Identification, Witnessing and Reaction to School Bullying Behaviour in Secondary Education

This study, carried out with 10,795 Compulsory Secondary Education students (Asturias, Spain), seeks to determine what behaviours are considered school bullying, to what extent they are witnessed, and what the typical reactions to bullying are. To gather the data, a 19-item questionnaire was used, with a high degree of reliability (α = 0.85). Regarding the results, most of the students have a concept of bullying generally related to the use of physical force, although the behaviours witnessed with the greatest frequency are insults, social exclusion, or damaging belongings. As for their reaction, students tend to assume the role of a proactive witness, the tendency being to ask an adult for help. Nevertheless, there are many students whose knowledge and recognition of bullying is scarce, and who, by choice, limit themselves to observing. The results are useful for designing educational actions aimed at the intervention and prevention of school bullying.


Introduction
It is undeniable that any school's positive and harmonious environment, regardless of the type, can be hindered by real problems like violence or abuse among peers, which often manifests as school harassment (or bullying). Accordingly, the 2018 PISA Report shows that 17% of Spanish students have suffered bullying at least once per month during the year prior to the test, positioning Spain below average in the violence due to the great exposure of the victim. To this, one must also add factors such as anonymity, which hinder attempts to localise and identify the aggressor, disinhibition, and the negative impact consequence of a greater number of witnesses (González Calatayud et al., 2020). Examples of cyberbullying include the publication or dissemination of information about, and images of, the victim (even intimate ones), identity theft, leaving offensive comments in forums, and sending threatening messages (Martínez, 2017).
In the preceding discussion of the concept, requirements and variants of bullying, the aggressor and the victim play a crucial role. Cuevas and Marmolejo Medina (2016) point out that most of the research on the subject has focused on victims, while the transcendental role of the events' witnesses has been relegated to the sidelines. This lack of attention is barely understandable if one considers the more significant number of witnesses compared to the rest of the actors, as well as the emerging concept of school bullying as a group phenomenon in which the active or passive involvement of everyone counts.
In this regard, the observation of an act of bullying implies participation in it, whether it be because the witness encourages a repetition of the aggression or because they take action to stop it (Armero Pedreira et al., 2011;Conde Vélez & Ávila Fernández, 2018;Sánchez Venteo, 2017). As such, one speaks of an active witness, who may or may not be a close friend of the aggressor, as someone who instigates the aggressor to continue; a passive witness as someone who does nothing to prevent the abuse and whose silence leads to a possible misinterpreted approval; and a proactive witness, as someone who neither encourages nor avoids the harassment and opposes it by defending the victim and turning to teachers or other adults for help. In the latter context, some studies demonstrate the importance of the proactive witness's role in minimising the emotional damage the aggressor can inflict on the victim (Sánchez Venteo, 2017).
Despite much discussion on the subject of school bullying and its impact on the present and future development in childhood and adolescence, it remains a complex phenomenon about which there is still much to be studied, to the point where it is considered to be little more than usual and characteristic mischief of that age (Gómez Tagle López, 2018). Furthermore, little research has been done from the witness's perspective (Cuevas & Marmolejo Medina, 2016), an actor that, as previously mentioned, is believed to play a significant part in the perpetuation of bullying. Given that adolescent student is at an age at which they are susceptible to being bullies or suffering bullying (Díaz-Aguado Jalón et al., 2013), this study aims to find out what they know about the subject, what behaviours they consider to be bullying, and whether the knowledge they possess reflects the reality, the notion and modalities addressed in this work. In addition, this study attempts to detect the volume of bullying behaviours in which the adolescent student assumes the role of witness and to analyse the active or passive reactions as witnesses.

Participants
10,795 pupils from 72 Compulsory Secondary Education of the Principality of Asturias participated in the study. Of the participants, 51.1% were girls, and 48.9% were boys, with a mean age of 13.9 (SD = 1.33). Regarding the classification of the centre in which the participants were enrolled, 54.5% were in public schools, and 45.5% were in semi-private schools. Concerning their academic status, 26.3% were in the first year, 27.2% in the second year, 25.2% in the third year, and 21.3% in the fourth year. Moreover, 82.8% of the students had never repeated a year, whereas 17.2% had repeated a year at least once. Lastly, regarding the family's level of education, 46.6% of mothers had obtained a university degree, 26.4% had obtained a high school diploma, 22.7% had completed their compulsory secondary education, and 4.3% had no formal education; in comparison, 39.7% of fathers had obtained a university degree, 28.0% had obtained a high school diploma, 26.5% had completed their compulsory secondary education, and 5.8% had no formal education.

Instrument
The instrument employed is the Self-Perception and Perception of Bullying in Adolescents Scale (SPB-A) (Álvarez-Blanco et al., 2022). Specifically, the study shows the results according to the following dimensions: Behaviours that are considered bullying, composed of 8 items; Bullying behaviours witnessed, composed of 8 items; and Reaction to bullying behaviours, composed of 3 items. This instrument, based on the Questionnaire for the Assessment of School Violence in the Preschool and Primary School (Cuestionario de Evaluación de Violencia Escolar en Infantil y Primaria, CEVEIP) (Albaladejo-Blázquez et al., 2013), presents good overall reliability, carried out by Cronbach's Alpha, of 0.85, similar to that reported for the original questionnaire (0.86). Regarding the scale used, a Likert-type scale was maintained, with values between 1 (never, or totally disagree) and 10 (always, totally agree), avoiding hence the tendency to a central value.
The factorial structure of the SPB-A was analysed using a cross-validation process with an exploratory factor analysis (EFA) and a confirmatory factor analysis (CFA) (Álvarez-Blanco et al., 2022). This allowed to identify the presence of the four factors obtained that explain the 51.25% of the variance and show an optimal fit of the model (χ2 = 3382.209, DF = 294, p < .000); CMIN/DF = 11.504; GFI = 0.955; RMSEA = 0.044; SRMR = 0.038; CFI = 0.958; IFI = 0.958; TLI = 0.950). Furthermore, it was verified that the factorial invariance of the structure according to the variables "classification", "sex", and "academic year" using multi-group confirmatory factor analysis (MGCFA). In the study carried out by Álvarez Blanco et al. (2022), the Cronbach's alpha reliability coefficient of the entire set of items was 0.86, and that of the resulting factors was 0.91 (F1), 0.90 (F2), 0.86 (F3), and 0.72 (F4).

Procedure
To carry out the data collection process, all the schools of the Principality of Asturias participating in the Director Plan for Coexistence and Improvement of Protection in Schools, Instruction 7/2013 of the Secretary of the State of Security, were contacted in collaboration with the National Police. The questionnaire was always delivered during official school hours to ensure that any doubts about the content or methodology could be resolved. The professional responsible for solving these doubts was the tutor (never the Police) in order not to condition the answers. The time taken to complete the questionnaire, including the explanation prior to its administration, was between 10 and 20 min.
In particular, to facilitate the data collection process, the questionnaire was computerised with the Google Forms tool so that the students could fill in the questionnaire directly in the schools' computer science classrooms.

Data Analysis
Once the data collection was finished, it was processed and analysed with the SPSS Statistics 24 software. Concerning the analyses carried out, these included descriptive analyses using the mean and standard deviation of each item considered. To analyse the possible statistically significant differences, a comparison of means was also carried out for cases where the variables are dichotomous (boy/girl and public/semiprivate), using the Student's t test. The effect size was calculated with Cohen's d, considering values of d lower than 0.20 to indicate a small effect size, between 0.20 and 0.50 a medium effect size, and above 0.50 a large effect size, although it must be kept in mind that even a small effect size may have a practical significance (Kirk, 1996). Where the contrast variable has more than two categories, as is the case for the academic year (1st /2nd /3rd /4th), the significant differences were analysed via an ANOVA, applying the Scheffé test post-hoc.

Results
For behaviours that students at secondary school consider school bullying (Table 1), physical aggression and isolating or ignoring the victim stand out, with means above 7 (values between 1 and 10). In contrast, those with lower means include bothering  Table 1 Descriptive results (means and standard deviation) of behaviours considered to be bullying the victim to prevent them from doing their work and taking away or hiding things from them, with means below 6 (out of 10). Concerning possible differences in what behaviours are seen as bullying, according to sex and classification of the centre (Table 2), the sex variable shows differences in all the behaviours considered, with medium effect sizes, and the classification of the centre in 4 of the 8 items, all of them with small effect sizes. In all cases, girls and students at semi-private schools tend to consider the relevant behaviours as bullying.
When taking the school year into account (Table 3), it is notable that all the items show statistically significant differences. In particular, regarding the behaviours considered to be bullying, the students of the 4th year demonstrate a greater consciousness of what constitutes bullying.
On the other hand, when asked about the bullying behaviours witnessed (Table 4), actions such as insulting, ignoring, or isolating, and taking away and hiding things are generally the most common, with means of above 4 (out of 10). On the other hand, the three forms of bullying linked to technology are the ones that are observed with the lowest frequency.
When sex and classification of the centre are considered (Table 5), the first variable shows 4 items with statistically significant differences, and the second variable presents 5 out of a total of 8, with small effect sizes in both cases. Mainly, girls have witnessed more incidences of ignoring, marginalising, and sending messages via social media, whereas boys have witnessed more cases of offensive or mocking photos and videos of someone on the Internet. Addressing the classification of the centre, public school students have seen more bullying cases in those items that presented statistically significant differences.
Again, when considering the school year (Table 6), all the items show statistically significant differences. When it comes to witnessing bullying behaviour, it is the 3rd year students that tend to demonstrate higher means (5 of the 8 items), followed by the 2nd years (2 of the 8 items); in 1 of the 8 items, the students of the 2nd and 3rd years show the same mean.  Finally, when a student becomes aware of an incidence of bullying (Table 7), they tend to a large extent to tell their families, followed by communicating it to a teacher. Likewise, the low mean of students that opt to do nothing is noteworthy.
In this case, all the items considered show statistically significant differences when taking sex into account, and none according to the classification of the centre (Table 8). Specifically, girls demonstrate a greater tendency to opt for communicating possible incidences of bullying to teachers and family members, while boys show higher means for the option of doing nothing, with both cases presenting small to medium effect sizes.
Regarding the actions taken when bullying is witnessed according to the school year (Table 9), statistically significant differences are observed in all 3 items considered. As the school year increases, a reduction is observed in communicating their awareness of the bullying event to a teacher or a family member, and they mainly opt for doing nothing.    Table 7 Descriptive results (means and standard deviation) of actions taken upon learning of incidences of bullying

Discussion and Conclusions
As stated in the introduction to this work, consulted international studies depict the Spanish educational system as more favourable than other countries in terms of school coexistence, while at the same time, there seems to be a reduction of bullying events after four years since the last HBSC report was published. Such observation can be transferred to Asturias, given that the percentage of students claiming to be bullied at school sometime during the two months before they participated in this study goes from 12.4% to 2014 to 10.8% in 2018 (Moreno et al., 2020b), a low percentage according to the means provided by this study. Despite this data, bullying is still a concerning phenomenon that generates an increasing interest among researchers. In this regard, the fact that bullying remains may be related to a lack of detection and early reporting of the events, which, in turn, encourages victims of bullying at school to isolate themselves and be rejected by the rest of the students, in contrast to bullies, whose behaviour is reinforced by the peer group (Hill, 2003;Sánchez Lacasa & Cerezo Ramírez, 2014). From this, and in line with the conclusions drawn by Martínez Sánchez et al. (2019) about the lower research attention that this issue is given in secondary education compared to primary education, we consider it necessary to address the problem that school bullying presents from the perspective of the adolescent student going population.
Thus, and aiming to shed more light on the matter, the main results reveal that, although the majority of the students have an understanding of school bullying in line with the theoretical scope of its concepts and types, the percentage of students that relate bullying to hitting, isolating, and ignoring is high, which is closely related to physical and relational bullying. However, behaviours that are also classified as bullying, including bothering the victim to keep them from working or taking away and hiding things from them, are less representative of bullying in the eyes of the stu- dents. This may be because indirect aggressions are incorrectly assumed to be more regular and characteristic of the age, which, in turn, corroborates the lack of information that the students have on, specifically, the plurality of forms that bullying can take (Martínez Pacheco, 2016). In this sense, the students must gain a more detailed understanding of what bullying entails, although the need for a greater understanding is high, as results demonstrated in the case of boys, public schools, and students between the 1st and 3rd year of Compulsory Secondary Education. This fact may be explained since the group with these features seems to be taking more action in bullying and underestimating this harassment (Cano-Echeverri & Vargas-González, 2018).
When asked what bullying behaviours they had witnessed, it is of interest that actions such as insulting and taking away or hiding things are in the first place. This agrees with the results of other studies that record a high incidence of verbal bullying (Avilés & Monjas, 2005;Oñate & Piñuel, 2007) and that rank it as the most common form of bullying (Nicolás Guardiola, 2011). In contrast, and despite the new scenario that technology provides, it is the three forms of cyberbullying considered in the questionnaire that are observed with the lowest frequency, a conclusion that is inconsistent with the prevalence registered by Garaigordobil (2011) and Save the Children (2016) but agrees with Herrera-López et al. (2017) andZych et al. (2015), who also highlight a lower number of cyberbullying cases against the defined as traditional bullying. This lack of agreement is widespread since many studies point it out due to the wide range of operationalisation about these phenomena (Chan et al., 2020). Nonetheless, the data gathered on this type of bullying should not be overlooked due to the severe repercussions for the adolescent students that Chamizo-Nieto and Rey (2020) attribute to it. From the group contrasts, we conclude that the observations made by boys and girls are similar, with boys highlighting hitting and girls ignoring or marginalising the victim. Furthermore, it can be concluded that most bullying events take place at public schools, as Herrera-López et al. (2017) also state in their study, and among students in their third year of Compulsory Secondary Education, which agrees with the increase of bullies that Avilés and Monjas (2005) observe in the ages of 13-15, and then start decreasing.
Concerning the witnesses, a role of utmost importance in the perpetuation of bullying (Sainio et al., 2010;Salmivalli, 2010), the primary reaction among students is to share the information with a family member, followed by sharing it with a teacher. This provides evidence for the predominance of proactive observers, active defenders of the victim, in strong contrast to the indifference generally associated with them (Cuevas & Marmolejo Medina, 2016). Even so, and given that a proportion of the sample indicates the opposite, that they do nothing about it, we consider it opportune to reiterate the necessity to educate adolescents on bullying and make them aware of the repercussions their role as witnesses can have. This necessity is particularly pronounced for boys in 3rd and 4th years since this study reveals a more passive attitude due to a lower understanding of bullying, greater violence suffered or witnessed, and the fear these events entail, according to the results of this and other studies already cited regarding these variables (sex and age).
From the prior discussion, it can be concluded that the knowledge of bullying that adolescents possess needs to be broadened, that the extent to which they observe the behaviours considered here is a clear indicator of the presence of this phenomenon in secondary education in Asturias, and that their roles as witnesses undoubtedly need to be improved in terms of proactivity. This, in line with the proposals of Baldry et al. (2017), raises interest in proposing the ongoing education of students, their families and teachers in this area from an early age. In this sense, the design and implementation of preventive programs that promote the rejection of bullying among students and encourage them to position themselves as proactive witnesses of the phenomenon, thus contributing to reducing or eradicating it, takes on particular relevance.
Finally, although it is considered that the present work constitutes an essential contribution to bridging the gap in the previous findings about the adolescent witness, it is important to highlight the interest in replicating this work in the future to deepen our understanding of the experience of those groups that showed the greatest need for support, such as boys in public schools. At the same time, the possibility of expanding the research to include other educational stages, regions, and the actors directly implicated has not been discarded. Likewise, the data provided serves as a guide for the design of plans, programs and projects aimed at preventing and reducing the effects of school bullying at any educational stage, on a national and local level. Furthermore, future lines of work include the implementation and evaluation, in particular, of evidence-based preventive programs aimed at the entire educational community.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.