The use of Internet and Online Social Networking Sites (SNS) is an increasingly popular leisure activity in many countries around the world (Kuss and Griffiths, 2011). It is part of today's lifestyle, especially in adolescence and youth (Johansson and Götestam, 2004; Devine and Lloyd, 2012). SNS constitute one of the most frequent communication tools among adolescents to establish and maintain new interpersonal relationships, allowing them to participate in social groups (Hearn and Foth, 2007). In fact, SNS have replaced other virtual communication tools, such as e-mail and text messages, as well as non-virtual ones (García et al., 2012). In the online domain, characterized by immediacy, anonymity and globalization, adolescents build their identity, strengthen existing social relationships and create new social bonds (Muñoz-Rivas et al., 2003). One of the most attractive features of SNS is precisely the possibility to be an active agent in the process of social interaction beyond geographical proximity (Echeburúa and de Corral, 2010).

However, the use of the Internet and the SNS is not without risks. Ease of access to the Internet as well as the immediacy and social interaction capabilities of these tools also seem to increase the risk of developing problematic Internet use (PIU) and, in particular, problematic use of SNS (O'Reilly, 1996; Wellman and Gulia, 1999; Preece, 2000), which in some cases can even be classified as addictive (Smahel et al., 2008; Fioravanti et al., 2012; Gómez Salgado et al., 2014). The concept of Internet addiction, however, has not been fully developed, and research in this area remains incipient and fragmented (Yang and Tung, 2007), and no consensus criteria have been established to differentiate between excessive and problematic use of the Internet or dependence.

This plurality of terms in scientific literature reveals that there are multiple conceptual approaches to this problem and a lack of a diagnostic category, such as Internet addiction (Fernández-Villa et al., 2015). All this hinders reaching a consensual definition of terms such as excessive use of the Internet, PIU, and dependency, that consider different conceptual nuances. Despite these limitations, we can say that both Internet addiction and the problematic or excessive use of this tool refer to an individual's inability to control his or her use of the Internet with negative effects deriving from this lack of regulation, such as psychological distress and functional impairment (Echeburúa and de Corral, 2010; Fernández-Villa et al., 2015). Meanwhile, these conceptual difficulties have been reproduced in the specific field of SNS. Andreassen and Pallesen (2014) define addiction to SNS as an excessive concern to connect to these interaction spaces dedicating such a large amount of time that other areas such as social activities, studies/work, interpersonal relationships, psychological health, and well-being are affected. However, just as with the concept of Internet addiction, other terms have been chosen in scientific literature, such as excessive or problematic use of the SNS, to avoid the diagnostic or clinical connotation proper to the term “addiction.”

The problematic use of the Internet and SNS usually affects populations who are vulnerable because of their age, such as adolescents (Griffiths and Wood, 2000; Pallanti et al., 2006; Puerta Cortés and Carbonell, 2014). Previous studies have indicated that 2 to 15% of adolescents use the Internet and SNS in an excessive and problematic way (Park et al., 2008; Durkee et al., 2012; Sasmaz et al., 2013). In Spain, 96.5% of children under 15 years usually access the Internet (Spanish Statistical Office, 2012), 11.5% of adolescents connect between 3–5 h a day and 5.5% use the Internet more than 5 h a week, a percentage that rises in late adolescence (Catalina et al., 2014). In addition, ~85% of adolescents are registered in at least one SNS (Bringué and Sádaba, 2009; Rial et al., 2014) and average age of initiation to Facebook has dropped to 12 years (García et al., 2012). Likewise, it has been estimated that between 3.7 and 10% of adolescents make excessive or problematic use of these virtual tools (Muñoz-Rivas et al., 2010; Carbonell et al., 2012). This high level of usage of SNS, at problematic frequency rates in adolescence, can be attributed to the fact that these virtual spaces constitute a scenario to explore and develop multiple identities away from the supervision of parents or other figures of formal authority (Mazzoni and Iannone, 2014; Andreassen, 2015).

Regarding gender, available empirical evidence offers inconclusive results. It has been observed that boys tend to have a higher PIU than women (McKenna and Bargh, 2000; Morahan-Martin and Schumacher, 2000; Anderson, 2001; Schumacher and Morahan-Martin, 2001; Durkee et al., 2012). However, in other studies, no significant differences were found on PIU or Internet dependency according to gender (Kim and Davis, 2009). One significant aspect allowing to examine gender differences in more detail is how these technologies are used. Thus, data obtained from different studies show that boys spend more hours per week using e-mail, online games and web page visits, while girls use chats and social networks more (Muñoz-Rivas et al., 2003; Andreassen et al., 2012). In fact, the problematic use of SNS has been found to be greater in girls, especially in young and adolescent girls (Ryan et al., 2014; Andreassen, 2015), probably because girls tend to engage more in activities involving social interaction than boys (Andreassen et al., 2012; Kuss et al., 2014; Van Deursen et al., 2015).

The consequences of the problematic use of the Internet and SNS in adolescents have aroused growing concerns in the scientific community and in society (Blaszczynski, 2006; Echeburúa et al., 2009). There is abundant empirical evidence that the PIU and the problematic use of SNS is associated with greater manifestations of adjustment problems in adolescence such as psychological distress (Andreassen and Pallesen, 2014), anxiety (Chabrol et al., 2017), depressive symptoms (Morrison and Gore, 2010; Cao et al., 2011; Pedrero et al., 2012), loneliness, and social isolation (Young and Rogers, 1998; Satici et al., 2014), and emotional problems (Caplan, 2010; Marengo et al., 2018). Furthermore, in a meta-analysis, Marino et al. (2018) found that the problematic use of Facebook was related to greater psychological distress, anxiety and depression, and low life satisfaction. The problematic use of the Internet and SNS has also been related to externalizing problems in adolescence (Rodríguez and Fernández, 2014), with higher levels of impulsivity (Young, 1999; Echeburúa et al., 2009), anger and hostility (Ko et al., 2008; Xiuqin et al., 2010; Adalier and Balkan, 2012; Carli et al., 2013), and poor impulse control (Grüsser et al., 2007). In the same way, adolescents who used Internet more frequently were found to be more likely to participate in aggressive and hostile behaviors (Ko et al., 2008; Xiuqin et al., 2010; Adalier and Balkan, 2012; Carli et al., 2013; Mak et al., 2014). More specifically, adolescents who spent more time connected to SNS showed to be more frequently involved in peer aggression behaviors such as bullying (Ko et al., 2012) and cyberbullying (Erdur-Baker, 2010; Navarro et al., 2013; Giménez et al., 2015; Kircabum and Bastug, 2017).

One limitation of these studies is that aggression is analyzed globally. However, not all aggressive behavior is due to high impulsivity. It is important to distinguish between different manifestations of peer aggression in adolescence, to be able to explain the variety of its causes and associated correlates (e.g. Crick and Dodge, 1996). In this line, Little's taxonomy (Little et al., 2003; see Card and Little, 2006; Card et al., 2008) is based on two axes: the forms (“whats”-overt vs. relational aggression) and functions (“whys” -reactive vs. proactive) (See Figure 1). Forms refer to the type of behavior intending to harm the victim: aggressive conduct may involve a direct confrontation with others (e.g., intimidating, pushing, hitting, threatening, or insulting-overt aggression) or, on the contrary, may imply damaging the social reputation or social status of victims and isolate them from their friends, going so far as to use their group of peers (e.g., social exclusion, social rejection, or spreading rumors; Xie et al., 2002; Little et al., 2003; Juvonen and Graham, 2014) With regard to functions, i.e., the aggression's purpose, this type of conduct has been found to be due possibly to reacting to real or perceived damage -reactive aggression-, or a means to obtain a desired result -proactive aggression- (Coie and Dodge, 1998). While reactive aggression is fundamentally emotional and impulsive (Card and Little, 2006), proactive aggression implies an intentional behavior directed toward fulfilling a desired goal and, therefore, depends on an evaluation of the consequences (Fontaine and Dodge, 2006; Frey et al., 2017).

By combining these two axes (forms and functions) four dimensions or subtypes of aggressive behavior are obtained: overt -reactive and proactive-, and relational –reactive, and proactive. This classification has been validated ecologically and these dimensions, although interrelated, have shown important differences (see Murray-Close et al., 2016). In addition, the conceptualization of forms and functions as subcategories contributes to a better understanding of the motives and socio-cognitive mechanisms underlying this behavior (Juvonen and Graham, 2014). Regarding victimization, defined as the experience of being subjected to physical, verbal and psychological aggression perpetrated by peers (Graham, 2006), adolescents with greater use of SNS have shown a greater probability of experiencing cyber-victimization experiences (Navarro et al., 2013; Giménez et al., 2015). However, it seems that no studies have yet analyzed victimization experiences in non-virtual contexts and the problematic use of the Internet and SNS.

The general goal of the present study was to analyze existing relationships between the problematic use of SNS, peer aggression and victimization in adolescents, taking into account gender and age. In addition, the forms (overt vs. relational) and the functions (proactive vs. reactive) of aggressive behavior were considered to evaluate peer aggression. Forms of victimization were considered as well. Overall, results confirmed that the problematic use of SNS was associated with greater involvement in all dimensions of analyzed peer aggression and with higher levels of overt (physical and verbal) victimization and relational victimization. Likewise, gender and age were shown to have an important role in the relationship between problematic use of SNS and relational victimization and verbal expression.

The first specific objective of this study was to examine the relationship between peer aggression and the problematic use of SNS, by gender and age. We hypothesized that adolescents with a problematic use of SNS would be more involved in all the subtypes of aggressive behavior measured. As expected, results showed that adolescents with greater problematic use showed a greater involvement in overt and relational aggression, with both a reactive and a proactive purpose. These results are compatible with those obtained in previous studies in which the PIU has been linked to Internet addiction with greater hostility and greater expression of aggressive behaviors in adolescents (Ko et al., 2008, 2012; Xiuqin et al., 2010; Adalier and Balkan, 2012; Carli et al., 2013).

However, results of the present study show a deeper analysis of these relationships: the greater the problematic use of SNS, the greater the manifestation of peer aggression, in its different forms -manifest and relational- and functions -reactive and instrumental. Our findings could be explained because high engagement in aggressive conducts, both proactive and reactive, can be attributed to the fact that both types of aggression are interrelated (Poulin and Boivin, 2000; see Kempes et al., 2005).

Previous research has linked PIU and Internet dependence to difficulties in controlling impulses (Grüsser et al., 2007; Echeburúa et al., 2009; Kim and Davis, 2009), which may explain the relationship between the problematic use of SNS and reactive aggression. However, proactive aggression, unlike reactive aggression, is a more elaborate, planned and, therefore, less impulsive behavior: it is thus linked less to these difficulties in impulse control. In fact, impulsivity and anger have shown to be key factors in distinguishing between both types of aggression (Raine et al., 2006). Conversely, proactive (or instrumental) aggression has been associated with a reduced ability to delay gratifications, so that the possible benefits are valued by the aggressors, regardless of the damage they may cause to the victims (Kempes et al., 2005; Ramírez and Andreu, 2006; Crespo-Ramos et al., 2017). Both traits, impulsivity and difficulties on delaying rewards are also linked to problematic use of Internet and SNS. Additional research, though, is needed to test whether these characteristics might mediate the association between Problematic use of SNS and peer aggression. In relation to the role of gender and age, results indicate that boys participate more frequently than girls in all analyzed dimensions of aggression, except in the relational reactive dimension, where equivalent scores were found. Regarding age, the only differences were in reactive relational aggression, where adolescents aged 11–13 were found to be more frequently involved in these behaviors.

Regarding the second specific objective, results from previous studies indicate that adolescents with greater use of SNS report greater cyber-victimization (Navarro et al., 2013; Giménez et al., 2015). The findings of this study indicated that, as hypothesized, the problematic use of SNS was associated with the experience of overt -physical and verbal- and relational victimization. In addition, gender and age seemed to play an important role in understanding these relationships. This conclusion was drawn from analyzing the interaction between problematic use of SNS, gender, and age with respect to victimization. Thus, in relation to gender, results indicated that boys and girls showed a similar degree of relational peer victimization in the problem use groups of the lower and middle SNS. However, in the high problematic use group, adolescents reported higher levels of relational victimization. We believe that this result is highly revealing, insofar that it is evident that adolescents with problematic use of SNS also present other adjustment problems that differ according to gender.

As results obtained in the univariate analyses show, boys were subject to relational victimization more frequently than girls, while reporting a higher problematic use of SNS. This result is compatible with that obtained by other authors (Bringué and Sádaba, 2009; Griffiths et al., 2014; Ryan et al., 2014; Andreassen, 2015). Girls seemed to use Internet tools more often than boys, primarily for communication, strengthening friendships and social interaction (Muñoz-Rivas et al., 2003; Espinar Ruiz and González Río, 2009; Andreassen et al., 2012). Nevertheless, when these variables were taken into account, girls with the most problematic use of SNS clearly reported greater relational victimization, mainly aimed at excluding adolescents from the peer group and eroding their social life.

In adolescence, SNS represent a platform to establish new relationships and to strengthen existing ones, whether friendly or romantic. Girls, more than boys, seem to find the online domain a safer and more protected space to initiate friendship and affective relationships (Subrahmanyam and Greenfield, 2008). However, a greater predisposition to online communication seems to entail a greater risk of suffering negative consequences derived from the problematic use of the Internet and SNS (Oberst et al., 2017). Results of the present study suggest that these risks are especially high in girls showing problems of relational victimization. Thus, it is possible that in a situation of victimization, girls participate more frequently in SNS and in online communication activities to strengthen existing friendship relationships and to create new links that minimize the adverse effects of the experience of victimization: this, in turn, may make these adolescents more vulnerable to developing a problematic use of SNS. This finding can be considered highly significant and these relationships should be analyzed more in-depth in future research.

Regarding the interaction between problematic use of SNS and age, results showed that adolescents aged 11–13 years old -early adolescence- who used the SNS in a problematic way, reported greater verbal and relational verbal victimization, compared to the rest of the groups. Previous research has shown that approximately between 3.7% and 10% of adolescents used the Internet in a problematic way. However, no data has been found regarding the problematic use of SNS at different stages of adolescence. These results are compatible with studies carried out in Spain, that found that 85% of adolescents were registered in at least one SNS (Bringué and Sádaba, 2009; García et al., 2012; Rial et al., 2014). The findings of the present study indicated that adolescents aged 14–16 years showed greater problems with the use of SNS. However, it was adolescents aged 11–13 years with high problematic use who reported greater overt verbal and relational victimization. These results could be explained by the fact that the highest prevalence of peer victimization occurs in early adolescence (Martínez-Ferrer et al., 2011; Povedano et al., 2012). In addition, at this developmental stage, they already have access to the SNS and, therefore, they can make a greater exploratory use of these tools, which could lead to a high and problematic use of the SNS. For adolescents, one of the most inviting aspects of these networks is that they enable building support networks, searching information and social support beyond geographical proximity, though often reduced to the virtual domain (Muñoz-Rivas et al., 2003; Echeburúa and de Corral, 2010). It is possible that the SNS constitute, in turn, a support-seeking tool for adolescents victimized by their peer group,

Despite its positive aspects, this study also presents some limitations. Positive aspects include the novelty of the field, which lacks almost completely in empirical evidence, in particular on the link between problematic use of SNS and peer victimization. In this sense, results of the present study contribute to scientific knowledge in this field. Another main contribution is related to the interaction between gender and problematic use of SNS. This result sheds light on the need to incorporate a gender perspective in future research focused on the continuity of the processes of aggression and victimization in the online and offline domains. Among its limitations, the cross-sectional nature of the study did not allow establishing causal relationships between studied variables, so longitudinal studies should test these relationships more in depth. In this sense, future longitudinal studies could examine these relationships in more detail. In addition, all sources were self-reports, leading to possible biases, especially in sensitive topics such as aggression and the problematic use of SNS. In this sense, the use of online surveys could minimize the aforementioned biases because adolescents might feel their identity more protected. Moreover, outliers and missing values could be easily identified. Furthermore, findings suggest that as pointed out by (Livingstone and Smith, 2014; Smith and Livingstone, 2017), offline and online continuity exists, so vulnerable adolescents in one area may also be vulnerable in the other, while those who take risks in one domain may also take them in others. Future research may find it revealing to incorporate measures of cyber-aggression and of cyber-victimization.

Despite its limitations, the results of this study allow addressing the extent of the problematic use of SNS in depth and connecting the offline and online settings. Many studies have found that the problematic use of Internet and SNS is associated with other existing adjustment problems such us internalizing symptoms and lower levels of life satisfaction (Caplan, 2010; Marengo et al., 2018). This study contributes to a better understanding of adolescents with a problematic use of SNS and its links to peer aggression and peer victimization. Findings of the present study are particularly important given that research on SNS and adjustment problems in offline settings is scarce. The constellation of adjustment problems is interrelated, so adolescents with more problems deriving from the use of these social networks are more frequently involved in -overt and relational- aggression and victimization. In addition, in girls and adolescents aged 11–13 years, the problematic use of SNS is linked to greater victimization suggesting social influences. These findings reveal the need for educational programs on Internet and SNS good practices so that adolescents develop a healthy use of these communication tools. Such programs could also help detect violence and peer victimization problems and, therefore, lead to initiatives to improve harmonious coexistence in the community. Moreover, at the public health level, identifying the problematic use of SNS among adolescents in the early stages could potentially lead to identifying improvements in preventing peer aggression and fostering positive social climates.

The present study was conducted in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical standards, with the approval of the management board of schools, the educational inspection services, and the Childhood Observatory of the Regional Government of Andalusia (Spain). All participants gave written informed consent. The required authorizations from the education authorities, the schools, and the children's families were obtained.

BM-F, DM, and GM had participated in the intellectual content, the analysis of data, and the writing of the work. BM-F, DM, and GM had reviewed the final version of the work and they approve it for publication.