Abstract
Digital platforms have reshaped childhood socialisation by redirecting role model preferences towards online content creators. This shift necessitates a re-examination of socialisation processes, particularly in non-Western contexts where algorithmic media intersect with family-based value systems. Drawing on qualitative data from open-ended questionnaires with 100 primary school children in Türkiye, this study examines children's role model choices and underlying motivations. The findings reveal a functional differentiation in children's social worlds. Family members are selected as moral anchors for qualities such as honesty, care, responsibility and compassion. In contrast, digital micro-celebrities are admired for performative and aspirational attributes, including appearance, wealth, popularity and entertainment. These results indicate that digitalisation and media do not result in a wholesale erosion of traditional values but instead produce a negotiated compartmentalisation of moral and aspirational dimensions within children's everyday socialisation.
Keywords
Introduction
For decades, the figures that children aspired to be were defined by the logic of broadcast media (Lotz, 2007). The television, where these figures appear offered a limited number of models and primarily enabled one-sided interactions within restricted time frames. Television screens presented distant, polished and often fictional role models – from superheroes to movie stars – whom children could admire but never touch (Meyrowitz, 1986).
The landscape of childhood socialisation is undergoing a seismic shift driven by the ubiquity of digital media. While socialisation has traditionally been understood as a process where individuals learn values and beliefs from immediate circles such as family and school (Grusec and Lytton, 1988), the rapid expansion of communication technologies has radically altered this environment. A child situated in any village globally with a mobile device and internet access can engage with millions of social media phenomena in any manner they choose, at any time they desire. Furthermore, another child with the same opportunities can broadcast live from their computer or mobile phone and engage in simultaneous online discussion with their viewers and followers on a range of topics, including their family, school, friends and leisure activities. Today, digital connectivity has become the default infrastructure of daily life (Boyd, 2014). As screen time rivals face-to-face interaction, the significant others in a child's life are no longer limited to physical proximity or TV schedules; they are increasingly algorithmically curated personalities (Gillespie, 2014) on platforms such as YouTube and TikTok.
This shift represents more than just a change in technology; it is a transformation in the nature of fame and authority. Unlike the unreachable celebrities of the mass media era, contemporary digital role models or ‘micro-celebrities’ (Abidin, 2018) build their influence on a performance of intimacy and relatable authenticity (Cunningham and Craig, 2017). As Lewis (2020) argues, YouTube creators explicitly differentiate themselves from the polished and elitist nature of broadcast television by weaponising relatability and authenticity. Influencers have evolved into primary ‘socialisation agents’ who offer orientation and simplify complex issues through perceived intimacy (Schmuck, 2026). Children do not just watch these figures; they feel a connection to them. Through vlogs, direct addresses and comments, YouTubers create a sense of parasocial friendship that traditional TV stars could never achieve. For a primary school child, a YouTuber is not a distant idol, but a digital peer who seems accessible, trusted and real (Stein et al., 2022).
Despite these profound changes, there is a need to understand how these digital shifts manifest in the specific role model preferences of primary school children. While existing literature acknowledges the risks and opportunities of digital media usage (Lemish, 2015; Livingstone, 2009), fewer studies have empirically examined how the traditional role models (parents, teachers) act in competition or conjunction with these new digital agents in the post-pandemic era. The blur between the real and the virtual necessitates a re-examination of who children want to be and why.
Furthermore, we still know little about how this global digital culture interacts with strong local traditions. Most research on digital childhoods originates from Western contexts. Türkiye offers a unique case study: it is a country with deeply rooted family values, yet it has one of the highest rates of digital media usage among children (TÜİK, 2024). This study, therefore, aims to explore a critical question: In a post-broadcast childhood, who wins the battle for the child's admiration? Does the global appeal of the YouTuber displace the local authority of the parent, or do children find a way to negotiate between these two worlds? By examining the role model preferences and underlying motivations of primary school children, this study documents the shifting power dynamics between traditional authority and algorithmic fame.
Conceptual framework
Role model
The concept of the ‘role model’ has long been central to understanding identity formation. Traditionally defined as individuals who occupy high social status and are perceived as worthy of emulation (Akbulut, 2016; Pleiss and Feldhusen, 1995), role models serve as behavioural blueprints that motivate individuals to adopt new habits or pursue ambitious goals (Morgenroth et al., 2015). However, the mechanism of this influence is not static; it evolves with the communication environment. While early research focused on direct personal contact or distant political figures (Bucher and Stelling, 1977), the digital age has introduced a new paradigm where role models are mediated through screens, creating a hybrid form of intimacy and distance.
In this context, Bucher and Stelling's (1977) typology of the ‘partial role model’ becomes particularly relevant for analyzing digital influencers. A partial role model exemplifies a specific skill or trait rather than a holistic lifestyle. Similarly, YouTubers and TikTokers often present a curated, performative version of reality – highlighting skills in gaming, humour or consumption – which allows children to emulate specific desirable fragments without access to the creator's full private life. This aligns with the concept of ‘wishful identification’, where children model themselves after media figures they aspire to be like, often driven by the figure's perceived success, wealth or popularity (Hoffner and Buchanan, 2005).
Role model characteristics
Social Cognitive Theory suggests that individuals gravitate towards models who exhibit similar characteristics such as race, age and gender (Hamlen and Imbesi, 2019; Zirkel, 2002). Nevertheless, some studies (Bagès et al., 2016) indicate that messages of achievement are more impactful than the gender of the role model. As posited by Allen and Collisson (2020), the influence of role models is enhanced when individuals perceive a similarity with them. However, if the degree of similarity is low or the achievements of the role models appear excessive, individuals may perceive these achievements as unattainable. Gartzia et al. (2021) demonstrated that approachable role models enhance individuals’ motivation and goal setting.
In the digital realm, this perceived similarity is amplified through parasocial interaction (PSI), the illusion of a face-to-face relationship with a media figure (Horton and Richard Wohl, 1956). Unlike traditional TV stars who remained distant, modern micro-celebrities use direct-to-camera addresses and interactive features to foster a sense of friendship and authenticity (Abidin, 2018). For a primary school child, a YouTuber is not just a distant idol but a digital peer whose perceived accessibility enhances the modelling process (Stein et al., 2022). Thus, the shift towards digital role models alters not only who children follow but how the relationship forms, moving from revering authority figures to engaging in interactive, parasocial connections.
The evolution of role models: from traditional authority to digital micro-celebrity
Historically, research on role model preferences has highlighted a duality between immediate authority figures and distant media icons. Early studies indicated that children primarily identified parents (Anderson and Cavallaro, 2002; White and O'Brien, 1999) or distant media figures such as sports heroes and TV stars (Biskup and Pfister, 1999; Duck, 1990) as their ideal selves. In this broadcast era, the relationship with media figures was characterised by vertical admiration; as Monnot (2010) describes, children acted as apprentices to distant pop stars, emulating their public personas from afar without any expectation of reciprocal interaction.
However, the proliferation of digital platforms has fundamentally disrupted this traditional hierarchy, shifting the focus from ‘distant stardom’ to accessible intimacy. Unlike the polished and unreachable celebrities of the twentieth century, contemporary micro-celebrities (YouTubers and TikTokers) build their influence on a curated public persona that mimics the spontaneity of real life (Archer and Robb, 2024).
This shift transforms the role modelling mechanism from simple admiration to a complex, algorithmically mediated relationship. While influencers appear to offer a transparent window into their private lives, they are actually engaging in a performance of authenticity designed to generate trust and relatability (Archer and Robb, 2024). Furthermore, this choice of role models is not entirely organic; it is heavily structured by data-driven interfaces. As Das (2023) argues, children's encounters with these figures are shaped by platform algorithms that prioritise engagement, effectively acting as invisible gatekeepers that curate who is worthy of admiration. Consequently, the modern role model is no longer just a figure of moral or professional excellence, but a digitally optimised entity that blends the authority of a star with the intimacy of a friend.
Present study
Role modelling patterns are highly sensitive to cultural contexts; however, the majority of research on digital childhoods originates from Western contexts. Consequently, there is a need for empirical evidence from diverse cultural settings like Türkiye to understand the global applicability of these shifts. As a country characterised by strong traditional family structures yet possessing one of the highest rates of digital media usage among children in Europe (TÜİK, 2024), it presents a distinct context where global algorithmic forces collide with local moral authorities. By providing data from the Turkish context, this study aims to facilitate a more comprehensive understanding of how global digital platforms interact with local socialisation patterns in the international literature.
Consequently, this study moves beyond simple frequency counts of ‘who children like’. Instead, it aims to deconstruct the functional stratification between traditional and digital agents. We posit that in a high-choice media environment, children do not simply replace parents with YouTubers but rather engage in a complex negotiation of identity. By focusing on the ‘process’ rather than just the ‘outcome’, this study seeks to understand how primary school children reconcile the moral authority of the home with the performative authenticity of the screen. This article contributes to communication scholarship by demonstrating that children's engagement with digital micro-celebrities does not necessarily entail a confusion or erosion of moral values. Instead, drawing on qualitative evidence from Türkiye, the study shows that children functionally compartmentalise their role models.
In contrast to previous studies in this field, the present study focuses on the process of role modelling, in addition to children's role model preferences and the reasons behind their role modelling behaviour. Consequently, the aim of this study is to understand in detail how primary school children balance traditional (parents, teachers) and digital (YouTubers, TikTokers) role models at a time when the boundaries between online and offline socialisation are increasingly blurred and digital connectivity has become the default infrastructure of daily life.
Method
Research design and participants
This study employed the basic qualitative research design. It seeks to answer three main questions about modelling: (1) Who do the participants take as role models? (2) Why do they take these people as role models? (3) What are the initial stages of the process by which the participants identified the individuals they considered to be role models? The study employed a typical case sampling technique, which is one of the purposive sampling methods. In this technique, individuals who are lacking in knowledge regarding the subject matter in question are brought together with individuals who possess a deeper understanding and can provide valuable insights to illuminate the nuances of the phenomenon under investigation (Patton, 2014). A total of one hundred third and fourth-grade students, aged between 9 and 10, participated in the study. Table 1 presents the information in question.
The participants of the study.
The participants came from a medium-sized Turkish city and were all from middle-class, Muslim families. Family and kinship ties are very strong in Türkiye, and families tend to be very protective. As there was no official evidence of other ethnic or religious groups living in the city, the sample was representative of the local population.
Instrument
The researchers employed an open-ended questionnaire as the primary data collection instrument. To prepare the interview form, a comprehensive literature review was conducted, similar studies were examined, and a conceptual framework was developed. The pilot interview form was submitted to experts for assessment of its suitability for participants, for correction of grammatical and spelling errors and for identification of any scientific inaccuracies. Following the feedback from the experts, the interview form was prepared for a pilot application. To facilitate a valid pilot application, schools that were like those where the actual application would be conducted were selected. In this context, the pilot application was conducted with a sample of 10 students from two primary schools. Following the pilot application, the data was subjected to analysis, the questions that students had difficulty in understanding were edited, and the interview form was finalised.
Procedure
An approval from the university's ethics committee was obtained for the purposes of this study. Once the requisite permissions had been obtained from the relevant authorities, the teachers were visited, and the requisite studies were initiated with regards to parental permissions. A consent form was dispatched to the parents who had indicated their willingness to participate in the study. Data was collected at school to minimise the potential influence of parental involvement on children's responses and to allow children to express themselves freely. Ethics committee permission was obtained from Gazi University with the number E-77082166-302.08.01-263528.
Data analysis and rigour
The data was analysed using a content analysis approach. The data regarding the role model preferences and reasons of the participants were presented in a descriptive manner, with the figures expressed as percentages and frequencies. The students’ responses were subjected to a process of coding and subsequent categorisation. For example, if respondents indicated a preference for individuals featured on television or YouTube as role models, they were classified within the ‘Celebrities’ category. Consequently, seven distinct categories of role models were identified. The findings for girls and boys are also indicated separately, using different colours. The category of reasons for choosing role models has been categorised based on the number of statements made by students. To ensure the trustworthiness of the analysis, a second researcher blindly coded 20% of the dataset. The inter-coder reliability was calculated using Miles and Huberman's (1994) formula (Reliability = Agreements / [Agreements + Disagreements]), resulting in a 92% agreement rate. Discrepancies were resolved through discussion to reach a consensus on the final thematic categories.
The participants were furnished with sufficient information during the data collection process. Given that the participants were of primary school age, it was assumed that they were participating in the questionnaire-filling process for the first time. Consequently, a sample questionnaire-filling demonstration was conducted. It was ensured that the participants were made aware that the questionnaire was not an examination. During the completion of the questionnaires, one of the researchers and the teaching staff helped and guided the students throughout the process. Thus, efforts were made to minimise the impact of external factors and potential sources of error that could compromise the validity of the study. To ensure the validity of the qualitative data, the researchers engaged in a collaborative analysis and the creation of categories.
Findings
Role model preferences
Figure 1 shows the role model preferences of 9–10-year-old children according to the number of role models they cited.
Participants’ role model preferences.
The most frequently cited source of role models for primary school students was celebrities (34%). Following the role models in the virtual environment, the most frequently modelled individuals were family members (27%). Family members were followed by teachers (13%), relatives (11%), peers (7%), historical figures (3%) and professionals (1%). It can be posited that third and fourth-grade primary school students view celebrities and family members as the most prominent role models. The frequency of seeing celebrities and family members as role models is close to each other. While there is no difference between boys and girls in terms of the role models they cite, there is a difference in the role models of relatives and teachers. Female students exhibited a greater proclivity for relatives and teachers as role models than their male counterparts.
From mainstream to micro-celebrities: the dominance of digital creators
Figure 2 shows who the celebrity role models are.
Participants’ preferred role model celebrities.
The analysis of children's preferences reveals a decisive shift in the landscape of admiration. As detailed in Figure 2, the traditional ‘distant hero’ (e.g., TV characters, movie stars) – once the staple of childhood socialisation – has been effectively displaced by digital content creators. The majority of participants (63%) identified YouTubers as their primary role models, while TikTok creators (16%) followed.
Qualitative responses suggest that this preference is driven by a perception of accessible intimacy. Unlike the fictional and polished narratives of television, children described micro-celebrities as extensions of their peer groups. For instance, when describing their favourite YouTubers, participants frequently used terms such as ‘funny’, ‘entertaining’ and ‘like a friend’. This indicates that for the digital generation, the criteria for admiration have shifted from extraordinary talent or heroic deeds to relatability and consistent digital presence.
Motivations for identification: authenticity, success and lifestyle
Figure 3 presents the primary school students’ reasons for taking role models.
Participants’ reasons for taking role models.
The most common reason for emulating role models is the observation of their behavioural characteristics (24%). The behavioural characteristics of role models included goodness, care, helpfulness and the avoidance of negative language. S64, who identified his parents as role models, articulated the rationale behind his choice of role models as follows: ‘They have been benevolent towards me since my birth, providing for my basic needs. That is why I consider them to be role models’. S46, who indicated that she views a YouTuber as a role model, stated: ‘The rationale behind this decision is that the individual in question is a paragon of virtue, dedicated to the betterment of children’. Although these interactions are mediated through screens, participants perceive digital figures as genuinely ‘caring’ or ‘virtuous’. This reflects the mechanism of parasocial interaction, where the curated intimacy of the YouTuber creates a sense of direct, personal connection similar to that of family members.
The second most frequently cited reason for the students’ selection of role models was the success characteristics of these individuals, representing 22% of the total responses. Characteristics commonly attributed to role models, such as diligence and resolve, were coded under this category. Related to this issue, S100 indicated that he had selected his elder brother as a role model. ‘I adopt him as a role model because he is an exemplary student. He instils in me a passion for the pursuit of knowledge’. S84, who indicated that he emulated TikTok creators, offered the following observation: ‘He is a successful TikToker; therefore, I consider him to be a role model. I greatly admire his videos’. For digital natives, ‘success’ is increasingly equated with visibility and content creation metrics. The admiration for a successful TikToker suggests that children engage in ‘aspirational identification’, viewing the visibility of these creators as a tangible achievement to be modelled.
Subsequently, character traits such as honesty, patience and respect were identified (15%). S16, who modelled her mother due to the latter's exemplary character traits, articulated the following: ‘I cite my mother as an exemplary of patience and honesty’. S97, who identified his friend as a role model due to his honesty, stated the following: ‘I aspire to become an individual who is truthful and possesses a clear understanding of their words and actions’.
An additional rationale for role modelling was the emulation of physical and social characteristics, such as beauty, wealth and popularity (15%). All but one of the participants who indicated that they took their role models due to their physical attractiveness also took role models in the virtual environment. S81, who indicated that she emulated a YouTuber named Cemre Solmaz and a TikToker, provided the following statement: ‘She is both physically attractive and affluent. I aspire to emulate her’. Similarly, all but one of the students who indicated that they emulated their role models due to their affluence also cited role models in the virtual domain as sources of inspiration. S38, who indicated a preference for his role model due to his affluence, offered the following observation: ‘Occasionally, they engage in activities that I am unable to undertake or that exceed my financial capabilities. Consequently, I emulate their conduct’.
Preferred role models according to the reasons for taking role models
Table 2 shows the role models chosen by the students according to their reasons for taking role models.
Participants’ preferred role models according to the reasons for taking role models.
Perhaps the most significant finding of this study is the ‘functional stratification’ observed in the children's reasoning. While digital figures dominate in terms of frequency, a qualitative analysis of the reasons for these choices reveals a sharp dichotomy between ‘family’ and ‘media’.
A significant proportion (40%) of individuals who are considered role models are family members, largely due to the behavioural characteristics they exemplify. Family members were followed by celebrities, with a figure of 24% being attributed to this category. Many participants (38%) identified family members as role models for success characteristics. Twenty percent of respondents indicated that they would prefer to model celebrities due to their success characteristics. The participants did not indicate a preference for celebrities as role models based on their character traits. In other words, the participants ascribed different reasons for their choice of celebrities as role models, which did not include their character traits. With regards to character traits, family members were the most preferred option (50%). Teachers were selected by the participants as role models according to their character traits at a rate of 25%.
In stark contrast, those who selected digital influencers focused almost exclusively on performativity and lifestyle. The justifications here centred on ‘wealth’, ‘popularity’ ‘gaming skills’ and ‘physical attractiveness’. As S81 noted about a YouTuber, ‘She is both physically attractive and affluent’. This discourse suggests that children are not confusing these two worlds; rather, they are actively assigning them different functions: parents are retained for ethical grounding, while micro-celebrities are adopted as guides for consumption and social status.
As with the role models selected based on their physical and social attributes, 75% of the role models with shared interests were celebrities. Of those who indicated that their preferred role model was funny and entertaining, 77% indicated that they modelled celebrities. The participants indicated that 70% of them had modelled their teachers because they constituted a source of learning. Subsequently, family members were cited as role models by 20% of participants, while 10% identified virtual environment role models.
Role modelling process
In the study, participants were asked to describe the circumstances that led to their identification as role models. The responses were classified according to the categories, as illustrated in Figure 4.
Role modelling process.
The commencement of the participants’ role modelling process was influenced by four key factors. The initial factor is that the role model and the individual undergoing modelling must possess identical values, and the former must possess the material or immaterial resources that the latter desires. The participants indicated that the process of role modelling commenced as a result of their admiration for the achievements of the individuals they perceived as role models, which prompted them to aspire to a similar level of success. In light of the fact that success is typically socially reinforced, it can be posited that the findings obtained align with the tenets of social learning theory. S51, who emulated his uncle's achievements, stated the following: ‘My uncle became a soldier when I was five years old, and following his success in that role, he was subsequently recruited for various assignments’.
When asked how they chose these figures, a recurring theme was the sheer volume of exposure. Many children conflated ‘time spent watching’ with ‘admiration’. Responses such as ‘I watch him every day’ or ‘I never miss a video’ suggest that algorithmic visibility is being internalised as a form of social bonding. This points to a mechanism where high-frequency interaction – driven by platform algorithms – is cognitively processed by the child as a genuine emotional preference.
Some participants indicated that their process of role modelling commenced when the individual they modelled exhibited favourable conduct. The participants indicated that the role modelling process commenced as a consequence of the individuals they perceived as role models demonstrating affection, virtuous conduct, benevolence, compassion, amicability, mirth and benevolence towards them. The participants who indicated that the process of identifying a role model commenced as a consequence of the role model's positive conduct towards them were students who lacked a role model within the virtual environment.
The participants also indicated that the act of role model imparting knowledge to the modelling person represents the initial catalyst for the modelling process. S30, who indicated that he emulated his parents, offered the following insights on the subject: ‘Those individuals [family] serve as my primary sources of guidance and instruction. As a result of the positive influence exerted by these individuals, I developed a sense of respect and compassion that aligns with theirs’.
Discussion and conclusion
Role model preferences
The present study aimed to investigate and elucidate the role model preferences, rationales and role modelling processes of primary school students aged 9–10 years in Türkiye context. The study is important because real people and media characters can be a source for identity development (Hamlen and Imbesi, 2019). The study mainly reveals that the third and fourth-grade primary school students predominantly select celebrities from the virtual environment as their role models. The majority of students (63%) indicated that they look up to YouTube creators, while 16% cited TikTok creators as a source of inspiration. In previous studies (Anderson and Cavallaro, 2002; Bricheno and Thornton, 2007; Karaboğa, 2019), participants have typically identified TV series and film actors, singers and football players as role models in the virtual environment. The findings of this study provide empirical evidence for a paradigm shift in childhood socialisation: the transition from the ‘television era’ to the ‘social media era’. While previous generations modelled themselves after fictional superheroes or distant TV stars (French and Pena, 1991), the current cohort predominantly selects micro-celebrities (YouTubers/TikTokers) as role models. This aligns with Lewi'’s (2020) argument that digital creators build trust through a performance of relatability, positioning themselves as anti-celebrities who are just like their audience. Unlike Superman or a movie star, a YouTuber presents a seemingly accessible, ordinary persona. This finding aligns with Abidin's (2018) concept of ‘calibrated amateurism’, where the purposeful rawness of the content creates an illusion of reality that is more appealing to digital natives than the polished perfection of traditional media stars. In recent years, with the emergence of new forms of child celebrity that go far beyond accidental viral fame (Feller and Burroughs, 2022), this phenomenon even influences children's career preferences, leading to a shift from traditional professions to emerging roles such as becoming a YouTuber (Hudders and Beuckels, 2024).
Although there is no certainty in measurements regarding ‘what children watch’ during the time they spend online (Livingstone and Local, 2017), the examination of how children's internet and social media usage habits have evolved in recent years may offer insights into the potential impact on their role model choices. The findings of the ‘Information and Communication Technology Usage by Children’, conducted by the Turkish Statistical Institute (TÜİK, 2024), an official organisation in the country where the study was conducted, indicated that 91.3% of children in the 6–15 age group utilise the internet. A total of 53.5% of children in the 6–10 age group indicated that they regularly utilise social media. Children use the internet mostly to watch videos; social media platforms such as YouTube and TikTok emerged as the primary mode of social interaction. In the same study, 25.5% of children in the 6–15 age group stated that they spend less time with their families because they spend more time in front of the screen. Besides, the 2019 Digital report indicates that YouTube is the most used social media platform in all age groups in Türkiye. The findings of Güney's (2020) study are also in alignment with these data. The study revealed that 85.9% of 11-year-old children use social media, with an average daily usage of up to two hours.
Additionally, international research has yielded comparable findings. A study conducted by Rideout et al. (2022) on behalf of Common-Sense Media in the United States revealed that children aged 8–12 spend approximately five and a half hours per day in front of a screen, with YouTube identified as the most frequently used social media platform. A further study conducted in the United Kingdom (Ofcom, 2022) revealed that the overwhelming majority (95%) of children aged 8–11 utilise video viewing and sharing platforms, with YouTube and TikTok being the most popular. The aforementioned studies demonstrate that children are spending an increasing amount of time in virtual environments, which may result in their adoption of digital platform content producers as role models. Indeed, one of the key findings of this study was that the frequency of time spent with a role model was found to be a significant predictor of the onset of the role modelling process.
Considering the findings of this study, the study conducted by French and Pena (1991) can be viewed from a different perspective. In their study on students’ role model preferences, the researchers compared their own findings, which were gathered during the pre-television era, with those of a similar study on hero (role model) preferences conducted in 1956. The findings of the study indicated that children's role model preferences shifted from individuals in their immediate surroundings to figures on television. The subsequent evolution of role model preferences after 2020, which could be termed the post-television or internet period, saw a further shift from individuals on television to those on the internet. This may be indicative of a gradual decline in the appeal of real-life role models.
In this study, participants indicated a preference for family members, including mothers, fathers and siblings, as role models. In some studies, on role model preferences in the literature, family members (Anderson and Cavallaro, 2002; Bricheno and Thornton, 2007; White and O'Brien, 1999) and celebrities (Biskup and Pfister, 1999; Duck, 1990; Karaboğa, 2019) were found to be the most preferred. Although mothers and fathers are physically the closest to their children, celebrities are, in fact, the people who are the farthest away. Two diametrically opposed scenarios can be identified. In contrast with the prevailing view, the proximity of role models to children is not a significant factor in determining their role model preferences. It may be the case that the duration of time spent with the role model, whether mutual or unilateral, is a more effective explanatory factor in the role modelling mechanism than physical proximity.
Reasons for role modelling
Many participants expressed admiration for their role models, citing their behavioural characteristics as the primary reason for this admiration. The participants indicated that they emulated their role models by observing their positive, caring, helpful and kind behaviours towards themselves, rather than towards others. Furthermore, 75% of the participants who emulated their role models due to behavioural characteristics selected individuals within their immediate social circle, including family members, educators, peers and relatives, as role models. In analogous studies (Anderson and Cavallaro, 2002; Bricheno and Thornton, 2007), it was determined that children predominantly selected their parents or other relatives as role models due to behavioural characteristics such as trust and care.
Secondly, the success characteristics of the role models were identified as a motivating factor for participants in selecting them as role models. The participants indicated that they emulated their role models due to their exemplary work ethic, success, determination and proficiency. In the studies conducted in the literature, it has been emphasised that characteristics such as success and hard work are important factors in participants’ role model preferences (Anderson and Cavallaro, 2002; Bricheno and Thornton, 2007; Çimen, 2019; Eskiler and Küçükibiş, 2019; Ivaldi and O'Neill, 2008; Karaboğa, 2019; Polat, 2012).
Participants prioritised positive character traits such as honesty, justice and compassion, notably excluding celebrities in this context. However, physical and social attributes – including appearance, wealth and popularity – were also influential factors. Role models selected based on these attributes were almost exclusively celebrities. Consistent with the literature, female students were particularly inclined to select role models based on these physical and social characteristics (Biskup and Pfister, 1999; Ivaldi and O'Neill, 2008; Karaboğa, 2019; Raviv et al., 1996; Türk and Atlı, 2022).
The study also examined role model preferences according to the reasons for taking role models. Mapping the reasons for selection onto the preferred role models reveals a clear distinction in how children organise their social world. The findings indicate a clear dichotomy: children predominantly select family members for their behavioural virtues, achievements and character traits, whereas digital celebrities are chosen for their physical appearance, entertainment value and shared interests. This reveals a clear separation of roles in how children build their sense of self. Family members are selected for ‘intangible’ qualities, acting as the ethical anchors who provide moral guidance and behavioural standards. In contrast, individuals in virtual environments are selected based on ‘tangible’ and performative attributes such as wealth, beauty and popularity. While Western-centric scholarship often portrays digitalisation as a process that undermines or reconfigures traditional value structures (Turkle, 2011; Twenge, 2017), the current findings from Türkiye suggest a more nuanced process of value negotiation and compartmentalisation. Rather than complete erosion, moral and character-based values remain anchored within the family unit, while digital figures occupy the aspirational and aesthetic domains of children's self-construction.
Consequently, the virtual role model appears to fulfil a specific psychological function: satisfying transient aspirations related to lifestyle and leisure. While children look to their parents to learn values (how to be ‘good’), they turn to YouTubers and TikTokers to engage in ‘wishful identification’ (Hoffner and Buchanan, 2005), satisfying desires for entertainment, gaming mastery, freedom and consumption. Thus, the digital realm does not necessarily replace the moral function of the family but rather monopolises the aspirational and aesthetic domain of the child's developing self. Wölfle and Schmuck (2025) demonstrated this distinction by revealing that young people do not view social media influencers as reliable sources of information during times of crisis.
Role modelling process
In the present study, the initial stages of the participants’ role modelling processes were also investigated. Four factors were identified as being effective in this process. The initial factor is that the role model and the individual undergoing modelling must possess identical values, and the role model must possess the material or immaterial resources that the modelling person desires. The notion that individuals perceived as role models are successful and hardworking is concurrent with the participants’ aspiration for success in accordance with their developmental stages. This shared understanding of success and the value of hard work is a crucial element in the initiation of the role modelling process.
Although the characteristics of role models may appear to differ, such as success, skills, determination to work, wealth, character traits, profession, popularity, freedom and pleasures, the results suggest that the underlying mechanism of role modelling is essentially consistent. This mechanism is founded upon the aspiration to obtain social or in-group approval. What has changed, however, is the currency of this approval. In the traditional sphere, social reinforcement is tied to communal values such as justice and courage. In the digital sphere, reinforcement is quantified through the metrics of popularity (likes, views, followers). Therefore, children are essentially rational social actors adapting to the ‘attention economy’ (Marwick, 2013). Thus, this is not a confusion of values but an adaptive response: children are essentially learning to speak two different social languages – one for the private sphere of the family, and another for the public sphere of the attention economy.
A critical finding of this study is the role of time spent in initiating the modelling process. In traditional sociology, time spent implies physical co-presence (e.g., with parents). However, in the digital realm, this manifests as ‘parasocial time’. The recurring consumption of a YouTuber's daily vlogs creates a sense of familiarity and routine. As Horton and Richard Wohl (1956) theorised, this ‘intimacy at a distance’ allows the media figure to enter the child's social circle as a ‘friend’. Our data suggests that when the frequency of digital interaction rivals that of family interaction, the parasocial bond becomes strong enough to function as a primary source of social learning. Rather than merely competing with traditional agents, these digital figures operate alongside them as a parallel socialisation force, exerting a comparable level of influence over the child's daily habits, interests and leisure consumption.
Finally, the mechanism of this influence is not entirely spontaneous. While children attribute their choices to the time they spend watching, this exposure is significantly shaped by how platforms function (Das, 2023). The ‘parasocial intimacy’ participants describe is often reinforced by algorithms that naturally suggest engaging content. Thus, a child's choice of a role model is not merely a solitary decision but an interplay between their personal preferences and the recommendations of the digital environment.
Conclusion
The transition from the broadcast era to the algorithmic age has fundamentally altered the landscape of childhood socialisation. As this study demonstrates, children in Türkiye are not passive recipients of digital culture but active agents who negotiate a functional stratification between traditional and digital authorities. They retain the family as a ‘moral anchor’ for ethical grounding while adopting micro-celebrities as ‘aspirational guides’ for lifestyle construction.
This dual structure suggests that the fears of a ‘complete erosion of values’ are overstated; however, it also highlights a new challenge. The ‘time spent’ heuristic observed in this study indicates that algorithmic visibility is increasingly conflated with social legitimacy. Therefore, media literacy education must evolve. Beyond technical skills, curricula should focus on ‘algorithmic awareness’, teaching children to recognise how platform architectures manufacture ‘intimacy’ and curate ‘fame’. Only then can they navigate the attention economy not just as consumers of content but as critical interpreters of influence.
Limitations and further research
Due to the methodology and cultural context of this study, there are a few limitations that need to be addressed for future research. The first of these is the cultural environment in which the study was conducted. Individuals’ role model preferences can be highly influenced by the cultural context they are in Yair et al. (2014). This study focused only on Turkish culture. In future research, comparative studies from different cultures can be conducted to reveal the effect of culture on role model preferences more clearly.
Secondly, the data was collected via a qualitative open-ended questionnaire (qualitative survey). While this allowed for a broader sample size (n = 100) compared to traditional interviews, it limited the opportunity for probing – asking immediate follow-up questions to explore the subconscious motivations behind children's answers. Future research using in-depth, one-on-one interviews or ethnographic observation could provide deeper insights into the nuances of these choices.
Thirdly, this study focused on children's stated preferences. However, we must acknowledge that these preferences are not entirely free; they are shaped by algorithms. Future research should investigate how platform algorithms (on YouTube and TikTok) push specific micro-celebrities to children, potentially narrowing their role model choices without them realising it.
Footnotes
Ethical approval
This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Gazi University (Date: 11. 01.2022/No: E-77082166-302.08.01-263528.).
Consent to participate
Written informed consent was obtained from the parents.
Author contributions
The authors co-contributed to the article.
Funding
The authors received no financial support for the research, authorship and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Data availability
The data and code necessary to reproduce the analyses presented here are publicly accessible, as are the materials necessary to attempt to replicate the findings. Analyses were also pre-registered. Data, code, materials and preregistration for this research are available from the first author.