Public Figures as Role Models of Serbian Adolescents: Who Are Idols and Why?

The work deals with public figures as role models of Serbian adolescents. We had three major goals: 1) to classify adolescents’ idols and compare data to previous studies; 2) to classify reasons for idols’ admiration and to relate them to specific role model types; 3) to test the relationship between idols’ preference and adolescents’ socio-demographic features. The sample was convenient and consisted of 1404 students from primary and secondary schools in 5 Serbian towns. The question regarding public role models and associated reasons was raised and pupils were asked to name a public figure they look up to and provide explanation why. The results show that: 1) Public figures from show business are the most popular; one new category (entrepreneurs) as a subcategory (you-tubers and bloggers within show business category) has been added to the earlier classification of role models, and there are changes in role model types preference indicating a value shift in Serbian adolescents; 2) Professional success of public figures is the most valued characteristic and sensible connection between types of idols and associated reasons is established by GLM ANOVA and logistic regressions; 3) the connection between idols’ preference and adolescents’ socio-demographic features is partly confirmed showing that those features are important for understanding adolescents’ role model choices. The results have been discussed in the light of previous research and the future directions of this topic investigation.


Introduction
In the field of social psychology role models are associated with observational learning, acquiring skills important for socialization, but also for obtaining behavioral patterns which could be positively or negatively valued by society (Aronson et al. 2001: 419-430, Bandura 1969aBandura 2004: 78;Rot 1973: 107-123). Authors divide role models into different categories and use various terms to label them. Bandura (1969a: 247-250;1969b: 145) differentiates live -actual models, with whom one has direct relation from symbolic provided in the media (real people or fictional characters). Popadić (1990: 301) uses the word authority (Serbian: uzori) for models with whom one has straight interaction and an emotional bond that serve as a basis for identification. There is no direct contact with idols and emotional connection is relatively weak (Popadić 1990: 301). The term figure is adopted for models available thorough media (Yancey et al. 2002: 58). One study (Anderson et al. 2001: 109) highlights differences between individuals known personally from sport stars, media personalities (musicians, actors, television personalities, fictional characters) and public figures like politicians.
In adolescence, role models outside the family play an important role in growing independence from parents and the process of identity formation (Anderson et al. 2001: 109;Abeyance et al. 2002: 55;Terkel 2002: 358-359). Bandura (1969a: 248) considers reasons that make adolescents receptive to influence of such models. Parents often cannot be the source of functional behavioral patterns to look up to, because of rapid technological and social changes. Besides, high social mobility together with increased educational opportunities and career choices steer adolescents towards other role models (Bandura 1969a: 248). The reason particularly relevant for this work is related to youngsters' tremendous interest in the media which are a powerful socialization agent. Bandura accurately predicted that in the near future, which is the time we live in now, the media would play a significant role in shaping behavior and social norms. He even argued that technology advances will make media models more influential than family, teachers and other real people (Bandura 1969a: 249).

Problem
This study deals with role models with whom adolescents have no direct relationship. There is no agreement on their effects on youth development. Popadić (1990: 301) associates idols' influence to imitation process which is not as deeply influential as identification. Although limited to external characteristics and relatively weak, imitation still leads to interiorization of assumed values and idols' attitudes. Similarly, Yancey and associates (2002: 56) claim that even though people frequently appearing on TV have fleeting identification effect they are often selected as role models by youngsters. Ivaldi and O'Neill (2008: 396) define musician models as persons whom adolescents admire but without necessarily holding aspirations to become like them. However, Campbell and Wolbrecht (2006: 233-245) found that the more that female politicians are visible on TV, the more likely girls are to indicate an intention to be politically active. According to them, male politicians are rarely models for boys because they are already aware that political power is open for men. Likewise, Lockwood and Kunda (1997: 92-102) conclude that relevant superstars provoke self-enhancement and inspiration when their success seems attainable but self-deflation when it seems unattainable. Engle and Kasser (2005: 264-265) state that celebrities idolization cannot be neglected in terms of developmental outcomes because it partially contributes to identity construction through adolescent interests in popular media.
Our first goal is to discover who youngsters' role models are and to which meaningful categories they can be assigned. The starting point is classification developed by Popadić used in numerous Serbian surveys (Popadić 1990: 309;1995: 130;2000: 194-199;Skoko 2000: 15-22;Stepanović et al. 2009: 407). This is a system of five categories: celebrities; athletes; politicians, scientists and artists; fictional characters. A similar categorization is used in a Taiwan study (Lin & Lin 2007: 578). In the previous research into celebrities' category, consisting of musicians and actors, one subclass was added -TV role models. It included TV presenters, fashion models and other people frequently appearing on TV (Stepanović et al. 2009: 408). This change may be treated as relatively minor but we considered it significant because it revealed specific media personalities who attract young people. Thus, in this study we are interested in potential new categories, or subcategories, of role models.
More importantly, we will compare our results with previous Serbian studies in order to detect preference changes implying a value shift in young people that can be associated with significant social and political affairs (Popadić 1995: 135-138;2003: 195-200;Stepanović et al. 2009: 411-414).
The second goal is to understand reasons for role model preferences. From the constructivist perspective it is significant to capture one's perception i.e. construction of proto-model since it can influence one's behavior (Eriksson-Zetterquist 2008: 259). Thus, modeling starts with a comparison between an image of oneself and that of a person who is a proto-model, then continues with a composition of the proto-model's characteristics resulting in an image of the role, and ends with a comparison between such an image and an image of oneself, leading to directives for specific action (Eriksson-Zetterquist 2008: 259). Moreover, youngsters' choices mirror associated interests and underlying value system (Popadić 1990: 301). We will pay special attention to possible connections between role model types and particular reasons to discover if a particular kind of reasons can be associated with a certain role model type as in Taiwan study (Lin& Lin 2007: 580-584).
Our third goal is to test possible relations between certain adolescents' characteristics (age, gender, social and economic status) and role model choices in order to better understand youngsters' preferences. Related studies show association between these variables (Lin & Lin 2008: 581;Popadić 1995: 127-138;Stepanović et al. 2009: 409-411;Yancey et al. 2002Yancey et al. : 59, 2011
Procedure. The investigation was part of the extensive study regarding adolescents' spare time. The question about idols contained the following instruction: "The next question is related to your role models.
Think about persons who are public figures, regardless of whether they are alive or not, as well as about their important features. Please finish the following sentence: I would like to be like ….. (name the person), because…… (provide a reason why)".
Data Analysis. Students' role models were classified in accordance with the classification mentioned above (Stepanović et al. 2009: 408). In the previous research we placed students' explanations in the following categories: intelligence, strong personality, humanity, communicativeness, prestige, physical appearance and professional success (Stepanović et al. 2009: 408). In the present study the reasons were similarly classified (Table 4). To test the relationship between role model types and reasons we transformed every reason category into separate dichotomous variable (level 0 -student did not provide that reason, level -1 student stated that reason) and used General linear model MANO-VA. In that way we tested differences between groups of students who have different idols types regarding each reason. We also transformed role model type variable into separate dichotomous variables (level 0 -student did not mention the idol from that category, level 1 -student mentioned the idol from that category) and applied logistic regression for each variable with dichotomous reasons variables as predictors. These regression analyses (for each role model type) will suggest if students who have idols from a certain category differ from those who do not in terms of the reasons stated. Crosstabs analysis is used to test the connection of idols preference with categorical socio-demographic variables (students' age, gender, school type) and role model gender. The relation of idol type with school achievement, parents' education level and economic status was investigated by one-way ANOVA.

Results
Public figures are perceived as role models by 40.3% students. Slightly more than 500 students named 333 public figures, which is even more than in the previous research (Stepanović et al. 2009: 407), suggesting there are no universal role models. The majority of students have male idols (71%), and nearly one third (29%) prefer females. We found gender differences regarding preference of male and female models (Cramer's V=0.59, χ²=203.205, df=1, p=0.000). Only 3.9% of male students have female role models, while 42.2% of girls have male idols. Đoković (Table  1) is the most admired role model, which is understandable knowing his success and frequent appearance in the media. Half of the most popular idols are athletes, and only one is female, which is in accordance with the mentioned prevalence of male models. Table 1: Twenty most popular public role models 1 Novak Đoković (66 students) 11.1%* 11 Angelina Jolie (6) 1.0% 2 Cristiano Ronaldo (17) 2.9% 12 Svetlana Ražnatović (6) 1.0% 3 Lionel Messi (12) 2.0% 13 Dan Bilzerian (6) 1.0% 4 Bill Gates (12) 2.0% 14 Beyonce (5) 0.8% 5 Zorana Jovanović (10) 1.7% 15 Bogdan Bogdanović (5) 0.8% 6 Ivana Španović (8) 1.3% 16 Jennifer Lopez (5) 0.8% 7 Kim Kardashian (8) 1.3% 17 Lana del Rey (5) 0.8% 8 Stephen Curry (8) 1,2% 18 Selena Gomez (5) 0.8% 9 Lebron ames (7) 1,2% 19 Boban Marjanović (4) 0.7% 10 Ronaldinho (7) 1.2% 20 Branislav Ivanović (4) 0.7% *the percentage is calculated as the ratio of of students who listed particular role models Even the most popular one, Đoković, was listed as the role model only by 11% of students, which is a trend also spotted in our earlier research. In the present study one new category (entrepreneurs) and one subcategory within the category of show business (you-tube stars and bloggers) appeared (Table 2). Public figures from show business and sport are the most admired ones. Namely, 19 out of 20 most popular public figures are from these two categories (Table 1). Other categories are considerably less present. For example, only 7 politicians/rulers are role models for 14 students (out of 567 who have idols). The most popular are domestic politicians Vojislav Šešelj (f=4), Aleksandar Vučić (f=3), Emperor Dušan (f=2). The most admired foreign politicians are Nelson Mandela (f=3). One student mentioned Che Guevara and another one Cleopatra. Table 3 shows a comparison between this survey and previous Serbian research on adolescents. It should be considered with reservations because the sample structure was not the same regarding age and school type variables, which are usually connected to idol preferences (Stepanović et al. 2009: 409). It is noticeable that entertainment and sport remain the most popular categories. Show business preference slightly declines, but sport preference grows. Scientists and artists remained low popularity among adolescents, like politicians whose rating is even lower than in 2007.
As previously mentioned, reasons for idol choices were classified in a similar manner as in earlier research (Stepanović et al. 2009: 409). Prestige category is now divided into 3 classes (fame, social power, wealth) and 1 category is added regarding adolescents' interest in a role model's profession as a future occupation (Table 4). Findings are very similar to the previous research (Stepanović et al. 2009: 409). Professional success is still by far the most frequent reason for adolescents' admiration of public figures. If prestige category remained undivided it would change the most, declining from 19% to 9%.
The novelty in this research is the attempt to find a connection between role model types and associated reasons. As mentioned earlier, we tested differences between groups of students who prefer different models for each reason by General linear model MANOVA. Unspecific explanations are excluded from the analysis and combinations of reasons are divided into separate reasons. The effect of role model type is statistically significant (Wilks' Lambda=0.556, p=0.000), which means that adolescents choose various role models for different reasons. Between subjects effects test reveals group differences in following reasons: intelligence (F=10.585, p=0.000), physical appearance (F=8.368, p=0.000), professional success (F=16.790, p=0.000) and social power (F=23.831, p=0.000). Post-hoc tests for these variables were conducted. They reveal that adolescents dominantly attribute intelligence to scientists and artists and seldom to other role model types (to entrepreneurs and fictional characters a bit more often than to show business and sport celebrities). Because of physical appearance students admire celebrities from show business, sometimes fictional characters and almost never others. Professional success is mostly associated with sport and much less with other role models. Adolescents value politicians and rulers because of social power other idols lack, except entrepreneurs at times.
The other way to associate role model preferences with specific reasons was logistic regression where the categories of reasons were predictors and role model types separate criterion variables. Regression models are statistically significant for majority of model types except fictional characters and entrepreneurs (Table 5). This analysis enables us to discover the main reasons for certain role model choice. For show business models, regression coefficients of physical appearance and professional success (negative value) are statistically significant. The other reasons coefficients are not significant meaning that it is equally likely that adolescents who admire celebrities give the same reasons as those with different preferences. Athletes are dominantly admired because of their results and humanity while beauty and intelligence are characteristics adolescents would not attribute to them (negative beta coefficients). Intelligence is the only reason associated with scientists/artists, while politicians/rulers are seen as powerful but not successful.
Idols preferences are related to socio-demographic variables: gender, age, school type, parents' education, economic status, school achievement. For the categorical variables chi-square analysis was applied and three are significant (Table 6). The strongest relationship is between role model preferences and gender. Three idol categories make the difference: show business, sport and entrepreneurs. Girls prefer show business celebrities, while boys admire athletes and entrepreneurs. School type is the next variable in terms of relation strength. Show business personalities are admired mostly by primary school students and less by grammar school pupils. Vocational school students prefer athletes, unlike students from grammar schools. Contrary to that, politicians and rulers are usually chosen by grammar school students, while primary school students have no such idols. Scientists and artists, together with fictional characters, are models dominantly for grammar school pupils. Younger students like show business idols more than older ones, while politicians are role models exclusively for older students. ANOVA shows than neither school achievement nor parents' education make any difference in students with various idol choices, although parents' education is near the significance limit. However, adolescents admiring particular models differ regarding economic status (F(545.5)=2.429, p=0.034). Post-hoc tests reveal that students who like entrepreneurs perceive their status as lower than those who admire other public figures. Moreover, pupils who appreciate scientist and artists regard their financial status as better than those preferring sportsmen, but same as those who like public figures from show business, politics and fictional characters.

Discussion
Our first aim was to identify role model types' prevalence and their changes over years indicating potential value shift. Nearly 40% of subjects have role models among public figures, which is less, but close to values from previous studies (Giuliano et al. 2007: 173;Stepanović et al. 2009: 407;Yancey et al. 2002: 57). As in other studies, male role models are far more favored than female ones, especially by boys (Engle & Kasser 2005: 265-280;Giuliano et al. 2007: 166, 174;Lin & Lin 2007: 577). Findings show that the most popular are people from show business, which was also revealed earlier (Lin & Lin 2007: 578-579;Popadić 1995: 130;Stepanović et al. 2009: 408). In this category we discovered a new subcategory including you-tubers and bloggers. New trends in the media obviously affected adolescents because, apart from people who are usually on TV, they follow those who appear exclusively on the Internet. High prevalence of athletic models can be considered as significant for the future lifestyle preferences since Giuliano and coworkers (2007: 183-184) report that youngsters having athletic idols tend to engage more in sport as adults. Politicians are rarely admired by adolescents, almost exclusively by male respondents. Only one woman was named by one girl, and it is a ruler from ancient history. This is in accordance with Campbell and Wolbrecht's (2006: 233-234) interpretation that male politicians rarely appear as role models for boys because, unlike girls, they have the impression that political power is reachable. Low politicians' status seems not to be specific for our society since Taiwan study revealed the same (Lin & Lin 2007:579). Artists and scientists are rarely idols for Taiwan adolescents as well. We detected entrepreneurs as a new category, which is in line with studies from 21 st century (Bosma et al. 2012: 410-422;Lin & Lin 2007: 578). Although not frequently chosen, people like Bill Gates, Steve Jobs and Mark Zuckerberg, started to draw adolescents' attention. Comparison between Serbian studies on idols over the years indicates that entertainment orientation is dominant, especially in the last decade. This is in accordance with adolescents' leisure survey, revealing that 3 out of 5 detected behavior patterns are focused on amusement (Stepanović 2011: 40). Sport is also very popular and its prevalence rises, especially in comparison to studies before 2000, which is understandable because of sport sanctions against Serbia in the 1990s (Popadić 1995: 132). This trend can be certainly linked to Novak Đoković's success, other athletes playing abroad, but also with recent good results of Serbia in team sports. Popularity of show business figures and athletes implies a strong media influence which is in accordance with Bandura's (1969: 249) predictions. Adolescents' orientations towards entertainment in the media is spotted in leisure studies as well (MacDonald & Shildrick 2007: 339-352;Treinor et al. 2010: 73-85, Zick 2010 and often stressed from the perspective of Stebbins's concepts of serious and causal leisure and their different developmental outcomes (Bartko & Eccles 2003: 233-240;Stebbins 1997: 17-25). Scientists' and artists' popularity decline suggests the drop in academic values. Our findings reveal that politicians' popularity has been in constant decline since 1990 (Popadić, 1995: 130). In that period political debates were quite new content on TV since multi-party political system had just appeared in Serbia. It is possible that politicians' reputation significantly decreased after war conflicts, numerous elections, and constant diagnosis of our society as transitional or in economic and political crisis (Bieber 2003: 86;Jakopin & Knežević 2009: 376;Petrovich et al. 2001: 183;Tomanović & Ignjatović 2006: 269;Šuković 2014: 43, 52;Veljanovski 2014: 244). Similarly, investigating changes in connotative meaning of political terms in adolescents, Kuzmanović and Petrović (2008: 423-433) concluded that they can be valued as negative and probably related to disappointment after great expectations associated with post-Milošević's era.
The second goal dealt with reasons for idol appreciation and their connection with particular model types. The distribution of adolescents' explanations is very similar to our previous study (Stepanović et al. 2009: 409), showing that professional success is the reason present far more than others. GLM MANOVA showed that adolescents admire particular role models because of different traits: intelligence, physical appearance, professional success and social power. Adolescents attribute intelligence almost exclusively to scientists and artists, whereas professional success is primarily associated with athletes. Physical appearance is mostly admired in actors and musicians. This result is in accordance with Taiwan study (Lin & Lin, 2007: 583-584) showing that media stars are notably described with "exterior" dimension including reasons as "good-looking, " "attractive dressing, " and "attractive body shape". Similarly, Ivaldi and O'Neill (2010: 179-195) argue that musicians are more valued as celebrities than because of their musical abilities. Adolescents associate social power with politicians/rulers, and sometimes with entrepreneurs. Logistic regressions were significant for show business celebrities, athletes, scientists/artists and politicians/ rulers. They provided a more detailed preview of the reasons structure around these idol types. Hence, actors and singers are perceived as attractive but not professionally successful. Contrarily, sportsmen are admired because of their achievements and humanity, but they are not seen as attractive (unlike in Taiwan study) and intelligent. Scientists and artists are solely described as intelligent. In keeping with the mentioned Kuzmanović's and Petrović's (2008: 423-433) conclusion, adolescents attribute social power to politicians, but not success.
The analysis of the relationship between adolescents' socio-demographic characteristics and their model preferences gives us additional insight into their choices. Gender differences are the most salient and expected (Campbell and Wolbrecht 2006: 233-245;Lin & Lin 2007: 581;Stepanović et al. 2009: 409-410;Yancey 2011, 39). Findings consistently indicate that girls prefer show business celebrities while boys favor sportsmen and politicians. Age differences were also noted and, as in Popadić's (1995: 133) study, imply that show business celebrities are more popular among primary school adolescents while politicians are exclusively idols to secondary school students. Popadić also found that scientists and artists are more admired by older adolescents while our results did not show this difference. Primary school students favor show business idols. As in our previous survey (Stepanović et al., 2009: 410) vocational school pupils prefer athletes, while those from grammar schools admire politicians, scientist and artists. We agree with Popadić's (1995: 136) comment that older adolescents' preferences, like ours from grammar schools, towards politicians/rulers and scientists/artists reflect their higher education and social awareness than in younger students or those in vocational schools. Students' economic status is related to their idols' preferences. Youngsters who value scientists and artists deem their financial status better than those who prefer athletes, which is in line with our previous result (Stepanović et al. 2009: 411). Interestingly, those who perceive their family income as modest opt for our new category -entrepreneurs more than others. We consider this finding significant as it implies that those young people are motivated to change their status and to eventually reach their models. Contrary to some other studies, we did not reveal a connection of role model preferences with school achievement and parents' education (Stepanović et al. 2009: 409-410;Yancey 2011:40).

Conclusion
Our results regarding great prevalence of male role models indicate the need for a more intensive promotion of female ones, especially in the politics domain (Campbell and Wolbrecht 2006: 244-245). Entertainment dominates adolescents' leisure time and, according to our findings, their values as well. Amusement is not necessarily negative since it is important for socio-emotional development and for the transition to adulthood (Stepanovic 2011: 42). However, the balance between activities associated with Stebbins's concept of "serious leisure" and others referring to "casual leisure" is pivotal for positive outcomes of adolescents' development (Treinor et al., 2010: 183-184). As in the previous study (Stepanovic et al., 2009: 414-415) we may conclude that school, but the media as well, have to do much more towards endorsing scientists and artists as role models for young people. Low popularity of politicians reflects adolescents' distrust towards traits associated with them, but also lack of youngsters' interest in politics related to political parties but not necessarily to political issues in broader sense since young people can be politically engaged in new ways not recognized by politicians (Farthing, 2010: 185-186). Adolescents most admire professional success in role models which suggests that they are important for youngsters' transition to the world of work, although this is more to do with abstract modeling (Bandura, 2005: 13) than their wish to be engaged in the same profession. The decline of reasons regarding fame, prestige and wealth implies that adolescents are not strongly directed towards external features, except in case of physical appearance which is understandable for their age. Social power does not attract young people which is consistent with the finding regarding politicians who are rarely models. As before we also established the significance of socio-demographic factors for role model choices, especially gender, age and school type.
We managed to associate type of students' explanations with certain role models. The discovered relations seem quite reasonable and provided us with deeper understanding of adolescents' preferences. However, further and more profound investigation is necessary. Having been investigating adolescents' role models for a long time, we find that the open-ended question estimating their preferences should be revised and improved with an appropriate scale allowing use of more powerful statistical analyses. Moreover, we plan to investigate adolescents' perception of role models' influence on different aspects of their identity and behavior. Our aim is to address and empirically test the previously mentioned dilemma regarding nature and strength of idols' influence on adolescents' development. Bibiliography: