Introduction

In many countries, children attend preschool/kindergarten or school between the ages of three and six years and this is true in Spain where 1.2 million children are registered in preschool (MECD, 2018). Research has indicated that interpersonal aggression is observable among this age-group, although these behaviours are still developing at this point. Interpersonal aggression in early childhood has become the focus of scientific interest in recent years, not only because of the associated negative consequences, but also because it may be a precursor of bullying at later ages (Barker et al., 2008; Rose, 2011).

A considerable number of studies have indicated that some children of this age behave aggressively towards others and that this behaviour differs both from rough play (so-called rough and tumble play, Smith & Boulton, 1990), and from bullying (Ortega & Monks, 2005). It has been argued that aggression during the early years does not meet the definition of bullying but is better considered as peer aggression and victimisation (Monks & O’Toole, 2021; Ortega & Monks, 2005; Vlachou et al., 2011). Bullying is defined as a form of intentional aggression in which one person, or several people intend to hurt, harass, humiliate, or exclude socially someone weaker, systematically and over time (Olweus, 1993). The definitional features of the power imbalance between bullies and victims and repetition over time, are key in distinguishing this phenomenon from general aggressive behaviour (Ybarra et al., 2014). Unlike bullying, which is observed during the primary school years, peer aggression and victimisation in the early years include aggressive behaviors where intentionality is not entirely evident. Peer aggression during the early years is characterized by temporal instability. Furthermore, these behaviours do not involve a clear power imbalance between the perpetrator and target of the aggression (Monks et al., 2003). This may be because at this stage of development, the power hierarchies within the peer-group are diffuse and the social roles are unstable (Schäfer et al., 2005). In this line, in later years it has been shown that aggressors tend to choose vulnerable victims (Veenstra et al., 2010). However, in the early school years peer-directed aggression has been found to be less selective, so it is likely that children who behave aggressively towards others change their targets frequently, meaning that experiences of victimisation tend to be shorter-lived for many children at this age (Monks et al., 2003; Murray-Close & Ostrov, 2009). Likewise, preschoolers who are victimised appear to also respond aggressively (Hanish et al., 2012). Furthermore, the intentionality of causing physical or psychological harm is not clear at this age, although some studies indicate that children aged 3 to 6 already understand what it means to harm others (Camodeca & Taraschi, 2015).

Types of Peer Aggression and Victimization and Gender Differences

Studies of peer aggression and victimisation in preschool have identified two types of behaviours: direct and indirect (Ortega et al., 2013). Direct behaviours occur in a direct confrontation between the perpetrator and target and may take physical, verbal or relational forms. This would include behaviours such as hitting, insulting or telling the other that they cannot be part of a game. Aggression can also be indirect and involves behaviours that may be mediated via a third party such as rumour spreading or speaking ill of others who are not present (indirect relational) (Cerda et al., 2012; Ortega et al., 2013). It has been shown that direct forms of aggression are more common that indirect forms during early childhood (Monks et al., 2011), likely because indirect forms require greater social understanding (Huitsing & Monks, 2018). Specifically, it has been found that physical aggression is the most frequent form, followed by verbal forms (Domenèch-Llaberia et al., 2008). The forms of aggression may vary by gender, with more involvement of boys in direct forms (Romera et al., 2021a). Gender differences in peer aggression and victimisation have been debated (Lee et al., 2021; Monks et al., 2021). However, research in preschool and early years of primary school have not found clear gender differences (Lee et al., 2016; Monks et al., 2011). Previous studies underscore that during early childhood, boys are more likely to behave physically aggressively towards their peers (Crick et al., 2006; Huitsing & Monks, 2018), while girls are more likely to exhibit relational and verbal aggression (Card et al., 2008; Douvlos, 2019). More recent research has indicated that girls experience less physical and verbal aggression than boys (Stubbs-Richardson et al., 2018), although some studies have observed that children were most likely to be victimized by children of the same sex (Monks et al., 2021). The literature does not seem to reflect gender differences in victimisation during this age (Lee et al., 2016; Vlachou et al., 2011).

Social Strategies in Peer Aggression and Victimisation

Research has indicated that young children use different behaviour strategies in peer interaction to meet different social and material resource related goals (Rudolph et al., 2011). According to Resource Control Theory (Hawley, 1999), individuals use different behavioural strategies to obtain and maintain resources including coercive and prosocial strategies (Persson, 2005; Roberts et al., 2020). Resource control in children may have immediate beneficial consequences, and these, in turn, may aid optimal development (Willer & Cupach, 2010). Coercive strategies, such as aggression have been widely recognized as a means to obtain control of social and material resources (Reijntjes et al., 2018). However, research increasingly recognizes that prosocial behaviour is also an adaptative means of obtaining resource control. While aggressive behaviour has been found to be the predominant resource control strategy used during the early years, research has indicated that children also employ prosocial strategies at around four to five years of age (Persson, 2005; Roberts et al., 2020).

Some children demonstrate socially insecure behaviour, an important peer interaction strategy, associated with the exhibition of hostile and antisocial actions, present from the early years (Mouratidou et al., 2019). Although aggressiveness, dominance, prosociality and social insecurity are typically identified during the school-age years (Crick & Dodge, 1994), to identify what behaviour strategies are linked with higher levels of peer aggression and victimisation, as well as which of them protect against its development is of practical and theoretical importance.

Aggressiveness is defined as an aggressive behaviour with two different mechanisms and refers to motive systems associated with a provocation or planned behaviour to reach a goal, that predict it: reactive and proactive aggressiveness (Dodge & Coie, 1987; Vitaro et al., 2006). In general, the studies suggest that aggressiveness predicts peer aggression and victimization (Giesbrecht et al., 2011; Ostrov, 2008), specifically with direct forms as physical aggression (Evans et al., 2019). In preschool, aggressiveness is mostly reactive to perceived threat (Vlachou et al., 2011).

Social dominance refers to the cognitive estimation of skills, abilities, and control behaviours over others (Charlesworth & Dzur, 1987). High levels of social dominance have been linked to aggression in early childhood, regardless of gender (Hawley & Geldhof, 2012; Huitsing & Monks, 2018; Perren & Alsaker, 2006). Regarding victimisation, research shows that victims are not lacking in dominance (Huitsing & Monks, 2018; Monks et al., 2011), although other research suggests that they lack assertiveness, which may place them at risk for repeated victimisation (Perren & Alsaker, 2006).

Prosocial behaviour, understood as a voluntary strategy behaviour, is recognized as a positive social skill that benefits others (Eisenberg et al., 2006; Marín, 2010), and is key to promoting positive social interactions (Padilla-Walker et al., 2015), although the underlying motivation can be selfish rather than altruistic (Hawley, 2014). In the preschool period, studies have shown that prosocial behaviours may be associated with both perpetration and victimisation, regardless of child gender (Ortega, et al., 2013), a relationship that may be mediated by attitudes and the perception of the morality of the behaviours of minors (Dereli, 2019).

Social insecurity refers to the demonstration of withdrawal, shyness, anxiety, and social avoidance. It has been recognized that insecurity in children is associated with maladaptive behaviours and relationships, including rejection and social isolation (Colonnesi et al., 2011; Crocker & Park, 2004; Ladd & Burgess, 1999). Perpetration of aggression in preschool has been characterized by lower social insecurity, while victimisation was not related to insecurity in this age-group (Huitsing & Monks, 2018; Mouratidou et al., 2019).

In summary, although the high level of aggressiveness, social dominance, and insecurity can place individuals at risk of peer aggression, prosociality may be a protective factor. Despite the advances in the relationship between social behaviour strategies and the involvement in peer aggression or victimisation, it is necessary to explore how they interact. For this, to identify different strategies that develop simultaneously may support a decline in peer aggression and victimisation. This study is particularly relevant because identifying precursors of aggression and victimisation patterns in early childhood, a key stage in the emotional and social development, may have implications for understanding development of aggression over the years and enable work to prevent aggressive behaviour escalating. For this, it is necessary to take a different approach adapted to their level of literacy. This led to the use of individual interviews supported by illustrations to elicit child reports of aggression by peers and self (Monks et al., 2003; Ortega & Monks, 2005).

The objective of this study was to: (a) analyze the social behaviour strategies associated with peer aggression and victimisation in preschoolers; (b) explore the interaction of gender and social behaviour strategies on involvement in physical and verbal peer aggression and victimisation.

The current study has the following hypotheses:

H1

Physical and verbal aggression are expected to be positively associated with aggressiveness and dominance and negatively with insecurity and prosociality.

H2

Physical and verbal victimisation are expected to be positively associated with aggressiveness and prosociality, and negatively associated with dominance.-+

H3

Involvement in physical and verbal aggression and victimisation is expected to be greater in boys than in girls.

Material and Methods

Participants

A total of 227 Spanish preschoolers and students from the early years at school (58.1% girls, n = 132) were involved, aged between 4 and 7 years (M = 5. 61; SD = 1.03; 4 years old = 16.7%, 5 years old = 30%, 6 years old = 29.1%, and 7 years old = 24.2%) from three different school year groups: 2nd year of infant class (35.7%; 66.3% girls—33,8% boys), 3rd year of infant class (27.3%; 47.6% girls—52.4% boys), 1st year of primary school (37%; 58.3% girls—41.7% boys). The classrooms contained between 7 and 26 pupils (M = 20. 61; SD = 4.77). Classes of between 7 and 15 students were 10.9% of the sample, classes with between 16 and 21 pupils made up 40.1% and those with between 22 and 26 pupils were 49% of the sample. An incidental non-probabilistic sampling approach was taken, in which a convenience sample was recruited. The participants belonged to five state-funded schools which were situated in a rural environment, with an average socio-economic level.

Instruments

Child reports on involvement in peer-aggression and victimisation were obtained using the interview technique described by Monks et al. (2003). There is evidence for the reliability and consistency of peer reports of aggression by young children (Monks & Smith, 2010). Children were shown stick-figure cartoons exemplifying different types of aggression: physical (hitting, kicking or pushing another) and verbal (shouting or saying unpleasant things to another) (see Fig. 1).

Fig. 1
figure 1

Stick-figure cartoons exemplifying physical and verbal aggression

Full size image

To identify the different types of aggression, children were asked to report whether any of their classmates behaved in these ways and whether any of their peers experienced these forms of aggression (victimisation) and were then asked to nominate peers for these behaviours. Children were able to nominate as many peers as they wished for each of the behaviours. Children were also asked to self-report the behaviours as well—either as taking that behaviour or not. Peer-reports received for each child for aggression or victimisation were summed and standardised by class size. Students were assigned to a role if they obtained a scored above the classroom mean on the scale for that role (z > 0).

Social behaviour strategies. The teacher version of Dodge and Coie's Reactive Proactive Aggression Questionnaire for Teaching Staff (1987) was employed (see Appendix). Four dimensions of this scale were selected for the current study—overall aggression (e.g., this child starts fights with peers) α = .91; ω = .88; dominance (e.g., this child thinks they are better than others) α = .89; ω = .82; insecurity (e.g., this child is too shy to make friends easily) α = .74; ω = .77; prosociality (e.g., this child is very good understanding the feelings of others) α = .67; ω = .75. The teacher was asked to complete a questionnaire for each pupil in their class using a 7-point frequency Likert scale. The results of the confirming factorial analysis were optimal for a single scale with a solution of 4 factors, being the indices for the full scale χ2 (28) = 166.387, p < .001, CFI = .95; NNFI = .95; RMSEA = .064 (90% confidence interval [IC] [.052, .074]). To analyze the variables, Freeman’s degree centrality was used for each dimension (Freeman, 1978), using standardized scores ranging from 0 to 100.

Procedure

The data collection was carried out just before the pandemic. Schools were approached for their participation. The child interviews were conducted individually during school hours in a quiet area outside of the classroom, without the presence of teachers or peers, in a single session of up to 30 minutes. Not all students in all classrooms participated, only those for whom parental/carer consent was obtained. As in Huitsing and Monks (2008), each child was shown stick-figure images of the different forms of aggression and were asked to identify who behaves in this way and to whom they did this. In this way, nominations were obtained for aggression and victimisation.

Teachers were asked to complete the questionnaires in their own time. All teachers in all classrooms participated.

Written consent was obtained from the participants' parents or legal guardians and oral consent was obtained from the children and teachers themselves. This study complies with the Helsinki Declaration on Confidentiality, Privacy and Informed Consent and has been approved by the Ethics Committee of Bioethics and Biosecurity of The University of Córdoba.

Data Analysis

Following the recommendations of Afifi et al. (2004), to determine whether an independent variable has a general effect on dependent variables, multivariate analyses were performed in three steps (DVs—physical aggression, verbal aggression, physical victimisation and verbal victimisation; IVs- gender, dominance, insecurity, aggressiveness and prosociality). First, the gender variable was added to the model. Second, the variables dominance, insecurity, aggressiveness and prosociality were added. Third, interactions between the psychosocial variables and gender were added. For parsimony, the final multivariate model only reflected significant interactions. To determine whether the addition of separate step variables contributed significantly to the overall multivariate linear model, we compared the models on the standard function ANOVA in R, to test for significant decreases in multivariate test statistics (e.g., Wilks’s lambda). As a final step, linear regression was used through the stepwise method with the resulting variables from the multivariate model. The correlated errors were examined by the Durbin-Watson statistic. The recommended range (1.5–2.5) indicates that errors are independent, so the absence of positive autocorrelation (values close to 0) and negative autocorrelation (values close to 4) (Durbin & Watson, 1971).

To illustrate the significant interactions resulting from the previous analyses, the simple slope test (Preacher et al., 2006) was performed. Encoding and analyzing data were performed with the SPSS version 22 statistical package. p < 0.05 significance levels were adopted.

Results

Prevalence and Sociodemographic Variables of Victimisation and Aggression in Preschoolers

Boys showed greater involvement in both peer-reported physical and verbal aggression, physical victimisation and teacher reports of general aggression. A positive relationship was observed between the peer-reported variables of physical aggression, verbal aggression, physical victimisation, and teacher-reports of dominance and aggressiveness with scores ranging from 0.14 to 0.51. Verbal aggression also correlated positively with verbal victimisation and aggressiveness with values of 0.60 and 0.14, respectively. Physical victimisation showed a positive relationship with dominance (0.28) and aggressiveness (0.44), and dominance with general aggression (0.73). Insecurity correlated negatively with prosociality (− 0.23) and aggressiveness (− 0.24) (see Table 1).

Table 1 Correlations, and descriptive, and gender differences for aggression and victimisation according to the typologies of physical and verbal behaviours
Full size table

Multivariate Regression of Physical and Verbal Peer Aggression and Victimisation

Analysis of the general multivariate linear model showed a significant association of gender with the dependent variables for Model 1 (see Table 2). When the social behaviour strategies were added in Model 2, the results indicated that gender, dominance, aggressiveness, and prosociality were significantly associated with the dependent variables. For Model 3, the results indicated that the interaction between gender and aggression was significantly related to the dependent variables.

Table 2 Multivariate general linear model for predictive variables of physical and verbal peer aggression and victimisation
Full size table

Linear Regression

The results of the linear regression analyses for physical and verbal aggression are presented in Table 3. All models are included for comparison.

Table 3 Linear regression model coefficients for physical and verbal aggression behaviours
Full size table

For physical aggression, the linear regression analysis showed statistically significant results (F = 22,299; df = 1, p < .001). Due to the significance of the change in R2, Model 2 was selected, with a value of R2 of 0.35. The Durbin-Watson statistic yielded a value of 1.97, confirming the absence of positive autocorrelation (values close to 0) and negative (values close to 4). Gender was found to be significantly related in all models, with boys being most associated with physical aggression. Social insecurity was negatively, and aggressiveness was positively associated with physical aggression (see Table 3).

For verbal aggression, the linear regression showed statistically significant results (F = 12,418; df = 1, p < .001). Model 3 was selected with a value of R2 of 0.257. The Durbin-Watson statistic yielded a value of 2.187. Gender was significant in all models, with boys being most related to this type of aggression. Insecurity was negatively associated, and aggressiveness positively associated (see Table 3).

The interaction of gender and aggressiveness was significant for verbal aggression. Simple slope analysis showed a positive association between verbal aggression and the aggressiveness in both boys (simple slope: B = 0.40; p = .019), and girls (simple slope: B = 0.50; p = .003). Follow-up analysis showed that the effect of general aggression on verbal aggression was not significantly different for boys and girls with high (simple slope: B = 0.02; p = .921) or low (simple slope: B = − 0.27; p = .092) levels of aggressiveness. These results indicate that in verbal aggression the positive relationship was stronger for girls compared to boys even though the gender effect was not significantly different at high or low levels of aggressiveness (see Fig. 2).

Fig. 2
figure 2

Graphical representation of the interaction between gender and aggressiveness in relation to verbal aggression in preschoolers

Full size image

For physical victimisation, linear regression analyses showed statistically significant results (F = 5078; df = 4, p < .001). Based on the significance of the change in R2, Model 2 was selected, with a value of R2 of 0.093 (see Table 4). The value of Durbin-Watson was 0.683, which confirmed the validity of the model. Dominance was negatively associated and aggressiveness positively to physical victimisation (see Table 4).

Table 4 Linear regression model coefficients for physical and verbal victimisation behaviours
Full size table

For verbal victimisation, linear regression showed statistically significant results (F = 3034; df = 1, p = .007). The value of R2 for Model 3 was 0.058. The value of Durbin-Watson was 1.649, so the validity of the model was confirmed. Gender was significant in all models, with boys being the most related to this type of victimisation. Prosociality and aggressiveness were both positively associated with verbal victimisation (see Table 4).

The gender by aggressiveness interaction was significant for verbal victimisation behaviours. Simple slope analysis showed a negative association between verbal victimisation and the pattern of aggression in boys (simple slope: B = 0.33; p = .002), but not in girls (simple slope: B = − 0.11; p = .236). Follow-up analysis showed that the effect of aggression on verbal victimisation was significantly different for boys and girls with high levels of aggressiveness (simple slope: B = − 0.57; p = .002), but not with low levels (simple slope: B = 0.05; p = .752). These results indicate that boys were more likely than girls to be involved in verbal victimisation at high levels of aggressiveness (see Fig. 3).

Fig. 3
figure 3

Graphical representation of the interaction between gender and aggressiveness in relation to verbal victimisation behaviours in preschoolers

Full size image

Discussion

This study contributes to the knowledge of peer aggression and victimisation during early childhood. The aim was to examine the behavioural strategies associated with the development of direct forms of perpetration/aggression and victimisation among young children.

Social Behaviour Strategies and Physical and Verbal Aggression

Teacher reported aggressiveness and social insecurity were important social behaviour strategies to explain the involvement in aggression (physical and verbal) in the early years of schooling. These results confirm findings from previous studies among older children (Vlachou et al., 2016), which point to aggressive behaviours, characterized by impulsivity and spontaneity, even in the presence of teachers.

Social insecurity was inversely associated with aggression, both physical and verbal. These results coincide with previous research that indicates that low levels of social insecurity characterize aggressors (Huitsing & Monks, 2018; Perren & Alsaker, 2006). This is consistent with studies that showed that the search for novelty, impulsiveness, and daring is associated with aggression among preschoolers rather than anxiety, insecurity, or withdrawal (O’Toole et al., 2017; Schmidt et al., 2002).

In the case of social dominance and, contrary to previous studies (Huitsing & Monks, 2018; Perren & Alsaker, 2006), it was found that this was not related to perpetration. This result suggests that there are not yet well-established individual differences in dominance within early childhood, which characterizes bullying at later ages (Romera et al., 2021b). It could be considered that in early childhood there is some instability in the involvement in aggression, there is not a clear intention to dominate peers, but rather to achieve instrumental objectives, as suggested by Crick and Dodge (1994).

Prosociality also did not show a link to involvement in aggression, either in its physical or verbal form. This can perhaps be explained because prosocial behaviour, which evolves with age, depends on various factors such as emotional support in the classroom (Johnson et al., 2013), perception and moral sensitivity (Dereli, 2019) or parenting styles (Gómez-Ortiz et al., 2019; Köster et al., 2016).

In short, it seems that perpetration is linked to general traits of aggressiveness and low levels of social shyness. Regarding prosociality and dominance, perhaps they are social behaviour strategies that require cognitive maturity to discern in which social situations to behave prosocially and when to exercise dominance in a strategic way, abilities that preschoolers have not yet acquired. Although, in line with Hawley (2014), some preschoolers do behave in a strategic way using prosocial behaviour and aggression interchangeably.

Social Behaviour Strategies and Physical and Verbal Victimisation

The second hypothesis was partially validated. Regarding social dominance, and in line with some previous studies (Perren & Alsaker, 2006), the results have shown that it is only negatively associated with physical victimisation, and not verbal victimisation. It is likely that it is force, physical strength, and other traits, which intimidate victims. On the other hand, verbal victimisation is likely associated with other aspects of the social context, as well the instability of the interpersonal relationships themselves (Lee et al., 2022). Social insecurity or shyness, as hypothesized, showed no connection to victimisation, whether physical or verbal. This result shows that victimisation in the early years has important differential characteristics compared to what happens in later years (Romera et al., 2020). Perhaps during early childhood, children do not choose to victimize their weakest companions or those who are socially isolated but display some randomness in the choice of their victims, hence there is no association between social shyness and victimisation (Huitsing & Monks, 2018). Previous research has found that prosociality is associated with both physical and verbal victimisation (Suárez-García et al., 2020). In the current study, there was a relationship between prosociality and verbal forms of victimisation. This result could be explained based on the nature of prosociality, which involves engaging in the problems of others to help them and provide more socially active responses, which perhaps exposes them to the aggression of others to a greater extent.

The third hypothesis was also partially confirmed. The results showed that while gender was associated with aggression and victimisation behaviours, only the interaction with verbal aggression was significant for both boys and girls. These data contradict previous studies that seem to indicate that both forms of aggression are more common among males as perpetrators (Douvlos, 2019; Huitsing & Monks, 2018; Romera et al., 2022). Regarding victimisation, the data reflected that the boys were more likely than girls to experience verbal victimisation, if they had obtained high scores on aggressiveness, according to teacher reports. In this regard, the previous literature does not appear to have established a clear relationship between gender and victimisation (Lee et al., 2016). Perhaps gender is only relevant depending on the level of aggression a child displays, which at these ages is very unstable (Hanish et al., 2012; Huitsing & Monks, 2018). These results suggest that preschool is a key point, for both girls and boys, in terms of promoting positive social and emotional development.

Conclusions

The conclusions of this study reinforce the importance of behavioural strategies (prosociality, aggressiveness, dominance, insecurity) along with the social perceptions of peers and teachers. It is clear, that the perception of both observers is moderately different from each other, but a certain salience of some traits, particularly aggression and social dominance, but also prosociality and to a lesser extent shyness or insecurity, may be reinforcing the roles of aggressor and victim, which at these ages are not fixed (Camodeca et al., 2015; Monks et al., 2003). Recognizing the differences between what is happening in early childhood compared to what is known about bullying in older children is important and opens a debate about what can and cannot be prevented in these years, in order to tackle bullying and cyberbullying later on.

Implications and Future Research

The results provide some guidance for interventions. According to aggression in early childhood an important implication for the field education is that in this behaviour the social security and aggressiveness are relevant factors. So, providing strategies to support children’s emotional and behavioral self-regulation may help children to establish positive peer relationships (Savina, 2021).

Other implication of this research is related to victimisation and the association with factors as dominance, aggressiveness and prosociality. It implies the design of education programs aimed to promote the development of effective strategies to minimize victimization (Mora-Merchán et al., 2021). Therefore, preschool educators should be encouraged to promote activities, based for example on communication or affective play, in which there are opportunities for interactions between children and their teachers to support positive emotional and social adjustment (Zych & Llorent, 2020).

A third implication of this study is related to results that connect the gender with aggression and victimisation behaviours. In this line, educators should be encouraged to pay attention to the types of activities that both boys and girls engage in order to address social interaction problems among peers that may lead to aggression.

In general, practitioners should consider social behaviour strategies, focused on aggression and victimisation behaviors, and establish intervention practices aimed at optimizing social competence and moral sensitivity from the early years of children’s education. So, for example, the practice of activities related with emotional intelligence and aspects such as self-control, motivation and emotional awareness may be useful (Kutnick et al., 2007). Based on the findings, the specific recommendations for today’s ECE classroom include the design and implementation of programs to guide social strategies and prevent the interpersonal violence at these ages. Creating social and emotional programs requires engagement from families, teachers, and preschoolers. Having clearly programs based on social strategies and training scholar community reflects what is important in early years. For example, creating programs focused on families around emotional and social management at home and providing education to preschoolers, teachers and supervisors. So, social and emotional programs as the center of early childhood curricula should imply the development of ethical children.

In terms of limitations, the cross‐sectional design of the study should be considered. A longitudinal design would provide further evidence on the social behaviour strategies model of peer-aggression in that it would make it possible to better understand the effects of this variables on this preschoolers’ behaviours. Furthermore, a longitudinal design would enable an examination of the effects of prolonged exposure to peer-aggression as well as the long‐term effects on other phenomena such as bullying. A further limitation is that evaluations made by teachers may not conform to children’s experiences in the classroom, although within the current study teacher-reports were combined with child reports of behaviour.

Future research should evaluate the effectiveness to reduce peer violence through training and promoting social and emotional evidence-based policies and programs. Also, further studies focused on preschoolers and other personal and contextual characteristics that could aid in our understanding of the developmental origins of aggression and victimisation behaviours should be considered as future lines of research. Finally, there are key developments that occur between the ages of four and seven years of age which were not addressed in the current study but would also be important lines of research to examine in the future.