Anger Communication in Bicultural Adolescents
Abstract
Little is known about bicultural adolescents’ emotional competence. The aim of the present study was to examine anger communication by comparing thirty-eight 16-year-old Moroccan-Dutch adolescents with 40 Dutch and 40 Moroccan peers using hypothetical anger-eliciting vignettes. Findings show that although Moroccan and Dutch adolescents were equally likely to feel angry, they differed in their anger communication in accordance with their cultural models: Moroccan adolescents were more likely to express their anger indirectly than their Dutch counterparts, whereas Dutch adolescents were more likely to react directly or aggressively. Critically, the anger communication styles of the Moroccan-Dutch youngsters fell in between the two monocultural groups.
The ability to appropriately communicate one's emotion within relationships is considered one of the central aspects of children's emotional competence (Saarni, 1999). Through the course of development, children learn how, when, and where to express their emotions according to what is culturally expected (Saarni, 1999). This process of emotion socialization seems already complicated for children who are brought up in one cultural setting. Yet, emotion socialization might be even more challenging for youngsters who are influenced by two cultures that could signal conflicting ideas about appropriate emotion communication (Matsumoto et al., 2008). The aim of the present study was therefore to examine bicultural adolescents’ anger communication styles by comparing Moroccan-Dutch adolescents who live in the Netherlands and have a Moroccan background, with their monocultural Dutch peers living in the Netherlands and their monocultural Moroccan peers living in Morocco.
Similar to other Western countries experiencing difficulties with one specific ethnic minority group, the behavior of Moroccan-Dutch adolescents has been the subject of discussion for many years in Dutch society, mostly stressing behavioral problems among this group. Empirical studies mainly tend to focus on youth delinquency and behavioral problems (e.g., Stevens, Vollebergh, Pels, & Crijnen, 2007; Eichelsheim et al., 2010) that contribute to the impression that Moroccan-Dutch adolescents, in particular, externally regulate and communicate their anger differently (i.e., more aggressively) than their Dutch peers, at least in situations outside the home. Yet, these studies have been focused on externalizing behaviors that apply to a small proportion of the population (Veen, Stevens, Doreleijers, & Vollebergh, 2011), and do not address different ways of communicating anger in common daily situations.
Studying anger communication styles in bicultural adolescents is also of theoretical relevance. Cross-cultural studies on emotions convincingly show that in individualistic cultures, such as the Dutch, negative emotions, such as anger, are more likely to be expressed in the presence of others than in collectivistic cultures, such as the Moroccan (e.g., Cole, Bruschi, & Tamang, 2002; Matsumoto, Yoo, & Chung, 2010). These differences are consistent with culture-specific models (Mesquita & Albert, 2007).
A typical Western model, such as the Dutch, is characterized by an independent and autonomous self, emphasizing the individual needs and concerns (Markus & Kitayama, 1991). Consequently, displaying one's anger when a personal concern is put at risk by someone else fits with the Western cultural model. Where separateness, uniqueness, and autonomy are characteristics of Western models, North African cultural models, such as the Moroccan, more strongly highlight relatedness, modesty, and obtaining honor and respect from others (Gregg, 2005). Expressing anger is incongruent with the cultural model that stresses attainment and maintenance of group membership, especially when honor is not the subject of the conflict. Consequently, direct anger communication might be less desirable because it might be perceived as disrespectful and might reflect a negative image upon one's family.
Moroccan-Dutch adolescents form an intriguing group in this respect, as they embody characteristics of the individualistic Dutch as well as the collectivistic Moroccan culture. Specifically, they are socialized according to norms, values, and goals of the parental culture at home, whereas teachers, peers, and media teach them how to behave according to the expectations of the dominant culture. Possibly, these distinct influences can be confusing to youngsters and might result in acting-out behaviors outside the home (Stevens et al., 2003). Conversely, based on the collectivistic cultural model, one would expect Moroccan-Dutch adolescents to avoid getting involved in heated interpersonal conflict situations, which could threaten their group membership and have negative impact on their social reputation and honor. Consequently, Moroccan-Dutch adolescents may be more likely to express their anger in a nonaggressive, direct and even indirect manner.
In an initial study, Novin and Rieffe (2009) compared Moroccan-Dutch and Dutch middle school children's spontaneous reactions to hypothetical peer-conflict situations. Consistent with Dutch and Moroccan cultural models, Moroccan-Dutch primary school children were less likely to report demanding and aggressive reactions to protect their personal needs than their Dutch peers were. The amount of reported neutral and assertive or direct responses was equal between the groups. Although these findings provide initial insight in bicultural children's anger communication, two important issues are still left unaddressed.
First, the findings of Novin and Rieffe (2009) failed to address how differences between the Moroccan-Dutch and Dutch children can be understood in terms of culture: do Moroccan-Dutch children's response styles correspond with a ”Moroccan way” or a “Dutch way” of responding, or do their response patterns differ from both cultural groups, leaving them in an ”in between-cultures“ position? To examine the cultural position of bicultural adolescents’ anger communication, it is essential to compare their responses to anger-evoking vignettes not only to responses by peers from the dominant culture, as did Novin and Rieffe in their study, but also to responses by peers from the parental culture.
Second, by focusing only on middle school children, it is unclear whether these differences hold when children reach adolescence. From a developmental perspective, children and adolescents are likely to differ in their anger regulation and communication. During adolescence, increases in negative affects become prevalent (Larson & Sheeber, 2008) and adequate regulation of negative emotions is critical due to physiological, psychological, and social changes (Furman & Buhrmester, 1992; Silk, Steinberg, & Morris, 2003). For bicultural Moroccan-Dutch adolescents, this transition phase may also mean that they are increasingly influenced by the Dutch culture in general, which might result in Moroccan-Dutch adolescents’ anger communication patterns similar to those of their Dutch peers as opposed to the varying patterns between Moroccan-Dutch and Dutch children. To attend to these two points, we extended the study of Novin and Rieffe by comparing Moroccan-Dutch 16-year-olds with their Dutch and their Moroccan peers.
The Present Study
Our interest in the cultural position of Moroccan-Dutch adolescents’ anger regulation strategies was based on an initial prediction that the Moroccan and Dutch groups vary in their self-reported anger communication styles to hypothetical conflict situations. Given the cultural descriptions described above, we expected that Moroccan adolescents would be more likely to report avoiding the situation (neutral style) or to express their anger subtly (indirect style) compared to their Dutch peers. Compared to their Moroccan peers, Dutch adolescents are more likely to report to explicitly express their dislikes (direct expression) or even more aggressively, because expressing one's true feelings in terms of honesty and expressing personal desires are more congruent with the typical Dutch culture than with the typical Moroccan culture.
Expectations about differences and similarities between the bicultural group and the Moroccan and Dutch groups are less straightforward, because no previous studies exist with this design. Similarities with the Moroccan group can be expected, especially in terms of fewer notifications of aggressive expressions than the Dutch group, which is consistent with the outcomes of Novin and Rieffe's study (2009). Alternatively, Moroccan-Dutch adolescents’ anger communication styles may show more similarities with the Dutch than with the Moroccan group, due to increasing Dutch influences in Moroccan-Dutch adolescents’ lives.
Finally, with respect to gender, it was expected that independent of cultural group, boys would report aggressive expression styles more often than girls, whereas girls would report neutral or indirect anger expression styles more often (e.g., Cheng, Mallinckrodt, & Wu, 2005; Zahn-Waxler, 2010).
Method
Participants
The sample consisted of 40 Moroccan-Dutch, 40 Dutch, and 38 Moroccan adolescents. All adolescents participated in an extensive project that examined emotional functioning of adolescents in the Netherlands and Morocco. The Moroccan-Dutch and the Dutch adolescents lived in the Netherlands at the time of the study and were recruited from schools in one of the three largest cities of the Netherlands (Amsterdam, Den Haag, Rotterdam). The adolescents in the Moroccan-Dutch group (35% boys; mean age 15 years and 4 months; SD = 7 months) had at least one parent who was born in Morocco (90% of the adolescents’ parents were both born in Morocco). The adolescents themselves were all born in the Netherlands or had immigrated before their first birthday. The adolescents in the Dutch group (48% boys; mean age 16 years and 1 month, SD = 4 months) and both their parents were born in the Netherlands. The other monocultural group consisted of 38 native Moroccan adolescents (50% boys; mean age = 16 years and 7 months, SD = 6 months) who lived and attended a school in Tetouan, a city in Northern Morocco. Tetouan is one of the cities in the Rif area where the Moroccan-Dutch population has its origin (Benali & Obdeijn, 2005). All Moroccan adolescents, as well as both their parents, were born in Morocco. Overall, the Moroccan-Dutch group was younger than the Dutch and the Moroccan groups, F(2, 117) = 66.29, p < .001, 
.
Instruments
Anger interview
The interview consisted of eight vignettes (http://www.focusonemotions.nl/index.php/culture), which were designed to be applicable in various Western and non-Western countries (Novin, Rieffe, Banerjee, Miers, & Cheung, 2010). Each vignette describes a conflict situation with a peer (friend or unknown peer) that was expected to evoke anger. Analyses revealed no differences between friend and unknown peer vignettes and therefore the analyses were conducted collapsing all eight stories. Consistent with previous studies, participants’ gender matched with the gender of the peer (e.g., Rieffe & Meerum Terwogt, 2006). To prevent order-effects, the vignettes were presented in random order. After reading aloud each vignette, it was stated that the participant would feel angry.
An example of a story is as follows: Together with a friend you are on your way to a party. Your parents have bought new clothes for you to wear to this party! You have dressed up and think you look nice. You feel good about yourself. On the way to the party your friend gets a can of coke out of his/her pocket. (S) he opens the can in a way that the coke spills on you. Your face and clothes are now covered in coke. You feel angry.
Following the presentation of each story, adolescents were asked five questions: (1) How angry would you feel? (Time 1) (2) What would you say to your friend (to the boy or girl)? (3) What would your friend (the boy or girl) say? (4) How angry would you feel now? (Time 2) (5) Would you still be friends? (This last question was only asked following the friend-vignettes).
The vignettes were presented in Arabic for the Moroccan group and in Dutch for the other two groups. The vignettes were translated from Dutch to Arabic by two native speakers who are fluent in both Arabic and Dutch and back-translated by a colleague with Ph.D. in Arabic. Inconsistencies between the languages were subsequently resolved through discussion.
Scoring
Anger intensity (Question 1 and 4)
Adolescents’ anger intensity was measured using a 6-point scale (from 0 = not angry at all to 5 = extremely angry).
Initial response (Question 2)
Initial verbal reactions to the peer were coded with reference to four categories based on previous studies focusing on anger reactions (O'Connor, Archer, & Wu, 2001; Kuppens, Van Mechelen, & Meulder, 2004; Novin & Rieffe, 2009): (1) neutral: not actively trying to reinstate own position of thwarted goals and plainly accepting the disadvantaged position (e.g., “Don't worry, it doesn't matter”); (2) indirect anger expression: showing that one is upset without being explicit about one's anger and without being verbally or physically aggressive (e.g., “now I have to go home and change”); (3) direct anger expression: explicitly expressing one's emotion, without being verbally or physically aggressive (e.g., “I don't like this”); (d) aggressive: accusing, using aggressive language or behavior, demanding a solution (e.g., “Now you have to buy me a new shirt”). Note that adolescents’ reactions could fall into more than one category.
Reaction from the peer (Question 3)
Adolescents’ responses to the third question were coded with reference to five categories: (1) positive: apologizing or giving a solution (e.g., “She would say that I could wear her dress”; (2) explanation: the peer would explain his or her actions (e.g., “He would say that he lost track of time”); (3) avoidant: the peer would distance himself or herself from the situation (e.g., “He wouldn't say anything”); (4) aggression: the peer would express verbal or physical aggression (e.g., “He would say it's my own fault”). Again, adolescents’ responses were not exclusive to one category.
Continuation of the relationship (Question 5)
Whether adolescents expected their friendship to continue was asked in the fifth question. This question was only asked after the vignettes with the friend.
All responses were tape-recoded and transcribed. The transcribed responses were coded afterward by two independent judges. Cohen's Kappas for the own and other responses ranged from 0.79 (own response, unknown peer vignette) to 0.92 (other response, unknown peer vignette). Disagreements were resolved through discussion.
Procedure
Adolescents in the Netherlands and in Morocco were individually interviewed by a Dutch and a Moroccan female interviewer, respectively. The interviews took place in a silent room in the adolescent's school during school hours. Before the stories were read aloud by the interviewer, adolescents’ were informed of the strict confidentiality and anonymity of their answers. This assessment took approximately 15 minutes. With the adolescents’ consent, all interviews were tape-recorded and subsequently transcribed by native speakers for coding.
Results
Degree of Anger Evoked
Adolescents’ reported anger intensity, in response to the vignettes, was analyzed by a 3 (cultural group: Dutch, Moroccan, and Moroccan-Dutch) × 2 (gender) analysis of variance (ANOVA). The analysis showed no significant main effect of cultural group F(2, 112) = 2.16, p = .12, 
. The main effect of gender, F(1, 112) = 4.79, p < .05, 
, indicates that girls reported a higher anger intensity than boys (Means [SDs] = 3.95 [0.57] and 3.74 [0.57], respectively). With a maximum of five, the vignettes evoked a reasonable amount of anger and thus appeared to be useful for the purpose of this study (Mean [SD] = 3.86 [0.58]).
Initial Verbal Responses in Anger-Evoking Stories
Adolescents’ initial verbal responses were counted with reference to the four categories (neutral, indirect anger expression, direct anger expression, aggression). Collapsing all eight stories, adolescents could thus receive a minimum score of 0 and a maximum of 8 for each category. A 3 (cultural group) × 2 (gender) × 4 (own response) repeated measures ANOVA revealed a main effect of own response, F(3, 110) = 396.85, p < .001, 
, indicating that adolescents were most likely to report expressing anger indirectly, whereas neutral responses were least likely to be reported.
With respect to cultural group differences, a main effect of cultural group was revealed, F(2, 112) = 5.64, p < .01, 
, which was qualified by a Cultural Group × Own Response interaction, F(6, 222) = 7.60, p < .001, 
. Largely consistent with our hypotheses, post-hoc testing with Bonferroni correction showed that Moroccan adolescents were more likely to report indirect anger expressions than their Dutch peers (Table 1). Dutch adolescents in turn were more likely to report direct or aggressive anger expressions than their Moroccan peers. Moroccan-Dutch adolescents’ responses did not differ from the two native groups, except for aggressive responses. Like the Moroccan group, Moroccan-Dutch adolescents were less likely to report aggressive expressions than their Dutch peers.
| Dutch | Moroccan | Moroccan-Dutch | Total | |
|---|---|---|---|---|
| Indirect expression | 3.952 (1.72) | 5.551 (2.27) | 4.6012 (1.61) | 4.69 (.1.98) |
| Direct expression | 3.081 (2.03) | 1.452 (1.29) | 2.2512 (1.43) | 2.27 (1.74) |
| Aggressive | 2.631 (1.98) | 1.423 (1.48) | 1.572 (1.58) | 1.88 (1.77) |
| Neutral | 0.431 (0.59) | 1.001 (1.31) | 0.951 (1.34) | 0.19 (0.51) |
| Total | 10.081 (1.67) | 9.421 (1.22) | 9.381 (1.29) |
Note
- Means in the same row that do not share subscripts differ at p < .05.
Furthermore, the analysis revealed a main effect of gender, F(1, 112) = 3.87, p < .05, 
, which was qualified by a Gender × Own Reaction interaction, F (3, 110) = 5.87, p < .001, 
. Post-hoc testing reveals that whereas boys were more likely to report aggressive reactions than girls, girls in turn were more likely to report either direct or indirect anger expressions (Table 2). Within-group analyses show that girls were more likely to be direct than aggressive in their responses, whereas there were no differences between these strategies for boys.
| Boys (N = 52) | Girls (N = 66) | |
|---|---|---|
| Indirect expression | 4.252 (0.23) | 5.031 (1.59) |
| Direct expression | 1.772 (1.58) | 2.671 (1.77) |
| Aggressive | 2.481 (2.11) | 1.412 (1.28) |
| Neutral | 0.841 (0.98) | 0.741 (1.29) |
| Total | 9.351 (1.34) | 9.851 (1.48) |
Note
- Means in the same row that do not share subscripts differ at p < .05.
Verbal Responses Expected from the Peer
Next, the expected reactions of the peer were assessed by a 3 (cultural group) × 2 (gender) × 4 (peer's response: positive, explanation, avoidant, aggression) repeated measures ANOVA. The analysis showed a main effect of peer's response, FGG(1.82, 204.03) = 216.41, p < .001, 
. In general, adolescents expected most often an apology and least often an avoidant response from the peer (Table 3).
| Dutch | Moroccan | Moroccan-Dutch | Total | |
|---|---|---|---|---|
| Positive | 4.931 (2.07) | 4.531 (1.93) | 4.901 (1.66) | 4.79 (1.88) |
| Explanation | 2.682 (1.67) | 4.601 (2.22) | 3.282 (1.54) | 3.50 (1.98) |
| Aggression | 1.201 (0.88) | 1.161 (0.79) | 0.981 (0.13) | 1.11 (0.85) |
| Avoidant | 0.331 (0.53) | 0.261 (0.64) | 0.131 (0.33) | 0.24 (0.52) |
| Total | 9.802 (1.98) | 11.051 (2.01) | 9.782 (1.77) |
Note
- Means in the same row that do not share subscripts differ at p < .05.
In addition, the analysis revealed a main effect of cultural group, F(2, 112) = 9.43, p < .001, 
, which was qualified by a Cultural Group × Peer's Reaction interaction, FGG (3.64, 204.03) = 5.14, p < .001, 
. Post-hoc analyses with Bonferroni correction show that Moroccan adolescents were more likely to expect an explanation from the peer than their Dutch and Moroccan-Dutch peers (Table 3). Moreover, Moroccan adolescents were equally likely to expect a positive response or explanation from the peer, whereas within the Dutch and Moroccan-Dutch group, positive responses were more often expected than explanations.
The Gender × Peer's Reaction interaction, FGG(1.82, 204.03) = 5.19, p < .01, 
, indicates that boys were more likely to expect a positive reaction from the peer than girls, whereas girls expected to hear an explanation more often than boys (Table 4).
| Boys | Girls | |
|---|---|---|
| Positive | 5.191 (1.77) | 4.472 (1.92) |
| Explanation | 3.082 (1.99) | 3.831 (1.93) |
| Aggression | 1.151 (0.89) | 1.081 (0.81) |
| Avoidant | 0.171 (0.43) | 0.291 (0.58) |
| Total | 10.231 (2.23) | 10.171 (1.80) |
Note
- Means in the same row that do not share subscripts differ at p < .05.
Reduction of Anger After Verbal Interaction
The degree to which adolescents expected their anger to be reduced after communicating with the peer was calculated by subtracting their reported anger intensity after communication (Question 4) from the reported intensity before communication (Question 1). A 3 (cultural group) × 2 (gender) univariate ANOVA revealed a main effect of cultural group, F(2, 112) = 3.40, p < .05, 
, indicating that Moroccan adolescents’ anger would be more likely to be reduced than Dutch adolescents’ anger (Means [SDs] = 1.03 [0.65] and 0.67 [0.52], respectively). Again, Moroccan-Dutch adolescents’ degree of anger reduction (Mean [SD] = 0.84 [0.64]) did not differ from either of the two native groups.
Expected Friendship Continuation
Whether adolescents expected the friendship to continue was counted over the four friend-vignettes (min = 0; max = 4). A 3 (cultural group) × 2 (gender) ANOVA showed a main effect of cultural group, F(2, 117) = 8.63, p < .001, 
. Post-hoc analyses with Bonferroni correction showed that although adolescents largely reported having positive expectations (92%), Dutch adolescents more often expected their friendship to be continued than their Moroccan peers (Means [SDs] = 3.93 [0.27] and 3.39 [0.72], respectively). Crucially, Moroccan-Dutch adolescents’ expectations (Means [SDs] = 3.53 [0.68]) did not differ from their Moroccan peers, but they expected to terminate the relationship more often than their Dutch peers.
Discussion
The aim of the current study was to examine Moroccan-Dutch adolescents’ anger communication in a peer-conflict situation. The initial comparison between Dutch and Moroccan adolescents provides a framework to understand the cultural position of Moroccan-Dutch adolescents’ anger communication. The results show that although adolescents’ self-reported anger level did not differ between the cultural groups, reported anger communication styles by Dutch adolescents are distinguishable from those reported by Moroccan adolescents as could be expected from Dutch and Moroccan cultural models. Crucially, Moroccan-Dutch adolescents’ reported anger communication pattern did not differ from those of both their monocultural peer groups.
Our interest in Moroccan-Dutch adolescents’ anger communication was initially triggered by the distinct cultural frameworks they are brought up with. Indeed, our results reveal that consistent with the Dutch cultural framework, Dutch adolescents were more likely to report speaking up for themselves than their Moroccan peers by explicitly letting the peer know about their negative feelings or even by demanding a solution from the peer that met their personal concerns. Although Moroccan adolescents also reported expressing their anger in these peer-conflict situations, they reported that they would do so more indirectly by calmly asking for an explanation or mentioning the consequences. In turn, Moroccan adolescents more often expected an explanation from the peer than the Dutch adolescents. This anger communication style seems to be consistent with the importance of behaving respectfully and responsible in an interpersonal situation, which is emphasized in the Moroccan culture (Pels & de Haan, 2007). Moreover, asking for an explanation takes into account the sentiments of other people that is described to be more significant in North African cultures than in Western cultures, where the individual's own internal standards are valued more strongly (Leung & Cohen, 2011).
Yet, the distinct cultural frameworks that are likely to play a role in the differences found between the Dutch and Moroccan group, both seem to influence Moroccan-Dutch adolescents’ reported anger communication. The findings in our study show that the Moroccan-Dutch adolescents’ reported anger communication styles overall did not differ from either their Dutch or Moroccan peers. Or, more precisely, the reported anger communication by the bicultural youngsters fell “in between” their parental and the societal dominant culture. Moroccan-Dutch adolescents’ reported anger communication did not include a high amount of either typical Dutch (direct and aggressive expression) or Moroccan (indirect expression) styles.
The possibility that these bicultural youngsters might report acting-out behaviors was not supported by these outcomes. In fact, the outcomes in our study do not seem to denote a misfit in either group. Instead, we could interpret this outcome in a positive way: the anger communication of the bicultural youngsters in our sample may not attract attention within either a Dutch or a Moroccan peer group. Expressing anger in a similar way as both monocultural peers may be highly beneficial for the bicultural group in terms of avoiding interpersonal difficulties. Our findings might suggest that being influenced by both Dutch and Moroccan culture as a bicultural adolescent leads to an integration of “typical” Dutch and Moroccan anger communication styles, resulting in less extreme cultural communication patterns. However, much more research is needed to understand the underlying mechanisms and specific cultural influences that could explain the differences and similarities between the cultural groups. Including individual differences in terms of acculturation patterns, ethnic identification, and cultural values might contribute to the understanding of the cultural position of bicultural adolescents’ anger communication styles.
The only difference in anger communication styles found between the Moroccan-Dutch and monocultural groups in this sample concerned the reported aggressive expressions. Consistent with a previous study focusing on children (Novin & Rieffe, 2009), our results imply that Moroccan-Dutch adolescents are less likely to report aggressive responses in a peer-conflict situation than their Dutch peers. Some scholars suggest that peers are especially important for Moroccan-Dutch youngsters, because they receive less support and understanding from their parents due to a lack of knowledge of the Dutch society (Pels & Nijsten, 2003). Responding aggressively could possibly jeopardize the relationship with their peers. Another possibility is that responding aggressively in the presented conflict situations might be considered unnecessary and even undesirable, as it might decrease the adolescents’ respect from others. In other more severe conflict situations where adolescents’ honor and respect are at stake, one could expect more aggressive expressions by Moroccan-Dutch adolescents to defend one's honor.
Consistent with a functionalistic perspective (Frijda, 1986; Gross, 1998), the outcomes suggest that anger communication would be beneficial for all groups in terms of intrapersonal as well as interpersonal consequences: all groups expected experiencing less anger after the communication than before, and the majority of the adolescents expected the friendship to be continued. However, communicating with the peer contributed more strongly in reducing anger in Moroccan than in Dutch adolescents. Dutch youngsters may need more time (alone) to downregulate their anger, because their first reaction might be to respond to individual concerns (Novin et al., 2010).
Furthermore, Dutch adolescents more often expected their friendship to continue after the conflict situation than their Moroccan peers. Personal damage done by a friend could be interpreted as more severe and less acceptable in collectivistic-oriented cultures where group harmony, especially within the in-group, is highly stressed. Personal damage done by a friend might be interpreted as more severe and a threat to one's honor and therefore might be less acceptable than in individualistic-oriented cultures. Interestingly, similar to their Moroccan peers, Moroccan-Dutch adolescents in this sample expected terminating the relationship more often than their Dutch peers. Norms about friendship and specific cultural values, such as obtaining honor and respect that could explain the differences, are factors that future research could attend to.
Surprisingly, the closeness of the relationship (friend or unknown peer) failed to influence adolescents’ reported anger communication style, despite previous studies suggesting otherwise. We can only speculate about the reasons for the lack of differences between anger communication toward friends and unknown peers across cultural groups in the current study. It might be that although adolescents reported similar answers in the friend and unknown peer condition, their underlying motivations may have differed. For example, one might express anger toward a friend indirectly to prevent damaging the relationship, whereas the indirect anger expression toward a stranger might be due to not wanting to make a public scene.
In addition to cultural differences, several gender differences were found. None of these gender differences interacted with the cultural groups in our sample. Although girls reported a higher level of anger intensity than boys, they were more likely to report expressing their anger indirectly or directly and to expect receiving an explanation from the peer than boys. Boys in turn were more likely to report responding aggressively than girls and expected more often a solution or apology from the peer. Consistent with previous studies, boys tended to express their negative emotions more overtly than girls (Garrett-Peters & Fox, 2007; McDowell, O'Neil, & Parke, 2000). However, our outcomes might suggest that the more frequently reported aggressive responses of boys are not necessarily unconstructive, because positive reactions might follow. Future research should examine whether gender differences in anger communication can be explained by differences in adaptiveness in terms of interpersonal consequences.
Limitations and Directions for Future Research
The present findings provide an initial framework for studying the cultural position of bicultural adolescents’ emotion communication. Yet, several limitations of the study should be noted. First, one should be cautious with generalizing our outcomes, due to the relatively small sample sizes and due to our specific focus on Moroccan-Dutch adolescents. So far, it is unclear whether investigation of another bicultural group would reveal similar effects. Future studies examining the causal effects that explain the group differences directly would make this matter more insightful (Chao & Otsuki-Clutter, 2011). This brings us to a second limitation that concerns the question of to what extent cultural models are accountable for the variation between the groups, instead of other noncultural factors such as age. In our study, the bicultural group was approximately 8 months younger than the Dutch and Moroccan group. One way to directly test the effects of cultural models on anger communication is to activate or to prime individualism, collectivism, or honor mindsets in each cultural group before they are presented with the hypothetical stories (see review of cultural priming in Oyserman & Lee, 2008). This would not only clarify which underlying (cultural) mechanisms are accountable for the differences, but may also give us insight in the behavior of Moroccan-Dutch adolescents in a Dutch context when primed by either a Moroccan or a Dutch mindset.
Further, using self-report by means of hypothetical vignettes may not reflect actual emotion responses in the real world (Parker et al., 2001). Instead, issues such as social desirability might have played a role. For example, Moroccan-Dutch youngsters might be more cautious in their responses than their Dutch peers, due to the fact that their reactions could more easily be interpreted as more severe behavior. They might not want to pour oil on the fire with regard to further damaging the image of the Moroccan-Dutch population in Dutch society. Moreover, the use of vignettes in cross-cultural research addresses another difficulty; language and translation equivalence cannot fully be guaranteed and interpretation differences due to variable cultural perspectives could play a role in adolescents’ responses. Direct observation of behavior in semi-structured or naturalistic situations would therefore be recommended for future research.
In conclusion, the bicultural adolescents in our study seem to navigate efficiently in the parental and dominant cultural contexts. This optimistic outcome calls for further examination, taking individual differences and specific cultural influences into account. For example, which of the bicultural adolescents use the differing cultures to their advantage and which adolescents are caught in between two worlds? Investigation of adolescents’ acculturation, migration factors, and parental cultural attitudes in relation to their behavior in the two cultural contexts might shed further light on this issue.
