliness, social support and attachment to the partner were assessed exactly as in Study 1, with highly similar internal consistencies for both men and women. In order to increase the comparability of relationship satisfaction with the life satisfaction measure, we used a one-item scale “How satisfied are you with your relationship in general” answered on an 11-point Likert scale ranging from 1 (not at all) to 11 20 ABC of social desires (a great deal). In order to increase longitudinal compliance, only a few variables including relationship status (still together as a couple or separated), the three ABC desires, and relationship satisfaction (if still together) were reassessed at Time 2. Results Descriptive statistics. Table 3 shows descriptive statistics of the main variables. As Study 2 relied on dyadic data, observations of male and female participants were not independent, and descriptive statistics and correlations are thus reported separately for the sexes. Replicating the findings of Study 1 for both men and women the desires for affiliation A and closeness C showed small positive correlations, and desire for being alone B correlated moderately negatively with C and negligibly with A. Also, the appetence and aversion subscales again showed relatively high negative correlations such that the internal consistency of each ABC scale was at least as high as the internal consistencies of its two subscales. In addition, the 1-year stabilities were high for both men and women considering the long retest interval (they should be interpreted as stabilities rather than reliabilities): The stabilities for the full scales were all above .70. Finally, between-partner correlations were all significantly positive but small. Sex and age differences. Sex differences in the means of the 9 variables of Table 3 were analyzed with t tests for dependent samples using stepwise Bonferroni correction to prevent alpha inflation. This time significant sex differences were found for the desires for affiliation and being alone both at the scale and the subscale level which can be mainly attributed to the facts that (a) the sample sizes for men and women were more than twice as large as compared to Study 1, and (b) the t test for dependent samples was more powerful than the t test for independent samples in Study 1. Women reported a higher desire for affiliation than men (t(577) = 4.29, p < .001, d = 0.25), particularly more appetence for affiliation (t(577) = 4.19, p < .001, d = 0.29) and less aversion of affiliation (t(577) = 2.95, p 21 ABC of social desires < .003, d = 0.18), and a higher desire for being alone (t(577) = 5.33, p < .001, d = 0.31), particularly more appetence for being alone (t(577) = 5.33, p < .001, d = 0.31) and less aversion of being alone (t(577) = 4.23, p < .001, d = 0.25). These sex differences were however generally small. The correlations of the 9 variables with age ranged from -.23 to .22 for women and from -.23 to .20 for men. For the full ABC desire scales, the correlations with age were -.20, p < .001 (affiliation), .19, p < .001 (being alone), -.23, p < .001 (closeness) for women and -.23, p < .001 (affiliation), .04, p > .05 (being alone), -.04, p > .05 (closeness) for men. Z-tests for correlations from independent samples using stepwise Bonferroni correction revealed that none of these correlations were significantly different from those in Study 1. Confirmatory factor analysis of the subscales. We extended the ABC model in Figure 1 to a model for couples by assuming the same measurement model with identical (co)variances for men and women, allowing for correlations between the two partners of a couple for each latent variable and each residual variable (see Figure 5.1 in Kenny, Kashy, & Cook, 2006, for the same approach). Considering the large sample size, this baseline model showed an acceptable fit, χ2(51) = 125.31, p < .001, CFI = .974, RMSEA = .050, but with low covariances a12, a13, a23, c1 (see Figure 1) similar to the baseline model in Study 1. Setting these covariances to zero significantly decreased the fit of the model, χ2(4) = 13.29, p = .01. Modification indices suggested two additional non-zero covariances between the partners of a couple, namely desire for being alone with partner’s desire for closeness for both men and women. Inclusion of these two covariances led to a significantly better fit, χ2(2) = 33.34, p < . 001, and the resulting final model showed an acceptable overall fit, χ2(53)=105.26, p < .001, CFI = .982, RMSEA = .041 (see Figure 3). The correlations within partners were highly similar to those of Study 1, and the between-partner correlations of the residuals and the latent variables were generally low (absolute correlations below .30) except for the slightly higher correlation between men’s and women’s desire for closeness to the partner (.34); a 22 ABC of social desires higher correlation is to be expected because desire for closeness to the partner is more relationship-specific than the other two desires. Notably, this final model for couples assumed identical variances and covariances for both men and women and was in its within-individual part identical with the final model for individuals in Study 1, fully replicating this individual model for both men and women. Apart from the between-partner covariances for the same latent variables and residual variables, two additional negative between-partner correlations across different latent variables suggested a within-couple linkage between the desires for closeness and being alone: The more one partner desired closeness, the less the other partner desired being alone. This between-partner linkage was less strong than the comparable within-partner link between the desires for closeness and being alone but of the same sign. Prediction of ABC desires. The same procedure as in Study 1 was applied and rendered highly similar results for men and women and highly similar results as in Study 1 with only few exceptions (see Table 4). First, men’s desire for closeness was most strongly associated among the personality traits with conscientiousness, not agreeableness. Second, men’s desire for being alone was as highly associated with support from parents as with support from partner, although both associations were not strong. Finally, among the Big Five, neuroticism, not openness, was the best predictor of desire for being alone in women and men. The small negative associations between neuroticism and all three desires are however plausible and were also found in Study 1. Considering the large number of associations tested, Study 2 replicated the associations between the ABC scales and the 17 predictors found in Study 1 very well. In a final analysis we studied the incremental contributions of sex, age, personality, satisfaction, support, and attachment to the ABC scales by hierarchical regressions. We entered these predictors blockwise from the most unspecific one (sex) to the most 23 ABC of social desires partnership-specific one (attachment to the partner), separately for the two studies and the three desires (see Table 5). Again, the results were highly similar across studies and sex within couples. Desire for affiliation was most strongly related to personality (mainly due to extraversion, see Tables 2 and 4); desire for being alone was most strongly related to attachment (mainly due to dismissing versus preoccupied attachment); and desire for closeness to the partner was most strongly associated with satisfaction (mainly due to relationship satisfaction). However, the other blocks of predictors also added significant incremental contributions to the prediction of each scale with only few exceptions. Discussion The results of Study 2 were largely consistent with Study 1, confirming, with the mentioned minor exceptions, all our expectations. The measurement model of the ABC desires was very well replicated in the dyadic sample as well as the convergent and discriminant validities of the three desires. Again, observed differences between male and female partners of a couple were small to negligible. Complementing the cross-sectional and individual design of Study 1, Study 2 rendered information on stability and dyadic similarity. All three desires were highly stable over a one-year period (rs > .70), given that participants remained in the same relationship. In addition, all three desires showed small to moderate associations between partners. Their high stabilities and modest dyadic similarities along with their differential relations to personality and relationship criteria suggest that the ABC desires are functions of stable differences between couples as well as between individuals. On the one hand, couples develop specific patterns of relationship regulation, which are held up over longer time periods and influence both partners’ desires likewise. This seems to be particularly true for the desire for closeness C, because it showed the highest relations to partnership-specific criteria, the highest dyadic similarity, and low associations with personality traits. On the other hand, individual differences affect the strengths of desires, and 24 ABC of social desires couple similarity may also be a result of assortative mating. This applies particularly to the desire for affiliation A, because it showed the highest associations with personality traits, low dyadic similarity, and negligible relations to partnership-specific criteria. Although the investigation of the relative influences of individual and dyadic differences would afford longitudinal studies, these results are in line with our proposition that the ABC desires are distinguishable by their degree of relationship specificity and agentic versus communal goal contents. The hierarchical regression analyses (Table 5) further corroborated these distinctions. First, A, the more relationship-general of the two communal desires, was best predicted by personality, particularly the sociability facet of extraversion, and by agreeableness. The partnership-specific desire C, on the other hand was best predicted by satisfaction, particularly satisfaction with the partnership. Second, the desire for being alone B was best predicted by romantic attachment, particularly the dismissing style, and was less related to personality and satisfaction than A and C, respectively. Thus, despite a modest negative correlation, B does not merely represent the opposite of communal motivation but an agentic motivational dimension in its own right. General Discussion We introduced a new approach to the structure and assessment of closeness and distance motivation in couple relationships. The triadic ABC model differentiates the desires for affiliation with friends (A), being alone (B), and closeness to the partner (C) along the lines of agentic versus communal goal states and high versus low relationship-specificity. The present studies found good support for the proposed three-dimensional structure and the differential validities of the desires. Based on principles of feedback control theory, the ABC desires were conceptualized as motivational set-points and could be validly and reliably assessed by the frequencies of 25 ABC of social desires appetitive and aversive experiences regarding specific goal states. The ABC scales thus differentiate affectively toned motivational experiences by goal contents. This approach contrasts the assessment of differential sensitivities for threat and reward or approach and avoidance motivation irrespective of goal contents (e.g., Elliot, Gable, & Mapes, 2006; Laurenceau, Kleinman, Kaczynski, & Carver, 2010). In the realm of couple relationships, the latter approach has been increasingly employed in recent years, with particular success in predicting how partners perceive their relationships and construe subjective relationship quality (e.g., Gable & Poore, 2008; Impett et al., 2008). However, in the context of couples’ distance regulation, the differentiation of goal contents seems crucial, because it is the goal content that defines whether behavior is directed towards proximity or distance to the partner. The high factor loadings and the low amount of shared variance between appetence/aversion subscales across goal contents in the ABC model confirmed that aversive and appetitive experiences are specific to goal contents. Thus, the proposed model constitutes a more differentiated motivational approach to dyadic closeness and distance than the established distinction of approach and avoidance motivation that does not account for different goal contents. The distinction of three desires in the ABC model was based on theoretical considerations about alternative goal contents relevant for couples’ distance regulation. However, the tripartite structure resembles inductively rendered taxonomies of personal goals. Chulef, Read, and Walsh (2001) conducted cluster analyses of 135 goals and identified 3 higher-order goal clusters that subsumed 30 lower-order clusters: an intrapersonal cluster comprising mainly agentic goals of freedom, personal growth, creativity etc., an interpersonal cluster subsuming relationship-general communal goals like belonging, friendship, social support, etc., and a cluster subsuming more relationship-specific goals related to marriage, family, sex, and romance. The ABC structure of goal contents is therefore consistent with 26 ABC of social desires previous findings outside the domain of couple relationships that confirmed (a) the basic distinctions of agentic versus communal goals and (b) the relationship-specificity of communal goals. In the following, we will discuss implications of the ABC model with regard to these two structural key features. Finally, we will outline limitations of this investigation and suggest directions for future research. Relationship-Specific Communal Desires The two studies gave strong support to the anticipated relationship-specificity of communal desires. These findings not only correspond with previous research on the structure of goals (Chulef et al., 2003), but also with research on adult attachment styles, which also show substantial within-person variability across different relationship types (e.g., Asendorpf & Wilpers, 2000; Cook, 2000).They also relate to studies on contextualized personality revealing that relationship-specific personality measures outperform global measures in the prediction of relationship outcomes (e.g., Slatcher & Vazire, 2009). Thus, it is not surprising that we found the desire for closeness C but not the desire for affiliation A to relate substantially to partnership satisfaction and the security of romantic attachment. The inclusion of communal desires applying to different relationships in the ABC model allows for differentiated analyses of the relationship-general and relationship-specific aspects of motivational processes in couples. In the specific research context of couples’ distance regulation, the role of A is presumably twofold. On the one hand, partners are usually introduced to one’s circle of friends, and both communal desires A and C may be satisfied simultaneously in joint activities. However, this may be possible only to a limited degree, because closeness to the partner involves many behaviors that usually afford dyadic privacy, such as intimate self-disclosure and sexuality. Thus, on the other hand, affiliation with friends and closeness to the partner are competing alternative goal states. 27 ABC of social desires The inclusion of A in our model of social motivation in couples therefore allows for the identification of distancing behavior due to communal motivation that is directed towards other relationships than one’s ongoing partnership. Such social alternatives are often neglected as sources of distancing behavior. In this first investigation, we focused on friendships. However, other relationship types, for instance family or work relations, can easily be added by replacing the word “friends” with other relationship partners in the proposed item formulations for A (see Appendix). Agentic versus Communal Desires The agentic desire for being alone was clearly distinguishable from both communal desires. Notably, the desires for being alone B and closeness to the partner C, although modestly correlated, differed markedly in their relations to the investigated criteria. B showed much smaller associations with variables of partnership quality than C and was primarily predicted by a specific attachment style (dismissing) rather than by the general security of attachment to the partner. These dissociations between B and C are plausible in the context of agency and communion as distinct motivational dimensions and challenge unidimensional conceptions of social motivation as apparent in, for instance, trait sociability or preferences for closeness over distance or vice versa (e.g., Burger, 1995; Christensen et al., 2006). Alternatively, the ABC model rests on the assumption that, whereas closeness and distance to one’s partner are opposite ends of a behavioral continuum, an individual’s motivation for closeness and distance can stem from different sources, and the kind of motivation behind the behavior is likely to affect its consequences. For instance, the differently sized correlations between relationship satisfaction and B or C suggest that distancing from the partner is more detrimental if due to low communal motivation than if due to high agentic motivation. In a qualitative interview study, Lavee and Ben-Ari (2007) reported that most of the interviewees were well aware of the distinction between getting 28 ABC of social desires “away from one’s partner” (i.e., low closeness motivation) and “getting close to one’s self” (i.e., high agentic motivation) as different motivational sources of dyadic distance (p. 653). The former motivation was associated with emotional detachment from the partner, while the latter primarily aimed at physical distance and was less associated with impaired relationship quality. The present quantitative studies corroborate this distinction and underline the necessity for a more differentiated view of distance motivation than just as the opposite of closeness motivation. Limitations and Future Directions A clear strength of the present investigation is the nearly perfect replication of findings across two large individual and dyadic samples that considerably differed in terms of age, relationship duration, marital status, presence of children, and sampling method. However, both sampling procedures involved self-selection of participants. Another limitation concerns the comprehensiveness of criterion variables. While the reported associations mark a first step towards the establishment of a nomological network of the ABC desires, this network needs to be complemented. In particular, relations to more specific measures of social motivation, such as approach and avoidance goals (Elliot et al., 2006; Laurenceau et al., 2010) and implicit social motives (Hagemeyer & Neyer, in press; McAdams, 1992), should be addressed in future studies. Also, the moderate internal consistencies of some criterion measures (particularly agreeableness) may have led to underestimations of their true associations with the ABC desires. However, this problem could not have affected the most relevant results, namely the relative sizes of correlations between a criterion and the three desires. Finally, the mainly cross-sectional analyses do not warrant causal interpretations of the reported associations. Thus, the distinction of determinants, consequences, and reciprocal relations of the ABC desires requires more extensive longitudinal investigations. 29 ABC of social desires The aim of the present study was the introduction and validation of a structural model of motivational desires that apply particularly to distance regulation in couple relationships. In future studies, we will examine the relations of the ABC desires with couples’ everyday distance regulation. While previous research has mainly concentrated on distance regulation at the micro-level of couples’ interaction styles in specific situations, the motivational conditions of between-couple differences in stable patterns of day-to-day distance regulation (e.g., amount of face-to-face contact, shared decision making, frequency of sexuality) have not received appropriate attention. Closely linked to everyday distance regulation is the investigation of unconventional living arrangements like long-distance and living-aparttogether relationships, which have become more prevalent in recent years and present markedly different challenges and opportunities to couples than traditional marriage or cohabitation (Asendorpf, 2008; Levin, 2004; Pistole, 2010). In fact, different living arrangements imply different distance-regulation styles. For instance, it is obvious that longdistance couples have less face-to-face contact than cohabitating couples. Do couples choose distance-regulation styles that match their needs? What are the consequences of good or poor matches between distance regulation and the partners’ needs for relationship quality and stability? The ABC model provides a comprehensive yet practically economic approach to the investigation of these questions. In previous research, subjective needs for closeness and distance have been assessed in multiple ways, ranging from standardized scales (often unidimensional) and idiographic assessments of relationship goals to ad-hoc inferences from qualitative interviews. This heterogeneity of measures naturally impairs the comparability of studies. The ABC model offers a standardized framework for the assessment of social desires in order to foster the comparability of results and the establishment of a consistent body of research on distance regulation in couple relationships. 30 ABC of social desires ReferencesArbuckle, J. L. (2010). Amos (Version 18.0). Chicago, IL: SPSS Inc.Aron, A., Aron, E. N., & Smollan, D. (1992). Inclusion of Other in the Self Scale and thestructure of interpersonal closeness. Journal of Personality and Social Psychology,63, 596-612.Asendorpf, J. B. (2008). Living Apart Together: Altersund Kohortenabhängigkeit einerheterogenen Lebensform [Living apart together: Ageand cohort-dependence of aheterogeneous life style]. Kölner Zeitschrift für Soziologie und Sozialpsychologie, 60,749-764.Asendorpf, J. B., & van Aken, M. A. G. (2003). Personality – relationship transaction inadolescence: Core versus surface personality characteristics. Journal of Personality,71, 629-666.Asendorpf, J. B., & Wilpers, S. (1998). Personality effects on social relationships. Journal ofPersonality and Social Psychology, 74, 1531-1544.Asendorpf, J. B., & Wilpers, S. (2000). Attachment security and available support: Closelylinked relationship qualities. Journal of Social and Personal Relationships, 17, 115-138.Atkinson, J. W., & Birch, D. (1970). The dynamics of action. Oxford, England: John Wiley.Bakan, D. (1966). The duality of human existence. Boston: Beacon Press.Bartholomew, K., & Horowitz, L. M. (1991). Attachment styles among young adults: A testof a four-category model. Journal of Personality and Social Psychology, 61, 226-244.Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonalattachments as a fundamental human motivation. Psychological Bulletin, 117, 497-529.31 ABC of social desires Baxter, L. A. (1990). Dialectical contradictions in relationship development. Journal ofSocial and Personal Relationships, 7, 69-88.Beck, A. T. (1967). Depression. New York, NY: Harper and Row.Ben-Ari, A., & Lavee, Y. (2007). Dyadic closeness in marriage: From the inside story to aconceptual model. Journal of Social and Personal Relationships, 24, 627-644.Berscheid, E., Snyder, M., & Omoto, A. M. (1989). The Relationship Closeness Inventory:Assessing the closeness of interpersonal relationships. Journal of Personality andSocial Psychology, 57, 792-807.Birtchnell, J. (1993). How humans relate: A new interpersonal theory. Westport, CT: PraegerPublishers/Greenwood Publishing Group.Bischof, N. (1993). Untersuchungen zur Systemanalyse der sozialen Motivation I: DieRegulation der sozialen Distanz--Von der Feldtheorie zur Systemtheorie [On theregulation of social distance – From field theory to systems theory]. Zeitschrift fürPsychologie, 201, 5-43.Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adultattachment: An integrative overview. In J. A. Simpson & W. H. Rholes (Eds.),Attachment theory and close relationships (pp. 46-76). New York, NY: Guilford Press.Brunstein, J. C., Schultheiss, O. C., & Grässmann, R. (1998). Personal goals and emotionalwell-being: The moderating role of motive dispositions. Journal of Personality andSocial Psychology, 75, 494-508.Burger, J. M. (1995). Individual differences in preference for solitude. Journal of Research inPersonality, 29, 85-108.Cabanac, M. (1971). Physiological role of pleasure. Science, 173, 1079-1168.32 ABC of social desires Cantor, N., & Malley, J. (1991). Life tasks, personal needs, and close relationships. In G. J.O. Fletcher & F. D. Fincham (Eds.), Cognition in close relationships. (pp. 101-125).Hillsdale, NJ: Lawrence Erlbaum Associates Inc.Christensen, A. (1987). Detection of conflict patterns in couples. In K. Hahlweg & M. J.Goldstein (Eds.), Understanding major mental disorder: The contribution of familyinteraction research. (pp. 250-265). New York, NY: Family Process Press.Christensen, A., Eldridge, K., Catta-Preta, A. B., Lim, V. R., & Santagata, R. (2006). Cross-cultural consistency of the demand/withdraw interaction pattern in couples. Journalof Marriage and Family, 68, 1029-1044.Christensen, A., & Shenk, J. L. (1991). Communication, conflict, and psychological distancein nondistressed, clinic, and divorcing couples. Journal of Consulting and ClinicalPsychology, 59, 458-463.Chulef, A. S., Read, S. J., & Walsh, D. A. (2001). A hierarchical taxonomy of human goals.Motivation and Emotion, 25, 191-232.Cook, W. L. (2000). Understanding attachment security in family context. Journal ofPersonality and Social Psychology, 78, 285-294.Craddock, A. E. (1997). The measurement of privacy preferences within maritalrelationships: The Relationship Privacy Preference Scale. American Journal ofFamily Therapy, 25, 48-54.Dewitte, M., & De Houwer, J. (2008). Proximity and distance goals in adult attachment.European Journal of Personality, 22, 675-694.Döring, N., & Bortz, J. (1993). Psychometrische Einsamkeitsforschung: Deutsche Neukonstruktion der UCLA Loneliness Scale (Psychometric loneliness research: Anew German version of the UCLA loneliness scale). Diagnostica, 39, 224-239.Dyrenforth, P. S., Kashy, D. A., Donnellan, M. B., & Lucas, R. E. (2010). Predicting33 ABC of social desires relationship and life satisfaction from personality in nationally representative samplesfrom three countries: The relative importance of actor, partner, and similarity effects.Journal of Personality and Social Psychology, 99, 690-702.Elliot, A. J., Gable, S. L., & Mapes, R. R. (2006). Approach and avoidance motivation in thesocial domain. Personality and Social Psychology Bulletin, 32, 378-391.Ernst, J. M., & Cacioppo, J. T. (1999). Lonely hearts: Psychological perspectives onloneliness. Applied and Preventive Psychology, 8, 1-22. Feeney, J. A. (1999). Issues of closeness and distance in dating relationships: Effects of sexand attachment style. Journal of Social and Personal Relationships, 16, 571-590.Gable, S. L., & Poore, J. (2008). Which thoughts count? Algorithms for evaluatingsatisfaction in relationships. Psychological Science, 19, 1030-1036.Gerlitz, J.-Y., & Schupp, J. (2005). Zur Erhebung der Big-Five-basiertenPersönlichkeitsmerkmale im SOEP (On the assessment of Big-Five-based personalitytraits in GSOEP). Berlin, Germany: Deutsches Institut für Wirtschaftsforschung.Goldsmith, D. (1990). A dialectic perspective on the expression of autonomy and connectionin romantic relationships. Western Journal of Speech Communication, 54, 537-556.Gottman, J. M. (1994). What predicts divorce? The relationship between marital processesand marital outcomes. Hillsdale: Erlbaum.Hagemeyer, B., & Neyer, F. J. (in press). Assessing implicit motivational orientations incouple relationships: The Partner-Related Agency and Communion Test (PACT).Psychological Assessment.Helgeson, V. S. (1994). Relation of agency and communion to well-being: Evidence andpotential explanations Psychological Bulletin, 116, 412-428.Hendrick, S. S. (1988). A generic measure of relationship satisfaction. Journal of Marriageand the Family, 50, 93-98.34 ABC of social desires House, J. S., Landis, K. R., & Umberson, D. (1988). Social relationships and health. Science,241, 540-545.Impett, E. A., Strachman, A., Finkel, E. J., & Gable, S. L. (2008). Maintaining sexual desirein intimate relationships: The importance of approach goals. Journal of Personalityand Social Psychology, 94, 808-823.Kantor, D., & Lehr, W. (1975). Inside the family. San Francisco, CA: Jossey-Bass.Kenny, D. A., Kashy, D. A., & Cook, W. L. (2006). Dyadic data analysis. New York, NY: Guilford Press.Kumashiro, M., Rusbult, C. E., & Finkel, E. J. (2008). Navigating personal and relationalconcerns: The quest for equilibrium. Journal of Personality and Social Psychology,95, 94-110.La Guardia, J. G., Ryan, R. M., Couchman, C. E., & Deci, E. L. (2000). Within-personvariation in security of attachment: A self-determination theory perspective onattachment, need fulfillment, and well-being. Journal of Personality and SocialPsychology, 79, 367-384.Larson, R., & Johnson, C. (1985). Bulimia: Disturbed patterns of solitude. AddictiveBehaviors, 10, 281-290. Laurenceau, J. P., Kleinman, B. M., Kaczynski, K. J., & Carver, C. S. (2010). Assessment ofrelationship-specific incentive and threat sensitivities: Predicting satisfaction andaffect in adult intimate relationships. Psychological Assessment, 22, 407-419. Lavee, Y., & Ben-Ari, A. (2007). Dyadic distance: From the inside story to a conceptualmodel. Journal of Social and Personal Relationships, 24, 645–655.Levin, I. (2004). Living apart together: A new family form. Current Sociology, 52, 223-240.35 ABC of social desires Long, C. R., Seburn, M., Averill, J. R., & More, T. A. (2003). Solitude experiences:Varieties, settings, and individual differences. Personality and Social PsychologyBulletin, 29, 578-583.Marsh, H. W., & O’Neill, R. (1984). Self-description questionnaire III: The construct validityof multidimensional self-concept ratings by late adolescents. Journal of EducationalMeasurement, 21, 153-174.McAdams, D. P. (1992). The intimacy motive. In C. P. Smith (Ed.), Motivation andpersonality: Handbook of thematic content analysis (pp. 224-228). New York, NY:Cambridge University Press.McAdams, D. P., Hoffman, B. J., Mansfield, E. D., & Day, R. (1996). Themes of agency andcommunion in significant autobiographical scenes. Journal of Personality, 64, 339-377.McClelland, D. C. (1985). Human motivation. Glenview, IL: Scott, Foresman & Co.Mikulincer, M., & Shaver, P. R. (2003). The Attachment behavioral system in adulthood:Activation, psychodynamics, and interpersonal processes. In M. P. Zanna (Ed.),Advances in experimental social psychology, Vol. 35. (pp. 53-152). San Diego, CA:Elsevier Academic Press.Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics, andchange. New York, NY: Guilford Press.Nestler, N., Back, M. D., & Egloff, B. (2011). Psychometrische Eigenschaften zweier Skalen zur Erfassung interindividueller Unterschiede in der Präferenz zum Alleinsein[Psychometric properties of two scales for the assessment of individual differences inpreference for solitude]. Diagnostica, 57, 57-67.Pistole, M. C. (1994). Adult attachment styles: Some thoughts on closeness-distancestruggles. Family Process, 33, 147-159.36 ABC of social desires Pistole, M. C. (2010). Long-distance romantic couples: An attachment theoreticalperspective. Journal of Marital and Family Therapy, 36, 115-125.Powers, W. T. (1973). Behavior: The control of perception. Oxford, England: Aldine.Prager, K. J. (1995). The psychology of intimacy. New York, NY: Guilford Press.Prager, K. J., & Buhrmester, D. (1998). Intimacy and need fulfillment in couple relationships.Journal of Social and Personal Relationships, 15, 435-469.Read, S. J., Monroe, B. M., Brownstein, A. L., Yang, Y., Chopra, G., & C., M. L. (2010). A neural network model of the structure and dynamics of human personality.Psychological Review, 117, 61-92.Reis, H. T., & Shaver, P. (1988). Intimacy as an interpersonal process. In S. Duck, D. F. Hay,S. E. Hobfoll, W. Ickes, & B. M. Montgomery (Eds.), Handbook of personalrelationships: Theory, research and interventions. (pp. 367-389). Oxford, England:John Wiley & Sons.Sanderson, C. A., & Cantor, N. (1995). Social dating goals in late adolescence: Implicationsfor safer sexual activity. Journal of Personality and Social Psychology, 68, 1121-1134.Sanderson, C. A., & Cantor, N. (2001). The association of intimacy goals and maritalsatisfaction: A test of four mediational hypotheses. Personality and SocialPsychology Bulletin, 27, 1567-1577.Schneider, M. E. (2001). Systems theory of motivational development. In N. J. Smelser & P.B. Baltes (Eds.), International Encyclopedia of the Social and Behavioral Sciences(pp. 10120-10125). Oxford, UK: Elsevier. Schultheiss, O. C. (2008). Implicit motives. In O. P. John, R. W. Robins, & L. A. Pervin(Eds.), Handbook of personality: Theory and research (3rd ed., pp. 603-633). NewYork: Guilford.37 ABC of social desires Slatcher, R. B., & Vazire, S. (2009). Effects of global and contextualized personality onrelationship satisfaction. Journal of Research in Personality, 43, 624-633.Steiger, J. H. (1980). Tests for comparing elements of a correlation matrix. PsychologicalBulletin, 87, 245-251.VanLear, C. A. (1998). Dialectic empiricism: Science and relationship metaphors. In B. M.Montgomery & L. A. Baxter (Eds.), Dialectical approaches to studying personalrelationships. (pp. 109-136). Mahwah, NJ: Lawrence Erlbaum Associates Publishers.Wiener, N. (1948). Cybernetics; or control and communication in the animal and themachine. Oxford, England: John Wiley..38 ABC of social desires Table 1Means, Standard Deviations, Internal Consistencies, and Intercorrelations of the ABC Scales (Study 1) IntercorrelationsDesire forM SD α Aap Aav B Bap Bav C Cap CavAffiliation A 5.43 0.80 .87 .09* .04 -.12* .16*** .13** -.18***appetence Aap 5.09 0.82 .77 -.73*** .10* .09 -.09 .14** .14** -.14**aversion Aav 2.23 0.90 .81-.07 .00 .12** -.16*** -.11* .19****Being alone B 4.28 1.05 .88-.39*** -.42*** .33***appetence Bap 3.99 0.97 .74-.67*** -.46*** -.45*** .44***aversion Bav 3.42 1.31 .88.27*** .32*** -.20***Closeness C 5.76 0.94 .93-appetence Cap 5.65 0.92 .86-.83***aversion Cav 2.13 1.05 .89N = 476 individuals. Trivial correlations between a scale and its subscales are not shown.* p < .05. ** p < .01. *** p < .001.39 ABC of social desires Table 2Prediction of the ABC Scales by Personality, Satisfaction, Support, and Attachment (Study 1) Zero-order correlationsUnique associations βExternal variableA B C A B COpenness.04 .13** .02 -.08 .16* -.01Conscientiousness .10* .08 .02 .03 -.08 -.01Extraversion.36*** -.01 .06 .36*** -.09 .06Sociability.49*** -.19*** .14** ---Agreeableness.16*** -.06 .14** .15*** -.10* .14**Neuroticism-.22*** -.11* -.11* -.15*** -.14** -.09Life satisfaction.23*** .09* .27*** .13* .10 -.03Global self-worth.24*** .16*** .20*** .04 .09 .09Loneliness-.30*** -.15*** -.17*** -.24*** -.15* .10Relationship satisfaction .10* -.13** .60*** -.06 -.26*** .63***Support from friends .35*** .04 .03 .36*** .05 .00Support from partner .04 -.20*** .57*** .03 -.20*** .57***Support from parents .01 -.07 .13** -.04 -.03 .00Secure attachment .13** -.04 .42*** .03 -.10 .35***Fearful attachment -.17*** .15** -.46*** -.16** .06 -.27***Preoccupied attachment -.11* -.28*** .02 -.06 -.29*** .19***Dismissing attachment .02 .45*** -.44*** .05 .38*** -.31***Note. N = 476 individuals. Unique associations refer to standardized regression coefficients βin multiple regressions of an ABC scale on all variables of a block (largest │β│printed bold).a Not included in the multiple regressions.* p < .05. ** p < .01. *** p < .001.40 ABC of social desires Table 3Means, Standard Deviations, Internal Consistencies, 1-year Stabilities, and Intercorrelations of the ABC Scales (Study 2) MenWomenIntercorrelationsDesire forM SD α rtt M SD α rtt A Aap Aav B Bap Bav C Cap CavAffiliation A 5.39 0.79 .86 .73 5.56 0.72 .84 .71 .23*** .07 .00 -.11** .19*** .14** -.21***appetence Aap 4.95 0.83 .75 .72 5.16 0.77 .74 .68 .18*** -.71*** .07 .04 -.08 .14*** .12** -.13**aversion Aav 2.17 0.87 .79 .65 2.05 0.81 .77 .63 -.67*** .26*** -.07 .04 .13** -.21*** -.13** .26***Being alone B 4.02 1.01 .83 .73 4.30 0.96 .84 .71 .11* .11** -.08 .21*** -.37*** -.40*** .30***appetence Bap 3.69 0.98 .73 .68 3.95 0.89 .68 .64 .01 .07 .05.22*** -.58*** -.47*** -.45*** .43***aversion Bav 3.64 1.29 .79 .66 3.35 1.27 .84 .68 -.15*** -.12** .16-.56*** .18*** .23*** .28*** -.14***Closeness C 5.98 0.75 .88 .73 5.98 0.79 .91 .72 .25*** .21*** -.24*** -.30*** -.40*** .17*** .35*** -appetence Cap 5.80 0.80 .82 .67 5.84 0.82 .83 .65 .23*** .22** -.21*** -.32*** -.38*** .21*** .33*** -.74***aversion Cav 1.85 0.80 .80 .68 1.89 0.86 .85 .71 -.24*** -.18*** .25*** .25*** .39*** -.11** -.79*** .32***N = 578 couples (393 for stability rtt). Correlations above the diagonal refer to men, below the diagonal to women, on the diagonal to between-partner correlations. Trivial correlations between a scale and its subscales are not shown.* p < .05. ** p < .01. *** p < .001.41 ABC of social desires Table 4Prediction of the ABC Scales by Personality, Satisfaction, Support, and Attachment (Study 2) MenWomenZero-order correlations Unique associations β Zero-order correlations Unique associations βPredictorsA B C A B C A B C A B COpenness.18*** .11* .06 .07 .16*** .02 .14*** .13** .09* .03 .16** .03Conscientiousness .09* -.16*** .30*** .03 -.17*** .26*** .10* -.01 .10* .06 -.01 .08Extraversion.35*** -.08 .12** .31*** -.16*** .08 .32*** .01 .14*** .27*** -.08 .11*Sociability.37*** -.27*** .20*** .41*** -.22*** .20*** -Agreeableness.12** .01 .21*** .11** -.01 .17*** .10* .01 .13*** .08 -.03 .11**Neuroticism-.19*** -.15*** -.09* -.10* -.20*** -.01 -.24*** -.16*** -.16*** -.18*** -.18*** -.13**Life satisfaction.15*** -.01 .24*** .05 .10* -.12** .21*** .05 .23*** .09 .04 -.06Global self-worth .21*** .00 .32*** .03 .04 .11* .28*** .16*** .29*** .06 .13* .11*Loneliness-.32*** .01 -.35*** -.30*** -.03 -.13** -.39*** -.18*** -.28*** -.34*** -.16*** -.10*Relationship satisfaction .07 -.23*** .58*** -.07 -.30*** .56*** .10* -.11** .54*** -.06 -.21*** .50***Support from friends .34*** .06 -.02 .32*** .11* -.07 .31*** .02 .04 .30*** .03 .03Support from partner .03 -.20*** .47*** -.01 -.17*** .47*** .10* -.14*** .46*** .06 -.14** .46***Support from parents .17*** -.18*** .12** .10 -.17*** .04 .15*** -.06 .10* .11** -.04 .00Secure attachment .09* -.12** .52*** .06 -.06 .36*** .17*** .07 .39*** .08 .10 .25***Fearful attachment -.08 .14*** -.50*** -.05 .02 -.28*** -.15*** .15*** -.44*** -.10* .14** -.25***Preoccupied attachment -.11** -.08 -.19*** -.09 -.14*** .01 -.19*** -.08* -.25*** -.13** -.13*** -.02Dismissing attachment .03 .41*** -.44*** .07 .41*** -.27*** .03 .36*** -.37*** .08 .34*** -.28***Note. N = 578 couples. Unique associations refer to standardized regression coefficients β in multiple regressions of an ABC scale on allvariables of a block (largest │β│printed bold).a Not included in the multiple regressions.* p < .05. ** p < .01. *** p < .001.42 ABC of social desires Table 5Hierarchical Regressions of the ABC Scales on Personality, Satisfaction, Support andAttachment Study 2Study 1MenWomenR2changeR2changeR2changePredictors A B C A B C A B CSex.01* .01 .00 -Age .01 .02*** .03*** .05*** .00 .00 .04*** .04*** .05***Personality .19*** .05*** .04*** .14*** .09*** .13*** .15*** .05*** .07***Satisfaction .03** .06*** .32*** .05*** .05*** .29*** .08*** .06*** .25***Support .08*** .02** .06*** .05*** .03*** .03** .05*** .01 .05***Attachment .02* .18*** .16*** .01 .12*** .12*** .02* .11*** .08***Total R2 .33*** .35*** .61*** .30*** .29*** .56*** .34*** .26*** .51***Note. N = 476 (Study 1), 578 (for men and women in Study 2). The largestR2change for eachdesire scale is printed bold.* p < .05. ** p < .01. *** p < .001.43 ABC of social desires Figure 1. The ABC model at the individual level.44 ABC of social desires Figure 2. Final model for individuals (Study 1). Presented is the standardized solution (allnon-zero βs are significant at p < .001).45 ABC of social desires Figure 3. Final model for couples (Study 2). Presented is the standardized solution (all non-zero βs within individuals and the non-zero between-partner correlations for the latentvariables are significant at p < .001).46 ABC of social desires AppendixABC scalesItems are answered on a 7-point frequency scale ranging from 1 to 7 (1 = never, 4 =sometimes, 7 = always). The original German items can be obtained from the first author. Desire for affiliation (A)Appetence subscale:1. In the presence of friends I feel relaxed2. I happily accept an invitation by friends3. I arrange meeting friends4. I like to be with friendsAversion subscale:1. I feel uncomfortable in the presence of friends (-)2. Being together with friends quickly gets to be too much for me (-)3. I avoid meeting friends (-)4. Being invited by friends is a nuisance to me (-) Desire for being alone (B)Appetence subscale:1. When I am alone, I feel relaxed2. I like to be completely alone3. I try to assure my personal freedom4. I prefer being aloneAversion subscale:1. I feel uncomfortable when I am alone (-)2. Being alone quickly gets to be too much for me (-)3. I avoid being completely alone (-)4. I prefer being not alone (-) Desire for closeness (C)Appetence subscale:1. In the presence of my partner I feel relaxed.2. I enjoy it when my partner wants to be close to me.3. I want to be close to my partner.4. I like being very close to my partner.Aversion subscale:1. I feel uncomfortable in the presence of my partner. (-)2. Being very close to my partner quickly gets to be too much for me. (-)3. I avoid being very close to my partner. (-)4. I prefer my partner being not very close to me. (-) Despite the opposite wordings of some appetence and aversion items, these items are notempirically redundant. Inspections of the inter-item correlations revealed that appetence andaversion items shared less than 40% of their variance.47