Individualism and collectivism as moderators of the association between attachment insecurities,coping,and social support

Abstract

In two studies, we examined the role of country and individualism and collectivism as moderators of associations between attachment insecurities (anxiety and avoidance) and coping and social support. Study 1 examined the ability of anxiety, avoidance, and country to predict coping and social support variables in Mexico and the US. Anxiety, avoidance, and country had significant effects on coping and social support variables, but there were no significant interactions between the attachment dimensions and country. Study 2 examined the same variables as in Study 1, but also included measures of individualism and collectivism. The results were similar and support the universality of attachment theory but also show that high individualism and low collectivism are associated with stronger relations between attachment dimensions, on one hand, and with diversion coping and perceived social support, perhaps because these variables are related to the most salient aspects of collectivism: duty to one’s group and passive coping.

Keywords

Attachment anxiety attachment-related avoidance coping individualism and collectivism social support

Attachment theory (Bowlby, 1973, 1980, 1982) has become one of the most influential frameworks for studying close relationships across the life span (Cassidy & Shaver, 2008). The study of adult attachment has led to many insights concerning adolescent and adult romantic relationships (see Mikulincer & Shaver, 2007, for a review). The theory was initially developed by Bowlby, a child psychoanalyst, to explain the formation and continuing significance of emotional bonds between human infants and their primary caregivers (usually the parents). The caregiver provides the infant with a “safe haven” in times of threat or distress and with a “secure base” from which to explore the world and develop intellectual, physical, and social skills. When threats or troubles arise, an infant typically signals or approaches a specific, significant caregiver (an “attachment figure,” in the theory’s terms) for protection, comfort, and assistance with emotion regulation. Over time, experiences and expectations established in these early attachment relationships become part of what Bowlby (1982) called internal working models of the self, relationship partners, and relationships. These working models guide how information about both significant and new persons and relationships is encoded, processed, interpreted, stored in memory, and eventually acted on (Bowlby, 1973; Collins, Guichard, Ford, & Feeney, 2006).

In the case of close relationships between adults, a similar emotional bond is formed as one person comes to rely on the other as a safe haven and secure base. The nature of secure and insecure attachments has been extensively studied in children (as reviewed in Cassidy & Shaver, 2008), adolescents (e.g., Allen, 2008), and adults (Mikulincer & Shaver, 2007). Attachment research has shown that people whose early caregivers were unavailable and unsupportive, called “insecure” by Bowlby (1982), are more likely than their secure counterparts to appraise their partners’ availability and responsiveness negatively (Mikulincer & Shaver, 2009, for a review). Disappointing or frustrating interactions with attachment figures erode a person’s reliance on support seeking and lead to one or both insecure attachment-related coping strategies: hyperactivation and deactivation. Hyperactivation strategies include energetic, anxious, controlling, or intrusive attempts to force a relationship partner to pay more attention and provide better care, whereas attachment system deactivation involves the suppression or inhibition of support-seeking tendencies (Waters & Waters, 2006).

People’s attachment orientations can be measured along two dimensions: attachment anxiety and avoidance (Brennan, Clark, & Shaver, 1998). The anxiety dimension indicates the degree to which an individual worries that a partner will not be available and supportive in times of need, which heightens efforts to demand and maintain closeness to relationship partners. The avoidance dimension indicates the extent to which a person distrusts relationship partners’ goodwill and capacity to help, which heightens efforts to maintain a safe degree of independence and self-reliance.

Each attachment insecurity dimension, anxiety or avoidance, is related to a particular way of coping with stress. Hyperactivation, which is the characteristic of people who score high on measures of attachment anxiety, intensifies negative emotional reactions to threats—reactions such as crying, sadness, anxiety, fear, and shame (Shaver, Mikulincer, & Chun, 2008). On the other hand, deactivation, which is the characteristic of avoidant individuals, includes inhibition of negative emotional responses because, when expressed, they can be interpreted as signs of weakness or vulnerability, contradicting the avoidant person’s sense of strength and self-reliance (Cassidy, 1994).

Much of the literature on adult attachment is based on the studies conducted in the US. As interest in the theory has spread around the world, interest has also grown in making cross-national and cross-cultural comparisons to see whether the major findings accumulated in English-speaking countries generalize. Here, we are particularly interested in the possibility that the US and Mexico may be culturally different in ways that affect relations between the attachment insecurity dimensions and other variables.

One way of conceptualizing the cultural differences is in terms of cultural syndromes. Triandis (1993) defined a cultural syndrome as a set of elements of subjective culture—shared attitudes, beliefs, categorizations, expectations, norms, roles, definitions, and values—organized around a theme. Triandis identified two such cultural syndromes, individualism, and collectivism.

In the case of individualism, the organizing theme is the centrality of the autonomous individual; in the case of collectivism, it is the centrality of the collective—family, tribe, work organization, or religious group. Later on, Triandis (1995) argued that there are four defining attributes of individualism and collectivism, respectively: (1) definition of the self, which can be either independent or interdependent (Markus & Kitayama, 1991); (2) personal goals that can take priority over in-group goals or vice versa (Triandis, 1990); (3) an emphasis on exchange rather than communal relatedness (Kim, Triandis, Kagitçibasi, Choi, & Yoon, 1994); and (4) the relative importance of attitudes and norms as determinants of social behavior. In individualistic cultures, attitudes are more important than norms, but in collectivistic cultures norms are given more weight than attitudes (Bontempo & Rivero, 1992; Davidson, Jaccard, Triandis, Morales, & Díaz-Guerrero, 1976).

The US is generally regarded as being a highly individualistic culture (Hofstede, 2001). In such cultures, people are expected to be self-reliant and their behavior is based largely on personal goals and interests, such as defending one’s personal space, acquiring goods and services for oneself, and establishing an independent identity. In contrast, collectivistic societies facilitate and encourage interdependent, cooperative, and group-oriented behavior. Groups have priority over personal goals, people conform to the norms of the group, and there is an emphasis on relationships (Triandis, 1995). If this reasoning is correct and individualism and collectivism affect people’s sense of relatedness to others, the associations between attachment insecurities and other aspects of relationships might be moderated by country (the US and Mexico) and by individualism and collectivism.

Although both individualism and collectivism, and tensions between them, probably exist within all large nations, there are likely to be differences in influence between the two syndromes in different countries (Kagitcibasi, 1997). Mexico, for example, has been found in some studies to be more collectivistic than the US (Hofstede, 2001), and the US has been found to be more individualistic than Mexico (Díaz-Loving, 2005). If these expectations are correct, there should be a greater emphasis on self-reliance in the US and a greater emphasis on interdependence and within-group harmony in Mexico, and these different emphases might qualify, or moderate, the expected associations between attachment insecurities and measures of coping and social support.

Only a few published studies of adult attachment have compared Spanish- and English-speaking countries. For example, Pierrehumbert et al. (2009) compared the attachment representations of preschool children from Switzerland, Spain, Italy, Chile, and Belgium. Friedman et al. (2010) explored the effect of collectivism on the associations between attachment-related avoidance and relationship problems in college students from Mexico, Hong Kong, and the US. Schmitt et al. (2004) explored the frequency of attachment styles across different nations, including Mexico, Argentina, Chile, and Peru. In general, these studies found that, despite attachment being a universal phenomenon, attachment-related processes exhibit some culture-specific features.

Some other cross-cultural studies have explored, for example, the effect of collectivism on Japanese and American mothers’ conceptions of attachment security (Rothbaum, Kakinuma, Nagaoka, & Azuma, 2007) and the level of intimacy and positive expectations about friendship in Korea and the US (Soon & Malley-Morrison, 2000). In general, these studies have shown that attachment security takes similar forms and has similar antecedents and consequences in different cultures; however, there are some cultural differences in variables associated with, or acting as moderators of the effects of, attachment security and insecurity. For example, Rothbaum, Kakinuma, Nagaoka, and Azuma (2007) found that there is more exploration associated with security and more anger and aggression associated with insecurity in the US than in Japan; Soon and Malley-Morrison (2000) found that secure Korean students reported less intimate relationships with friends than secure Americans reported. None of the existing studies has explored the association between kinds of attachment insecurity, on one hand, and coping and social support, on the other hand. Exploring those issues is the purpose of the present article.

Attachment insecurities and individualism and collectivism as predictors of coping and social support

According to attachment theory and research, people who have developed a sense of attachment security through interactions with generally available and responsive attachment figures are likely to have developed skills to deal effectively with stressful events. Insecure individuals, either highly anxious or avoidant or both, are likely to lack these skills and have developed less effective coping strategies. Highly attachment-anxious individuals, for example, tend to rely on emotion-focused coping strategies, appraise events in more threatening ways, and appraise themselves as less capable of coping with adverse situations; whereas highly avoidant individuals tend to use distancing coping and to seek less social support (Mikulincer & Florian, 1995; Mikulincer, Florian, & Weller, 1993). When measuring coping strategies with the Multi-Situational Coping Scale (Góngora & Reyes Lagunes, 1998), we expected attachment anxiety to predict passive acceptance of negative emotions (e.g., getting and staying sad or agonizing when facing a problem) and diversion or distraction (e.g., trying not to think about the problem). We also expected avoidance to be inversely related to reappraisal (e.g., looking on the bright side of a situation) and to be positively related to diversion.

Several studies have found evidence that early interactions with attachment figures affect how people experience social support (see Mikulincer & Shaver, 2009, for a review). In our study, we explored four different areas of such experience: (1) seeking social support, (2) expectations or beliefs about others’ socially supportive behavior (positive and negative), (3) perceiving actual social support, and (4) emotional reactions to receiving social support (positive and negative). People who experienced inconsistent social support in childhood (those who then develop an anxious, vigilant, clinging attachment pattern) and those who experienced parental distance or lack of care (and therefore developed a more avoidant, self-reliant attachment pattern) should have opposite approaches to seeking social support (Shaver & Mikulincer, 2002). Based on the documented tendency for highly anxious individuals to hyperactivate their attachment systems as a means to maintain proximity with their attachment figures (Mikulincer & Shaver, 2003; Shaver & Mikulincer, 2002), attachment anxiety was expected to correlate positively with seeking social support. Avoidant individuals tend to cope with stressful events in the opposite way. They tend to deactivate their attachment systems by denying the importance of their problems and dealing with them alone (Mikulincer & Shaver, 2003; Shaver & Mikulincer, 2003); therefore, avoidance was expected to correlate negatively with seeking social support.

Research has shown that, based on their negative early experiences with attachment figures, both anxiety and avoidance are associated with negative views of others. When compared with their secure counterparts, people who score high on either anxiety or avoidance are less likely to see others as trustworthy, dependable, and altruistic (Collins & Read, 1990) and are more likely to attribute hostile intent to others in anger-eliciting situations (Mikulincer, 1998). Therefore, we expected both anxiety and avoidance to correlate negatively with positive beliefs about social support and positively with negative beliefs about others’ ability and willingness to provide social support.

Attachment insecurities may also color perceptions of receiving social support, especially when the support provider’s behavior is somewhat ambiguous (Collins & Feeney, 2004). Relative to secure individuals, insecure ones (whether anxious or avoidant) tend to perceive less support, even when the support provider’s actual behavior is statistically controlled (Besser & Neria, 2011; Collins & Feeney, 2004). Therefore, both forms of insecurity were expected to correlate with perceptions of receiving less than optimal support. Finally, attachment-anxious individuals seem to be less benefited by an attachment figure’s supportive behavior than their less anxious counterparts (Campbell, Simpson, Boldry, & Kashy, 2005), perhaps due to their doubts about their attachment figures’ motives and willingness to provide support. Therefore, we expected anxiety to be negatively correlated with positive reactions to social support and positively correlated with negative reactions.

However, besides attachment security, individualism and collectivism influence coping and social support-related behavior as well. Our main hypothesis is that the expected relations between attachment and coping and social support are present everywhere, but the magnitude of such relations is likely to be weaker in more collectivistic and less individualistic cultures. The sense of agency, which is related to individualism (Brewer & Chen, 2007; Waterman, 1984), is an important determinant of coping behavior. Collectivistic individuals may lack this sense of agency over their lives (Sastry & Ross, 1998) and tend to rely more on emotion-focused coping strategies (Kuo, 2010), despite their levels of attachment security. Thus, differences in attachment orientation may have less influence on coping strategies in a collectivistic society.

Similarly, although attachment anxiety and avoidance are generally associated with low perceived social support, cultural differences in social relationships may affect this association. For example, a sense of intense emotional connectedness and a constant exchange of responsiveness between people are characteristics of collectivistic societies (Sinha & Verma, 1994), which may allow even anxious and avoidant members of a collectivistic society to perceive themselves as having adequate social support. With regard to social support seeking, individuals who score high on collectivism may be less willing to seek social support if they perceive this as self-serving and likely to disturb group harmony (Kim, Sherman, Ko, & Taylor, 2006), which could weaken the effect of anxiety and avoidance on social support seeking. Thus, we would expect the associations between attachment anxiety and avoidance, on one hand, and beliefs about others’ socially supportive behavior and perception of receiving social support, on the other hand, to be weaker in a collectivistic society.

A collectivistic orientation may favor sharing of stressful events and seeking social support, because the self is viewed as fundamentally connected to others (Markus & Kitayama, 1991). The tendency to avoid seeking social support from others and the experience of negative emotions, typical of highly avoidant people, could be partially eliminated by the collectivistic tendency to rely on others to cope with stressful events and to consider this coping mechanism appropriate. Thus, we expect associations between attachment avoidance and seeking social support and experiencing negative emotions after receiving social support to be weaker in a collectivistic sample. As mentioned below, Mexico has been found to be more collectivistic than the US (Hofstede, 2001) by some measures, and the US has been found to be more individualistic (Díaz-Loving, 2005). Therefore, we collected data from Mexican and US university students to test our hypotheses.

There is also some indirect evidence that individualism and collectivism are related to differences in attachment. There are suggestions in the cross-national study by Schmitt et al. (2004), for example, that there are national differences in scores on Bartholomew and Horowitz’s (1991) brief measure of attachment styles, with North America and Western Europe being more avoidant than East Asia. The Mexican participants in Schmitt et al.’s (2004) study scored lower on measures of the secure, preoccupied, and fearful attachment styles and higher on dismissing than “Western US” participants (the sample in that study that most closely resembles the Californians in ours). However, these participants were college students, whereas the Mexican participants in that study were volunteers—a community sample. According to Rothbaum, Weisz, Pott, Miyake, and Morelli (2000), the greater emphasis on social harmony, rather than individuation, typical of collectivistic cultures might affect some Western developmental tasks linked to attachment security in infancy. These authors claim that some collectivistic nations, such as Japan, discourage exploration behavior and the development of autonomy because they do not approve of differentiating self from others. (This article, however, was heavily criticized by several attachment researchers, including Chao, 2001; Gjerde, 2001; Kondo-Ikemura, 2001; and Posada & Jacobs, 2001.) Despite the mixed evidence obtained to date, we expected Mexican students to score higher on attachment anxiety and US students to score higher on attachment-related avoidance.

We conducted two separate studies to test these hypotheses, one to examine country differences and the other to replicate findings from the first study, while including measures of individualism and collectivism.

Study 1

We collected data in the US, from students attending the University of California (UC), Davis, and in Mexico, from students attending the Autonomous National University of Mexico in Mexico City. The goal of the study was to see whether there are significant cultural differences in associations between the attachment insecurity dimensions and measures of coping and social support. We expected the direction of associations in both samples to be compatible with the existing literature on attachment, but weaker for Mexicans, whom we expected to be more collectivistic. In other words, we expect both attachment anxiety and avoidance to be better predictors of passive coping and negative experience of social support (negative beliefs about other’s behavior, low perception of receiving social support, and negative emotions) in a Mexican sample, when compared with an American sample.

Method

Participants

We sampled 116 Mexican students (97 women and 19 men) between 18 and 25 years of age (M = 20.42, SD = 1.55). Most of them (N = 114; 98.3%) were single, one (0.9%) was married and one (0.9%) was cohabiting. All of them were undergraduate students. Regarding religious beliefs, 1.7% (2) were Protestant or other non-Catholic Christian, 60.3% (70) were Catholic, 29.3% (34) were atheistic or agnostic, 1.7% (2) were Buddhist, and 6.9% (8) reported another religion. The sample was homogeneous with respect to ethnic background. All Mexican participants were residents of Mexico City, undergraduate students at the Universidad Nacional Autónoma de México (Autonomous National University of Mexico).

We also sampled 92 American students (72 women and 20 men) between 18 and 27 years of age (M = 19.21, SD = 1.52). Most of them were single (94.7%, 89); one (1.1%) was married and 2.1% (2) were cohabiting; 16% of them (15) were Protestant or other non-Catholic Christian, 27.7% (26) were Catholics, 5.3% (5) were Jewish, 2.1% (2) were Muslim, 11.7% (11) were Buddhist, 21.3% (20) were atheistic or agnostic; 13.8% (13) reported some other religion, and 2.1% (2) did not provide information about religion. Finally, 29.8% (28) were European Americans, 29.8% (28) were Asian Americans, 9.6% (9) were Latin Americans or Hispanics; 11.7% (11) reported some “other” ethnic background, and 19.1% (18) provided no information about ethnic background. This ethnic diversity is the characteristic of UC Davis.

Materials

Attachment insecurity dimensions

Participants completed the Experiences in Close Relationships Questionnaire (ECR; Brennan et al., 1998), which assesses two kinds of attachment insecurity: anxiety (18 items concerning fear of rejection and abandonment; e.g., “I worry about being rejected or abandoned”) and avoidance (18 items concerning discomfort with closeness and interdependence; e.g., “I find it difficult to allow myself to depend on others”). Both sets of items are evaluated on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree). Factor analyses revealed the same two-dimensional structure in both the samples. Cronbach’s alphas for the anxiety scale were .84 for Mexicans and .92 for Americans; alphas for the avoidance scale were .89 for Mexicans and .91 for Americans.

Coping styles

The Multi-Situational Coping Strategies Scale (MSCSS; translated and adapted for this study based on Góngora & Reyes Lagunes, 1998) includes 91 items concerning coping with problems in the domains of health, partner, family, friends, and work. For the sake of clarity and parsimony, the scale structure for this study was reduced to three general factors using a second-order factor analysis. Those three factors are: reappraisal (e.g., “I try to see the bright side”), passive acceptance of negative emotions (e.g., “I get depressed,” “I agonize”), and diversion or distraction (e.g., “I pretend that nothing is happening”). The items in each of these categories were submitted to factor analysis in both the American and Mexican samples to determine the nature of possible subscales. In each case, the results were very similar for the two samples. Good reliability coefficients for both the samples on all scales provide further evidence of similar within-scale coherence. The reappraisal scale includes 43 items (α = .92 for Mexicans and .95 for Americans); passive acceptance includes 25 items (α = .72 for Mexicans and .78 for Americans); and diversion includes 23 items (α = .76 for Mexicans and .77 for Americans). All items were evaluated on a 6-point scale ranging from 1 (never) to 6 (always).

Social support

The Social Support Scale (SSS; Frías & Díaz-Loving, 2011), based on attachment theory, includes 45 items related to: (a) beliefs about others’ socially supportive behavior (e.g., “People help because they are generous”), (b) seeking social support (e.g., “When I am in trouble or in need, I ask my significant others to comfort me”), (c) perception of receiving social support (e.g., “When I am in trouble or in need, my significant others comfort me”), and (d) emotional reactions to receiving social support (e.g., “When I receive support, I feel indebted”). As with the copying styles measure, the items in each of these categories were submitted to factor analysis in both the American and Mexican samples. In each case, the results were very similar for the two samples. Good reliability coefficients for both the samples on all scales provide further evidence of similar within-scale coherence. The items assessing beliefs about others’ socially supportive behavior were divided into two subscales: positive beliefs, which includes nine items (α = .73 for Mexicans and .79 for Americans), and negative (skeptical, suspicious) beliefs, which includes four items (α = .78 for Mexicans and .77 for Americans). The seeking social support subscale includes 11 items (α = .82 for Mexicans and .84 for Americans). The perception of receiving social support subscale includes eight items (α = .85 for Mexicans and .92 for Americans). Finally, the emotional reactions to receiving social support items were divided into two subscales, one concerning positive emotional reactions, with nine items (α = .81 for Mexicans and .90 for Americans), and the other concerning negative emotions, with four items (α = .79 for Mexicans and .82 for Americans). All of the items were evaluated on a 6-point Likert-type scale ranging from 1 (disagree strongly) to 6 (agree strongly).

Procedure

The ECR was translated into Spanish, and the translation was evaluated using a back-translation technique. The SSS and MSCSS were translated into English and checked by the same method. Mexican and American participants completed questionnaires through the Internet Survey System at UC Davis. All participants were invited to take part in the study in exchange for credit in a psychology course at their respective universities. All signed a consent agreement and were informed that they could withdraw from the study at any point.

Results

Because of the number of statistical tests conducted, a p value of .01, rather than .05, was chosen to distinguish statistically significant from nonsignificant results, in both Study 1 and Study 2. The means and SDs for key variables in both countries in Study 1 are shown in Table 1, along with t tests comparing the means for the two countries. Attachment anxiety and avoidance scores were significantly higher in the American sample than in the Mexican sample, suggesting that the Mexican participants feel more secure on average than the American participants. And the Americans scored higher on passive acceptance of negative emotions, diversion or distraction as a coping mechanism, and perception of receiving helpful social support. Mexicans scored significantly higher than Americans on coping by reappraising and having positive reactions to receiving social support.

Table 1.

Means and SDs for key variables in both countries and t tests comparing means between countries.

	Mexico			US			t	df	p*
	Mean	SD	N	Mean	SD	N	t	df	p*
Anxiety	3.35	1.04	116	4.17	1.05	92	5.58	206	.00
Avoidance	3.04	.89	116	3.60	1.07	92	4.10	206	.00
Reappraisal	4.28	.62	116	3.94	.74	93	3.63	207	.00
Passive acceptance	2.40	.48	116	2.99	.50	93	8.66	207	.00
Diversion	2.39	.48	116	2.92	.51	93	7.62	207	.00
Positive beliefs about OSSB	4.24	.64	116	4.24	.56	92	.00	206	.99
Negative beliefs about OSSB	3.80	.98	116	4.00	.83	92	1.54	206	.12
Seeking social support	4.34	.83	116	4.20	.76	92	1.21	206	.22
Perception of receiving social support	4.43	.64	116	4.51	.82	92	.79	170.71	.00
Positive emotional reactions to social support	5.10	.60	116	4.72	.75	92	3.93	171.71	.00
Negative emotional reactions to social support	2.74	.68	116	3.18	.90	92	3.87	165.60	.00

OSSB: other’s socially supportive behavior.

*Probability levels are rounded to two decimal places.

A series of hierarchical regression analyses was conducted to determine how the attachment insecurity variables and country, considering both unique main effects and interactions, were related to the coping and social support variables. The results are summarized in Table 2, which also shows which variables and interactions were entered as predictors on each step of the analysis. Attachment anxiety significantly predicted passive acceptance of negative emotions and coping by diversion or distraction, seeking social support, and reacting negatively to receiving social support. Avoidance significantly and positively predicted diversion and negative reactions to social receiving social support. Furthermore, avoidance was negatively associated with reappraisal, seeking social support, perception of received social support as helpful, and having positive reactions to social support.

Table 2.

Regression coefficients predicting coping and social support variables from country and attachment anxiety and avoidance.^a

	Reappraisal	Passive acceptance	Diversion	Positive beliefs about OSSB	Negative beliefs about OSSB	Seeking social support	Perception of receiving SS	Positive reactions to SS	Negative reactions to SS
	Ad R ² = .16	Ad R ² = .34	Ad R ² = .30	Ad R ² = .01	Ad R ² = .02	Ad R ² = .20	Ad R ² = .14	Ad R ² = .18	Ad R ² = .29
Anxiety	−.10	.25***	.20***	−.09	.01	.26***	−.06	.06	.36***
Avoidance	−.31***	.13	.19**	−.02	−.05	−.47***	−.38***	−.37***	.26***
Country	.12	−.38***	−.34***	−.05	−.03	.05	−.18**	.19**	−.06
	Ad R ² = .15	Ad R ² = .35	Ad R ² = .30	Ad R ² = .01	Ad R ² = .05	Ad R ² = .19	Ad R ² = .13	Ad R ² = .18	Ad R ² = .28
Country × anxiety	.03	.15	.14	.08	.11	−.01	.03	.08	.00
Country × avoidance	.04	.00	−.00	−.18	−.23	−.03	−.10	−.13	−.09

OSSB: other’s socially supportive behavior; SS: social support; Ad R ²: explained variance.

^aCountry values: US = 0, Mexico = 1.

Some country effects were significant, even with attachment anxiety and avoidance statistically controlled. Americans scored higher on passive acceptance, diversion, and perception of receiving social support, while Mexicans scored higher on positive reactions to receiving social support. These correlational results are similar to t tests in Table 1, except that the effects of the attachment variables on reappraisal and negative reactions to receiving social support became nonsignificant in the regression analysis (when country was taken into account). Finally, there were no significant interactions (at a p value of .01 or lower) between attachment insecurities and country.

Discussion

Attachment anxiety and avoidance scores were significantly higher in the American sample than in the Mexican sample, suggesting that the Mexican participants felt more secure, on average, than the American participants. The Americans scored higher on accepting negative emotions, using diversion or distraction as a coping mechanism, and perceiving social support received as helpful. The Mexicans scored significantly higher than the Americans on coping by reappraising and having positive reactions to receiving social support.

Most of these country differences remained significant even when anxiety and avoidance were statistically controlled: Americans scored higher on passive acceptance, diversion/distraction, and perception of receiving social support, while Mexicans scored higher on positive reactions to receiving social support.

Although the country differences were not what we expected, the findings for individualism and collectivism were in line with our expectations. Accepting negative emotions, using diversion as a coping strategy, and perceiving social support as useful (in this study, all characteristics of the American sample) are consistent with the tendency to use passive coping strategies, which has previously been associated with collectivistic cultures. Using reappraisal as a coping style and having positive reactions to receiving social support (in this study observed in the Mexican sample) are consistent with the individualistic tendency to cope with stressors in a more active manner, both behaviorally and cognitively. They are also consistent with considering the seeking of social support to be an acceptable way to deal with problems, despite its potential for disturbing group harmony. This pattern of results leads us to believe that Mexicans (at Mexican college students) may not be as collectivistic and Americans (or American college students in a large university in California) may not be as individualistic as we thought. We explored this possibility in Study 2.

With country differences statistically controlled, attachment anxiety significantly predicted passive acceptance of negative emotions. This result fits with the tendency of attachment-anxious individuals to perceive negative emotions as congruent with attachment goals and to sustain or even exaggerate them in order to attract attachment figures’ attention in hopes of receiving protection and support (Cassidy & Kobak, 1988; Mikulincer & Shaver, 2003). The tendency of more anxious participants to get distracted or divert their attention from problem solving is also consistent with the previous literature. Problem solving works against the anxious person’s self-construal as helpless and incompetent, because being competent and self-reliant might result in a loss of attention and support from attachment figures (Mikulincer & Shaver, 2009).

The results for social support are also consistent with previous findings. Attachment anxiety predicted social support seeking, which fits with anxious individuals’ previously documented energetic, intrusive, and controlling attempts to induce a relationship partner to pay more attention and provide better care (see Mikulincer & Shaver, 2009). On the other hand, their tendency to react negatively to receiving social support fits with Main’s (1995) observation that either because they are overwhelmed in conflict settings, which reduces their capacity to monitor and notice their partners’ actions, or because they are doubtful about their partners’ goodwill, anxious individuals may be more inclined to deny, dismiss, or discount their partners’ positive behavior.

Attachment-related avoidance was negatively associated with reappraisal and positively associated with diverting attention or distracting oneself from problematic situations. These results are in line with avoidant individuals’ tendency to deny or suppress emotion-related thoughts and memories and to avoid anything that might activate their attachment systems (Kobak, Cole, Ferenz-Gillies, Fleming, & Gamble, 1993). Avoidance also predicted passive acceptance of negative emotions. Because avoidant individuals tend to inhibit overt emotional expressions and fail to seek social support (Mikulincer & Shaver, 2003), they may not use effective problem-focused efforts and may continue to experience negative emotions when facing a problem that overrides their defenses. In line with this interpretation, avoidance was negatively associated with seeking social support and perceiving received social support as helpful. These results fit with Bowlby’s (1973) concept of “compulsive self-reliance,” which refers to the avoidance of support seeking based on a long history of disappointing interactions with attachment figures.

In general, all of the significant associations of coping and social support variables in both countries were compatible with previous research, providing support for the hypothesized universality of attachment-related processes. One of the goals of Study 2 was to identify culture-specific variations in this universal phenomenon.

Study 2

Study 1 indicated that the expected effects of attachment insecurities on coping and social support processes occurred in our Mexican sample and that there were mean country differences in several variables. In that study, the cultural differences were assumed to be aligned with country differences, and we imagined that the Mexican sample had a more collectivistic orientation than the (presumed) more individualistic American sample. But these assumptions might not apply to samples of university students, because they are relatively well educated in both countries and tend to be relatively well off economically, which increases social independence (Freeman, 1997; Triandis, McCusker, & Hui, 1990). Both factors, education and economic resources, have been shown to be associated with decreased collectivism and increased individualism (Freeman, 1997; Oyserman, Coon, & Kemmelmeier, 2002) regardless of country. In Study 2, we included measures of individualism and collectivism to assess country differences in those orientations and to see whether the orientations affect associations between attachment insecurities and the coping and social support variables.

Method

Participants

We sampled 158 Mexican university students (116 women and 42 men) between 18 and 32 years of age (M = 20.74, SD = 1.86). Most of them (99.4%, 157) were single; only one (0.6%) was married. Regarding religious beliefs, 7% (11) were Protestant or Other non-Catholic Christian, 68.4% (108) were Catholic, 19% (30) were atheistic or agnostic; and 5.7% (9) reported another religion. The sample was homogeneous in terms of ethnic background.

We also sampled 248 American university students (180 women and 68 men) between 18 and 30 years of age (M = 20.11, SD = 1.81). Most of them (99.8%, 235) were single, 4.8% (12) were cohabiting, and 0.4% (1) provided no information about marital or relationship status. Regarding religious beliefs, 19.4% (48) were Protestant or other non-Catholic Christian, 21% (52) were Catholic, 2.8% (7) were Jewish, 4% (10) Muslim, 10.1% (25) Buddhist, 29.4% (73) atheistic or agnostic, 12.9% (32) reported another religion, and 0.4% (1) provided no information about religion. Finally, 20.6% (51) were European American, 33.9% (84) were Asian American, 2% (5) were African American, and 10.9% (27) were Latin American or Hispanic; 12.5% (31) reported some other ethnicity or combination of ethnicities, and 20.2% (50) provided no information about ethnicity. As mentioned in Study 1, this ethnic diversity is the characteristic of UC Davis.

Materials

We used the scales described in Study 1 to measure the attachment insecurity dimensions, coping strategies, and social support variables. Cronbach’s alphas were as follows: attachment anxiety: .89 and .91, for Mexico and the US, respectively; avoidance, .84 and .90. Alphas for the coping subscales were: reappraisal, .92 for Mexicans and .95 for Americans; passive acceptance, .84 and .85; and diversion, .79 and .84. Finally, alphas for the social support subscales were: positive beliefs about other’s socially supportive behavior, .72 for Mexico and .84 for US, respectively; negative beliefs about others’ behavior, .78 and .73; seeking social support, .80 and .86; perception of receiving social support, .91 and .92; positive reactions to social support, .88 and .89; and negative reactions to social support, .87 and .86.

We added a self-report measure of individualism and collectivism (Singelis, Triandis, Bhawuk, & Gelfand, 1995). This scale includes 32 items divided into two subscales: (a) individualism with 16 items (e.g., “I enjoy being unique and different from others in many ways;” Cronbach’s α = .74 for Mexicans and α = .77 for Americans) and (b) collectivism with 13 (e.g., “It is important to maintain harmony within my group;” Cronbach’s α = .73 for Mexicans and α = .76 for Americans). These moderate alpha coefficients are typical for this measure (Singelis et al., 1995). Items in these subscales were evaluated on a 6-point scale ranging from 1 (disagree strongly) to 6 (agree strongly).

Results

Means, SDs, and mean differences for key variables in both the countries are shown in Table 3, along with t tests comparing the country means. Attachment anxiety scores were significantly higher in the American sample than in the Mexican sample, as in Study 1, whereas attachment-related avoidance scores were only nonsignificantly higher in the American sample. (In Study 1, this difference was significant.) As in Study 1, the results suggest that Mexican participants were more secure, on average, than American participants. Also as in Study 1, Americans scored higher on passive acceptance of negative emotions, diversion or distraction, negative beliefs about others’ socially supportive behavior, and negative emotional reactions to social support. Mexicans scored significantly higher on reappraisal coping, seeking social support, perception of receiving social support, and positive emotional reactions to social support. Of special interest in Study 2, Mexicans scored significantly higher on individualism and Americans scored significantly higher on collectivism, the opposite of what might be expected based on cultural stereotypes. Thinking that this might be attributable to the number of Asian Americans in the US sample, we checked to see whether Asian and non-Asian participants differed from the other American participants. In fact, there were no significant differences in individualism or collectivism between Asian and non-Asian participants in the US sample: individualism t (247) = .80, p = .42 (M = 4.05, SD = .52 for Asians; M = 4.11, SD = .55 for non-Asians); collectivism t (247) = 1.35, p = .17 (M = 4.31, SD = .50 for Asians; M = 4.21, SD = .53 for non-Asians).

Table 3.

Means and SDs for key variables in both countries and t tests comparing means between countries.

	Mexico			US			t	df	p*
	Mean	SD	N	Mean	SD	N	t	df	p*
Anxiety	3.60	1.09	158	3.98	1.06	248	3.46	404	.00
Avoidance	3.21	.96	158	3.38	.94	248	1.74	404	.08
Reappraisal	4.24	.62	158	4.04	.70	248	2.96	404	.00
Passive acceptance	2.47	.57	158	3.15	.57	248	11.63	404	.00
Diversion	2.47	.52	158	3.07	.58	248	10.57	404	.00
Positive beliefs about OSSB	4.29	.67	158	4.23	.65	248	.97	404	.33
Negative beliefs about OSSB	3.79	1.04	158	4.11	.79	248	3.36	270.60	.00
Seeking social support	4.55	.76	158	4.25	.81	248	3.80	404	.00
Perception of receiving social support	5.08	.82	158	4.60	.82	248	5.60	404	.00
Positive emotional reactions to social support	5.36	.64	158	4.82	.70	248	7.92	355.44	.00
Negative emotional reactions to social support	2.57	1.10	158	3.32	.94	248	7.25	404	.00
Individualism	4.48	.59	158	4.09	.54	248	6.78	404	.00
Collectivism	3.99	.62	158	4.25	.52	248	4.41	404	.00

OSSB: other’s socially supportive behavior.

*Probability levels are rounded to two decimal places.

Bivariate correlations between the main study variables are shown in Table 4. Attachment anxiety was correlated with avoidance, passive acceptance of negative emotions, diversion, negative beliefs about others’ socially supportive behavior, seeking social support, negative reactions to social support, and collectivism. Attachment anxiety was also inversely correlated with perception of receiving social support. Attachment-related avoidance was significantly correlated with passive acceptance, diversion, negative reactions to receiving social support, and individualism. Avoidance was also inversely correlated with reappraisal, positive beliefs about others’ socially supportive behavior, seeking social support, perception of receiving social support, positive reactions to social support, and collectivism. Regarding the two cultural syndromes, individualism was significantly correlated with avoidance (as just mentioned), reappraisal, negative beliefs about others’ socially supportive behavior, and positive reactions to receiving social support. Collectivism was significantly correlated with the attachment dimensions and all of the outcome variables. Individualism and collectivism were not significantly correlated with each other.

Table 4.

Pearson correlations between main study variables.

	1	2	3	4	5	6	7	8	9	10	11	12	13
1. Anxiety	1	.21**	−.08	.30**	.27**	−.01	.19**	.19**	−.19**	−.11	.38**	.00	.24**
2. Avoidance		1	−.43**	.19**	.22**	−.24**	−.01	−.32**	−.43**	−.36**	.33**	.13**	−.13**
3. Reappraisal			1	−.05	−.07	.29**	.05	.39**	.46**	.50**	−.19**	.13**	.26**
4. Passive acceptance				1	.97**	−.02	.19**	−.06	−.26**	−.29**	.46**	−.09	.16**
5. Diversion					1	−.03	.15**	−.11	−.28**	−.30**	.45**	−.08	.15**
6. Positive beliefs OSSB						1	−.00	.29**	.34**	.38**	−.10	.02	.24**
7. Negative beliefs OSSB							1	.10	.01	−.00	.23**	.16**	.14**
8. Seeking social support								1	.61**	.52**	−.17**	.07	.20**
9. Perception of SS									1	.70**	−.33**	.07	.17**
10. Positive reactions to SS										1	−.34**	.13**	.26**
11. Negative reactions to SS											1	.03	.14**
12. Individualism												1	.03
13. Collectivism													1

SS: social support; OSSB: other’s socially supportive behavior.

**p < .01.

As in Study 1, hierarchical regression analyses were conducted to determine how the attachment insecurity variables and country, considering both main effects and interactions, were related to the coping and social support variables (the steps of the analysis are shown in Table 5). First, regression analyses were computed that included the attachment insecurity dimensions, country, and interactions between them as predictors. Most of the results from Study 1 were replicated in Study 2 (see Table 5). As in Study 1, attachment anxiety significantly predicted passive acceptance of negative emotions, diversion, seeking social support, and negative reactions to social support. Moreover, in Study 2 anxiety also predicted negative (skeptical) beliefs about others’ socially supportive behavior. As in Study 1, avoidance predicted reappraisal, diversion, seeking social support, perception of receiving social support, and positive and negative reactions to receiving social support. In Study 2, avoidance was also inversely related to positive beliefs about others’ socially supportive behavior.

Table 5.

Standardized regression coefficients predicting social support and coping.

	Reappraisal	Passive acceptance	Diversion	Positive beliefs about OSSB	Negative beliefs about OSSB	Seeking social support	Perception of receiving SS	Positive reactions to receiving SS	Negative reactions to receiving SS
Step 1	Ad R ² = .26	Ad R ² = .30	Ad R ² = .27	Ad R ² = .10	Ad R ² = .10	Ad R ² = .22	Ad R ² = .28	Ad R ² = .33	Ad R ² = .28
Anxiety	.01	.21**	.17**	.03	.17**	.26***	.09	−.03	.21***
Avoidance	−.41***	.09	.16**	−.22**	−.10	−.42***	−.46***	−.33***	.24***
Country	.11	−.47***	−.43***	.06	−.25**	.21***	.25***	.39***	−.31***
Individualism	.12**	.02	.01	.00	.24**	.04	.01	.01	.07
Collectivism	.25***	.03	.04	.23**	.03	.13**	.19***	.33***	.05
Step 2	Ad R ² =.28	Ad R ² =.31	Ad R ² =.28	Ad R ² =.09	Ad R ² =.10	Ad R ² =.22	Ad R ² =.31	Ad R ² =.34	Ad R ² =.29
Anxiety × country	−.11	.02	−.00	−.07	−.00	−.02	−.03	−.06	.17**
Avoidance × country	.00	.00	−.02	−.02	.06	.10	.12	.06	−.03
Anxiety × individualism	.12**	.19**	.16**	.12	.06	.07	.02	.04	.04
Anxiety × collectivism	.07	.07	.10	.04	−.02	−.08	−.01	.02	.07
Avoidance × individualism	.01	−.09	−.11	−.13	.06	−.00	−.01	.01	.01
Avoidance × collectivism	−.03	.06	.06	−.02	−.02	.07	.22***	.15***	.04
Step 3	Ad R ² =.27	Ad R ² =.32	Ad R ² =.29	Ad R ² =.12	Ad R ² =.10	Ad R ² =.22	Ad R ² =.31	Ad R ² =.35	Ad R ² =.31
Anxiety × country × individualism	.01	−.19**	−.14	−.15	−.10	−.01	−.01	.01	−.23***
Anxiety × country × collectivism	−.00	−.02	−.05	−.04	.04	.05	−.11	−.12	.02
Avoidance × country × individualism	.02	.03	.03	.23**	.00	.02	−.05	.03	.05
Avoidance × country × collectivism	.04	−.11	−.11	−04	.00	.01	.06	.13	−.11

SS: social support; OSSB: other’s socially supportive behavior; Ad R ²: explained variance.

**p <.01, ***p <.001.

As in Study 1, there were significant country effects on passive acceptance and diversion, with Americans scoring higher, and in perception of receiving social support and positive reactions to receiving social support, with Mexicans scoring higher. In Study 2, there were also country effects on negative beliefs about others’ socially supportive behavior and negative reactions to receiving social support, with Americans scoring higher, and on seeking social support, with Mexicans scoring higher. These results, like the results of Study 1, suggest that Mexicans cope with problems more constructively and experience social support more positively.

There were no significant interactions between attachment insecurities and country in Study 1, but in Study 2, anxiety and country interacted significantly to predict negative reactions to receiving social support. A simple slopes test revealed that anxiety predicted negative reactions to receiving social support somewhat better in Mexico (β = .34, p < .001) than in the US (β = .27, p < .001), but the positive direction and significance levels of the β-coefficients were similar. There were no other significant interactions between the attachment scales and country.

Next, we added individualism and collectivism, and interaction terms including those variables, to the regression model. Individualism predicted reappraisal and negative beliefs about others’ socially supportive behavior; collectivism predicted reappraisal, positive beliefs about others’ socially supportive behavior, seeking social support, perception of being socially supported, and positive reactions to receiving social support (see Table 5). All significant effects of the attachment dimensions and country remained significant when individualism and collectivism were in the model. Additionally, the effect of country on reappraisal became significant when the cultural variables were included. These results suggest that country differences cannot be explained by differences in individualism and collectivism.

There were also significant interactions between the attachment insecurity variables and the cultural variables (Table 5). We used simple slopes tests to examine the significant interactions, following the guidelines proposed by Aiken and West (1991). The interaction of anxiety and individualism predicting reappraisal was significant. A simple slopes test revealed that attachment anxiety significantly predicted reappraisal (inversely) when individualism levels were high (β = −.16, p = .02) but not when individualism levels were low (β = −.02, p = .69; Figure 1).

Figure 1.

Interaction between attachment anxiety and individualism predicting the use of reappraisal coping strategies. The units on the Y-axis are on standardized scores.

The interaction between anxiety and individualism predicted passive acceptance of negative emotions; a simple slopes test revealed that in line with our expectations, attachment anxiety predicted passive acceptance coping better when individualism levels were high (β = .39, p < .001) than when individualism levels were low (β = .18, p = .008), although both associations were in the same direction and both were significant (Figure 2). Finally, there was a significant interaction between anxiety and individualism when predicting diversion as a coping strategy. A simple slopes test revealed that, as expected, anxiety predicted diversion coping better when individualism levels were high (β = .35, p < .001) than when individualism levels were low (β = .16, p = .022) (Figure 3), although again both β-coefficients were significantly positive.

Figure 2.

Interaction between attachment anxiety and individualism predicting the use of passive acceptance coping strategies. The units on the Y-axis are on standardized scores.

Figure 3.

Interaction between attachment anxiety and individualism predicting the use of diverting coping strategies. The units on the Y-axis are on standardized scores.

There was a significant interaction between attachment-related avoidance and collectivism predicting perception of receiving social support. A simple slopes test revealed that in line with our hypothesis, avoidance predicted perception of receiving social support better when collectivism levels were low (β = −.60, p < .001) than when collectivism levels were high (β = −.25, p < .001; Figure 4), although both coefficients were significantly negative. There was also a significant interaction between attachment-related avoidance and collectivism predicting positive reactions to perceived reception of social support. A simple slopes test revealed that, as expected, attachment-related avoidance was inversely associated with positive reactions to receiving social support more strongly when collectivism levels were low (β = −.46, p < .001) than when collectivism levels were high (β = −.21, p < .001) (Figure 5).

Figure 4.

Interaction between attachment-related avoidance and collectivism predicting perception of receiving social support. The units on the Y-axis are on standardized scores.

Figure 5.

Interaction between attachment-related avoidance and collectivism predicting positive reactions to receiving social support. The units on the Y-axis are on standardized scores.

There were also three significant three-way interactions (Table 5). First, there was an interaction between anxiety, country, and individualism in predicting passive acceptance of negative emotions. To examine the form of the interaction, we conducted separate regression analyses for each country, checking to see if the two-way interaction between anxiety and individualism was significant. It proved to be significant in the US (β = .18, p = .002) but not in Mexico (β = −.07, p = .371). A simple slopes test indicated that, in the US, anxiety significantly predicted passive acceptance when individualism levels were high (β = .44, p < .001) but not when individualism levels were low (β = .08, p = .283). The second significant three-way interaction involved the same three variables—anxiety, country, and individualism—in predicting negative reactions to receiving social support. Breaking this down by country, we found that the anxiety by individualism interaction was significant in Mexico (β = −.23, p = .005) but not in the US (β = .05, p = .317). A simple slopes test indicated that, in Mexico, unexpectedly, anxiety was a better predictor of negative reactions to receiving social support when individualism levels were low (β = .67, p < .001) than when individualism levels were high (β = .27, p = .001), although both β-coefficients were significantly positive. Finally, there was a three-way interaction between avoidance, country, and individualism in predicting positive beliefs about others’ socially supportive behavior. Further analyses revealed that the interaction between avoidance and individualism was significant in predicting these positive beliefs in Mexico (β = .22, p = .008) but not in the US (β = −.09, p = .115). A simple slopes test revealed that, in Mexico, unexpectedly, avoidance was inversely related to positive beliefs about others’ socially supportive behavior when individualism levels were low (β = −.52, p < .001) but not when individualism levels were high (β = −.09, p = .304).

Discussion

Means of individualism and collectivism by country

In this study, American students were more collectivistic and Mexican students were more individualistic (when measured with Singelis et al.’s, 1995, subscales). For a number of reasons, it is possible that collectivism is lower among college students in Mexico than in the broader Mexican population. University education may reduce adherence to collectivist norms and values (Mishra, 1994; Oyserman et al., 2002) because it encourages forming personal opinions and achieving individual success. It also provides opportunities to join new groups with different norms and values than those held by one’s family of origin (Triandis et al., 1990). College students also tend to be higher in socioeconomic status than nonstudents, and this may be associated with greater individualism and lower collectivism (Hofstede, 2001; Triandis et al., 1990).

Although previous studies have found that some Mexican students score relatively high on collectivism and relatively low on individualism, our results are compatible with the fact that urban residence may weaken, or condition, collectivistic orientations and enhance individualistic orientations (Matsumoto, Kudoh, & Takeuchi, 1996; Mishra, 1994). For example, Shkodriani and Gibbons (1995) reported higher levels of collectivism in a sample of Mexican students, when compared with a group of American students, but the Mexican participants were residents of Puebla, a provincial, somewhat conservative city (about 7% the size of Mexico City, where our study was conducted). Despite the effect of university education on increasing individualism and decreasing collectivism, students living in smaller cities may be different from students living in Mexico City.

Interactions of attachment insecurities, individualism and collectivism, and country

Some of the interactions between attachment variables and cultural syndromes were as expected. In the case of reappraisal, passive acceptance of negative emotions, and using diversion as a coping strategy, the association between these variables and attachment anxiety was stronger, as expected, when individualism was higher. Moreover, in the case of the perception of receiving social support and reacting to social support received, the associations between these variables and avoidance were stronger when collectivism was lower. These results are in line with our idea that the attachment variables might have stronger effects on coping and social support when individualism levels were high and collectivism levels were low. There were, however, a number of other interactions involving individualism and collectivism that did not prove significant, and one (the association between attachment anxiety and reappraisal) that was opposite to our prediction. Although we do not have a confident explanation of this pattern of results, the coping and support variables that worked as expected are the ones that other studies have found to be most closely related to individualism and collectivism, such as duty to the group (Oyserman et al., 2002) and passive (or acceptance) coping (Díaz-Guerrero, 1967). This possibility should be followed up in future research.

Of special interest, the unexpected results occurred mainly in Mexico. For example, although attachment anxiety significantly predicted passive acceptance of negative emotions when individualism levels were higher in the US, there was almost no difference in Mexico. This is compatible with claims in the literature (Díaz-Guerrero, 1967) that active coping is more common in the US than in Mexico, although this was not the case in our university samples. Perhaps, American students’ attachment anxiety causes them to adopt coping strategies that are not the common ones in their country. Attachment anxiety was a better predictor of negative reactions to receiving social support when individualism levels were low only in Mexico. This may have occurred because students in Mexico expect to be socially supported, and the more anxious and less individualistic students may not be coping well autonomously and are especially upset that other people are not helping them enough (Kim et al., 2006). Similarly, attachment-related avoidance was inversely related to positive beliefs about others’ socially supportive behavior when individualism levels were low only in Mexico, where people generally expect to be socially supported.

General discussion

Our main hypothesis was that the association between attachment dimensions and coping and social support would be weaker in a sample of high collectivistic and low individualistic university students, when compared with less collectivistic and more individualistic university students. Results from both studies moderately supported this prediction. Since most studies of adult attachment have been conducted in the US, we first had to demonstrate that the basic associations between attachment and coping and social support are significant in a sample of non-English-speaking Mexican participants.

Both studies provided support for hypotheses derived from attachment theory in both countries. Attachment anxiety was expected to be associated with a general pattern of hyperactivation of the attachment system. And in fact, more anxious participants reported that when coping with stress, they tended to be overwhelmed by the emotional consequences of the stressful event and to divert their attention away from the problem. When being stressed (and probably even when not being stressed; Mikulincer, Gillath, & Shaver, 2002) they sought support from others, and although this should generally lead to a reduction in distress, anxious individuals remained upset.

Avoidant participants’ self-reports suggested, as expected, a pattern of deactivation of the attachment system. They diverted attention from problems and tended not to use reappraisal as a coping strategy, which would require acknowledging threats that needed to be reappraised. They neither sought social support nor perceived that they were being supported. When their need for social support was expressed, and they received support from others, they did not experience themselves as benefitting from it and felt that their sense of independence had been compromised. These patterns were uniform across the two countries, as expected from Bowlby’s (1982) claim that the attachment behavioral system is an evolutionary adaptation common to all human beings. (For further evidence of cross-cultural universality of basic attachment processes, see van IJzendoorn & Sagi-Schwartz, 2008.)

Although Triandis (1995) claimed that individualism and collectivism are not confined within particular geographical borders, in cross-cultural studies it has been common to assume that people from a particular society tend to show similar levels of these cultural variables. Based on the previous literature, in Study 1, we expected Mexican students to be more collectivistic and less individualistic than US students. If that were the case, and if individualism enhances and collectivism decreases the effect of attachment insecurities on our dependent variables, we would expect such effects to be weaker in the Mexican sample. Results from this study did not provide support for our expectation, however, and in general the ability of attachment anxiety and avoidance to predict coping and social support variables was similar in the Mexican and US student samples.

Besides country, other factors can affect the levels of individualism and collectivism, such as education level and urban residence (Mishra, 1994), and these two variables distinguish Mexico City students from the Mexican norm. According to the Instituto Nacional de Estadística, Geografía e Informática (Mexican National Institute for Statistics, Geography, and Computing) (INEGI, 2011), the average Mexican has 8.1 years of formal education, compared with at least 12 years of formal education for university students.

Therefore, we replicated and extended Study 1 by adding individualism and collectivism measures to assess these cultural characteristics at the level of the participants in our study. Results from Study 2 showed that Mexican students scored higher on individualism and lower on collectivism than US students. Other studies have found similar unexpected patterns; for example, Matsumoto, Kudoh, and Takeuchi (1996) compared university students from Japan and the US and found that the Japanese students scored higher on individualism and the American students scored higher on collectivism.

Moreover, when comparing Japanese university students and working adults, students scored higher on individualism than working adults, especially when interacting with colleagues (rather than friends and family members). Matsumoto et al. argued that culture is not a rigid or static entity, but a dynamic complex of forces within large societies that can change over time and over situations and roles. They suggested that changes in family affluence, communication with other groups, and mobility within and between countries contribute to changing adherence to particular cultural syndromes. We were able to test our main hypothesis about differences in the predictive power of attachment insecurities as affected by individualism and collectivism, regardless of country.

There were a few significant effects of individualism and collectivism, in the expected directions. For participants in both the countries who scored higher on individualism, the associations between attachment anxiety, on the one hand, and using diversion coping but not reappraisal were stronger. Also in both countries, for participants who scored lower on collectivism, the inverse associations between avoidance, on the one hand, and perception of receiving social support and positive reactions to social support, on the other, were stronger. These results are in line with our hypothesis that the ability of attachment insecurities (which are within-individual processes) to predict coping and social support would be enhanced by individualism and decreased by collectivism, because defining oneself as independent and giving priority to personal goals over in-group goals, as assessed by the Singelis, Triandis, Bhawuk, and Gelfand (1995) measure, should increase the influence of personality variables on emotion regulation strategies.

However, in most cases, individualism and collectivism did not significantly affect the basic associations between attachment insecurities and our dependent variables, suggesting that the effects of attachment insecurities on coping and social support were uniform in student samples from Mexico and the US. Our results show that both individualistic and collectivistic individuals who are insecure with respect to attachment hold negative expectations about others’ motives for providing support, and these expectations affect their seeking of social support. However, the collectivistic norm to support others promotes a sense being helped and feeling good about it, despite these negative attachment-related expectations. On the other hand, the lack of a sense of personal control, related to collectivism, may also promote the use of diversion coping strategies, regardless of attachment insecurities. This and other alternative explanations should be tested in future research. Finally, in some cases, associations involving individualism and collectivism were moderated by country, suggesting that holding atypical values for one’s culture may affect an individual’s ability to relate positively to others and cope with stress in healthy ways.

In sum, many aspects of adult attachment theory (Mikulincer & Shaver, 2007) were supported in student samples from both the US and Mexico. There were, however, some interactions between variables that involved differences in individualism or collectivism, and some interactions with country that could not be accounted for by individualism and collectivism, at least as measured in our study. The results are generally compatible with van IJzendoorn and Sagi-Schwartz’s (2008) summary of the cross-cultural literature on children’s attachment to caregivers: enough of the results are similar across cultures to support Bowlby’s (1982) claim that attachment is a psychobiological process that occurs everywhere, and that individual differences in attachment patterns affect other psychological processes, such as emotion regulation, everywhere. But there are also some effects of culture, as would be expected in the case of any universal psychobiological process (e.g., mating) that is affected by cultural values and rules.

The two studies reported here have some limitations. All of the data were based on self-report measures, and the two samples of students may not be representative of all similarly aged adults in Mexico and the US. In fact, we would expect college students to be different in various ways, including education and socioeconomic status, but also in personality characteristics, than noncollege adults of the same age. Nevertheless, it is important to study these potentially influential members of their respective societies. In both countries, attachment insecurities affect, or are associated with, other potentially important variables that will influence mental health and relationship quality across the life span (Mikulincer & Shaver, 2007). Future studies should examine in more detail national characteristics that are not well captured by existing measures of individualism and collectivism, and examine how individual differences affect actual behavior, not just self-perceptions and self-reports.

Footnotes

Funding

This research was funded in part by the Mexican National Council for Science and Technology (Consejo Nacional de Ciencia y Tecnología, CONACyT; to M.T.F.C.).

References

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Allen

(2008). The attachment system in adolescence. In Cassidy

Shaver

P. R.

(Eds.), Handbook of attachment: Theory, research, and clinical applications (2nd ed., pp. 419–435). New York, NY: Guilford Press.

Besser

Neria

(2011). When home isn’t a safe heaven: Insecure attachment orientations, perceived social support, and PTSD symptoms among Israeli evacuees under missile threat. Psychological Trauma: Theory, Research, Practice, and Policy, 4, 34–46.

Bontempo

Rivero

(1992). Cultural variation in cognition: The role of self-concept in the attitude-behavior link. Paper presented at the meeting of the American Academy of Management, Las Vegas, NV.

Bowlby

(1973). Attachment and loss: Vol. 2. Separation: Anxiety and anger. New York, NY: Basic Books.

Bowlby

. (1980). Attachment and loss: Vol. 3. Loss: Depression and sadness. New York, NY: Basic Books.

Bowlby

(1982). Attachment and loss: Vol. 1. Attachment (2nd ed.). New York, NY: Basic Books.

Brennan

K. A.

Clark

C. L.

Shaver

P. R.

(1998). Self-report measurement of adult attachment: An integrative overview. In Simpson

J. A.

Rholes

W. S.

(Eds.), Attachment theory and close relationships (pp. 46–76). New York, NY: Guilford Press.

Brewer

M. B.

Chen

Y. U.

(2007). Where (whow) are collectives in collectivism? Toward conceptual clarification of individualism and collectivism. Psychological Review, 114, 133–151.

10.

Campbell

Simpson

J. A.

Boldry

Kashy

D. A.

(2005). Perceptions of conflict and support in romantic relationships: The role of attachment anxiety. Journal of Personality and Social Psychology, 88, 510–531.

11.

Cassidy

(1994). Emotion regulation: Influences of attachment relationships. In Fox

(Ed.), The development of emotion regulation. Monographs of the Society for Research in Child Development, 59, 228–283.

12.

Cassidy

Kobak

R. R.

(1988). Avoidance and its relationship with other defensive processes. In Belsky

Nezworski

(Eds.), Clinical implications of attachment (pp. 300–323). Hillsdale, NJ: Erlbaum.

13.

Cassidy

Shaver

P. R.

(Eds.). (2008). Handbook of attachment: Theory, research, and clinical applications (2nd ed.). New York, NY: Guilford Press.

14.

Chao

(2001). Integrating culture and attachment. American Psychologist, 56, 822–823.

15.

Collins

N. L.

Feeney

B. C.

(2004). Working models of attachment shape perceptions of social support: Evidence from experimental and observational studies. Journal of Personality and Social Psychology, 87, 363–383.

16.

Collins

Guichard

Ford

Feeney

(2006). Responding to need in intimate relationships: Normative processes and individual differences. In Mikulincer

Goodman

(Eds.), Dynamics of romantic love: Attachment, caregiving, and sex (pp. 149–189). New York, NY: Guilford Press.

17.

Collins

Read

S. J.

(1990). Adult attachment, working models, and relationship quality in dating couples. Journal of Personality and Social Psychology, 58, 644–663.

18.

Davidson

A. R.

Jaccard

J. J.

Triandis

H. C.

Morales

M. L.

Díaz-Guerrero

(1976). Cross-cultural model testing: Toward a solution of the etic–emic dilemma. International Journal of Psychology, 11, 1–13.

19.

Díaz-Guerrero

(1967). The active and passive syndromes. Revista Interamericana de Psicología, 1, 263–272.

20.

Díaz-Loving

(2005). Emergence and contributions of a Latin American indigenous social psychology. International Journal of Psychology, 40, 213–227.

21.

Freeman

(1997). Demographic correlates of individualism and collectivism: A study of social values in Sri Lanka. Journal of Cross-Cultural Psychology, 28, 321–341.

22.

Frías

Díaz-Loving

(2011). Desarrollo y validación de una escala para medir apoyo social en adultos con perspectiva de apego. Revista Iberoamericana de Diagnóstico y Evaluación Psicológica, 32, 9–34.

23.

Friedman

Rholes

W. S.

Simpson

J. A.

Bond

Díaz-Loving

Chan

(2010). Attachment avoidance and the cultural fit hypothesis: A cross-cultural investigation. Personal Relationships, 17, 107–126.

24.

Gjerde

P. F.

(2001). Attachment, culture, and amae. American Psychologist, 56, 827–827.

25.

Góngora

Reyes Lagunes

(1998). El enfrentamiento a los problemas y el locus de control. In La Psicología Social en México (pp. 165–172). Mexico City, Mexico: Asociación Mexicana de Psicología Social.

26.

Hofstede

(2001). Culture’s consequences: Comparing values, behaviors, institutions, and organizations across nations. Thousand Oaks, CA: Sage.

27.

INEGI. (2011, February 7th). II Conteo the Población y Vivienda 2005. Retrieved from http://www.inegi.org.mx/Sistemas/temasV2/Default.aspx?s=est&c=21702

28.

Kagitcibasi

(1997). Individualism and collectivism. In Berry

J. W.

Segall

M. H.

Kagitcibasi

(Eds.), Handbook of cross-cultural psychology: Vol 3. Social behavior and applications (pp. 1–49). Boston, MA: Allyn & Bacon.

29.

Kim

H. S.

Sherman

D. K.

Taylor

S. E.

(2006). Pursuit of comfort and pursuit of harmony: Culture, relationships, and social support seeking. Personality and Social Psychology Bulletin, 32, 1595–1607.

30.

Kim

Triandis

H. C.

Kagitçibasi

Choi

S. C.

Yoon

(1994). Individualism and collectivism: Theory, method, and applications. Newbury Park, CA: Sage.

31.

Kobak

R. R.

Cole

H. E.

Ferenz-Gillies

Fleming

W. S.

Gamble

(1993). Attachment and emotion regulation during mother-teen problem solving: A control theory analysis. Child Development, 64, 231–245.

32.

Kondo-Ikemura

(2001). Insufficient evidence. American Psychologist, 56, 825–826.

33.

Kuo

B. C. H.

(2010). Culture’s consequences on coping: Theories, evidences, and dimensionalities. Journal of Cross-Cultural Psychology, 42, 1084–1100.

34.

Main

(1995). Attachment: Overview, with implications for clinical work. In Goldberg

Muir

Kerr

(Eds.), Attachment theory: Social, developmental, and clinical perspectives. Hillsdale, NJ: Analytic Press.

35.

Markus

H. R.

Kitayama

(1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224–253.

36.

Matsumoto

Kudoh

Takeuchi

(1996). Changing patterns of individualism and collectivism in the United States and Japan. Culture & Psychology, 2, 77–107.

37.

Mikulincer

(1998). Adult attachment styles and individual differences in functional versus dysfunctional experiences of anger. Journal of Personality and Social Psychology, 74, 513–524.

38.

Mikulincer

Florian

(1995). Appraisal and coping with a real-life stressful situation: The contribution of attachment styles. Personality and Social Psychology Bulletin, 21, 406–414.

39.

Mikulincer

Florian

Weller

(1993). Attachment styles, coping strategies, and posttraumatic psychological stress: The impact of the Golf War in Israel. Journal of Personality and Social Psychology, 64, 817–826.

40.

Mikulincer

Gillath

Shaver

P. R.

(2002). Activation of the attachment system in adulthood: Threat-related primes increase the accessibility of mental representations of attachment figures. Journal of Personality and Social Psychology, 83, 881–895.

41.

Mikulincer

Shaver

P. R.

(2003). The attachment behavioral system in adulthood: Activation, psychodynamics, and interpersonal processes. In Zanna

M. P.

(Ed.), Advances in Experimental Social Psychology, 35, 53–152.

42.

Mikulincer

Shaver

P. R.

(2007). Attachment in adulthood: Structure, dynamics, and change. New York, NY: Guilford Press.

43.

Mikulincer

Shaver

P. R.

(2009). An attachment and behavioral systems perspective on social support. Journal of Social and Personal Relationships, 26, 7–19.

44.

Mishra

(1994). Individualist and collectivist orientations across generations. In Kim

Triandis

H. C.

Kagitcibasi

Choi

S. C.

Yoon

. (Eds.), Individualism and collectivism: Theory, method, and applications (pp. 225–238). Thousand Oaks, CA: Sage.

45.

Oyserman

Coon

H. M.

Kemmelmeier

(2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128, 3–72.

46.

Pierrehumbert

Santelices

M. P.

Ibañez

Alberdi

Ongari

Roskam

… Borghini

. (2009). Gender and attachment representations in the preschool years: Comparisons between five countries. Journal of Cross-Cultural Psychology, 40, 543–566.

47.

Posada

Jacobs

(2001). Child–mother attachment relationships and culture. American Psychologist, 56, 821–822.

48.

Rothbaum

Kakinuma

Nagaoka

Azuma

(2007). Attachment and amae: Closeness in the United States and Japan. Journal of Cross-Cultural Psychology, 38, 465–486.

49.

Rothbaum

Weisz

Pott

Miyake

Morelli

(2000). Attachment and culture: Security in the United States and Japan. American Psychologist, 55, 1093–1104.

50.

Sastry

Ross

C. E.

(1998). Asian ethnicity and the sense of personal control. Social Psychology Quarterly, 61, 101–120.

51.

Schmitt

D. P.

Alcalay

Allensworth

Allisk

Ault

Austers

… ZupanÈiÈ

, (2004). Patterns and universals of adult romantic attachment across 62 cultural regions: Are models of self and of other pancultural constructs? Journal of Cross-Cultural Psychology, 35, 367–402.

52.

Shaver

P. R.

Mikulincer

(2002). Attachment-related psychodynamics. Attachment and Human Development, 4, 133–161.

53.

Shaver

P. R.

Mikulincer

Chun

D. S.

(2008). Adult attachment theory, emotion regulation, and prosocial behavior. In Vandekerckhove

von Scheve

Ismer

Jung

Kronast

(Eds.), Regulating emotions: Culture, social necessity, and biological inheritance (pp. 121–145). Malden, MA: Blackwell.

54.

Shkodriani

Gibbons

(1995). Individualism and collectivism among university students in Mexico and the United States. The Journal of Social Psychology, 135, 765–772.

55.

Singelis

T. M.

Triandis

H. C.

Bhawuk

Gelfand

M. J.

(1995). Horizontal and vertical dimensions of individualism and collectivism: A theoretical and measurement refinement. Cross-Cultural Research, 29, 240–275.

56.

Sinha

J. B. P.

Verma

(1994). Social support as a moderator of the relationship between allocentrism and psychological wellbeing. In Kim

Triandis

H. C.

Kagotcibasi

Choi

S. C.

Yoon

(Eds.), Individualism and collectivism: Theory, method, and applications (pp. 267–275). Thousand Oaks, CA: Sage.

57.

Soon

Malley-Morrison

(2000). Young adult attachment styles and intimate relationships with close friends: A cross-cultural study of Koreans and Caucasian Americans. Journal of Cross-Cultural Psychology, 31, 528–534.

58.

Triandis

H. C.

(1990). Cross-cultural studies of individualism and collectivism. Nebraska Symposium on Motivation: Cross-cultural perspectives (pp. 41–133). Lincoln, NE: University of Nebraska.

59.

Triandis

H. C.

(1993). Collectivism and individualism as cultural syndromes. Cross-Cultural Research, 27, 155–180.

60.

Triandis

H. C.

(1995). Individualism and collectivism. Boulder, CO: Westview Press.

61.

Triandis

H. C.

McCusker

Hui

C. H.

(1990). Multimethod probes of individualism and collectivism. Journal of Personality and Social Psychology, 59, 1006–1020.

62.

van IJzendoorn

M. H.

Sagi-Schwartz

(2008). Cross-cultural patterns of attachment: Universal and contextual dimensions. In Cassidy

Shaver

P. R.

(Eds.), Handbook of attachment: Theory, research, and clinical applications (2nd ed., pp. 880–905). New York, NY: Guilford Press.

63.

Waterman

A. S.

(1984). The psychology of individualism. New York, NY: Praeger.

64.

Waters

H. S.

Waters

(2006). The attachment working models concept: Among other things, we build script-like representations of secure base experiences. Attachment & Human Development, 8, 185–197.