Introduction

Presently, it is almost a common place to point that adolescents and emerging adults —people between 12 and 25 years of age— are part of the “digital generation”: the age group that was born in the middle of the expansion of the Internet and digital technologies, and which has grown in a world where the daily use of instant messaging services, videocalls and social media are increasingly common. To the extent that the use of devices for digital communication and social media massifies at global level, especially among the youngest, an increasing body of academic literature has approached issues such as: the effects of time of exposure to digital communications media —including problematic use or Internet “addiction” ^{(Machimbarrena et al., 2019)} — as well as the effects on mental health, with mixed results ^{(Ellis, 2019}; ^{Dickson et al., 2019)} ; the emergence of various modalities of violence between adolescents on digital media ^{(Zhu et al., 2021)} , or else from adults toward children and adolescents, frequently under the form of victimization in the sphere of sexuality ^{(Machimbarrena et al., 2018)}; the appearance of young subcultures and communities on the Internet, characterized by the virtuality of relationships among their members, or the growing “youtuber” culture and its influence on adolescence ^{(Aran-Ramspott et al., 2018)} , for some recent references.

In this scenario, one side of the studies has become interested in the effects the use of such devices has on the way adolescents and youth establish interpersonal relationships with their peers, and the impact this has on the fundamental development processes of adolescence such as those related to identity and sexuality.

The seminal work by ^{boyd (2014)} raised questions on the ways digital technologies have changed the form this age group communicates with their peers and with the adult world. ^{boyd (2014)} argues that processes such as socializing with others, establishing personal limits and self-expression, which are proper to adolescence and the beginning of adult life, have been affected by new technologies, as these have modified the possibilities for relating, introducing (and potentially expose) to the others, expanding them much beyond the space-time restrictions that marked the development processes of previous generations. The ways in which some characteristics proper to digital media —such as the possibility of an anonymity, asynchronous communication or the simplicity to access, share or make contents viral— have been the goal of increasing research interest.

The present article intends to contribute with the existing literature, focusing on a central aspect in terms of terms of the needs and development goals of adolescence: the formation of a personal identity ^{(Pfeifer and Berkman, 2018)} . At the age of digital communication media, adolescents have to mark their autonomy from the adult world, deal with the goals of acceptance and admiration by their peers and undertake the task of forging a coherent personal identity in a context that comprises a mixture of online and offline interactions, whose limits are increasingly blurred ^{(Machimbarrena et al., 2019)} . The construction of a personal identity —which also considers self-representation before the others, particularly the reference group— increasingly involves the use of digital tools and the construction of profiles on social media. We enquire on the logics that support the construction of such profiles and digital identities associated in a sample of secondary students in Talca (N=892), centre shouth of Chile, by means of adapting a validating the instruments proposed by ^{Davis (2013)} and ^{Sohier and Brée (2017)} . The research question of this work was: how do children and adolescents use the technologic resources offered by social media to appear before the others?

Interpersonal relationships in the digital era

As their peers around the world, Chilean children and adolescents (ChA) are enthusiastic users of digital technologies and social media. According to a recent studio ^{(Criteria and VTR, 2021)}, 9 out of 10 Chilean ChA have access to mobile phones with internet, and this takes place as of 10 years of age; more than 90% habitually uses a social medium. The most used are YouTube, WhatsApp, TikTok and Instagram, and they use them to watch videos (96%) and interact with friends as (88%), even before the Covid19 pandemic ^{(León and Meza, 2018)}. Within this context, distinctions between online and offline spaces are increasingly diffuse, particularly for those who were born in a world already populated by virtual interactions. ^{Feixa et al. (2016)} referred to the generation that has grown up in the digital era are the “generation #”, characterized by frequent use of the Internet and mobile devices, for which social relationships are a mix of co-presence and constant virtuality.

This hybrid nature of relationships in the digital era would be a distinctive characteristic of the social experience of children and adolescents, and which poses the demand of redefining issues such as personal intimacy, closeness and feeling of belonging to a peer group. This makes it necessary to re-elaborate the social codes that regulated face-to-face interactions in the past ^{(Granic et al., 2020}; ^{Pfeifer and Berkman, 2018)} , “re-mediating” social relationships, where the dual meaning of the concept defined by ^{Lasén and Casado (2012)} ; on one side, it provides a new medium or support for the expression of a practice already established in culture, whereas on the other, the opportunity to re-signify or re-elaborate a practice of the offside world, as with the notions of intimacy or closeness in friendship and partner relations. The very notion of intimacy —a key element in close interpersonal relationships— would have been “remediated” in this sense ^{(Machimbarrena et al., 2019)} . Traditionally, intimacy has entailed a certain level of physical and emotional proximity, where the involved individuals share, for instance, personal information, interests and experiences from which other, non-intimate subjects, are excluded. In the virtual world, physical closeness is not a requirement any longer, and new codes are created for emotional closeness ^{(Miguel, 2018)} . This supposes that the limits between the public, the private, and the intimate are in constant renegotiation, for the broadening of potentialities known beyond the time-space limits that characterized face-to-face relationships, it is necessary to decide how information will be presented (and how much), for example, in a social-media profile. The information must be sufficient to show an attractive and relatively reliable image of the subject behind the profile ^{(Ellison et al., 2011}; ^{Casimiro, 2015)} , at once, however, they must take the inherent risk of offering information to a potentially infinite audience into consideration ^{(boyd, 2014}; ^{Duffy and Chan, 2019}; ^{Miguel, 2018)}.

These considerations may widely vary according to experience and emotional maturity of the users; according to a study by ^{Movistar Chile and Fundación Nativo Digital (2020)}, 34% of the users between 8 and 18 years has profiles and public accounts on social media, this proportion increases to 42 in the subgroup between 15 and 18 years. For youth, smartphones have become the essential device: eight of ten surveyed individuals state that the first they do in the morning and the last at night is to check on their phone; moreover, 78% of the respondents between 8 and 18 years states the Internet is a “safe place”. In this same age group, 74% is confident that their data are safe on the Internet; this figure reaches 87% for children between 8 and 14 years ^{(Movistar Chile and Fundación Nativo Digital, 2020)}.⁴

In addition to considering the potential risks of publishing personal information on the Internet, the practices of the youth in the construction of social-media profiles have been object of interest from the standpoint of their role in one of the central evolutionary tasks of adolescence: the affirmation of personal identity ^{(Pfeifer and Berkman, 2018)} . This, due to the plasticity they offer the users, in terms of the image that is sought to be displayed to the world ^{(Antón-Cuadrado and Levratto, 2021)} , and because of their insertion in a digital culture characterized by the proliferation of potential identification and socialization modes ^{(Aran-Ramspott et al., 2018)} . In this process, the task of feeling recognized and/or accepted in a peer group, central up to 14 years, and the later task of reasserting their proper values and priorities before their peers and the world (particularly significant between 14-15 and 18 years), involve self-presentation and construction processes of an “identity narrative” ^{(McAdams and McLean, 2013}; ^{McAdams, 2018)} which may be re-mediated, that is, taken to a new technologic support which might imply the re-elaboration of social codes and meanings.

The concept of identity narrative —understood as the active integration of significant experiences and personal values with the image the individuals have of themselves, in constant elaboration guided by the search for meaning and internal coherence as regards “who I am” ^{(McAdams, 2018)} — implies that identity is built connecting events of life by means of causal links and / or attributions with ex post facto meaning, connecting the past with the present and projecting it toward the future as aspirations and goals in life. The process would start in childhood, though in adolescence, it become particularly crucial for later developments ^{(Nelson and Fivush, 2019)} .

For digital-era adolescents, this process of identity development is accompanied by platforms such as Facebook, Instagram or TikTok, among others, which stress the hybrid nature of the experience, allowing: a) the extension of the possible identity references, which are no longer limited to space-time proper to face-to-face relationships; and, b) the possibility of more resources for self-expression and self-presentation before the others, since social media allow the users to “build” the image of themselves they want to portray via the images, photographs and even short stories, by means of which the virtual audience receives an elaborated image, before which they may react (for example, following or liking). In this way, adolescents “start learning and building collectively the codes of the desired and the ignored, the celebrated and the rejected” ^{(Basile and Linne, 2014: 210)} . And by doing so, they leave a “digital fingerprint” whose temporality may be ordered as an identity narration, or a reflexive self, and potentially reach adult life, exposing them to the critical look of education institutions and even future employers ^{(Duffy and Chan, 2019)} .

Si even if in the world of face-to-face interactions there are also multiple referents, social media enable greater plasticity as regards the image that can be presented, stressing the performative dimension. Accepting someone as a friend on social media, sharing thoughts and images and becoming part of a virtual community of interest stimulate a feeling of belonging and connection with the others through platforms that were explicitly designed for such purpose, and that enable ChA to build and manage much more extensive or elaborate networks than possible without these platforms ^{(Granic et al., 2020)} .

At once, the technologic possibilities of social media put forward important issues regarding self-presentation such as the limits between showing our best photograph and alter the proper image in order to respond to the expectations of the peer group that is the “audience”. By having a greater control on the impression they make on the others on social media, ChA may introduce themselves to their generation peers and/or to an “imagined audience” with a highly selective version of themselves ^{(Duffy and Chan, 2019)} . Although this process has developed in parallel to the expansion of the use and access to digital devices, this does not mean that the use of these resources on the web lacks shared rules: on the contrary, the existence of “rules of etiquette” (frequently unwritten) in the use of various social media and technologic media has been documented in the literature ^{(Gangneux, 2018}; ^{Pangrazio, 2019)} .

In this way, participants in online interactions not only share communication codes such as images and emojis, but also certain learning as regards appropriate and inappropriate behaviors on the web, depending even on the sort of platform or digital medium utilized and on the possible “audience” in each of these platforms ^{(Duffy and Chan 2019)} .

In a qualitative study in Chile, ^{Del Prete and Redón-Pantoja (2020)} provide evidence on virtual social networks emphasizes the acceptance and recognition of others. This may induce the desire of modelling their own subjectivity to the demands of an increasingly demanding audience and might restrict the development of an autonomous moral (and thereby, a more “authentic” self-representation) in the youngest social media users. Nevertheless, in an extensive revision on this topic, ^{Granic et al. (2020)} also point at the fact that online interactions and some videogames may provide ChA with a space for important issues regarding their own identity and the importance of asserting their values and personal beliefs, deciding to differentiate from the group, with positive effects for mental health and the self-esteem of boys and girls. ^{Aran-Ramspott et al. (2018)} provide evidence on the capability of adolescents to distinguish the media and entertainment role of influencers and youtubers, of their characteristics as potential identity models to follow, valuing authenticity as a desirable feature.

Here, it is important to point out that this construction process of an online identity narrative, as the counterpart developed in face-to-face contexts, is marked by social stratification and differentiation processes such as gender or socioeconomic level (SEL): virtual environments where the online profiles are created are neither socially nor culturally neutral, but crossed by sociocultural codes and interpretations that guide the social practices that occur in them ^{(Basile and Linne, 2014}; ^{Del Prete and Redón-Pantoja, 2020}; ^{Tello-Navarro and Gómez-Urrutia, 2021)} . Within this logic, the norms and codes adolescents use in the digital space frequently reproduce the power relations and asymmetries of resources originated in face-to-face interactions ^{(Mannerström et al., 2018)} .

Methodology

Our research focuses on secondary-school students in the city of Talca, Maule Region (center-south Chile). At first, the research team contacted the authorities of three schools (two public, one private) that serve adolescent populations (secondary) and which were interested in participating in the study with their entire school populations. The public/private distinction was interesting because private schools serve a population with greater access and use of internet, which might impact on the youth’s use skills. The goals of the research were explained and the necessary authorization were obtained; after this, an open invitation was made to the students, their parents or legal guardians, who received an information minute with the details of the project and an informed consent format they had to sign in order to authorize the adolescents to answer anonymous survey. The invitation was made by the research team in authorized visits to the classroom and teachers who supported in the recruiting phase without being part of the study, for the survey was applied before the Covid-19 pandemic (March 2020).

As inclusion criteria, we considered being between 14 and 19 years, having their parents’ or legal guardians’ authorization and being regular users of social media that need to create a personal profile. Those who decided to participate received an informed consent form, in which the voluntary and confidential nature of their participation. Although it was a non-probabilistic sample, similar numbers of men and women were recruited. The instruments were applied at the time of Course council so as not to interrupt the usual education activities.

Instruments: students were asked to report information such as age, sex and the grade they were studying, as well as data on their use of social media. Following, they were given two scales on digital identity, adapted from proposals by ^{Davis (2013)} , and ^{Sohier and Brée (2017)} . Davis’s scale (2013) proposes two dimensions with 11 items on a four-point Likert-type scale: online peer communication and online identity expression and exploration. Por its part, the scale by ^{Sohier and Brée (2017)} put forward four dimensions with 14 items on a four-point Likert-type scale: multiplication of the self, which refers the possibility of building multiple identities in the web; self-quest; virtual sociability; and virtual reputation. The original English-language scales were translated by means of back translation; linguistic correction and the adequation of the translation was revised by two child psychologists and two professors specialized in teaching Spanish language for the target group of the survey and two professors specialized in teaching Spanish language for the target group of the survey with a view to ensuring the adequacy of the items and that they could be understood by target population. Over this process, some of the original items proposed by ^{Sohier and Brée (2017)} underwent important changes as they were deemed unclear or very difficult to understand, and are summarized in Table 1.⁵ Answering the instruments took them approximately 40 minutes. Once processed and reviewed, the surveys, there were 892⁶ (N) valid surveys.

As analysis techniques, central tendency and dispersion measures were used, while for correlational analysis, item-test correlations and polychoric correlations ^{(Lara et al., 2018)} ; for multivariate analysis, exploratory factor analysis and confirmatory factor analysis ere considered. Three software programs were used: SPSS 22 for the analysis and selection of elements from each subsample, plus general statistics; FACTOR for the exploratory factor analysis; and RStudio, for the confirmatory factor analysis. In the analysis of internal reliability, both scales had good indicators: in the case of ^{Davis (2013)} , reliabilities were 𝛼 = .840 and ω = .846. There were adequate behaviors in terms of kurtosis, since all the items were within asymmetry ranges, whereas kurtosis within ±1. As regards item-test correlation coefficients, all were over 0.3, save item 10, which was dropped. For its part, ^{Sohier and Brée’s (2017)} had reliabilities of 𝛼 = .824 and ω = .826, with all the items within asymmetry ranges and kurtosis within ±1. To verify the existence of the original dimensions in the Chilean application, exploratory (EFA) and confirmatory factor analyses (CFA) were run, randomly dividing the total sample into two same-size subgroups. CFA results are reported for each scale:

Davis’ scale: the determinant of the correlation matrix was 0.0048. Kaiser-Meyer-Olkin (KMO) measure of sample adequacy was 0.905, considered very good ^{(Hair et al., 1999)} . Additionally, Bartlett’s sphericity test also produced a suitable result for the questionnaire, 𝜒2(45)=2355.3, 𝑝<.00001. Polychoric correlation matrixes were used for ordinal variables ^{(Lara et al., 2018)} . The correlation matrix indicated two values, concentrating 72.81% of the explained variable. Later on, exploratory and confirmatory analyses were carried out, extracting one- and two-factor solutions. For FCA, the diagonally weighted least squares (DWLS) was used, since the multivariate normality criterion was not met ^{(Lloret-Segura et al., 2014)} , adding to the ordinal nature of the items. As indicators of goodness of fit, chi-squared test (X ²), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), comparative fit index (CFI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean Square Residual (SRMR). Both EFA and CFA demonstrated that the two-factor structure is the most suitable for this sample, as observed in Table 2, concurring with the original scale.

Sohier and Brée’s scale (modified): the determinant of the correlation matrix was 0.0178. Kaiser-Meyer-Olkin (KMO) sample adequacy measure had a value of 0.866, which is considered very good ^{(Hair et al., 1999)} . Bartlett sphericity test also presented a result suitable for the questionnaire, 𝜒2(91)=1778.6, 𝑝<.00001. Being ordinal variables, polychoric correlation matrixes were used ^{(Lara et al., 2018)} . CFA was carried out on the basis of models from one up to four factors. DWLS was used because the multivariate normality criterion was not met ^{(Lloret-Segura et al., 2014)} , plus items are measured on an ordinal scale. As indicators of goodness of fit, chi-squared test (X ²), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), comparative fit index (CFI), Root Mean Square Error of Approximation (RMSEA), and Standardized Root Mean Square Residual (SRMR). According to Table 3, the values obtained indicate that the four-factor model has the best indicators regarding the rest; that is to say, it has an almost perfect goodness of fit. Once the psychometric properties of the scales were analyzed, the data were compared by means of central tendency and dispersion statistics. In like manner, comparisons were made according to sex and SEL, by means of Student’s t-test.

Results

As a first approximation, the study asked the respondent about some habits and practices in online interactions, summarized in Table 4. All the respondents had a profile on a social media site at least. As noticed in the table above, the respondents reported high levels of authenticity in their self-presentation, even though almost a third stated tried to appear happier that actually feeling. As regards hiding problems or fears, more than a half of the girls (53.9%) declared doing it regularly or always, while almost 38% of the boys admitted to doing the same. This may be because the literature reports that in general the girls are more aware than boys of the risks entailed by disseminating sensible information on the Internet ^{(Gómez-Urrutia and Tello-Navarro, 2020)} , due to the perception of more likelihood of being victims of predatory behavior on the web. About a quarter of both sexes reported telling real and personal information, showing a rather careful management of personal information. A low percentage admitted inventing information so that it seems that their lives are more interesting, with significant differences by sex.

However, consulted about their perception on what other people do on the internet, 51.9% of girls and 56.7% of boys declared believing that others lie so that they are followed. This contrasts with their own report on the authenticity of online interactions, as displayed in Table 5. Under 20% of the participants stated increasing followers and receiving likes as goals of their interactions, and more than 80% of the girls and 66% of the boys declared publishing actual images. The difference is statistically significant, which suggests that girls tend to present more images with them, as compared with the increasingly common practice of using icons or animals, cartoons or fictional characters as part of the repertoire of self-presentation, more common among boys.

This is consistent with other research works that show that personal appearance if a key factor in the girls’ presentation of the self; while boys underscore more personal qualities ^{(Basile and Linne, 2014}; ^{Gómez-Urrutia and Tello-Navarro, 2020)} which in this case may be represented via these other images that are not real, though not deceptive in nature. In the same line, 48% of the girls and almost 25% of the boys declared improving the appearance of the photographs and images using filters especially designed for such purpose. For both sexes, more than 60% reported allowing their friends to share photographs with them, as well as verifying the information they receive before sharing it —an increasingly frequent practice in adolescence.

In this study no differences were detected by SEL, but by sex; contributing to the important set of literature that points that even in the youngest generations, gender norms model online interactions in a very similar way to what occurs in the face-to-face world. On the ^{Davis’ scale (2013)} , statistically significant differences were noticed on favor of the women for online identity exploration dimension (t=-2.20, p=.028), which assesses the degree at which participants are encouraged to express and explore various aspects of their personality and preferences on the web (for example the item “I like using the Internet to try various forms to express myself”). For the peer online communication dimensions, which assesses the way the participants use the web to establish and maintain friendship networks (for example, the item “on the web, I feel I am more involved with my friends’ activities), there were statistically significant differences verified in favor of men (t=2.22, p=.027) (Table 6).

In ^{Sohier and Brée’s modified scale (2017)} , there were statistically significant differences were noticed by sex only for Virtual reputation dimension (t=3.07, p=.002), s noticed in Table 7. On this scale, such dimension refers to the use of the web as a way to present an attractive and interesting image (for example, the items “I’d do everything to have a good reputation” or “I like to show that I am someone interesting”), which in spite of reaching middle-high scores for both sexes, seems to be particularly relevant for boys.

In this scale, the highest scores were for Virtual sociability dimension, which points at the use of the web as an element that facilitates establishing more fluid interactions with other people (items such as “On the web, I am able to express more easily that in real life ” and “I feel freer to speak to other people when I’m on the web”). The other two dimensions had average scores. We discussed the implications of these results, in the context of the existing literature, in the following section.

Discussion

The result verify that the presentation of a personal identity and image carefully constructed is very important for ChA. Though, the resources and strategies utilized are different according to sex: for girls, the exploration and presentation of identity is more relevant and within this context, physical appearance has a very important role. This is consistent with gender representations that exalt a certain beauty stereotype as a central characteristic to define female attraction. Likewise, it would explain the difference in favor of women for Online expression and identity exploration (t=-2.20, p=.028) on ^{Davis’ scale (2013)} , as well as the fact that women tend more than women to present images of them on social media and use filters to improve them. Consistently with other research works, this finding reinforces the idea that producing a digital identity is a management task or even branding that requires specific competences ^{(Huerta et al., 2021)} , but it does not escape from gender asymmetries —and with it, those of power-, which imply that women have to be more careful with their self-presentation. Women have to be more vigilant in terms of the image they present to their peers and other imagined audiences in the virtual world —for example, avoiding images in which they appear behaving in a way considered undesirable for girls such as drink alcohol, or wearing clothes deemed “revealing”—. This pressure is important for both sexes, but it is much heavier for girls ^{(Åberg et al., 2020}; ^{Lowe-Calverley et al., 2019)} . ^{Duffy and Chan (2019)} speak of this self- surveillance or “monitoring” as an increasingly internalized feature in the culture of profiles and social media.

Nevertheless, the technologic possibilities offered by digital media have also opened roads to subvert norms and gender codes. Among these characteristics, the possibility of anonymity or partial or complete concealment of identity, and the asynchronous nature of communication are factors that make negotiation possible, and up to a certain point, to subvert gender expectation. This is particularly noticeable in the case of women, since traditionally women’s self-expression (including the expression of their bodies and sexuality) has been more socially controlled than men’s ^{(Åberg et al., 2020)} . Hence, digital media would provide girls with a broader margin to explore various faces of their own personality, decreasing the risk of being socially sanctioned as they do not behave by gender rulers.

^{Meenagh (2015)} points out that one of the strategies used by women to explore their own identity and preferences is to keep a “good girl” appearance (this is, following the existing regulations and expectations regarding women’s behavior) in the offline world, while a nontraditional identity and/or behavior in online environments, however, it is still an authentic presentation, even if exploratory, of their identity. This would be possible, indeed, because mediated interactions enable the user to have greater control on the information delivered, being feasible to alter or hide key characteristics of the proper identity, and by doing so, avoid some of the social rules of the face-to-face world ^{(boyd, 2014)} .

In a similar line, ^{Duffy and Chan (2019)} report that the fact of being able to produce more than one identity in the virtual sphere enables the youth to present an image of themselves especially produced for certain audiences, showing themselves in one way to their peers and in another to the adult world. To do so, privacy filters and/or multiple accounts on social media —as the case of “Finstas” (fake + Instagram), alternative Instagram accounts (not necessarily deceptive) created mainly to show a more spontaneous version of the self— would also be functional for this identity exploration process, but following the logic of minimizing overexposure risks. In this way, even if adolescents intend to increase their visibility level, which is related to the reach of their posts, they would also manage their image for different “audiences” and differences levels of intimacy and closeness. Even if Internet allows subverting certain logics, for example, the risk of having a “bad rap” is also greater, since in the web, information may go beyond their close circles ^{(Duffy and Chan, 2019)}.

Some statistically significant differences were found in favor of men for Online peer communication (t=2.22, p=.027) in ^{Davis’ scale (2013)} ; similarly, in ^{Sohier and Brée’s (2017)} modified scale, statistically significant differences were found by sex for Virtual reputation dimension (t=3.07, p=.002), also in favor of men. This suggests that for boys the possibility to accomplish their own recognition as interesting people, particularly before the peer group, is one of the main attractions of online interactions. In this last scale, as observed, the highest scores were for Virtual sociability, which refers to the greater ease some individuals have to interact with others and to express on the Internet, when compared with face-to-face meetings. The affinity and closeness the users establish in their digitally mediated conversations may be understood from the idea of “disclosure” ^{(Belli et al. 2010}; ^{Tello-Navarro and Gómez-Urrutia, 2021)} , defined as the fascination of speaking to someone not well acquainted over a screen and that allows saying things that would not be said face to face. This is because the possibility offered by the digital media to control the times and flow of communication and which enable, for instance, thinking well an answer or ponder a reaction in a way that is impossible in the real world, as well as the nonverbal aspects (blush, babble) which may make people feel more exposed than initially intended. Online interactions, being perceived to be more “under control” by those involved, make adolescents feel more comfortable and less exposed in this modality. Similarly, receiving virtual feedback and almost instantaneously —such as likes or gaining followers— or the possibility to be tagged in other peoples’ posts produces a feeling of belonging or being in a common space with other whom interests are shared with, but not the problems of everyday coexistence and where what is approved or not by the virtual environment is quickly known.

Conclusions

The results of this study try to contribute with the existing body of evidence about the way in which the construction of digital profiles is approached by ChA as an important source of resources to present and interact with their peers. To the extent that digital communication technologies become more sophisticated and accessible, the hybrid nature of youth’s social relationships (online/offline) tend to stress, underscoring the importance of enquiring the way new generations use technologic possibilities to explore forms of sociability. From this standpoint, our results also reinforce the idea that a significant part of social codes of the offline world —such as gender expectations and regulations— is assumed and re-elaborated in digital contexts, frequently helping the reproduction of opportunity inequalities and power (^{Mannerström et al., 2018}) that may be defining in the processes of identity definition in adolescence.

Differences found by sex in this work are in this regard, as a central part of the process of identity formation is constituted by the adoption or rejection of conventional rules —inherited from previous generations— about gender and sexuality. As pointed out, from this perspective, digital communication technologies display an ambivalent nature: even if the possibility of anonymity, asynchrony of communication and exposure to multiple identity models pose physical and psychic for adolescents, they can also be a resource to subvert traditional gender norms and logics. This would allow youth to explore ways of being in an environment that, by not requiring physical presence, provides them with a wider range of resources —as well as control, but also risks— to unfold their interactions.

Understanding these processes is fundamental for parents and educators can accompany them and minimize the risks, as on the other side, overexposure to virtual audiences may have negative impacts on self-esteem and the search for identity of youth or exposing them to cyberviolence of the undesired dissemination of information or personal images. In the face of this challenge, the present article intends to contribute with recent data and for the Latin American context to understand the way youth explore their identity and relate with their peers in the virtual world, and by doing so they build new forms of sociability. It is worth bringing to mind that a limitation of this study is to be based on a non-probabilistic sample, and in this way, it is not statistically representative of the universe. The results have to be considered within the context of such limitation. All in all, data do not allow an approximation to the construction of adolescent identities over digital media in Latin America. Likewise, they propose relevant issues to continue researching aspects such as the construction of differentiated profiles according to the audience (peers v. parents and /or educators, for example), which may be an important part of the development of youth autonomy, but they also put forward key issues about control and adult supervision of these processes.