--> Europe's Journal of Psychology ejop.psychopen.eu | 1841-0413 Research Reports Attachment Style and Rejection Sensitivity: The Mediating Effect of Self-Esteem and Worry Among Iranian College Students

The present study evaluated the relations between anxious attachment styles and rejection sensitivity, and the potential mediating role of self-esteem and worry. A sample of 125 Iranian college students completed surveys assessing rejection sensitivity, attachment style, worry and self-esteem. Structural Equation Modeling (SEM) analyses were conducted. Results show that there is a significant positive relationship between anxious attachment styles and rejection sensitivity. The study suggests that a higher score in anxious attachment styles is associated with a higher level of worry and lower level of self-esteem and it is also associated with higher level of rejection sensitivity. Furthermore, there is a positive significant relationship between worry and rejection sensitivity and there is a negative significant relationship between self-esteem and rejection sensitivity. Results indicate that self-esteem and worry mediate the relationship between anxious attachment styles and rejection sensitivity.

rejection are also part of the defining criteria for several psychiatric diagnoses, including social phobia, avoidant personality disorder and borderline personality disorder (American Psychiatric Association, 2000). The rejection sensitivity model draws from interpersonal theories of personality (Ayduk, Zayas, et al., 2008).
According to these theories, rejection sensitivity originates from childhood rejection and underlies interpersonal difficulties (Downey & Feldman, 1996). Horney (1937) attributes maladaptive relationships to basic anxiety about desertion, abuse, humiliation and betrayal. She believes that this anxiety is underlying a sensitivity "to any rejection or rebuff no matter how slight, [for example] a change in an appointment, having to wait, failure to receive an immediate response" (Horney, 1937, pp. 135-136). Also, the core of Bowlby's attachment theory is the proposition that we share with other species an evolutionary-based instinct to seek protection from a parent (or carer) when we experience danger. Based on observation of children who have experienced separations and losses, Bowlby first articulated his ideas about attachment theory in terms of the damaging effects that maternal deprivation can have on a child. According to Bowlby children in orphanages and other institutions where they are separate from their parents, especially their mothers, can demonstrate emotionless or sociopathic identities (Bowlby, 1969;Dallos, 2006). The notion that people have a fundamental desire to gain acceptance and to avoid social rejection forms the basis of attachment theory (DeWall et al., 2012. Based on this theory, the aim of this study was to create and test a model that illustrates how anxious attachment styles may influence the development of RS directly and indirectly. Rejection Sensitivity and Attachment Style. According to Bowlby (1980), children's mental models of themselves and of relationships influence their future relationships. The core of these models are expectations about whether significant others will satisfy their needs or be rejected. These expectations derive from the reliability with which their primary caregiver meets their needs in early childhood. The theory suggests that when caregivers tend to meet children's needs sensitively and consistently, children develop secure working models that incorporate the expectation that others will accept and support them. Conversely, when caregivers tend to react to children's need with rejection, children develop insecure working models that incorporate doubts and anxieties about whether or not others will accept and support them. Insecure working models are thought to underlie mistrustful or ambivalent orientations to adult relationships (Hazan & Shaver, 1994). Bartholomew (1990) developed a four-category model of attachment styles based on the four combinations obtained by dichotomizing the subject's abstract image of the self into positive (low dependence) or negative (high dependence) on one axis, and dichotomizing the subject's abstract image of the other into positive (low avoidance) or negative (high avoidance) on an orthogonal axis. This yielded the four attachment categories denoted as secure (positive self, positive other), preoccupied with relationships (negative self, positive other), dismissing of intimacy or dismissing-avoidant (positive self, negative other), and fearful of intimacy or fearful-avoidant (negative self, negative other). Bartholomew and Horowitz (1991) tested this four category model of attachment by developing the instrument Relationship Questionnaire, containing one prototype paragraph for each of the four attachment styles that describes the characteristics of that style (Chotai, Jonasson, Hagglof, & Adolfsson, 2005).
Secure individuals, who have received consistent and responsive care-giving, tend to be comfortable with intimacy and are confident that they are valued by others and are able to rely on others for support (Collins & Feeney, 2000). Such individuals are not threatened by closeness, and maintain high level of self-esteem. Preoccupied persons, who have experienced inconsistent and insensitive care-giving early in life, have an amplified desire for closeness and dependence, but their concern about rejection is also heightened. Such people believe that they are not worthy of being loved and cared for. Preoccupied individuals are highly dependent, demand attention and become distress if the significant other is not readily available in intimate relationships. They tend to be hyper-vigilant and overly concerned with the quality of their relationships (Beyder-Kamjou, 2004;Turner, 2008). Preoccupied individuals are sensitive to significant other's response to them and interpret any disagreement from them as an indication that they desire separation and want to reject them (Pistole & Arricale, 2003). Fearful individuals have experienced uncaring and unavailable caregivers. Such individuals desire intimacy, acceptance and approval by others; however, they are typically distrustful and uncomfortable with their intimate relationships and tend to avoid closeness because of their fear of rejection. Their desire to closeness leads them to establish intimate relationships but they keep significant others at a distance because they fear being rejected (Collins & Feeney, 2000). Dismissing individuals have a history of rejecting or unresponsive attachment figures. Such individuals tend to be suspicious about their intimate relationships and value independency, achievement, self-reliance and self-sufficiency. They do not trust others to get close to them and tend to refrain from getting close to them. Such people see their close relationships as relatively unimportant (Beyder-Kamjou, 2004;Turner, 2008).
Based on Bowlby's theory and the conducted researches in this field, parents who react to their children's needs in a rejecting way raise children who are sensitive to rejection Downey & Feldman, 1996;Pachankis, Goldfried, & Ramrattan, 2008). It seems that this sensitivity continues well into the children's adulthood by means of internalizing past experiences of rejection . Some studies have explored the relationship between adult's attachment styles and RS (Downey & Feldman, 1996;Kennedy, 1999).
For example Erozkan (2009) found that RS levels of students who had authoritarian parents were higher than RS levels of students who had experienced other parenting styles. Also the RS levels of the group of students who had fearful and preoccupied attachment styles were significantly higher than those of others.
The RS model (Downey & Feldman, 1996) argues that people high in RS not only anticipate rejection from others but are also highly concerned and anxious about its actual occurrence. Thus anxious expectations of rejection form the core component of high RS (Ayduk, Gyurak, & Luerssen, 2008). As mentioned before, people with secure attachment styles have more safe and reliable interpersonal relationship and experience more satisfaction and those with dismissing attachment styles do not respect interpersonal relationship. Instead they have only positive self image and avoid experiences of emotions (Beyder-Kamjou, 2004). But the important point is that these two attachment styles experience less anxiety and those who are afraid of being rejected would experience more anxiety in interpersonal relationship. So probably secure and dismissing attachment styles, by comparison to anxious attachment styles, would be less sensitive of being rejected. Thus, in this study it was hypothesized that anxious attachment styles (preoccupied & fearful) would be positively associated with RS. (Only these two styles were studied in the proposed model).
Mediators: Worry and Self-Esteem. Individuals with insecure attachment styles (particularly preoccupied and fearful) experience high levels of anxiety that reflect the tendency to be hyper-vigilant to potential signs of separation from, loss of, and rejection by attachment figures. This kind of tendency is a source of psychological or emotional concern and is an important social component of worry (Wilson, 2006;Allison, 2003;Mellin, 2008). On the other hand, the main purpose of self-esteem is to monitor social relations and detect social rejections. In such a view, self-esteem is a sociometer which activates negative emotions when signs of exclusion appear (Leary & Downs, 1995). Seemingly, individuals with low self-esteem feel to be rejected and are disapproving of others. They may lack self-confidence and social skill required for initiating and developing relationships. Considerable research indicates that secure attachment is associated with higher self-esteem level (Foster, Kernis, & Goldman, 2007) Khoshkam,Bahrami,Ahmadi et al. 365 and individuals with fearful and preoccupied styles of attachment have the lowest self-esteem of the attachment style groups (Huntsinger & Luecken, 2004). Also there is evidence showing that the most important source of daily stress that adults experience consists of negative social interactions and the self-esteem system has a monitoring function for social acceptance such that interpersonal rejection and social exclusion (real or perceived) are associated with lower self-esteem .
In the present study, anxious attachment styles (fearful and preoccupied) were hypothesized to be positively associated with worry and negatively associated with self-esteem, and in turn, positively related to rejection sensitivity. Also, it was hypothesized that worry is positively related and self-esteem negatively related to rejection sensitivity. Worry and self-esteem mediate the anxious attachment styles.

Method Participants
Participants were 125 undergraduate students (63 female and 62 males), recruited from introductory psychology classes at university of Isfahan, Iran. The age range was 18-30 years old with mean age of 20.76 years old and a standard deviation of 1.94 years old. After the purpose of the study and general procedures were described, students were consent to participate in this study, provided with a set of questionnaires. All of them were volunteers, and were not given any payment.

Measures
Rejection Sensitivity Questionnaire (RSQ; Downey & Feldman, 1996). This scale includes 18 hypothetical scenarios in which an individual makes a request to a significant other that makes him/her vulnerable to rejection. Participants are required to rate their agreement with each scenario on a 6-point Likert scale. They indicate (a) their request (anxiety over anticipated rejection; scale: 1, very unconcerned; 6, very concerned) and (b) their subjective likelihood estimate that the person(s) in each scenario will actually respond positively to their request (expectations of acceptance; scale: 1, very unlikely; 6, very likely). Downey and Feldman (1996) report an internal consistency coefficient (Cronbach's α) of .83, test-retest reliability coefficients of .83 and .78 for 3 weeks and 4 months retest intervals, respectively, and supportive construct validation evidence from three additional studies (Ayduk, Zayas, et al. 2008). In this research the standard "forward -backward" procedure was applied to translate the questionnaire from English into Persian (Iranian language). Two independent professionals translated the items and a provisional version was provided. Subsequently it was back translated into English and after that a cultural adaption of the final version was provided. The internal consistency coefficient (Cronbach's α) of the RSQ in the present study was found to be .84.
Attachment Style Questionnaire (ASQ). Attachment styles were measured by the two scales from the Attachment Style Questionnaire (Hofstra, Van Oudenhoven, & Buunk, 2005) that were used to assess fearful and preoccupied attachment for all respondents. A 5-point Likert scale was used, ranging from strongly disagree (1) to strongly agree (5). An example of an item from the 4-item fearful scale is: "I am afraid that I will be deceived when I get too close to others"; from the 5-item preoccupied scale: "I often wonder whether people like me". According to Hofstra et al., (2005), the internal consistency estimates (Cronbach's α) were .68 for the fearful scale and .80 for the preoccupied scale. Also the construct validity of the ASQ was satisfactory. The fearful and preoccupied attachment styles were negatively correlated with self-esteem and positively correlated with distrust. In the present study by using a standard "forward -backward" translation procedure, the English language version of the questionnaire was translated and adapted into Persian (Iranian language) and the internal consistency (Cronbach's α) was found to be .81 for the fearful scale and .89 for the preoccupied scale.
Rosenberg Self-esteem Scale. The 10-item Rosenberg Self-esteem Scale used in this study. Participants used a 4-point Likert scale (1= strongly disagree, 4= strongly agree) to rate items such as "I feel that I have a number of good qualities". According to Martin-Albo, Nunez, Navarro, & Grijalvo (2007) the internal consistency of the scale (Cronbach's α) was .88 and test-retest correlation was .84. This scale was adapted for Iranian participants by Shapurian, Hojat, & Nayerahmadi (1987). In this study the Persian version was used and its internal consistency (Cronbach's α) was .81.

The Penn State Worry
Questionnaire is a 16-item measure designed to assess the tendency to worry excessively ("I am always worrying about something" and "My worries overwhelm me"). Each item is followed by a Likert-type response scale ranging from 1 (Not at all) to 5 (Absolutely true). A total score is derived from the score on all items. Dastgiri

Results
According to the hypothesis of the present study, the RS score was set as the endogenous variable. Two attachment style variables (fearful and preoccupied) were set as exogenous variables which were the central parts of the structural equations that have direct effects on the RS. Also, worry and self-esteem have direct effects on the RS, mediating the attachment styles; neither is independent of the other. In this study data were analyzed via Structural Equation Modeling (SEM). Regarding the assumption of multivariate normality, the hypothetical model was tested using Maximum Likelihood (ML) estimation procedures in LISREL 8.51. Table 1 presents the correlation matrix, means, standard deviations and internal consistency coefficients for the study variables. All of the correlations between RS and proposed variables (anxious attachment styles, worry and self-esteem) are significant. To achieve the overall goal outlined in the present study, the structural equation model was tested. Fig.1 shows the construction of the proposed model in which all the connections and interactions of the variables are presented.
The path coefficients given in this model are shown in  Khoshkam,Bahrami,Ahmadi et al. 367 with self-esteem. Worry has a positive significant relationship with RS and self-esteem has a negative significant relationship with RS.  According to Table 2, All path coefficients of the variables in this model are significant. Therefore, we verify that anxious attachment styles have a positive significant relationship with RS through worry and self-esteem. The fearful attachment style, through mediators such as worry and self-esteem, has an indirect effect of .14 on RS (Table 3) and a direct effect of .29 (Table 2), for a total effect of .43. Also, the preoccupied attachment style, through mediators such as worry and self-esteem, has an indirect effect of .17 on RS (Table 3) and a direct effect of .28 (Table 2), for a total effect of .45. Therefore, their positive significant effects on RS increase comparing to indirect and direct effects in this model. The goodness-of-fit test yielded a chi-square of 3.21, TLI =.97, ECV =.03, CFI =.98, GFI =.98, AGFI = .97, and RMSEA = .001. Table 4   Note: Df = degree of freedom; p = probability value; ECV = expected cross validation; GFI = goodness of fit index; AGFI = adjusted goodness of fit index; TLI = Tucker-Lewis Index; CFI = comparative fit index; RMSEA = root mean square error of approximation.

Discussion
This study tested a model of the associations between insecure anxious attachment styles (fearful & preoccupied), worry, self-esteem and rejection sensitivity. The results show that there is a positive significant relationship between anxious attachment styles and worry, and also negative significant relationship between anxious attachment styles and self-esteem. Also rejection sensitivity appears to be significantly positively correlated to fearful and preoccupied attachment styles and worry, and significantly negatively correlated to self-esteem. The proposed model shows that fearful and preoccupied attachment styles contribute to development of a high level of worry and a low level of self-esteem. Also, worry leads to increase RS and self-esteem leads to reduce RS. Finally, fearful and preoccupied attachment styles contribute to development of RS through the development of worry and reduction of self-esteem. These paths, by giving the large total effects (.43, .45) are the effective paths toward the increasing level of RS.
These findings are consistent with the theoretical perspective of attachment that outlines two dimensions of working models: anxiety and avoidance. The anxiety dimension leads individuals to reflect to cues signaling rejection or separation from significant others sensitively. Anxiously attached people worry excessively about losing their relationship and support of significant others (Wilson, 2006). The individuals with higher levels of attachment anxiety seek out and anxiously anticipate threats in their relationships. They have negative self-view and would doubt their ability to deal effectively with threats on their own and therefore the levels of worry increase for them.
It is thus expected that individuals with insecure anxious attachment styles (preoccupied and fearful) would be prone to worry. This expectation is supported by the findings of this study and consistent with the findings reported by Allison (2003), Kang, Shin, Hwang, Lee, and Song (2008), Brown, Whiteside (2008) and Muris, Meesters, Morren, and Moorman (2004).
Also, contrary to those who have insecure attachments, those who have secure attachments have a feeling of confidence in the available support and love of attachment figures and are motivated to feel good about themselves.
Feeling good about oneself protects against threats and feeling of vulnerability. Self-esteem is predicted based on attachment experiences. Individuals with insecure attachment styles have lower self-esteem. As it was expected, in this study there is a negative significant relationship between anxious attachment styles and self-esteem that is consistent with the results presented by Downey & Feldman (1996), Collins & Read (1990), Hart, Shaver, & Goldenberg (2005), Huntsinger and Luecken (2004).
Attachment theory focuses on the impact that parenting has upon an infant and how it would affect individual's relationships (Bowlby, 1969). The patterns of communication and negotiation of need satisfaction formed between an infant and his or her caregiver are relatively stable in a persons' lifetime and healthy or unhealthy types of attachment depend on these patterns. Based on this assumption, Downey & Feldman (1996) suggested that anxious expectations of rejection are formed as a result of repeated rejection from parents or significant others.
It seems that individuals with anxious attachment styles have higher rejection sensitivity. The results of this study confirmed the hypothesized association between anxious attachment styles and rejection sensitivity that is also supported by the finding of Erozkan (2009) and Downey & Feldman (1996).
Interpersonal self-efficacy in rejection sensitive people is low and may be a factor of difficult relationships. As rejection sensitivity increases, both confidence and ability in social interactions decrease, particularly on the occasion of meeting new people where the chances of rejection are highest (Butler et al., 2007). Interpersonal difficulties in rejection sensitive people impair their sense of personal worth and efficacy, leading to depression and decreased self-esteem in a feedback loop (Leary, 1999). Therefore, these individuals may tend to be particularly concerned with preventing rejection. As it was hypothesized, self-esteem and worry were significant mediators of the relationship between anxious attachment styles and rejection sensitivity. Finally, the proposed model suggests that anxious attachment styles and RS not only have a direct relation, but one that is mediated by worry and self-esteem.
In addition, these findings support the importance of focusing on individuals' styles of attachment in development of rejection sensitivity interventions. It seems that, providing interventions in order to improve anxious attachment styles may be useful for rejection sensitive individuals. Further research is needed to explore these suggestions.
Several limitations of this study should be noted. This study relied on a modest size and convenience sample of undergraduates, only from the University of Isfahan in Isfahan city. Therefore, findings may not be generalized to the larger population; further study is needed in which larger samples are selected from several cities. Cross-cultural studies can also be considered in future works. The other limitation is the cross-sectional nature of this investigation which renders causal conclusions less convincing. A longitudinal approach may contribute to our understanding of causal relationships between attachment style and RS. Despite the limitations, all of the fit indices are acceptable, suggesting a good fit between the proposed theory and our data. Hence, it can be claimed that this study contributes to our knowledge of the origins of RS.