1 Introduction

In recent years, a considerable increase has been noted in the number of reports of bullying by young people who use new information and communication technologies (ICTs) (Martínez-Monteagudo et al., 2020). Expanding on the definition of traditional bullying, cyberbullying has been defined as aggressive behavior between students carried out repeatedly through electronic media by a group or an individual toward a victim who cannot easily defend themselves (Smith et al., 2008). In this regard, cyberbullies look to emotionally hurt their victims through threats, insults, malicious jokes, spreading rumors, and encouraging social exclusion, among other behaviors (Yudes et al., 2020). It usually starts in adolescence with the use of mobile phones and computers by minors. In Spain, the start of massive use of ICTs begins around 11 years of age (Andrade et al., 2021). In terms of participation, there are different profiles: cyberbullies, cybervictims, those not involved or spectators, and victimized cyberbullies (young people who, after having been direct victims of bullying, end up becoming cyberbullies). In this sense, some victimization processes, such as the anonymity given by technological resources, can predict the aggressive response of many adolescents (Cuadrado et al., 2019). The prevalence of cyberbullying, which was studied using a sample of 14,000 Spanish adolescents, shows that, in the sample, 8.8% were cybervictims, 3.1% were cyberbullies, and 4.9% were victimized cyberbullies (Rodríguez-Hidalgo et al., 2020). The consequences of cyberbullying are serious as the victims are being found to be increasingly vulnerable in terms of experiencing depression, anxiety, self-harm, suicide attempts, and antisocial behavior (Skilbred-Fjeld et al., 2020). Moreover, the increase of cyberbullying cases in recent years has increased the interest for this problem to be studied in the context of higher education (Bernardo et al., 2020), as well as the consequences that such a phenomenon can generate both academically (poor school performance), socially (rejection by peer groups) (Tippett & Wolke, 2014), and emotionally (high rates of anxiety and depression) (Eden et al., 2016; Heiman et al., 2016; Kaur et al., 2022).

Regarding the study of the different profiles of cyberbullying, most studies that have analyzed cyberbullying profiles have focused on secondary education students, therefore, there is a gap in the literature regarding bullying profiles or classes in university students. In the case of university students, we must keep in mind that the transition from high school to university is an exciting experience for some, but it is also an overwhelming and stressful experience for others (Sennett et al., 2003). The period of adaptation to the university environment could be considered a time of vulnerability for cyberbullying among young people, as it is a time of stress when new friendships are sought and there are changes in the use of ICT (now without parental supervision) that put them at risk of suffering or engaging in cyberbullying-related behaviors (Khine et al., 2020). The increasing use of mobile phones and overexposure to social media has led to a culture of cyberspace (Dennehy et al., 2020). It is a new context, with new responsibilities and demands. Starting a university career creates a stressful environment that often leads to poor academic performance, psychological, physical, personal and social adjustment problems (Bernardo et al., 2022).

Prevalence data in a study of 439 US college students indicated that 38% of students knew someone who had been cyberbullied, 21.9% had been cyberbullied, and 8.6% had cyberbullied someone else (MacDonald and Roberts-Pittman, 2010). Furthermore, the number of profiles obtained in the research varies from three to four depending on the study.

There are few studies of latent profile or latent class for cyberbullying in university students and therefore the main findings with secondary school students and adolescents are offered to contextualize the study with university students. Therefore, it is important to have research that analyzes profiles in young people for their vulnerability to develop or suffer cyberbullying behaviors due to their independence in the use of ICT and the spatially stressful situation in the establishment of a social and support network in the transition to university.

In this regard, the study by Hu et al. (2021) outlined three cyberbullying profiles (traditional and cyberbullying perpetrator-victims, traditional bullying victims, and minimal involvement) in a sample of 3,675 Chinese adolescents. Yoo (2021), using a latent class analysis on a sample of 3,656 South Korean adolescents, also found three cyberbullying profiles (high risk, transient, and low risk). However, most research has established four profiles (cyberbullies, cybervictims, victimized cyberbullies, and those not involved) regarding cyberbullying in adolescent samples (Cho et al., 2020; Guo et al., 2021).

Using a recent sample of university students, the work by Lee et al. (2021) studied how empathy and attachment moderate the relationship between patterns of bullying participation and depressive and anxiety symptoms in 835 university students in Singapore. Through a latent class analysis, four bullying profiles were found: bully-victims, cyberbully-victims, relational and verbal victims, and persons uninvolved in bullying.

The aim of this study is to analyze latent profiles with consideration of aggression and cyberbullying victimization behaviors in a sample of Spanish university students.

Research has reported the negative consequences of cyberbullying, including social anxiety, which involves a fear of negative evaluation, and general and specific social avoidance of new situations or individuals (La Greca, 1999). Social anxiety consists of the development of an excessive fear toward specific or general social situations and can impair an individual's ability to interact with others (American Psychiatric Association, 2013). Regarding the relationship between cyberbullying and social anxiety, different studies note a positive relationship between victimization and social anxiety (Suárez-García et al., 2020) in adolescents. Some studies conclude that social anxiety increases the probability of being a victim, while others note that being a victim (Pabian & Vandebosch, 2016) increases the probability of having social anxiety (Coelho & Romao, 2018). The study by Ruiz-Martín et al. (2019) focused on a systematic review of studies with the aim of analyzing the relationship between social anxiety and cyberbullying experiences. The results show that most studies found social anxiety to be a risk factor for cybervictimization, and to a lesser extent, a consequence of it. Likewise, Martínez-Monteagudo et al. (2020) analyzed the relationship between cyberbullying and social anxiety in a sample of 1,412 Spanish adolescents using a latent class analysis (LCA). The results showed that the students with a high cyberbullying (bullies and victims) profile presented high scores in social anxiety compared to the other profiles. In a recent study by Núñez et al. (2021), whose aim was to identify, also through an LCA, victimization profiles using a sample of 3,120 Spanish adolescents and to analyze them in relation to social anxiety, the results showed that the higher the level of cybervictimization, the higher the level of social anxiety.

Regarding the relationship between social anxiety and victimized cyberbullies, the results are contradictory, as they may present characteristics of victimization such as anxiety, and characteristics of cyberbullying such as aggression. In this sense, the empirical evidence has not shown a mediating effect of anxiety between cybervictimization behaviors and subsequent cyberbullying (Pabian & Vandebosch, 2016). In this sense, although anxiety is related to engaging in cyberbullying behaviors in the case of perpetrators of cyberbullying in the context of university (Wei et al., 2022), it is less clear that it acts directly in the case of victimized cyberbullies.

Moreover, cyberbullying is an intentional aggressive behavior that is repeated by a single individual or a group against someone who cannot easily defend themself and is characterized using different electronic forms of contact (De Pasquale et al., 2021). In this regard, many studies have demonstrated the existing relationship between a high level of aggressiveness and bullying among peers. A study by Martínez-Monteagudo et al. (2019) with a sample of 1,102 Spanish students showed that as physical aggressiveness and anger increased so too did the probability of being a victim, bully, or victimized bully. The recent study by Peker and Yildiz (2021), using a sample of 333 adolescents, also demonstrated the relationship between aggressive behavior and cyberbullying. The study found that aggressiveness was a precursor to cyberbullying behaviors and that self-control played a mediating role between aggressive behavior and cyberbullying. Similarly, the study by Tanrikulu and Erdur-Baker (2021) with a sample of 598 university students in Turkey and using a structural equation analysis, found a positive relationship between aggressiveness and peer cyberbullying.

Although the previous empirical evidence has shown the relationship between cyberbullying and different personal variables such as social anxiety (DeSmet et al., 2019; Wang et al., 2019) and aggressiveness (Schade et al., 2021), there is a lack of studies that specifically examine the relationship between the different cyberbullying profiles, social anxiety, and aggressiveness in university students. Moreover, a review of the literature shows that most studies find four cyberbullying profiles, however, other studies find only three. Therefore, the aim of this research is twofold: (1) to identify the different cyberbullying profiles while considering cybervictimization and cyberaggression behaviors, and (2) to verify whether there are statistically significant differences in social anxiety and aggressiveness among the profiles identified with respect to cyberbullying in a sample of Spanish university students.

Considering the review of the previous literature, four profiles are expected to be found regarding cyberbullying, as well as differences in cyberbullying profiles according to different dimensions of social anxiety and aggressiveness. Specifically, the following hypotheses have been posed: (1) four profiles are expected to be found with the following characteristics: cyberbullies, cybervictims, victimized cyberbullies, and those not involved; (2) the profile of students that are not involved is expected to present significantly lower scores in social anxiety and in the different dimensions of aggressiveness (physical, verbal, hostility, and anger); (3) the profile of cyberbully students is expected to present significantly higher scores in social anxiety and in the different dimensions of aggressiveness (physical, verbal, hostility, and anger); (4) the profile of victimized cyberbully students is expected to present significantly lower scores in social anxiety and significantly higher scores than the previous profiles in the dimensions of hostility and anger.

2 Method

2.1 Participants

The reference population was undergraduate university students at the Universities of Valencia and Alicante (Spain). Two-stage random cluster sampling was conducted. In the first stage, three public universities were randomly selected in Valencia and Alicante. Once the universities were selected, in the second stage of sampling, eight classes were randomly selected from each university. Once the classes were selected, a random selection was carried out and 1,404 students were chosen from three universities, of which 36 were eliminated due to omissions or errors in the tests. Due to the random sampling method, the socioeconomic status and ethnic compositions of the overall sample are assumed to be representative of the community in terms of key variables (e.g., ethnicity, academic performance, etc.). Therefore, a total of 1,368 university students (494 males; 36% and 874 females; 64%) participated in the research in the following academic years: 1st year (45%), 2nd year (21.9%), 3rd year (12.1%), and 4th year (20.9%). The mean age of the participants was between 18 and 49 years (M = 21.34; SD = 4.45).

By means of the Chi-square test, used to analyze the homogeneity of the frequency distribution, it was found that there were no statistically significant associations between the sex of the participants and the academic year (χ2 = 18.44; p > 0.05).

2.2 Instruments

2.2.1 European Cyberbullying Intervention Project Questionnaire (Brighi et al., 2012; Del Rey et al., 2015)

The Spanish version of the European Cyberbullying Intervention Project Questionnaire (ECIPQ) (Del Rey et al., 2015) was used to assess cyberbullying. It is a scale consisting of 22 Likert-type items with five response options, with a scoring system between 0 (never) and 4 (always). It has two dimensions: Cybervictimization and Cyberaggression. For both dimensions, the items refer to actions like saying mean things ("someone has called me names or insulted me using the Internet or cell phone messages"), excluding individuals or spreading rumors about them ("I have been left out (excluded/ignored) or blocked from a social network platform or chat"), impersonating ("someone has created a fake email account or profile on social networks to impersonate me"), etc. These items refer to actions that happen on electronic media and refer to a time interval of the last two months. The scale has adequate reliability indices for this study (α cybervictimization = 0.80, α cyberaggression = 0.88).

2.2.2 Social Anxiety Questionnaire for Adults (SAQ-A30; Caballo et al., 2010)

This instrument consists of 30 items that are scored on a five-point Likert-type scale, from 1 = Not at all or very little discomfort, tension, or nervousness, to 5 = a lot or extreme amounts discomfort, tension, or nervousness. It evaluates five dimensions of social anxiety: 1) Speaking in public/Interacting with people of authority ( "a teacher in class or a superior in a meeting asked me a question"), 2) Interacting with strangers ("having a conversation with a person I just met"), 3) Interacting with the opposite sex ("asking an attractive person of the opposite sex to go out with me"), 4) Assertive expression of annoyance, displeasure, or anger ("expressing my anger at a person who is picking on me"), and 5) Being embarrassed or ridiculed ("greeting a person and it not being reciprocated"). The scale has adequate reliability indices, ranging between 0.74 and 0.87 for the dimensions scores.

2.2.3 Aggression Questionnaire (AQ; Buss & Perry, 1992; adapted by Andreu et al., 2002)

This is an instrument consisting of 29 items that refer to aggressive behaviors and feelings and are coded through a 5-point Likert-type scale (Completely false for me = 1; Completely true for me = 5). It consists of four scales: Verbal Aggressiveness ("my friends say I argue a lot"), Physical Aggressiveness ("I often find myself getting into fights"), Hostility ("sometimes I am quite envious"), and Anger ("I feel angry, as if I am going to explode"), which assess the three components of aggressiveness: motor/behavioral (physical and verbal aggression), cognitive (hostility), and physiological-emotional (anger). The AQ was developed using the Hostility Inventory (Buss & Durkee, 1957) and the Spanish adaptation of it was carried out by Andreu et al. (2002). The internal consistency coefficients of the AQ scores in this study were acceptable: Physical Aggressiveness (α = 0.77), Verbal Aggressiveness (α = 0.74), Anger (α = 0.64), Hostility (α = 0.75), and total AQ score (α = 0.90).

3 Procedure

First, once the universities had been selected, a meeting was held with the management team of the faculties to explain the objectives of the research work and the evaluation instruments to be used to request their permission and encourage their collaboration. The questionnaires were completed voluntarily and were done collectively during a class session, ensuring the anonymity of the participants by means of identification numbers on the answer sheets. The researchers were present during the completion of the tests so to clarify possible doubts and verify that correct administration had been done. Emphasis was placed on the total completion of the tests, with an average time for each questionnaire of approximately 15 min being used to do them. The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board (or Ethics Committee) of the University of Alicante (UA-2018–02-21).

3.1 Statistical analysis

In order to identify cyberbullying profiles, a latent profile analysis (LPA) was used. In accordance with the work by Ondé and Alvarado (2019), this analysis method is a procedure that is adequate in establishing profiles when a wide sample of participants is used. The profiles were defined using the different combinations of the two main cyberbullying behaviors: victimization and aggression. The choice of the necessary number of profiles to identify an adequate representation of the data was carried out using the lowest indicator of the Bayesian information criteria (BIC) and the information from the Akaike criterion (AIC), as well as the value closest to one for the entropy, as the fit indices. Likewise, the LRT (Vuong-Lo-Mendel-Rubin likelihood-ratio test) and the BLRT (bootstrap likelihood ratio test) indicators with values close to 0 were used. The statistical analyses were developed using the MPlus version 8 program.

Once the different groups were established using the LPA, a multivariate analysis of variance (MANOVA) and an analysis of variance (ANOVA) were carried out with the purpose of analyzing the differences between profiles in social anxiety and aggressiveness dimensions. To understand the magnitude and the effect size of these differences, the eta2 index was found. For those analyses in which the differences were statistically significant, post hoc tests were then carried out to identify the groups between which the differences were established. The Bonferroni test was used as this test does not require sample sizes to be the same. Finally, the size of the d effect (Cohen, 1988) was calculated to subsequently calculate the magnitude of the differences found. Its interpretation is as follows: a small effect size is found at values of 0.20 ≤ d ≤ 0.49, moderate between 0.50 ≤ d ≤ 0.79, and large at values d ≥ 0.80. Analyses were carried out using SPSS version 26.0 statistical software.

4 Results

4.1 Cyberbullying profiles

From the LPA, and using the cyberaggression and cybervictimization scores, the model that obtained the best fit for the BIC, AIC, and entropy indicators was the three-profile model (see Table 1).

Table 1 Fit indices of the latent profile analysis (LPA) values in bold showing the best model fit

The first profile, no cyberbullying, consisted of a total of 1,198 students (87.58%) who presented low scores in victimization and aggression, and is identified as those "not involved". The second profile consisted of a total of 137 students (10%) who presented moderate scores in victimization and aggression and is identified as "victimized cyberbullies". The third profile, high cyberbullying, consisted of a total of 33 students (2.42%) who presented high scores in aggression, and is identified as "cyberbullies". Figure 1 shows the results from the LPA and includes the z-scores for victimization and aggression.

Fig. 1
figure 1

Graphic representation of the LPA solution

4.2 Intergroup differences in social anxiety

Regarding social anxiety, the MANOVA results did not show significant statistical differences about the three cyberbullying profiles (Wilks' Lambda = 0.99; F(10,1365) = 1.039, p = 0.40, η2p = 0.00).

Although it was observed that victimized cyberbullies were those who showed the highest scores in social anxiety in the subscales relating to fear of being embarrassed, interacting with the opposite sex, and interacting with strangers, this difference was not significant (see Table 2).

Table 2 Means and standard deviations of social anxiety between classes and statistical significance

4.3 Intergroup differences in aggressiveness

Regarding the displays of aggressiveness, the MANOVA analysis found that there were differences between the latent profiles of cyberbullying (Wilks' Lambda = 0.95; F(8.1365) = 8.69, p < 0.00, η2p = 0.02).

Table 3 shows the ANOVA results, comparing each aggressiveness dimension among the three cyberbullying profiles.

Table 3 Means and standard deviations of aggressiveness between classes and statistical significance

The analyses present significant statistical differences in all aggressiveness dimensions. Specifically, the “victimized cyberbullies” student profile showed higher scores in physical and verbal aggressiveness, anger, and hostility than the group of “not involved” students.

Moreover, the “cyberbullies” group showed scores that were significantly higher in physical aggressiveness dimension in comparison to the “not involved” group.

However, there is no significant difference between the “cyberbullies” group and the “victimized cyberbullies” group in terms of aggressiveness scores.

The effect size for the differences found in the comparisons between the not involved group and the victimized bully group, as well as in the comparisons between the not involved group and the cyberbully group, was moderate in terms of physical aggressiveness dimension (Table 4). Regarding the dimension of hostility, anger, and verbal aggressiveness, the effect size for the differences found between the not involved group and the victimized bully group was small (see Table 4).

Table 4 Cohen’s d index to post hoc Bonferroni contrast between the mean scores and the three classes in the factors of aggressiveness

5 Discussion

The objective of this work was twofold. Firstly, it intended to identify the different cyberbullying profiles and, secondly, analyze the differences in aggressiveness and social anxiety in the different cyberbullying profiles in a sample of university students. Unlike previous research, this study analyzes the importance of aggressiveness and social anxiety in the educational context using the analysis of different profiles of university students in terms of cyberbullying. In addition, and unlike previous works, this research has contemplated such a relationship considering the analysis of effect sizes to determine the magnitude of the differences found, which was recommended by different authors (e.g., Cohen, 1988; García et al., 2011).

The first specific aim of the study was to identify the different cyberbullying profiles by comparing the cyberaggression and cybervictimization scores. The analysis of the latent profiles allowed for three cyberbullying profiles to be identified: the first profile, with low scores in victimization and aggression, identified as those “not involved”; the second profile, with moderate scores in victimization and aggression, identified as “victimized cyberbullies”; and the third profile, with high scores in victimization and aggression, identified as “cyberbullies”. Our starting hypothesis expected to find four cyberbullying profiles based on the traditional understanding of bullying (bullies, victims, victimized bullies, and those not involved). The existence of these three profiles may be because in the case of cyberbullying as opposed to traditional bullying, it is not so common to find a cybervictims profile in university students exclusively due to the mediating effect of other variables such as emotional intelligence and self-esteem (Núñez et al., 2021). These profiles obtained in the present investigation are in line with the results found in previous research (Delgado et al., 2019; Hu et al., 2021; Martínez-Monteagudo et al., 2020; Yoo, 2021) whose results note the existence of three cyberbullying profiles using the latent class analysis technique. Specifically, and in line with the results of other studies, the “not involved” profile was also the profile with the highest percentage (87.58%) with respect to the other profiles, followed by the "victimized cyberbullies" profile (10%) and the "cyberbullies" profile (2.42%). This latter profile received higher scores in previous research (Martínez-Monteagudo et al., 2020; Yoo, 2021). The study by Martínez-Monteagudo et al. (2019), with a sample of 1412 Spanish adolescents, revealed the existence of three profiles with respect to cyberbullying: a first profile (42.70%) with very low scores on the victimization, aggression, and aggression-victimization subscales, identified as "not involved", a second profile, high cyberbullying (30. 02%) with high levels of victimization, aggression, and aggression-victimization, identified as "bully-victims", and a third profile, low cyberbullying (27.26%), with moderately low scores on the three analyzed subscales of cyberbullying, identified as "rarely victims and bullies". As in our study, the highest percentage corresponds to the "not involved" group, followed by the "cyberbullying victims" and the "cyberbullies" profile. The study of Yoo (2021), with a sample of 3656 adolescents, aimed to analyze profiles according to victimization and cyberbullying using the LCA. It identified three profiles of cyberbullying victimization: increasing high-risk group (3.9%), transient group (6.0%) and low-risk group (90.1%). Considering previous research findings and in line with our study, it is important to pay attention to the profile of increasingly victimized cyberbullies, as the experience of victimization increases the risk of cyberbullying perpetration. Based on current research, young people involved in bullying perpetration are more likely to engage in cyberbullying perpetration later in life (Giumetti et al., 2022).

There are very few studies on cyberbullying profiles in university students. In this sense, it is necessary to analyze them due to the contradiction in the results in some cases and to provide new data on the subject to advance in the knowledge of how the phenomenon of cyberbullying varies over time. The profiles are dynamic and may be modified over the years or by the characteristics of the samples studied. In addition, the emergence of new technologies, applications and social networks should lead us to a constant updating of this problem.

Regarding the second specific aim of the study, the results show significant statistical differences between the cyberbullying profiles found and the different dimensions of aggressiveness, but not with social anxiety, which partially confirms the second and third hypotheses of the study. This result may be because in adolescence there is a high level of social anxiety related to social status, sense of belonging to the group, and fear of negative evaluation (Henricks et al., 2021; Ruiz et al., 2021), however, in the university setting, this does not seem to have as much pressure on students. In this sense, previous research has suggested that social anxiety is associated with victimization and perpetration of (cyber)bullying in adolescents, but there are no studies that relate social anxiety and different cyberbullying profiles in university students. This absence of relationship can be explained, according to Fahy et al. (2016), because the self-esteem of cyberbullies may be strengthened as there are no repercussions for their behaviors and they do not see the face-to-face reactions of the victims (Kowalski et al., 2012). Specifically, the students with the “not involved” profile show lower scores in physical and verbal aggressiveness, anger, and hostility than the “victimized cyberbullies” and “cyberbullies” student group. These results are in line with those of other authors who note a lower level of aggressiveness in not involved or spectator students (Ferreira et al., 2021; Leung, 2021). Moreover, the “victimized cyberbullies” group shows significantly higher scores in all dimensions of aggressiveness than the “not involved” group. This result is in line with the study by Cimke and Cerit (2021) who also found, in a sample of university students, that victimized cyberbullies showed higher levels of aggressiveness when compared to pure cyberbullies. Furthermore, this result is especially important because it constitutes a group that may show higher rates of reactive aggressiveness than even the cyberbullying group (Pascual-Sánchez et al., 2021; Tanrikulu & Erdur-Baker, 2021). It is important to note the relationship between offline aggressiveness and engaging in cyberbullying behaviors (moderate or high) vs. not involved, and the practical implications these findings may have in universities: in designing tutorial action plans, teaching strategies for managing aggressive behaviors in students who are simultaneously victims and aggressors on the internet. These students can manifest a lot of anger and emotional repercussions that can reinforce the appearance of cyberbullying behaviors. Therefore, it is important that they feel supported by the administrations and to provide them with the necessary training to ask for help. In this sense, this study is novel because it provides evidence in the scarce explanation of the phenomenon of cyberbullying in young university students.

Finally, the “cyberbullies” group of students presents more physical aggressiveness than the “not involved” group, which supports the results of previous research (Valdés-Cuervo et al., 2021). This result has significant relevance from an educational perspective as the “not involved” group is still not particularly studied and this group may reinforce cyberbullying for the cyberbully or may choose to stop cyberbullying by calling attention to an incident, by helping or defending the victim, or by ceasing to share or comment on cyberbullying incidents (Leung, 2021).

This study has some limitations. Firstly, although the type of sampling used guarantees the representativeness of the sample obtained, future studies should confirm whether the results found in university students are maintained in other educational stages. Most of the research focuses on adolescents, so these results can be very useful for understanding the phenomenon of cyberbullying at university level. It would also be advisable for future work to use longitudinal designs to provide more conclusive data on the causal relationships between the variables under study. It is also important to note that the non-involved group may be made up of students who hold different attitudes towards bullying and should be analysed in more detail. Finally, this research studies the relationship between aggression and social anxiety with cyberbullying. Future studies should consider the personal and social variables that may influence the emergence and maintenance of the phenomenon of cyberbullying in university students.

6 Conclusions

This study provides novel information of great practical utility with respect to the study of cyberbullying at the university level. On the one hand, this study confirms the existence of different cyberbullying profiles through the consideration of different aggression and victimization behaviors, and not considering the profiles typically studied for bullying. Specifically, three cyberbully profiles were found in the student sample analyzed: a first profile with low cyberbullying and cybervictimization behaviors (not involved), consisting of 87.58% of college students who have low scores on cyberbullying; a second profile with moderately high cyberbullying and cybervictimization scores (victimized cyberbullies), consisting of 10% of college students who have higher victimization scores and, in addition, have a behavioral profile with higher scores in social anxiety and aggression; and a third profile with high scores on cybervictimization and very high scores in cyberaggression (cyberbullies), consisting of 2.42%, who have higher aggression scores than those “not involved” but lower scores in social anxiety and aggressivity than “victimized cyberbullies”. Moreover, significant statistical differences were found between the different cyberbullying profiles for the displays of aggressiveness. Specifically, it is worth paying special attention to the “victimized bullies” students profile, whose scores on the different aggressiveness scales were higher than those of the not involved group.

On a practical level, the results of this study support the effectiveness of programs aimed at adaptively channeling aggression in university students (Hamzah et al., 2021) and improving social competence and preventing cyberbullying situations in a comprehensive way (Feijoó et al., 2021). The results of this study point to the importance that working with different prevention strategies with students and educating them in the responsible use of new technologies, as well as training those in their surroundings, may help to reduce the number of cyberbullying victims both in and out of the educational context (Ruiz-Martín et al., 2019).

Finally, the results show the need for educational intervention with the victimized cyberbullies students profile since this group, when compared with those not involved, presents higher rates of aggressiveness and bullying and are direct targets of verbal, physical, and indirect bullying as well as cyberbullying (Tong & Talwar, 2020). In this sense, it is necessary to train these students in coping strategies that help them to manage stress and anxiety and to offer them assertive behavior alternatives and reduce the level of aggressiveness. Considering the consequences that problematic use of new technologies can have on the well-being of young people, we must address this stage and detect possible behaviors that can end up being very negative. In this sense, early detection, and a comprehensive approach (Méndez et al., 2020) to the problem by analyzing the most at-risk profiles can help to design the best guidelines for educational action.