Susceptibility to peer pressure and attachment to friends

The role of attachment to friends in the explanation of adolescents’ susceptibility to peer pressure was explored, regarding the way these two constructs are measured. In Study 1, 475 high school students (194 boys and 281 girls) were given Susceptibility to Peer Pressure Questionnaire, and their attachment to friends was measured with Modified Experiences in Close Relationships Inventory and Relationship Scales Questionnaire. One month later, 80 boys and 80 girls participated in Study 2, where they completed the same Susceptibility to Peer Pressure Questionnaire in a chat-room simulation, convinced that they can see other students’ answers and that their own answers could be seen by others. When susceptibility to peer pressure was measured by self-report questionnaire, the level of avoidance proved to be a significant predictor for boys, while the level of anxiety and the model of others were significant predictors for girls. When susceptibility to peer pressure was measured experimentally, the results showed that attachment dimensions predict only girls’ susceptibility and that the only significant predictor is their model of others.

Over the past four decades, researchers have almost completely neglected susceptibility to peer pressure as a process.Past research dominantly focused on situations in which peer pressure occurs or on individual differences of those adolescents who conform to or resist the peer pressure (Brown, Bakken, Ameringer, & Mahon 2008).Therefore, Brown et al. (2008) consolidate results from previous research in the field and, based on these findings, the authors propose conceptual model of the peer influence process.The proposed model of peer influence represents the first theoretical framework in the field of adolescents peer influence and within this model it is possible to observe different modes of peer influence as well as of peer pressure.
The conceptual model of peer influence process assumes that some events trigger peer pressure on group or dyad member, and an adolescent can respond to that pressure in various ways.Adolescents respond to the peer pressure by accepting it and conforming to their peer's norms, expectations or demands or ignoring it, and by confronting it with a counter influence.When this basic sequence is activated, there are numerous factors that determine an adolescent's reaction to peer pressure.Brown et al. (2008) group all these variables into six categories.Individual and contextual factors have direct effect on adolescent's response to peer pressure.Out of all possible individual differences, the most likely to be included in research are gender and age.Contextual factors refer to broader social context such as school, neighborhood or community.On the other hand, relation between individual or contextual factors and susceptibility to peer pressure can be moderated with four factors -adolescent's openness to influence, salience of peers who exert pressure, adolescent's ability or opportunity to perform, and relationship dynamics.
Openness to influence refers to adolescent's perception of peer's behavior as urging and encouraging him or her to act in expected way but also it refers to adolescent's disposition to conform to peer pressure in terms of personality trait.Studies have mainly focused on gender and age differences in susceptibility to peer pressure.However, in these studies susceptibility to peer pressure has been treated as expected behavior reaction to a specific situation rather than a personality trait.These findings have shown that boys are more prone to engage in deviant behavior as a response to peer pressure (e.g.Brown, Clasen, & Eicher, 1986;Lebedina Manzoni, Lotar, & Ricijaš, 2008a;Lotar, 2011a;Pardini, Loeber, & Stouthamer-Loeber, 2005;Sim & Koh, 2003;Steinberg & Silverberg, 1986) and also that adolescents' openness to influence decreases with age (e.g.Brown, Clasen, & Eicher, 1986;Steinberg & Monahan, 2007).
Relationship dynamics, as one of the moderators between individual factors and adolescent's response to peer pressure, includes several different indicators and can be operationalized in numerous ways.For example, nature of the relationship could be determined as friendship quality, relationship duration or as power dynamics between adolescents and their peers.Research findings have shown that adolescents are more likely to conform to acquaintances than to close friends (Burton, Ray, & Mehta, 2003;Lightfoot, 1997).Relationship dynamics could also be determined by adolescent's attachment to a friend.Baumeister and Leary (1995) emphasize the importance of the need to belong as fundamental human motivation for the whole line of different behaviors.The need to belong is manifested through person's desire for interpersonal attachment and frequent contact or interaction with other people.When it comes to disagreement in attitudes and behavior among an adolescent and his/her group of peers, adolescent who conforms to the group is motivated by his/her tendency to avoid confrontation and negative emotions.Peers often exert pressure by threatening with negative emotions as a reaction to mockery, rejection or nonacceptance.Anxiety provoked by the fear of relationship loss can be powerful encouragement for adolescents to change their behavior and adjust it to peers' expectations.However, based on results of research in a field of attachment it is arguably to conclude that adolescents differ according to their need to belong.
While some adolescents have a great need for other people and become anxious by the thought that they could lose important person, there are also those who feel self-sufficient.It is plausible to conclude that people with dismissive attachment style don't have high need to belong as opposed to those with preoccupied attachment style.Accordingly, we can expect the difference in susceptibility to peer pressure among adolescents with those two attachment styles.Allen, Porter, McFarland, McElhaney, & Marsh (2007) had similar hypothesis.Authors assume that there is greater possibility for adolescents with secure attachment to their parents to have formed and maintain friendships with low level of peer pressure.In this study Allen et al. (2007) examined relation between secure attachment to parents and extent to which close friends tried to influence adolescent to engage in negative behaviors (e.g. to pick a fight, smoke, cut classes, make fun of others, get bad grades).Results have shown that attachment security is linked to lower levels of peer pressure experienced by adolescents.One of the possible and probably the most likely explanation of these results is that adolescents with secure attachment style are less susceptible to peer pressure and therefore their friends exert less pressure.The results of the study conducted in Croatia are in favor of such an explanation because they have shown that adolescents with secure attachment style reported lower susceptibility to peer pressure than those with insecure attachment (Lotar, 2011b).
The results of several studies favor the assumption that susceptibility to peer pressure concerning misconduct could be higher among adolescents with insecure attachment.Those studies mostly explore relation between attachment to peers or parents and risky or delinquent behavior.There is no direct examination of the relation between attachment and susceptibility to peer pressure.The attachment studies have pointed out association between adolescents' selfreport of insecure attachment to peers or parents and involvement in delinquent behavior (McElhaney, Immele, Smith, & Allen, 2006;Nelson & Rubin, 1997).Furthermore, insecure attachment to peers or parents proved to be associated with alcohol consumption, cigarette smoking, and drug abuse among adolescents (e.g.Bailey & Hubbard, 1990;DeFronzo & Pawlak, 1993;Lee & Bell, 2003).On sample of Croatian adolescents, Brašnić, Ajduković, & Ručević (2009) have shown that the attachment security in relation with significant adult and with peers is negatively correlated to psychopathic tendencies and self-reported risky and delinquent behavior.In mentioned research, attachment was measured with Inventory of Parent and Peer Attachment (Armsden & Greenberg, 1987).
There are several issues that need to be addressed in research of susceptibility to peer pressure and its determinants.The conceptual model of peer influence process (Brown et al., 2008) resulted from consolidation of prior research results in the field, and it has not been systematically tested so far.Moreover, there are two prominent methodological approaches to studying susceptibility to peer pressure.First line of research on conformity among children and adolescents was inspired by Asch's experiment (Mayeux & Cillessen, 2007).Second line of research started in 1970's and in that research line susceptibility to peer pressure was measured by self-report measures (e.g.Berndt, 1979;Brown, Clasen, & Eicher, 1986;Clasen & Brown, 1985).Recent studies are oriented to experimental methodology again, and those studies (e.g.Cohen & Prinstein, 2006;Prinstein, Brechwald, & Cohen, 2011), in comparison to studies that use self-report measures of susceptibility to peer pressure, show higher susceptibility among adolescents.
The main goal of this research is to examine the role of the attachment to friends in the explanation of adolescents' susceptibility to peer pressure.The other, more specific goal is to find out whether the relation between the adolescents' attachment to friends and their susceptibility to peer pressure differs regarding the way these two constructs are measured.If the attachment to friends is an important determinant of susceptibility to peer pressure, it will be a significant predictor of susceptibility, regardless of the constructs' operationalizations.Considering that the attachment is in recent research usually operationalized through dimensions of anxiety and avoidance (Brennan, Clark, & Shaver, 1998) or through model of self and model of others (Bartholomew, 1990), in this research the attachment is operationalized both ways.As both low anxiety and positive self-model indicate adolescent's self-acceptance, positive self-image, and low level of fear of rejection, it is expected that those variables are connected with low susceptibility to peer pressure.As opposed to those attachment dimensions, low avoidance and positive model of others indicate positive perception of others and care for others' opinions, so it is assumed to be connected to higher level of susceptibility to peer pressure.Along with these main effects of attachment dimensions, the interaction effect is also expected, meaning that those adolescents with the preoccupied attachment style (high anxiety and low avoidance/negative model of self and positive model of others) would show the highest level of susceptibility to peer pressure.Finally, we expect that the aforementioned effects will show regardless susceptibility to peer pressure is measured by self-report in a private condition or experimentally in a public condition.
In order to achieve these goals the present research consists of two studies.We will present the method and results for each of the studies separately, and then discuss all the results together.

STUDY 1
In Study 1, we examined the role of the attachment to friends in the explanation of adolescents' susceptibility to peer pressure measured by their anonymous self-report.

Method
Participants.Participants were 475 high school students from Zagreb, the capital of Croatia.There were 194 (41%) boys and 281 (59%) girls ranging in age from 15 to 17 years (M=16.02;SD=0.33).They all attended the 2 nd grade of grammar high schools.(Kamenov & Jelić, 2003) is a shortened version of the ECRI (Brennan, Clark, & Shaver, 1998) retaining the same psychometric properties as the original scale.It is an 18-item self-report inventory that measures attachment to friends.Ratings are made on a 7-point scale ranging from 1 (I strongly disagree) to 7 (I strongly agree).The inventory consists of two subscales, measuring two attachment dimensions -anxiety and avoidance.Each subscale consists of 9 items, strongly correlating with the underlying factor.The two subscales are orthogonal (r = .08)and acceptably reliable (Cronbach's alphas are .83for Avoidance and .75 for Anxiety).

Instruments. Modified Experiences in Close Relationships Inventory
Relationship Scales Questionnaire (Griffin & Bartholomew, 1994) contains 30 short statements drawn from Hazan and Shaver's (1987) attachment measure, Bartholomew and Horowitz's (1991) Relationship Questionnaire, and Collins and Read's (1990) Adult Attachment Scale.On a 5-point scale ranging from 1 (not at all like me) to 5 (very much like me), participants rate the extent to which each statement best describes their characteristic style in close relationships.According to authors, the RSQ can be worded in terms of general orientations to close relationships, orientations to romantic relationships, or orientations to peer relationships.In order to assess participants' attachment to friends, we changed the wording in all the items that referred to "romantic partners" and replaced it with "friends".Five statements contribute to the secure and dismissing attachment patterns and four statements contribute to the fearful and preoccupied attachment patterns.Scores for each attachment pattern are derived by taking the mean of the four or five items representing each attachment prototype.By using the scores from the four prototype items it is possible to create linear combinations representing the self and other-model attachment dimensions (Kurdek, 2002).The two subscales are slightly correlated (r = .22).
Susceptibility to Peer Pressure Questionnaire (Lotar, 2012) is a self-report measure that consists of descriptions of seven hypothetical situations concerning peer pressure to misconduct (smoking, alcohol consummation, smoking marihuana, stealing, cutting classes, abandoning parents' rules, risk sexual behavior).Participants need to imagine themselves in each situation and choose one of the four answers that best describes their reaction in that situation.Answers are scored from 0 to 3, with 0 meaning refusal to conform to peer pressure and 3 meaning complete conforming to it.Total score can be computed as linear combination of scores for all seven situations, with higher scores indicating greater susceptibility to peer pressure.Considering that each item deals with peer pressure to a different type of misconduct, the SPP questionnaire proved to be acceptably reliable (Cronbach's alphas are .73for boys and .67 for girls).
Procedure.Participants were approached in their classrooms and were invited to participate in a study on interpersonal relationships among peers.Parents received written information about the study and all the parents who objected their child's participation had to return signed refusal.Adolescents who agreed to participate in the study signed an informed consent form and received a set of randomly ordered self-report scales described earlier in the Instruments section.Participants completed the instruments anonymously, but they had to write down the same code on each instrument in order to enable researchers to put together the scales completed by the same person.

Results
From presented means in Table 1 it can be seen that we found significant gender differences on all included measures.As expected, boys report higher susceptibility to peer pressure (SPP) concerning risky behavior than girls.Boys also show higher avoidance in relations with friends.Also, their model of self and the model of their friends are more positive than those that girls report.On the other hand, anxiety in friendship is more pronounced among girls than boys.Having in mind the obtained gender differences, in order to examine relation between self-reported susceptibility to peer pressure and attachment dimensions, correlations were calculated separately for boys and girls (Table 2).It was expected that susceptibility to peer pressure, regardless of gender, will be positively related to anxiety and model of others and negatively to avoidance and model of self.Results have shown that when it comes to boys, susceptibility to peer pressure has low negative correlation only with avoidance.Boys with lower avoidance in friendships reported higher susceptibility to peer pressure.Correlations for girls on the other hand, show that higher self-reported susceptibility to peer pressure is related to higher anxiety and more positive model of others.Although the results are in line with our hypotheses, obtained correlations between self-reported susceptibility to peer pressure and dimensions of attachment are lower than expected.
Predictive value of attachment in explanation of adolescents' susceptibility to peer pressure was tested with hierarchical regression analysis.We included both dimensions of attachment in the first step of the analysis and the interaction of those dimensions in the second step.Different regression analyses were conducted for two different operationalizations of attachment.As results have shown significant gender differences on all variables included in this research, separate regression analyses were conducted for boys and girls.When attachment was operationalized through anxiety and avoidance, results have shown that avoidance is a significant predictor of boys' susceptibility to peer pressure, while for girls anxiety proved to be a significant predictor (Table 3).Boys with lower avoidance and girls with higher anxiety report to be more susceptible to peer pressure.Although it would be expected, there is no significant interaction of anxiety and avoidance on susceptibility to peer pressure.Anxiety and avoidance explain only 3.7% of boys' susceptibility to peer pressure and only 2.8% of girls' susceptibility.
When attachment is operationalized through model of self and model of others, results of hierarchical regression analyses showed that only model of others is a significant predictor of girls' susceptibility to peer pressure (Table 4).Girls with more positive model of their friends report higher susceptibility to peer pressure.There is no significant interaction effect of self-model and othermodel on participants' susceptibility to peer pressure.Model of others explain more variance (5.1%) of girls' susceptibility to peer pressure than anxiety.

STUDY 2
In Study 2, we examined the role of the attachment to friends in the explanation of adolescents' susceptibility to peer pressure measured experimentally in a simulation of a chat-room.In order to do that, special computer software was created, based on the idea presented in Prinstein et al. (2011).

Method
Participants.A sample of 80 boys and 80 girls was randomly chosen from the pool of participants described in Study 1.The age of participants ranged from 15 to 17 years (M=16; SD=0.34).Materials and procedure.Approximately one month after completing self-report measures described in Study 1, an experimenter, who was blind to participants' attachment style or their susceptibility to peer pressure, invited participants to volunteer for testing a new measuring procedure that uses Internet chat-room.There were eight students invited from each classroom (four boys and four girls).
Groups of four students simultaneously entered a spaced room and each participant was placed in front of a laptop.Laptops were placed in a way that students could not see the monitors of other students' laptops.All the students were given the instruction how to log in the chat-room and explained the task.They had to use the same code they used on self-report measures in Study 1 as a password, and they had to enter their full name, classroom code and their gender.
Each student participated in a chat-room simulation together with three other students from the school (confederates).They were told that they participate in a research that tests the possibility of applying psychological instruments via online chat-room and that they have to answer the questions in given order to enable the computer to register those answers accurately.One by one hypothetical situation from the SPP Questionnaire was presented on a monitor and participants could see the answers given by other three students before he or she could choose his or her own answer.Figure 1 presents how the screen looked like while participants were doing this task.There is no correlation between boys' susceptibility to peer pressure in experimental situation and any dimension of their attachment to friends.As opposed to boys, girls' susceptibility to peer pressure in experimental situation is related only to their model of others (r=.28, p<.05).Girls with more positive model of others are more susceptible to peer pressure in experimental situation.
To answer the question about predictive value of attachment in explanation of susceptibility to peer pressure in experimental situation, hierarchical regression analyses were conducted in the same way as in Study 1.In the first step of the analysis we included dimensions of attachment and in the second step their interaction.Separate regression analyses were conducted for different attachment operationalizations and for each gender.
Anxiety and avoidance dimensions of attachment do not predict participants' reactions when they are exposed to peer pressure in experimental situation (Table 5).Interaction effects of anxiety and avoidance are not significant either.Hierarchical regression analyses which examined the prediction value of self-and other-model on susceptibility to peer pressure in experimental situation have shown the same results as in Study 1.Only model of others proved to be a significant predictor of girls' susceptibility to peer pressure (Table 6).Girls with more positive model of others are more susceptible to peer pressure and this dimension explains almost 8% of susceptibility to peer pressure variance.Interaction effects of self-and other-model on susceptibility to peer pressure in experimental situation are not significant.

DISCUSSION AND CONCLUSION
Our research showed that attachment to friends is a significant predictor of adolescents' susceptibility to peer pressure for socially unacceptable and risky behaviors, which is in accordance with our initial assumptions and previous research that showed relation between adolescents' self-report of insecure attachment to peers or parents and involvement in delinquent behavior (McElhaney, Immele, Smith, & Allen, 2006;Nelson & Rubin, 1997).It was expected that susceptibility to peer pressure, regardless of gender, will be positively related to anxiety and model of others and negatively to avoidance and model of self.Results have shown that the role of attachment to friends in explanation of susceptibility to peer pressure differs for boys and girls and depending on the way both constructs are operationalized.
Firstly, we have to acknowledge that boys both self-reported and experimentally showed higher susceptibility to peer pressure concerning risky behaviors than girls.These results reinforce the findings of numerous studies that when it comes to risky or delinquent behaviors, boys are more willing to engage in those behaviors if their peers urge them to do so (e.g. Brown et al., 1986;Lebedina Manzoni et al., 2008;Lotar, 2011a;Pardini et al., 2005;Sim & Koh, 2003;Steinberg & Silverberg, 1986).It is possible that boys are more susceptible to peer pressure in domain of misconduct, but it is also possible there is greater social acceptance of boys performing misconduct then girls.In most societies, including Croatia, these types of behaviors are commonly seen as a typical and expected phase of boys' development and partly construe their masculine role.Therefore, boys generally view those types of behaviors as more acceptable to them than girls (Lotar, 2012).Although our results clearly show that boys are more susceptible to peer pressure then girls, the specific type of behavior that is more acceptable to boys could partially contribute to susceptibility to peer pressure.
In Study 1, participants on average reported that they wouldn't engage in presented behaviors even if their peers urge them to, but they would worry about what their peers will think of them because of that.Considering that the susceptibility to peer pressure for misconduct reaches its peak at 14 and 15 years of age (Berndt, 1979), susceptibility to peer pressure obtained in this study is very low.It is important to note that low susceptibility to peer pressure could be a result of the selected methodology approach.In favor of this assumption are results of studies that use different methodological approaches when it comes to measuring susceptibility to peer pressure.Studies using self-report measures of susceptibility to peer pressure (e.g.Lebedina Manzoni et al., 2008;Lotar, 2011a), show lower susceptibility in comparison to those that use experimental methodology (e.g.Cohen & Prinstein, 2006;Prinstein et al., 2011).Therefore, we decided to conduct the Study 2, in which we exposed the participants to an actual peer pressure.As we can see, this resulted with higher level of susceptibility for both boys and girls, although boys' susceptibility was still higher than girls'.
When susceptibility to peer pressure is measured by self-report questionnaire (Study 1), the results almost entirely confirm our hypotheses about the role of attachment to friends, but show different effects for each gender.The level of avoidance proved to be a significant predictor for boys, while the level of anxiety and the model of others were significant predictors for girls.
Let us first consider the effects of avoidance and anxiety dimensions.Boys with lower avoidance in friendships reported higher susceptibility to peer pressure.Results for girls on the other hand, show that higher self-reported susceptibility to peer pressure is related to higher anxiety.Different results for boys and girls could be more easily explained if we consider gender differences obtained for both attachment dimensions.In accordance with results from many other studies (e.g.Kamenov & Jelić, 2005;Picardi, Caroppo, Toni, Bitetti, & DiMaria, 2005;Marušić, Kamenov, & Jelić, 2006;2011), boys report higher avoidance then girls, and girls report higher anxiety then boys.In other words, there are boys with higher level of avoidance in the sample but there are no girls with that level of avoidance, so it is not surprising that there is no effect of avoidance to susceptibility to peer pressure among girls.Accordingly, boys do not have high level of anxiety so among boys anxiety cannot provoke greater susceptibility to peer pressure.
On the other hand, we can assume that mean values for boys' and girls' susceptibility to peer pressure represent the normative behavior for each gender and therefore confirm that boys are more willing to engage in misconduct then girls if their friends ask them to do so.This finding could also be considered for interpreting different effects of avoidance and anxiety for boys and girls.Since boys are generally more willing to engage in these behaviors under the peer pressure, only the boys who are more avoidant (who are cool and maintain an emotional distance from others, and who do not care what others think of them) show lower susceptibility to peer pressure.The opposite could be true for girls: since girls are generally less willing to engage in misconduct under the peer pressure, only the girls who are more anxious about their relations with friends (who desperately seek the company and attention of others and tend to be highly dependent on others for self-esteem) show higher susceptibility to peer pressure.This interpretation is confirmed with the results of second regression analysis, where model of others proved to be important predictor only for girls' susceptibility.
One of our assumptions was that adolescents with the preoccupied attachment style (high anxiety and low avoidance/negative model of self and positive model of others) would show the highest level of susceptibility to peer pressure.However, results show no significant interaction effects of attachment dimensions on susceptibility.This could be explained with relatively high self-models of our participants, regardless of their gender.Having in mind that the participants are regular 2 nd grade students of grammar high schools in Zagreb, the capital of Croatia, this result is not unexpected.Other research on this population has also shown relatively high self-esteem and positive selfmodels (e.g.Jelić, 2008;Lebedina-Manzoni, Lotar, & Ricijaš, 2008b;Lebedina-Manzoni, 2011).Since enrolling criteria for these schools are very high, only the best students meet those criteria and they are usually coming from complete and well-functioning families and have well-educated parents.Of course, the individual differences exist, but not as much as in vocational high schools.Therefore, we could assume that the most of our participants were securely attached to their parents and have developed positive self-models.No significant interaction effects of attachment dimensions on susceptibility to peer pressure could therefore be explained with the small number of students with different attachment styles other than secure one.
A comparison of the results from both studies enables us to answer the research questions more completely.When susceptibility to peer pressure was measured experimentally (Study 2), the results showed that attachment dimensions predict only girls' susceptibility and that the only significant predictor is their model of others.Girls with more positive model of others proved to be more susceptible to actual peer pressure in experimental situation.This result is expected and supports the result from Study 1, but the lack of other effects obtained in Study 1 is unexpected.It drives us toward conclusion that the relation between attachment and susceptibility to peer pressure depends on the way the latter is measured.It could be the case that in a survey situation, when they fill out the self-report questionnaires, adolescents reflect more about how they typically behave and what kind of person they are, so the relations between different variables could more easily show.Or the relations could simply be explained with the same methodology used for assessing all the variables.On the other hand, when they find themselves in an actual peer pressure situation, adolescents don't have the time to reflect about themselves and they behave more susceptible if they care for others' opinions more.The reason why the effect was shown only for girls was explained earlier, with the type of behaviors that were used in our research, that are generally more acceptable for boys.
Finally, we must refer to the amount of susceptibility to peer pressure variance that was explained with the attachment dimensions.It varies from 0 to 8%, depending on measures, which is lower than it was expected.But this result could be seen as an indirect confirmation of the conceptual model of the peer influence process (Brown et al., 2008), according to which there are multiple predictors and mediators of susceptibility to peer pressure.Our research shows that attachment is only one of them and future research should consider many others that are proposed within this model.

Figure 1 .
Figure 1.Chat-room simulation seen by participants on a computer monitor.

(
Note: names shown in Figure 1 are fictional and do not respond to actual participant or confederates) 120

Table 1 .
Means and standard deviations of self-reported susceptibility to peer pressure (SPP) and attachment to friends **p<.01

Table 2 .
Correlation between self-reported susceptibility to peer pressure (SPP) and attachment to friends *p<.05 **p<.01

Table 3 .
Results of hierarchical regression analysis with self-reported susceptibility to peer pressure as a criterion and anxiety and avoidance in friendship as predictors

Table 4 .
Results of hierarchical regression analysis with self-reported susceptibility to peer pressure as a criterion and model of self and model of others as predictors

Table 5 .
Results of hierarchical regression analysis with susceptibility to peer pressure in experimental situation as a criterion and anxiety and avoidance in friendship as predictors

Table 6 .
Results of hierarchical regression analysis with susceptibility to peer pressure in experimental situation as a criterion and model of self and model of other as predictors