Abstract
The paper investigates reputational status dynamics (coolness and popularity) in a Hungarian primary school sample. The data derives from an emergent mixed methods research study, where one wave of a panel data collection (N = 754) was complemented with focus group interviews (N = 144) among sixth grade students (age 12–13). The quantitative analysis applied multilevel regression models, while the qualitative part relied on thematic analysis. Our results, mostly in line with the international literature, underlined the importance of athletic abilities, verbal aggression, and physical appearance, while some interesting gendered and ethnic patterns also emerged. For instance, verbal aggression was associated with status for girls and non-Roma boys, while physical aggression was only associated with status for Roma girls. The novelty of the research study lies in the mixed approach, while the paper also contributes to our knowledge about status dynamics in the Hungarian, and Eastern European, context.
The investigation of peer status is one of the central areas of peer relations research. An extensive body of literature has shown that preoccupation with status is one of the prime concerns for many adolescents (e.g., Coleman, 1961; LaFontana & Cillessen, 2010), while, simultaneously, status dynamics can have a serious impact on students’ development and well-being including such factors as school engagement (e.g., Troop-Gordon et al., 2011) or participation in risk behavior (e.g., Mayeux et al., 2008). Additionally, students with high peer status, popularity in particular, have been found to be profoundly influential in setting peer norms (Brechwald & Prinstein, 2011). However, while studies investigating salient aspects of status (acceptance, popularity, coolness) have proliferated in the West European and North American literature, less is known about these dynamics in “non-Western” contexts (although some results on Chinese pupils are also available in English, e.g., Lu et al., 2018; Niu et al., 2016; Schwartz et al., 2009). Additionally, most research studies on peer status take either an ethnographic or a sociometric approach, while the integration of qualitatively and quantitatively informed research is less typical (for a few exceptions see Eder & Kinney, 1995; Garner et al., 2006). The present paper intends to address both of these gaps by presenting the results of an emergent mixed methods research study conducted on a Hungarian primary school sample. This sample contains a high proportion of ethnic Roma students as well as students with disadvantaged social background. While the social and educational integration of disadvantaged students is a concern globally, the integration of disadvantaged ethnic minorities is also a salient issue in many countries. In Hungary, the Roma, the largest ethnic minority group in the country, faces multiple social disadvantages that involve residential and educational segregation, low educational outcomes, low levels of labor market participation, and widespread social prejudice (e.g., Kemény et al., 2004; Keresztes-Takács et al., 2016). Moreover, the Roma minority is the most disadvantaged ethnic group in several other European countries, thus the present paper can also contribute to the better understanding of peer dynamics involving Roma students in general.
The formerly socialist Central and Eastern European countries (e.g., Hungary, Poland, or Slovakia) may provide a cultural context different both from the “Western” and the Chinese contexts. According to the research on cultural dimensions by Hofstede and his colleagues, Hungary has similarly low score on power distance and high score on individualism as the Northwestern European and “Anglo” countries, while it scores much higher (similarly to many Eastern European countries) on uncertainty avoidance and restraint, closer to the score of most Asian countries (Hofstede et al., 2010). As the individualism-collectivism dimension has been assumed to be important for peer status (see for instance the cross-cultural studies by Li et al., 2012; Owens et al., 2014), we may expect behavior related to competition (e.g., aggression, athleticism) to be salient in Hungary, similarly to the “Western” results. On the other hand, Hungary’s score on the indulgence-restraint dimension suggests that Hungarian pupils may experience stronger pressure to adhere to strict social norms than their “Western” peers. Indeed, among Hungarian adults, a survey by Eurobarometer found that the level of gender stereotypes was one of the highest within the European Union (European Commission, 2017).
The goal of the paper is to explore reputational status dynamics (popularity and coolness) in a Hungarian primary school sample, drawing on focus group interviews and a survey database that contains variables, to a great extent, comparable to international findings. In our analysis, we will explore both general trends and gender and ethnic differences. Additionally, the application of a mixed methods integration scheme, the “simultaneous bidirectional” framework (Moseholm & Fetters, 2017), will serve for the integration of qualitative and quantitative findings so far less typical in the peer status literature.
Status, Popularity, and Coolness
The contemporary literature tends to distinguish two dimensions of peer status: a reputational one, typically measured by popularity, and a dimension related to social preference, usually measured by liking nominations (acceptance) or the difference of liking and disliking nominations (preference). While acceptance and preference are related to the extent one is liked by their peers, popularity is related to social power, prestige, and visibility (Cillessen & Marks, 2011). As the two dimensions have been shown to be only moderately correlated, they are generally considered as two distinct forms of status (Mayeux et al., 2011; Parkhurst & Hopmeyer, 1998). In addition to popularity, some researchers have been experimenting with the alternative construct of “coolness” to measure the reputational dimension (e.g., Bellmore et al., 2011; Jamison et al., 2015; Kiefer & Wang, 2016). Since the focus of the present paper is the reputational dimension of status, the rest of the section will briefly review the relevant findings only related to popularity and coolness.
One of the most widely discussed findings in the literature is the positive association between different forms of aggression and popularity/coolness (e.g., Cillessen & Mayeux, 2004; Kiefer & Wang, 2016; Parkhurst & Hopmeyer, 1998; Rodkin et al., 2006). Evolutionary theorists have proposed that aggression is often used strategically in order to gain status, and “bistrategic youth,” that is, youth who use both aggression and prosocial behavior have the highest level of popularity (Pellegrini et al., 2011), which proposition has been supported by findings from non-evolutionary researchers as well (e.g., Dijkstra et al., 2009; Faris & Felmlee, 2011). In order to get a more refined picture, it is useful to distinguish overt/direct and relational/indirect aggression. Relational/indirect aggression refers to behavior that intends to damage another person’s social relationships or social position through manipulation, for example, by spreading gossips or sabotaging the target person’s friendships or romantic relationships (Card et al., 2008; Crick & Grotpeter, 1995). Researchers comparing the two types of aggression have found that relational aggression has a stronger association with popularity (e.g., Cillessen & Mayeux, 2004; Prinstein & Cillessen, 2003), while some studies suggest that after controlling for relational aggression the effect of overt aggression (e.g., hitting, pushing, mocking) becomes insignificant (e.g., Rose et al., 2004; Waasdorp et al., 2013). Additionally, some quantitative studies have found the association between overt aggression and popularity/coolness to be stronger (e.g Waasdorp et al., 2013) or only significant (e.g., Kiefer & Wang, 2016) for boys, while relational aggression to be more strongly (e.g., Cillessen & Mayeux, 2004) or only associated (e.g., Kiefer & Wang, 2016) with popularity for girls. However, these studies typically do not distinguish between overt verbal (e.g., teasing, yelling, mocking) and physical aggression. On the other hand, the qualitative tradition of peer relations research does make such distinction. They also underline the importance of “toughness” and the ability to physically and verbally intimidate/dominate others in boys’ popularity (e.g., Adler & Adler, 1998; Adler et al., 1992; Eder et al., 1995) as well as the role of social manipulation and verbal intimidation in girls’ popularity (e.g., Adler et al., 1992; Eder, 1985; Merten, 1997). Additionally, in the United States, several studies have found a stronger association between aggression and popularity for African American than European American students, in particular in Black-majority settings (e.g., Luthar & McMahon, 1996; Waasdorp et al., 2013). In the Chinese cultural context, findings have been more controversial, some studies showing positive association between overt (Niu et al., 2016) and relational (Schwartz et al., 2009) aggression and popularity, while other studies found this association to be negative (Tseng et al., 2013).
In addition to aggression, athletic abilities have also been consistently found to be positively associated with all forms of peer status (e.g., Coleman, 1961; Shakib et al., 2011). However, some gendered and ethnic patterns have also been found: it consistently contributes to male status more than to female status (e.g., Eder & Kinney, 1995; Holland & Andre, 1994), and according to earlier studies, it contributes to the status of African American male students more than for other groups (e.g., Kennedy, 1995), a finding that most recent studies usually do not support (e.g., Chase & Machida, 2011; Shakib et al., 2011). Research studies among Chinese adolescents have also found positive association between popularity and athleticism (e.g., Niu et al., 2016).
The role of academic performance seems to be more controversial. While students tend to believe that it impacts their peer status negatively (e.g., Juvonen & Murdock, 1995), most research studies have found a nonsignificant impact of academic performance (grades) on popularity (e.g., Boyatzis et al., 1998; Meijs et al., 2010). while some studies did actually find a negative association (e.g., Gorman et al., 2002). Ethnographic investigations by Adler and colleagues found that “extreme deviation” (being “nerdy” or “dummy”) had a negative effect on boys’ popularity, while academically well-performing girls did not suffer from similar stigma (Adler & Adler, 1998; Adler et al., 1992). However, the importance of cross-cultural differences needs to be emphasized here: in the Chinese context, studies have actually found positive association between academic achievement and popularity (Li et al., 2012; Niu et al., 2016).
It is useful to distinguish academic performance from engagement. Academic engagement can be understood as a multidimensional construct that can be divided into subcategories including behavioral (e.g., following the rules, involvement in learning), emotional (e.g., sense of belonging and appreciation) and cognitive (investment in learning) engagement (Fredricks et al., 2004). Peer relations researchers have typically investigated some forms of behavioral (dis)engagement (e.g., unjustified school absences, disruptive behavior during classes) and found a negative association between engagement and popularity/coolness (e.g., De Laet et al., 2015; Kiefer & Wang, 2016; Troop-Gordon et al., 2011). With regards to racial and ethnic differences, the “acting white” hypothesis (Fordham & Ogbu, 1986), which proposes that in some settings academically well-performing African American pupils suffer social sanctions from their Black peers, has been the most influential in the last few decades. However, most large scale quantitative studies did not find support for the widespread prevalence of such dynamics (e.g., Ainsworth-Darnell & Downey, 1998), while a few studies could demonstrate such effects (e.g., Fryer & Torelli, 2010).
Furthermore, popularity has been positively associated with risk behavior (e.g., Franken et al., 2017; Mayeux et al., 2008), physical attractiveness (e.g., Adler et al., 1992; Vaillancourt & Hymel, 2006), involvement in romantic relationships (e.g., Carlson & Rose, 2007), prosociality (e.g., Dijkstra et al., 2009), and extraversion (e.g., Wolters et al., 2014). Research with Chinese children and adolescents also found positive association between popularity and physical appearance (Li et al., 2012), dating (Niu et al., 2016), prosociality (Li et al., 2012; Niu et al., 2016), and assertiveness (Schwartz et al., 2009).
In the Hungarian context, the few available research studies on peer status have focused on interethnic relations between Roma and non-Roma students, typically addressing the “acting white” hypothesis. However, these studies did not measure reputational status but social preference (Habsz & Radó, 2018), friendship and adversary nominations (Hajdu et al., 2019), or bullying (Kisfalusi, 2018). Although none of these studies found evidence for the presence of an ethnic “oppositional culture,” one vignette experiment (Keller, 2020) found that Roma students in classes with high ethnic diversity rated hypothetical peers with high GPA as less “cool.”
Method
The research took an emergent mixed methods design, where one wave of a longitudinal survey data collection was complemented with focus group interviews. The survey questionnaire was collected by a research group in Central and Northern Hungarian primary schools following up the same pool of students for two academic years in a total of four waves, once per semester (students who arrived to the classes between waves were also involved if parental consent was obtained). The focus group research complemented the last wave of the survey research in the spring of 2015 when students attended grade six (age 12–13). The main objectives of the survey were to explore ethnic segregation and interethnic social relations of students, and to examine the interrelated status hierarchies and social dynamics in the class. Since the ethnicity of students is not registered by Hungarian institutions, a sample representative of the ethnic proportion of schools in the selected region was not feasible. However, in line with the research objectives, schools with a “known” higher proportion of ethnic Roma students were overrepresented in the sample. The survey predominantly contained questions involving sociometric peer nomination procedures, which were complemented by a variety of other questions involving sociodemographic information, academic motivation, or attitudes toward competition. Due to the sociometric nature of the data collection, the goal was to collect information on full closed groups, therefore all students from all relevant classes (grade 5 in waves 1–2 and grade 6 in waves 3–4) were invited to participate. In the first wave, the sample involved 1183 pupils from 61 classes in 35 schools, whereas in the fourth wave it involved 1054 students from 53 classes in 34 schools. For the cross-sectional quantitative analysis presented in this paper, a limited version of the wave four database will be used (see below). The wave four data were collected in March and April 2015, while the focus group research was conducted in May and June. The focus group questions were predominantly organized around students’ perception and interpretation of “popularity” and its dynamics in the class, and to a lesser extent around their interpretations of ethnicity and friendship in general. The analysis followed a concurrent, “simultaneous bidirectional” framework (Moseholm & Fetters, 2017, pp. 7–8); quantitative and qualitative data were collected and analysed separately, and both results framed the final interpretation. The qualitative data were analysed using qualitative thematic analysis, while for the quantitative data multilevel regression models were applied.
Quantitative Data
Participants
As mentioned above, the original sample for wave four involved 1054 students in 53 classes from 34 schools (51% male, 35% ethnic Roma, and 31% with low SES background). However, for the present analysis, classes with no or only one ethnic Roma student were excluded, leaving a sample of 754 students in 41 classes from 30 schools (51% male, 47% Roma, 39% low SES). In this reduced sample the average proportion of Roma students was 42% per class (ranging from 8% to 83%), while 59% of Roma and 21% of non-Roma pupils had low SES. The average size of the classes were 19 students, ranging from 10 to 25 pupils.
Procedure
Students completed the self-administered survey on tablets during regular classes under the supervision of trained research assistants. The data collection took maximum one school class (45 minutes) in each classroom. As respondents were underaged, parental consents were also obtained. Active consent (written permission) was asked before wave one for all the four waves; however, parents of non-participating and new pupils had the chance to give consent before each wave. In wave four, parental consent rate was 97% and the response rate 90%. In addition to the student questionnaire, form teachers were also asked to complete a questionnaire about the pupils using similar sociometric nomination procedures (they could select from the list of all students those pupils who met a certain criterion).
Peer nominations
Coolness
Students were provided with a list of all classmates and were asked to select those peers they considered “cool”. Incoming nominations were summed and divided by the number of respondents, thus creating a score between 0 and 1. In the quantitative analysis this coolness score was used as the dependent variable representing reputational status.
The same procedure was applied to calculate the scores for the following independent variables:
Friendship
Students were provided with a list of all classmates to indicate their relationship with them. Relations were measured on a five-point scale including: “I would never be friends with him/her” (−2), “I don’t like him/her” (−1), “He/she is indifferent to me” (0), “I like him/her” (1), “He/she is my good friend” (2). Since this question measures both friendship and likeability, calculating an “acceptance/preference score” could be problematic. Consequently, the “He/she is my good friend” nominations were recoded to 1 and all the other nominations to 0, and a “friendship score” was calculated following the procedure described above.
Self and teacher reports
Roma
Students could choose from the following four categories of ethnicity: “Hungarian,” “Roma,” “both Hungarian and Roma,” or “member of another ethnicity.” Those students who identified themselves as “Roma” or “both Hungarian and Roma” were coded as Roma (1) the others as non-Roma (0) (there were only 28 pupils who selected “other ethnicity” in wave four, they were coded as “non-Roma” for the current analysis). For students with missing data (74 pupils), ethnicity was imputed based on the previous waves: if they had selected “Roma” or “both Hungarian and Roma” earlier, we coded them as Roma, otherwise as non-Roma. In addition to individual-level ethnicity, we also involved the proportion of Roma students as a class-level control variable (
Smoker
Students were asked whether they smoked. They could select from the following options: “Yes, regularly,” “Yes, but only in company,” “No, but I have tried it,” “No never.” Those who selected either of the first two options were coded as “smokers.”
Disadvantaged
According to the Hungarian Child Protection Act (Child Protection Act, 1997), children are considered “disadvantaged” if their parents have low educational level or are long-term unemployed, or if they live under insufficient housing conditions or in a segregated environment. Those children who meet more than one of these criteria are considered “multiply disadvantaged.” Schools have administrative data on disadvantaged students as pupils belonging to these categories are entitled to certain benefits (e.g., free school lunch). In the teachers’ questionnaire, form teachers were asked to select “disadvantaged” and “multiply disadvantaged” pupils from the list of all students. Those students who were selected as belonging to either of the two categories were coded as “disadvantaged” for the present analysis. Additionally, the proportion of disadvantaged students (
GPA
End-of-semester grades were obtained from schools. The GPA was calculated from the grades students obtained at the end of the previous semester in the following subjects: Hungarian literature, Hungarian grammar, mathematics, and history. The Hungarian school system uses a five-point grading scale ranging from 1 (fail) to 5 (excellent). We used two variables in our analysis, the class-level mean GPA (
Descriptive Statistics of the Variables Used in the Multilevel Models.
Qualitative Data
Participants
There were 21 focus group interviews conducted in ten classes from eight schools including altogether 144 students. All the 171 pupils in the ten classes were invited to these interviews (84% response rate) and active parental consent was also obtained for the participation in the qualitative research. Although sociodemographic information was not collected directly during the interviews, according to the quantitative database, 54% of this sample were male, 68% ethnic Roma, 19% had low SES; thus classes with a higher proportion of Roma students were somewhat overrepresented compared to the quantitative database. Ten of the groups were fully gender segregated, three boy groups included only one girls, while the remaining groups were somewhat more mixed; however, these groups often also involved the “remaining” pupils. Overall, there was a strong tendency toward gender segregation. Although ethnicity was not registered directly, based on the interviewers’ perception and students’ ethnicity-related remarks, groups in mixed-ethnicity classes typically seemed to be mixed ethnically (11 groups), whereas groups in schools with a very high proportion of Roma students were, naturally, ethnically segregated (10 groups).
Procedure
The interviews were administered by a research team of five (moderators and co-moderators/observers), typically in two simultaneous groups at a time, on the premises of the school during regular classes, but in spaces assigned specifically to the interviews (e.g., library room, unused classrooms). Before the interviews students were asked to self-organize themselves into groups between the size of five and eight, thus letting students decide on the group composition. Parental consent for participation in the interviews was also obtained. The interviews started with a warm-up session when each participant could express their opinion about the survey questionnaire. Then students were asked to individually write those characteristics that they believed made someone popular in the class on a post-it note (one characteristic per note), and after a few minutes the moderator collected these notes. These characteristics were then ranked and discussed by the whole group. There was a list of characteristics (being good at sports, having good/bad grades, having cool gadgets, drinking alcohol and smoking) that moderators were asked to discuss in case any of them did not come up during ranking. This section was followed by a few questions about ethnicity and conceptualizations of friendship at the end of the interviews. On the transcripts of the interviews thematic coding was applied by the author of the present study following a “reflexive” approach (Braun et al., 2018); the final codes were the result of an iterative coding process and the construction and interpretation of themes were made by the author.
Results
Qualitative Results
When students listed and ranked characteristics they believed contributed to popularity in their class at the beginning of the interviews, they almost exclusively ranked physical strength, good academic performance, being good at sports, and in groups where it was listed, being funny/humorous at the first three places. However, as in-depth discussions of popularity-related traits evolved, the importance of negative behavioral patterns such as the lack of respect toward teachers, verbal and physical confrontation with peers, and “arrogance” also emerged. Simultaneously, good academic performance was rarely mentioned later, and was mostly considered unimportant or somewhat negative from the perspective of popularity. The following sections will briefly present some results related to the most salient themes of the interviews.
Athleticism
As expected, being good at sports was reported to be a salient component of school popularity. In the case of boys, this appeared to be the most important component in every group, although some forms of sports, soccer in particular, seemed to be more important than others. Additionally, boxing seemed to be the second most important sport among boys, which was probably related to the importance of physical strength (see below). In the case of girls, a wide variety of sports were mentioned (e.g., volleyball, handball, but also football). However, in some girl groups some “resentfulness” toward sports was also observable.
(Mixed ethnicity boy group, rural area)
[All quotations were translated by the author, original names were replaced with English pseudonyms to ensure anonymity. Hungarian pronouns have no gender.]
Physical strength and aggression
Physical strength, the ability to protect oneself and one’s friends, was a recurring theme in most boy groups. While the initiation of verbal and physical confrontation was almost exclusively presented in a negative light, the ability to “protect oneself” from such aggression appeared to be highly desirable. Verbal aggression was mentioned both in the case of boys and girls, while physical confrontation, with a few exceptions, was predominantly related to boys. Verbally aggressive and popular peers, especially girls, were sometimes described as “arrogant.” Typically they were widely disliked by group members and their popularity was associated with denigrating and mocking others.
(Mixed gender, mostly Roma group, small town)
Importantly, while “authentic” physical strength was admired among boys, aggressive but physically weak boys were looked down on.
(mixed-ethnicity boy group with one girl, capital city)
Kindness, openness, and resilience
During the interviews, the importance of kindness, friendliness, and helpfulness were frequently emphasized. However, they were often verbalized as the negation of non-desirable characteristics (e.g., not mocking others, not being rude), thus sometimes they might have been part of pupils’ desires rather than actual characteristics contributing to status. The ability of “being able to talk to everyone”’ and some forms of resilience to teasing (not being “too touchy”) also frequently came up, typically in combination with the prosocial traits mentioned above.
(mixed-ethnicity boy group with one girl, capital city)
Bad behavior
The role of “bad behavior” (a term pupils frequently used) was more controversial. It seems that “softer” forms of bad behavior (e.g., jests aimed at teachers and classmates) could greatly contribute to one’s popularity in many classes, while “too much” bad behavior could have had the contrary effect. Importantly, students were typically really open about this issue and its (assumed) contribution to popularity.
(Only-Roma girl group, rural area)
Smoking
Substance use is one form of risk behavior that is often discussed in the literature. Interestingly, while pupils were quite open about other forms of “bad behavior,” they tended to become visibly defensive and tense when asked about smoking. They heavily denied that anyone smoked and often controlled each others’ answers to ensure that no one mentioned names. Students occasionally mentioned that parents were going to “kill” them in case they found out they were smoking, while such concerns never came up with regards to physical confrontations with peers or disrespect toward teachers. However, after being repeatedly assured that no-one would learn about their smoking habits, students often admitted that smoking could make someone more popular among peers.
(Only Roma girl group, rural area)
Good academic performance
Most students claimed that school performance did not affect popularity. However, peers with good academic performance were sometimes considered “antisocial” or “boring” and students with good grades had to prove they were not “nerds.” Interestingly, hostility toward academically engaged students was most typically verbalized in some of the only-Roma girl groups of the sample.
(Only-Roma girl group, rural area)
Additionally, in many cases, strong resentfulness was verbalized toward those peers who were perceived as the “teacher’s favorite” for getting “special treatment” or for being “too kind” toward the teacher.
Quantitative Results
In the quantitative phase of the analysis, random intercept multilevel models were built for coolness (Table 2). In each model, intercepts were allowed to vary across classes, and standard errors were clustered at the class level. In Model 1, the overall effects of individual- and class-level variables were observed, while in Model 2 gender and ethnic interactions were added. In Model 3 some of the nonsignificant effects and interactions were removed, while in Model 4 peer-perceived “smartness’ was used instead of GPA to capture the academic dimension (the two variables are highly correlated: 0.77), otherwise the model is identical to Model 3. Information on GPA and disadvantaged social status was not available for all classes, thus in Models 3–4 the case number also increased.
Multilevel Models of Coolness.
p
We can see that in the model without interactions (Model 1), athletic abilities, verbal aggression, physical appearance, having a lot of friends, gender, and ethnicity are statistically significantly associated with coolness, while GPA, smoking, disadvantaged social background, and being considered smart or the teacher’s favorite are not associated with coolness in any of the models. After adding interactions (Models 2–4), some gendered and/or ethnic patterns appeared in the case of sports, physical appearance, and verbal and physical aggression. In order to have a better understanding of these differences, the coefficients were calculated and joint significance tests carried out for each demographic group (Table 3). Our results show that sports were significantly associated with coolness for Roma and non-Roma boys alike, and in two of the models, to a slightly smaller extent, for Roma girls as well. Verbal aggression was very strongly associated with coolness for non-Roma girls, and to a smaller extent but statistically significantly for Roma girls and non-Roma boys as well. On the other hand, surprisingly, physical aggression was statistically significant only in the case of Roma girls and this association was strong. Interestingly, physical appearance was more strongly associated with coolness for boys than girls, in particular for non-Roma boys.
Joint Coefficients and Significance Tests (1. Non-Roma Girls; 2. Roma Girls; 3. Non-Roma Boys; 4. Roma Boys).
Discussion
The paper investigated reputational status dynamics (coolness and popularity) in a Hungarian primary school sample taking a mixed methods approach. Most of our results are in line with the findings of the international literature, while the research also uncovered some interesting ethnic and gendered patterns. The interviews and the multilevel regression models both underpinned the importance of athleticism for boys, while it seemed less important for girls. In Hungary, Roma students have much lower grades on average than their non-Roma peers (in the sample the mean GPA for Roma girls was 2.74, for non-Roma girls 3.62, for Roma boys 2.40, for non-Roma boys 3.23), which could incentivize them to excel in other areas for instance in sports. However, our results show no important ethnic differences in the case of boys, although some of the multilevel models showed statistically significant association for Roma but not for non-Roma girls. This may be due to the fact that, contrary to the presence of many high profile African American athletes in the US, there are very few widely known Roma athletes in Hungary that could serve as role models. More in line with Roma cultural traditions, the role of music and arts could serve such role; however, unfortunately, the present research did not focus on this area. Additionally, the interviews also revealed a prestige hierarchy of sports, with soccer and boxing being on top. This combined with the more limited availability of other sports in the disadvantaged and/or rural areas most of the sample came from might also structurally hindered most girls from gaining status through sports, as boxing and soccer are traditionally considered as “masculine” sports.
Academic performance, as measured by school grades, did not seem to play any role in peer status according to the multilevel models, which was mostly confirmed by the interviews as well. However, the strong resentment pupils voiced during the interviews toward academically engaged students who showed “too much kindness” to teachers was not observable in the multilevel models. Of course, it is possible that students’ interpretation of the “teachers’ favorite” question in the survey also involved pupils who were good at sports or were kind and sociable, that is, students who were also popular among their peers, in addition to academically engaged students. Importantly, the multilevel models showed no ethnic or gender differences in the relationship between academic performance and status, contrary to some of the predictions of the international literature. As we have seen, during the interviews some girls in only Roma groups voiced strong resentment toward academically engaged peers, which happened in two out of the three almost exclusively Roma classes. However, the multilevel models show that such “oppositional culture” was not prevalent in the full sample, in spite of the Roma students’ much lower GPA and the social marginalization of the majority of the Roma population in Hungary.
The results regarding the different forms of aggression are also interesting. Although the survey contained items that could be associated with relational aggression (e.g., “Who do you think talks about you behind your back?”), only a few per cent of respondents selected any names, thus the statistical analysis of such answers was not feasible. On the other hand, the survey distinguished two forms of overt aggression, verbal and physical, which allows us to draw more refined conclusions about the different gendered and ethnic patterns. Although some of the literature associated overt aggression more strongly with boys and with some disadvantaged ethnic minorities, we saw that in our sample, according to the multilevel models, verbal aggression was most strongly associated with status for non-Roma girls, and this association was statistically nonsignificant only for Roma boys. On the other hand, the association with physical aggression was only statistically significant for Roma girls, which partly confirms the expected ethnic but not the gendered patterns. The qualitative data also underlined the important relationship between verbal aggression (mocking, denigrating others) and status in the case of girls, while the role of physical aggression in the case of Roma girls was a result largely unexpected based on the interviews.
The case of smoking was an interesting example that demonstrated the benefits of using a mixed methods approach. Although only eight per cent of the survey respondents (57 pupils) said that they smoked sometimes or regularly, the interviews found that this activity may have been more prevalent. Additionally, the interviews uncovered that pupils’ smoking was strongly stigmatized by their parents and teachers, which made students hesitant to admit this activity even during the group interviews. Contrary to the multilevel models, the group interviews did imply that pupils could gain status by trying out smoking. However, there were regular smoker among both the high and low status students, which can also explain the nonsignificant quantitative results.
Naturally, the results presented in this paper have several limitations. First, the data derive from a relatively small sample that is not representative of the Hungarian early adolescent population. Although the main priority of sociometric peer relations research is to obtain data from full or nearly full closed groups and our survey did well with its 90% response rate, future research would certainly benefit from the investigation of larger Hungarian and Eastern European samples, possibly involving more regions or even countries. Second, group composition and interviewer characteristics (sex, ethnicity, age, etc.) could have had significant impact on the focus group results, for instance, the self-selection of pupils into groups could lead to participants performing certain social roles in front of their friends. Although this is a general concern for all focus groups in similar settings, we found that letting students self-organize the groups led to students typically being more comfortable and self-confident during the interviews. Since more extensive ethnographic fieldworks have been successful in exploring social roles and motives for public behavior, mixed methods researchers in the future could also try to combine ethnographic and survey methods. Third, the focus group results could have been affected by the questions of the preceding survey research. Indeed, answers provided by pupils at the initial listing and ranking of characteristics at the beginning of the group interviews were often fairly similar to the options and phrasing of the survey questionnaire. However, during the rest of the interviews several topics that were not part of the survey (e.g., physical strength, kindness, friendliness, humour, “bad behavior”) were raised and discussed in details by pupils. Finally, the “reflexive approach” (Braun et al., 2018) followed for the thematic analysis means that only the author coded and interpreted the interviews. This might limit the scope of possible interpretations, as other analysts may interpret the same data differently. However, in our view, the benefits of applying a “truly qualitative” approach overweight this disadvantage, as the person who is familiar with the goals and background of the research as well as with the atmosphere and dynamics of the interviews has a more complex understanding of the observed phenomena than an analyst who only codes the transcripts without this background knowledge.
In spite of these limitations, in our view, the study presented here has several strengths. First, to our knowledge, this is the first paper that focused on reputational status dynamics in a Hungarian early adolescent sample involving a wide range of variables/factors and one of the few studies conducted outside of the “Western” context. Second, the application of a mixed approach, which is also rare in the peer status literature, enabled a more complex and refined interpretation of the results. Third, the focus on gender, ethnicity, and the intersections of the two dimensions underlined some interesting patterns, such as the role of physical aggression in Roma girls’ reputational status or the larger role of verbal aggression in girls’ status. Finally, the distinction between physical and verbal overt aggression in the survey data also provides an important addition so far less typical in quantitative analyses.
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was funded by the “Lendület” program of the Hungarian Academy of Sciences (project name: “Competition and Negative Networks: The Origin, Dynamics, and Harmful Consequences of Negative Relations”).
