Bidirectional longitudinal associations of perpetration and victimization of peer bullying with psychosocial problems in adolescents: A cross-lagged panel study

Abstract

Research on school bullying often focuses on the directional path of bullying and/or victimization leading to psychosocial problems, while such one-dimensional views have been shown to be too simplistic. Furthermore, recent research has shown that patterns of bullying at school differ for boys and girls, which makes gender a particularly relevant factor in exploring the causes and consequences of bullying. Therefore, the present study explored the bidirectional, longitudinal associations of bullying and bullying victimization on several psychosocial problems via a longitudinal cross-lagged panel study in 1243 adolescents in the Netherlands, while taking into account potential gender differences. Data were collected in September 2011 and 2012. Results showed that both bullied boys as well as girls reported more conduct problems at follow-up. Both boy and girl bullies reported less pro-social behavior and more peer problems at follow-up, but boys also reported more conduct problems at follow-up, while girls did not. Furthermore, in girls, emotional problems were associated with more victimization at follow-up, while inattention-hyperactivity problems and less pro-social behavior were related to increased chances of being a perpetrator of bully at follow-up. Conversely, in boys, baseline inattention-hyperactivity problems were not associated with being a bully later on, but rather with increased chances of being a bullying victim at later times. These results can help to tailor future anti-bullying interventions at schools.

Keywords

Adolescent being bullied bullying psychosocial problems school

Many people have long considered bullying to be a harmless part of growing up, while this is a serious misconception (Copeland, Wolke, Angold, & Costello, 2013). Nowadays, it is increasingly perceived as a serious public health threat with severe consequences for victims’ health, happiness, and life-long psychosocial and psychiatric problems (Copeland et al., 2013; Esbensen & Carson 2009; Kaltiala-Heino, Rimpela, Rantanen, & Rimpela, 2000; Olweus 1993; Wienke Totura, Green, Karver, & Gesten, 2009). Bullying is defined here as intentional and repetitive negative actions (e.g. physical aggression, verbal abuse or social exclusion) characterized by an imbalanced power between the victim and bully or perpetrator (Olweus 1993; Wienke Totura et al., 2009). Recent studies in adolescents showed that psychosocial, behavioral problems such as conduct problems, aggressive behavior, hyperactivity, poor peer relations and lacking pro-social behavior skills are all related to bullying and being bullied (Stavrinides, Georgiou, Nikiforou, & Kiteri, 2011; Stavrinides, Georgiou, & Theofanous, 2009; Veenstra, Lindenberg, Munniksma, & Dijkstra, 2010; Wolke, Woods, Bloomfield, & Karstadt, 2001). Much research on the topic often did not utilize a longitudinal design (Hawker & Boulton, 2000) and/or was not performed specifically in adolescents (Wolke, Woods, Bloomfield, & Karstadt, 2000). Studies that were performed via longitudinal design in adolescents showed significant effects of bullying and being bullied on different psychosocial problems (Esbensen & Carson, 2009; Fekkes, Pijpers, Fredriks, Vogels, & Verloove-Vanhorick, 2006; Kaltiala-Heino et al., 2000; Stavrinides et al., 2010; Stavrinides et al., 2011).

However, should bullying be seen a cause or as a consequence of such psychosocial problems? Or should it be seen as both? While in some studies bullying and victimization were demonstrated to lead to emotional and behavioral problems (Arseneault, Bowes, & Shakoor, 2010; Nansel et al., 2001) others claim there is only evidence for the opposite direction of these effects (Hogdes & Perry, 1999; Schwartz, 2000; Veenstra et al., 2005). As highlighted by Fekkes et al. (2006), many studies only address a single directional path in the analyses of bullying and psychosocial problems, while increasing evidence shows complex, bidirectional effects are more likely to be at play in case of both internalizing as well as externalizing psychosocial problems. However, despite these advances, recent studies state that more longitudinal evidence is needed to unravel the complex, bidirectional associations between bullying behaviors and particular psychosocial problems (Arsenault et al., 2010; Reijntjes et al., 2011). Reijntjes et al. (2011) also indicated this in their recent meta-analysis on the bidirectional associations between bullying and externalizing psychosocial problems. As Copeland et al. (2013) illustrated, these effects are not limited to the bullied, but bullies themselves are also at risk for, e.g. depression and antisocial personality disorder. Similar conclusions were drawn in another recent meta-analysis that studied the bidirectional associations between bullying and internalizing psychosocial problems (Reijntjes, Kamphuis, Prinzie, & Telch, 2010).

Gender differences

Recent research has shown that patterns of bullying at school differ for boys and girls. Such studies show that boys are not only more often involved in bullying, e.g. a higher perpetration frequency, but also that they are involved in different types of bullying than girls (Camodeca, Goossens, Terwogt, & Schuengel, 2002; Nansel et al., 2010; Wang, Ianotti, & Nansel, 2009; Young & Sweeting, 2004). Furthermore, these studies show that coping strategies and particular psychosocial problems associated with bullying also differ between boys and girls. It was for example shown that girls are more at risk for internalizing symptoms and more likely to use avoidance strategies such as ignoring and not doing anything to stop the bullying, while boys are more likely to use physical aggression or revenge as a coping strategy, or humor as an attempt to distract and stop the bullying (Baldry, 2004; Pepler & Craig, 2008). Such findings make gender a particularly relevant factor in exploring the causes and consequences of bullying (Nansel et al., 2001; Veenstra et al., 2010; Wang et al., 2009).

The present study

Despite the advances in recent research on the effects of school bullying on adolescents’ psychosocial well-being and vice versa (Reijntjes et al., 2010; Reijntjes et al., 2011), more longitudinal studies are needed to explore the bidirectional paths of these associations (Arsenault et al., 2010; Reijntjes et al., 2010; Reijntjes et al., 2011). Therefore, the main purpose of this study is to assess such bidirectional paths, while differentiating between particular psychosocial problems and taking into account potential gender differences.

Methods

Participants

A convenience sample of high school students in The Netherlands was asked to complete a survey in September 2011 (baseline) and September 2012 (wave 1) based on the Dutch Health Behavior in School-Aged Children (HBSC) questionnaire (van Dorsselaer et al., 2009). The schools participated out of intrinsic motivation, without being provided with funds or other incentives to participate. All schools are situated in, to Dutch standards, suburban areas of middle-large to large cities and included students from both urban and rural background.

Procedure

The survey was completed individually in class. Survey procedures were designed to allow students to participate voluntarily and anonymously. Students were informed about the nature of the survey beforehand, as were their parents. The fact that all answers would be confidential and that participation was voluntary was also made clear in advance. At any time, student and/or parent could decide to withdraw the student from participation; there were no parents who withdrew their child/children from the study. All schools were assisted with the questionnaire procedures by the research team. No specific intervention(s) that focused on bullying were implemented in the period of time that the surveys spanned. The medical ethical committee of the University Medical Centre Utrecht has approved to the procedures of this study after reviewing its ethical considerations.

Instruments

Demographic characteristics

The collected demographic characteristics were age, gender, ethnicity, socio-economic status (SES), which school and class a student attended and his/her current educational level. All these characteristics were surveyed similarly as in the Dutch HBSC questionnaire. Students’ SES was measured via the Family Affluence Scale (FAS), as integrated in the HBSC survey (FAS-score: 0–2 = low; score 3–5 = medium; 6–9 = high) (Currie et al., 2008).

Bully behaviors

Dan Olweus’ conceptual definition of bullying was used, meaning that the following three criteria need to be met: (1) physically harming, making fun of, excluding and/or spreading rumors (gossiping) about someone; (2) the behavior must be carried on over time repeatedly; and (3) the bully and victim do not possess equal strength or power in the interactions (Olweus, Limber, & Mihalic, 1999). This was questioned by asking whether students had been bullied in the recent three months, and then also more specifically whether this happened with regard to physical aggression, verbal abuse, gossiping/rumor spreading and/or via the Internet or mobile phone. Similar questions were also asked with regard to being the perpetrator of the bullying behavior. This operationalization of the measurement of bully behavior was based on the Olweus Bully Score and the Olweus Bully Victim Score (Solberg & Olweus, 2003). The two scores distinguish bullies and victims from respectively non-bullies and non-victims with a validated cut-off of at least two times a month. In previous European studies, these measures’ validity and reliability were demonstrated, stating that these scores allow for prevalence estimates of bullying and being bullied to be obtained conveniently, that they have a reasonably well-defined meaning, and that they are easily and unambiguously understood by users and researchers (Kyriakides, Kaloyirou, & Lindsay, 2006). Although this does not guarantee that this will be the case outside of Europe, it does provide confidence that the questions were understood and interpreted well in our Dutch study. According to Kyriakides et al. (2006), the Olweus Bully Score and the Olweus Bully Victim Questionnaire has satisfactory psychometric properties to measure bullying and bully victimization in a general population of adolescents such as was used in the current study. They concluded that the conceptual design of the instrument, construct validity (using the Rasch Model; cumulative D = 82%) and internal consistency (r > 0.88) were satisfactory. In addition, Kyriakides et al. also performed these psychometric property checks on a gender-stratified sample and found that the properties remained similar (Kyriakides et al., 2006). Therefore, the Olweus Bully Score and the Olweus Bully Victim Questionnaire can be noted an acceptable instrument to meaure bullying in the current study.

Lastly, to prevent ambiguity, those who are the perpetrators of bullying will be referred to as bullies, while those being victimized will be referred to as victims or being bullied. Other studies sometimes also identify a third category, namely those who are simultaneously the perpetrators and victims of bullying (the so-called bully-victims). These were not singled out in the current study due to sample size considerations. However, they were taken into account in the statistical analyses by determining the relationships between bullying and being bullied with psychosocial problems in the same model. Since the behaviors of bullying and being bullied are highly correlated, they are likely to share a certain variance in their associations with psychosocial problems, which may lead to confounding issues when not being taken into account in the same model. Therefore, all studied associations were integrated into the same model.

Psychosocial problems

Psychosocial problems were measured with the Strengths and Difficulties Questionnaire (SDQ) (Goodman, Meltzer, & Bailey, 1998), which measures behavioral characteristics on the topics emotional symptoms, inattention-hyperactivity, peer relationship problems, conduct problems and pro-social behavior (Goodman et al., 1998). The SDQ’s reliability and validity were recently demonstrated in a comparable sample of Dutch youth to measure these five types of problems with appropriate distinction between them (Muris, Meesters, & van den Berg, 2003). In their study on the validity and reliability of the SDQ, Muris and colleagues, similar to our current study, used a normal Dutch community sample of adolescents. Their SDQ analysis yielded five factors that were in agreement with the existing SDQ subscales of hyperactivity-inattention, emotional symptoms, peer problems, conduct problems, and prosocial behaviour. In addition, they found acceptable results in terms of the SDQ’s internal consistency (Crohnbach’s alpha = 0.64) and test–retest stability (ICC = 0.59). Lastly, in regard to the concurrent validity of the SDQ, in the study of Muris et al, its scores correlated in a theoretically meaningful way with other measures of psychopathology (YSR, r = 0.74; RCMAS, r = 0.58–0.74; CDI, r = 0.64; ADHDQ-C, r = 0.46–0.66) (Muris et al., 2003). Therefore, it can be concluded that the psychometric properties of the self-report version of the SDQ are satisfactory to use the current study.

The subscales emotional symptoms, inattention-hyperactivity, peer problems and conduct problems, each composed of five items scored on a three-point Likert-scale (0 = ‘not true’; 1 = ‘somewhat true’; 2 = ‘certainly true’), together add up to a total SDQ-score of maximum 40 points over the total of 20 items. A score of 0–15 points is considered a ‘normal’ score, whereas a score higher than 15 is categorized as ‘(potentially) problematic’. The pro-social behavior score is left out of this total score and analysed separately, as is common practice (Goodman, 1997; Muris et al., 2003).

Data analysis

To assess the longitudinal, bidirectional associations between bullying and victimization with the latent construct of psychosocial problems, a cross-lagged panel analysis was performed on two data waves via structural equation modeling (SEM) in Mplus (Muthén & Muthén, 2007). This way, relationships of bullying on psychosocial problems were analysed with a one-year lag time, while controlling for reverse relationships. The same was done regarding victimization and psychosocial problems. This design was chosen because cross-relationships over time enable examining bidirectional relationships and a temporal order in these relationships. In order to control for the possible confounding effect of age, ethnicity, educational level and socioeconomic status, those variables were included in each model. Another advantage of the chosen analytical approach and statistical package was that all analyses were done within the same model, so that the problem of multiple testing (and a possibly inflated p-value) was averted.

As the latent scores on psychosocial problems were not distributed normally, the assumption of multivariate normality was not fully met. However, in SEM, multivariate normality concerns a sufficient but not necessary condition for applying the usual tests of statistical significance. Especially with larger samples, the parameter estimates are not likely to be biased when the normal-distribution assumption is violated (Bollen, 1989). Therefore, maximum likelihood estimations with robust standard errors (MLR) were used (Byrne, 1998).

As stated, different associations between bullying and victimization with psychosocial problems were hypothesized for boys and girls, based on existing literature. Therefore, all analyses were performed on a gender-stratified sample.

Results

A total of 1242 participants completed the survey at both time points T0 and T1. Within each wave of data collection approximately 80% of the students completed the survey; those who did not were absent from school on the day that the survey took place or they were not given the survey by their teacher (due to conflicting scheduling and time issues). Students who were in their last year of high school at baseline and those in their first year at wave 1 could, by design, only complete time point of data collection and were thus excluded from the analyses. The number of students who did not complete the survey when presented with it was negligible. Of the surveys, 95% was fully complete on all used variables; no methods for data imputation were used.

Of the 1242 participants, 504 were boys and 738 were girls. Their mean age at baseline was approximately 13-years-old (Mean 13.17, SD 1.21). Their educational level was representative for a Dutch sample of adolescents, while their socioeconomic status (FAS score, Mean = 7.12 (SD = 1.49)) was somewhat higher than that of their peers in The Netherlands (van Dorsselaer et al., 2009). Of the girls, 2% reported having been bullied recently, while 6% indicated having been bullied recently at baseline. At follow-up 2% were perpetrators and 4% victims. For boys, 8% were recently bullied and 6% indicated having been bullied by peer(s) recently. At follow-up 5% were victims and 4.5% perpetrators. At baseline psychosocial problems were present in 15% of the boys and 12% of the girls. At follow-up this was respectively 21% for girls and 15% for boys.

Firstly, the results of the ‘general model’ present the bidirectional associations between bullying and victimization with overall psychosocial problems (Table 1). Bullied boys and girls were more likely to also be bullied at follow-up (respectively OR 5.03, 95% C.I. 3.00–8.44 and OR 5.26, 95% C.I. 3.42–8.08). Following being bullied at T0 furthermore led to psychosocial problems at T1, i.e. a problematic SDQ score, in girls (OR 2.00, 95% C.I. 1.29–3.12). Adolescent boys who reported bullying others at baseline were more likely to report psychosocial problems at follow-up (OR 1.72, 95% C.I. 1.01–2.91), but this was not the case in girls. Furthermore, those that reported being bullies at baseline were also more likely to report this behavior than those who did not bully at baseline. Overall, being a victim of bullying had a somewhat more pronounced effect on the psychosocial problems of girls than it did in boys (Table 1). In contrast, bullying had more effects on, which seemed to indicate more externalizing problem behavior in boys following being a bully than it did in girls. This was illustrated both by the significant effects on the total SDQ scale in boys (whereas this was not the case in girls), as well as on stronger effects on, especially, externalizing problem subscales of the SDQ such as conduct problems (Table 2).
Table 1.
The longitudinal associations of bullying and victimization with the development of general psychosocial problems in adolescents, stratified by gender (N = 1242).

Longitudinal associations from baseline (T0) to time 1 (T1) Odds Ratio^a (95% CI)

Boys Girls

Being bullied at T0 > Being bullied at T1 5.03* (3.00–8.44) 5.26* (3.42–8.08)

Being bullied at T0 > Psychosocial Problems at T1 1.49 (0.84–2.66) 2.00* (1.29–3.12)

Bullying at T0 > Psychosocial Problems at T1 1.72* (1.01–2.91) 1.12 (0.62–2.04)

Bullying at T0 > Bullying at T1 2.86* (1.66–4.93) 3.90* (2.13–8.04)

Psychosocial Problems at T0 > Psychosocial Problems at T1 10.61* (6.40–17.60) 5.26* (3.42–8.08)

Psychosocial Problems at T0> Being bullied at T1 1.23 (0.62–2.46) 1.64 (0.88–3.08)

Psychosocial Problems at T0 > Bullying at T1 1.31 (0.59–2.18) 0.79 (0.37–1.68)

a
Alpha of 0.05 is used as significance level; CI = Confidence Interval; * = significant.

Table 2.
The longitudinal associations of bullying and victimization with the development of specific psychosocial problems in adolescents, stratified by gender.

Longitudinal associations
Odds Ratio^a (95% CI)

Baseline (T0) Follow-up (T1) Boys Girls

Being bullied Emotional problems 0.87 (0.36–2.07) 1.57 (0.88–2.79)

Bullying Emotional problems 0.83 (0.36–1.87) 1.02 (0.49–2.14)

Being bullied Conduct problems 1.75* (1.03–3.00) 2.31* (1.41–3.78)

Bullying Conduct problems 2.00* (1.23–3.27) 1.54 (0.83–2.88)

Being bullied Peer problems 1.09 (0.64–1.86) 1.46 (0.88–2.42)

Bullying Peer problems 1.84* (1.12–3.02) 2.26* (1.27–3.99)

Being bullied Inattention-Hyperactivity 1.20 (0.70–2.06) 0.70 (0.43–1.12)

Bullying Inattention-Hyperactivity 0.88 (0.51–1.51) 1.29 (0.74–2.24)

Being bullied Pro-social problems 1.02 (0.62–1.70) 1.12 (0.64–1.97)

Bullying Pro-social problems 1.98* (1.24–3.15) 3.84* (1.95–6.21)

Emotional problems Emotional problems 5.67* (2.22–14.44) 16.60* (10.45–23.40)

Being bullied 0.36 (0.12–1.10) 2.18* (1.17–4.06)

Bullying 0.37 (0.11–1.33) 0.78 (0.33–1.88)

Conduct problems Conduct problems 4.59* (2.94–7.19) 1.91* (1.00–3.64)

Being bullied 1.21 (0.61–2.38) 1.05 (0.46–2.37)

Bullying 1.59 (0.86–2.92) 1.58 (0.62–4.02)

Peer problems Peer problems 6.34* (4.07–9.88) 3.79* (2.38–6.04)

Being bullied 1.10 (0.60–1.99) 0.98 (0.49–1.97)

Bullying 1.10 (0.57–2.11) 0.84 (0.29–1.03)

Inattention-Hyperactivity Inattention-hyperactivity 5.82* (3.91–8.66) 9.91* (6.99–14.05)

Being bullied 2.98* (1.78–4.97) 1.28 (0.72–2.28)

Bullying 1.27 (0.77–2.08) 2.15* (1.24–3.72)

Pro-social problems Pro-social behavior 3.67* (2.47–5.50) 6.67* (3.85–11.57)

Being bullied 1.21 (0.70–2.09) 0.34 (0.10–1.21)

Bullying 1.98* (1.20–3.26) 1.97* (1.00–3.99)

a
Alpha of 0.05 is used as significance level, CI = confidence interval, * = significant.

Furthermore, adolescents who reported psychosocial problems at baseline were more likely to report them at follow-up. However, reporting psychosocial problems at baseline did not make either boys or girls more likely to report being bullies or being bullied than those without reports of such problems at baseline. Despite a lacking relationship between a problematic general SDQ score at T0 and being bullied or being a bully at T1, Table 2 does illustrate a longitudinal relationship between specific SDQ sub-scales and bullying and/or being bullied (Table 2). Further analyses exploring the bidirectional associations of bully behaviors with specific psychosocial problems are summarized in Table 2 and reported on in the following section.

Bullied adolescents and their psychosocial problems

Boys and girls who were bullied at baseline reported more conduct problems at T1 (OR 1.75, 95% C.I. 1.03–3.00 for boys and OR 2.31, 95% C.I. 1.41–3.78 for girls). Victimization at baseline had no significant effects on psychosocial problems other than on conduct problems (Table 2).

Adolescents who reported psychosocial problems at baseline were more likely to report being victimized at T1 than those who indicated no such problems at baseline. However, these effects differed by gender. Boys reporting attention-hyperactivity problems at T0 were significantly more likely to be victimized at T1 (OR 2.98, 95% C.I. 1.78–4.97), while for girls this was true for those with emotional problems (OR 2.18, 95% C.I. 1.17–4.06).

Bullies and their psychosocial problems

Boys who were bullies at baseline reported more conduct problems (OR 2.00, 95% C.I. 1.23–3.27), peer problems (OR 1.84, 95% C.I. 1.12–3.02), and pro-social problems (OR 1.98, 95% C.I. 1.24–3.15) at T1 than those who were not (Table 2). Girls who were bullies at baseline also reported more peer problems (OR 2.26, 95% C.I. 1.27–3.99) and lacking pro-social behaviors (OR 3.84, 95% C.I. 1.95–6.21) at T1 than those who were not bullies at baseline.

The only psychosocial problems at T0 that were associated with being a bully at T1 in boys were lacking pro-social behavior. This association was significant for both boys (OR 1.32, 95% C.I. 1.12–1.56) and for girls (OR 1.35, 95% C.I. 1.11–1.63). In girls, additionally, the relationship between inattention-hyperactivity problems at T0 and being a bully at T1 was also significant (OR 2.15, 95% C.I. 1.24–3.72).

Discussion

This study explored the bidirectional longitudinal associations of bullying and bully victimization with psychosocial problems in a sample of Dutch high school adolescents via a cross-lagged panel design. Bullying and victimization at baseline were both associated with higher odds of reporting specific psychosocial problems a year later. Conversely, reporting certain specific psychosocial problems at baseline was also associated with victimization of bullying and/or with being a bully at T1. These specific effects differed for boys and girls.

When looking at the subscales of the SDQ, boys who reported bullying their peers at baseline reported more conduct problems, peer problems and less pro-social behavior at follow-up than those who did not. Among girls, those who bullied their peers at baseline also reported more peer problems and less pro-social behavior at follow-up than girls who did not. These findings were consistent with previous research (Arsenault et al., 2010; Copeland et al., 2013; Craig, 1998; Fekkes et al., 2006; Nansel et al., 2001; Stavrinides et al., 2011).

Furthermore, only boys who reported attention-hyperactivity problems at baseline were more likely to be victimized at follow-up. For girls emotional problems at baseline were significantly associated with being bullied at follow-up. For both boys and girls, less pro-social behavior at baseline was associated with being a bully at follow-up.

The design of the current study allowed for distinguishing between the directions of the different studied associations. Generally, in most previous studies, bullying/being bullied and psychosocial problems were seen only as one being the consequence of the other (Stavrinides et al., 2011). However, by such an approach one is more prone to make too simplistic inferences about true effects and mechanisms. The results of the current study indicated that certain specific psychosocial problems were associated with adolescents being more likely to become bullying victims and/or perpetrators of bullying themselves, while vice versa being a perpetrator or victim were also associated with specific psychosocial problems. These findings indicated that more research on this subject is needed with a focus on these specific dynamics. Also, differences of these effects between boys and girls were also apparent, which may indicate, e.g. different coping mechanisms and/or responsiveness with regard to bullying behavior.

Are adolescents with psychosocial problems more often bullying perpetrators and/or victims?

In line with previous studies, the current study showed that emotional problems and inattention-hyperactivity problems were significantly associated over time with being bullied (Hodges & Perry, 1999; Schwarz, 2000; Veentra et al., 2005). In addition, Veenstra et al. (2005) demonstrated significant effects of conduct problems on being a bully later on, while this association was not significant in the current study. A reason for this could be that Veenstra et al. (2005) explored these relationships in younger children instead of adolescents. However, these inconsistencies are an interesting research topic for future studies.

Overall, based on our results, certain psychosocial problems were related to future bullying and bullying victimization. These associations differed for boys and girls. Among girls, especially emotional problems were associated with being a victim of peer bullying, while inattention-hyperactivity problems seemed to be the types of problems to make boys prone to being victimized. These findings indeed seem to point to different mechanisms between bully behaviors and psychosocial problem development in boys and girls. In line with previous research, in a general sense, more externalizing problem behaviors were associated with bullying in boys, while in girls bully behaviors were more related to internalizing problems (Baldry, 2004; Pepler & Craig, 2008; Young & Sweeting, 2004).

Are bullies more likely to develop psychosocial problems?

Psychosocial problems at T1 showed the strongest association with reporting such problems at baseline, which is in line with several previous studies (Bond, Carlin, Thomas, Kerryn, & Patton, 2001; Craig & Pepler, 2003; Kaltiala-Heino et al., 2000). Also consistent with other recent studies, we found being a perpetrator of bullying at T0 to be associated with developing psychosocial problems at T1, in particular to more conduct problems and poor peer relations (Craig, 1998; Fekkes et al., 2006; Nansel et al., 2001; Reijntjes et al., 2011; Wienke Totura et al., 2009; Wolke et al., 2000). In contrast with other studies, the associations of bullying with emotional and inattention-hyperactivity problems were not significant (Wienke Totura et al., 2009; Wolke et al., 2000). However, the studies that did report significant relationships of all distinct problems that are measured by the SDQ were studies that only took into account the effects of physical aggression as a measure for bullying. In contrast, the current study used a broader measure for bullying that also includes other kinds of damaging behavior such as gossiping or verbal abuse as types of bullying, based on Solbeg & Olweus (2003). Since physical aggression can be seen as the most extreme form of bullying, it is more likely to also lead to stronger effects on psychosocial problems. Further reasons to explain the possible differences with previous research are highlighted in the next section.

The associations of bullying with developing peer problems and lacking pro-social behavior as reported in the current study were in line with those in comparable studies (Pepler, Jiang, Craig, & Connolly, 2008; Reijntjes et al., 2011; Stavrinides et al., 2010). Bullying was associated strongest to peer problems, and was also significantly associated with less pro-social behavior.

Are victimized adolescents more likely than others to develop psychosocial problems?

Victims of bullying were overall more likely than others develop psychosocial problems, consistent with results from other research (Dao et al., 2006; Reijntjes et al., 2010; Reijntjes et al., 2011; Wolke et al., 2000). The somewhat differing definitions of bullying used in some earlier studies, as explained in the previous section, might explain the somewhat smaller and less often significant effects of such associations in the current study.

Also consistent with previous research, the psychosocial problems that were related strongest with being bullied were conduct problems (Juvonenet, Graham, & Schuster, 2003; Nansel et al., 2001; Reijntjes et al., 2011; Siegel et al., 2009; Veenstra et al., 2010; Wienke Totura et al., 2009; Woods, Done, & Kalsi, 2009). However, in the current study it was not found that being bullied resulted in more emotional problems, peer problems and lacking pro-social behavior, whereas some other studies these associations were found to be significant (Schwartz, 2000; Wienke Totura et al., 2009). These differences with other literature should be studied further in future research.

An explanation for why other studies reported different results than the current study might be that some used a particular problem (such as emotional problems or peer problems) as a single main outcome. This might lead to an overestimation of the associations when certain psychosocial problems are significantly inter-correlated, such as those of the SDQ. As a consequence, when they are not being analysed together in comprehensive analyses, this inter-relatedness might lead to confounding. Therefore, since all analyses in the current study were performed within the same statistical model, this problem was averted, as was the potential problem of multiple testing and an inflated p-value. Thus, the current study presents relatively conservative findings. Furthermore, as stated, some studies used other measures to assess bullying victimization and/or perpetration, which could have led to different results. For example, Wienke Totura et al. (2009) used a multiple informant strategy, meaning that they included peer- and teacher indicators in addition to using self-reported surveys to identify bullying victims and perpetrators. Lastly, some past studies were performed in younger children than adolescents, which may have led to different effects, although it is hard to speculate about in what specific way this might have impacted findings.

Although gender differences were observed and effects were found more often to be significant for girls than boys, these differences were fairly limited, which is in line with two recent meta-analyses (Reijntjes et al., 2010; Reijntjes et al., 2011) and other research (Siegel, Greca, & Harrison, 2009), but less prominent than expected from earlier literature (Baldry, 2004; Camodeca et al., 2002; Pepler & Craig, 2008; Veenstra et al., 2010; Wang et al., 2009; Young & Sweeting, 2004). These contradictions make a case for more gender-specific studies in this field (Baldry, 2004; Camodeca et al., 2002; Pepler & Craig, 2008; Veenstra et al., 2010; Wang et al., 2009; Young & Sweeting, 2004). In future studies more distinction should be made between the types of bullying perpetration and victimization, which would make it more possible to identify differentiating patterns of effects between boys and girls, as illustrated by Carbone-Lopez, Esbensen, and Brick (2010). Additionally, the found differences might arise from the fact that boys and girls perhaps cope differently with events such as being bullied, resulting in different effects on psychosocial problems.

Strengths and limitations

This study is one of the first to have quantified the associations of different psychosocial problems with bullying and victimization in adolescents via a cross-lagged panel design, while also exploring gender differences. Most previous studies that entailed a longitudinal design only took into account a one-way effect from bullying to psychosocial problems or vice versa. In addition, many such studies did not specifically explore gender differences, rarely distinguished between different types of psychosocial problems and/or often explored such associations in primary school children and not adolescents. In addition the utilized cross-lagged panel model allowed for the analyses to be performed in a single model, which prevented confounding by correlated psychosocial problems and it averted the problem of multiple testing and an inflated p-value. This adds significantly to the reliability of our findings.

A limitation of the current study is the use of self-report instruments, although it is the predominant methodology in current research on the topic. If this could be combined with parallel data collection from parents, teachers and/or peers in a multiple informant strategy it could lead to more reliable and valid estimates that are less prone to inflated correlations as well as to recall and self-presentation biases (Goldweber, Waasdorp, & Bradshaw, 2013; Stavrinides et al., 2011; Wienke Totura et al., 2009). However, online, confidential self-report measures in this context have been shown to be sufficiently reliable and valid in previous research (Wang et al., 2005). Also, the study duration was one year, while a longer follow-up time would provide the findings with more strength. Thirdly, even though longitudinal relationships were presented in this study, it does not mean that such relationships are causal, since third factors, such as adverse parenting practices or a bad school environment, could possibly lead to both. Fourthly, recent research increasingly shows that online bullying (or cyber bullying) might also significantly affect children’s and adolescents’ psychosocial well-being. The current study did not specifically look at this in depth, which should be considered a limitation. Lastly, the results from this relatively small study should be replicated in larger studies to increase the reliability of the findings. However, based on previous research (Wissmann, Toutenburg, & Shalabh, 2008) issues such as multicollinearity due to small sample size were unlikely to have affected the results of this study, since the category with the largest number of cases was chosen as the reference category in all models.

Conclusions

The current study shows significant associations between bullying perpetration and victimization with psychosocial problems in adolescents. To the authors’ knowledge no previous studies reported on the bidirectional paths of these particular psychosocial problems and peer bullying, while exploring specific gender differences, via a longitudinal cross-lagged panel study design in adolescents. Bullying perpetration and victimization were associated with a greater likelihood of particular psychosocial problems at follow-up as well as the vice versa effects; this means that having psychosocial problems at baseline lead to being more likely to be bullied and victimized at later times. Results illustrated significant gender differences in these relationships, in line with recent meta-analyses, but less pronounced than those found in earlier studies. These findings may have important implications for future studies as well as for parents, educators and professionals, since interventions can now be targeted and fine-tuned more specifically.

Overall, bullying and being bullied both led to more psychosocial problems and vice versa, their associations to psychosocial problem development seem to form vicious circles once started. Studies showed that these consequences can be serious and long-lasting (Bond et al., 2001; Craig & Pepler, 2003; Kaltiala-Heino et al., 2000; Olweus et al., 1999). As Kaltiala-Heino et al. (2000) illustrated: ‘being a victim is known to be a long-lasting situation that is even repeated in new surroundings’. This suggests that interventions to prevent bullying are vital to break this dark cycle (Hodges & Perry, 1999). Since adolescents spend most of their time in school and also experience an important part of their psychosocial development and socialization there, schools seem to be the ideal place for such intervention. As Richard, Schneider, and Mallet (2012) stated, a whole-school approach (WSA) setting is probably the most effective approach to tackling school bullying. Via HLM techniques they determined that between-school effects, with the climate variables of school security and the quality of student-teacher relationships were the strongest predictors of school bullying. Woods et al. (2009) added to this by concluding that the fact that ‘bullying is very much a group process that bolsters the argument for a systemic approach to its correction’. This is in line with research by Huitsing and Veenstra (2012) who also show and plead for more in-depth research into classroom relations, social structures, and social network perspectives to better understand the dynamics behind peer bullying in order to design better interventions to address it. The Health Promoting School’s whole-school approach presents such an approach and more studies should therefore aim to research its effectiveness in reducing bullying victimization and perpetration.

Longitudinal associations from baseline (T0) to time 1 (T1)	Odds Ratio^a (95% CI)
Being bullied at T0 > Being bullied at T1	5.03* (3.00–8.44)	5.26* (3.42–8.08)
Being bullied at T0 > Psychosocial Problems at T1	1.49 (0.84–2.66)	2.00* (1.29–3.12)
Bullying at T0 > Psychosocial Problems at T1	1.72* (1.01–2.91)	1.12 (0.62–2.04)
Bullying at T0 > Bullying at T1	2.86* (1.66–4.93)	3.90* (2.13–8.04)
Psychosocial Problems at T0 > Psychosocial Problems at T1	10.61* (6.40–17.60)	5.26* (3.42–8.08)
Psychosocial Problems at T0> Being bullied at T1	1.23 (0.62–2.46)	1.64 (0.88–3.08)
Psychosocial Problems at T0 > Bullying at T1	1.31 (0.59–2.18)	0.79 (0.37–1.68)

Longitudinal associations	Odds Ratio^a (95% CI)
Being bullied	Emotional problems	0.87 (0.36–2.07)	1.57 (0.88–2.79)
Bullying	Emotional problems	0.83 (0.36–1.87)	1.02 (0.49–2.14)
Being bullied	Conduct problems	1.75* (1.03–3.00)	2.31* (1.41–3.78)
Bullying	Conduct problems	2.00* (1.23–3.27)	1.54 (0.83–2.88)
Being bullied	Peer problems	1.09 (0.64–1.86)	1.46 (0.88–2.42)
Bullying	Peer problems	1.84* (1.12–3.02)	2.26* (1.27–3.99)
Being bullied	Inattention-Hyperactivity	1.20 (0.70–2.06)	0.70 (0.43–1.12)
Bullying	Inattention-Hyperactivity	0.88 (0.51–1.51)	1.29 (0.74–2.24)
Being bullied	Pro-social problems	1.02 (0.62–1.70)	1.12 (0.64–1.97)
Bullying	Pro-social problems	1.98* (1.24–3.15)	3.84* (1.95–6.21)
Emotional problems	Emotional problems	5.67* (2.22–14.44)	16.60* (10.45–23.40)
	Being bullied	0.36 (0.12–1.10)	2.18* (1.17–4.06)
	Bullying	0.37 (0.11–1.33)	0.78 (0.33–1.88)
Conduct problems	Conduct problems	4.59* (2.94–7.19)	1.91* (1.00–3.64)
	Being bullied	1.21 (0.61–2.38)	1.05 (0.46–2.37)
	Bullying	1.59 (0.86–2.92)	1.58 (0.62–4.02)
Peer problems	Peer problems	6.34* (4.07–9.88)	3.79* (2.38–6.04)
	Being bullied	1.10 (0.60–1.99)	0.98 (0.49–1.97)
	Bullying	1.10 (0.57–2.11)	0.84 (0.29–1.03)
Inattention-Hyperactivity	Inattention-hyperactivity	5.82* (3.91–8.66)	9.91* (6.99–14.05)
	Being bullied	2.98* (1.78–4.97)	1.28 (0.72–2.28)
	Bullying	1.27 (0.77–2.08)	2.15* (1.24–3.72)
Pro-social problems	Pro-social behavior	3.67* (2.47–5.50)	6.67* (3.85–11.57)
	Being bullied	1.21 (0.70–2.09)	0.34 (0.10–1.21)
	Bullying	1.98* (1.20–3.26)	1.97* (1.00–3.99)

Footnotes

Author biographies

Vincent Busch, PhD, has a background in biomedical sciences and health sciences. He obtained his PhD in public health epidemiology at the University Medical Center Utrecht in the Netherlands. His main research interests are in the field of health behavior of children and adolescents and public health intervention research. He is currently involved with implementing and performing research on the innovative Jump-in project of the municipal health service of Amsterdam.

Lydia Laninga-Wijnen is a PhD student at the Department of Interdisciplinary Social Sciences at Utrecht University. There, she performs research on social relationships of children and adolescents and the associations with psychosocial health.

Tom Albert van Yperen, PhD, is a Social Scientist and a former Special Professor at Utrecht University for research and education on the topic of improving the effectiveness of child and youth care. Since September 2011 he has been special Professor at the State University of Groningen. His research and teaching activities there are focused on the issue of how the effectiveness of child and youth care can be monitored in a way that professionals consider the monitor as non-bureaucratic, stimulating and relevant for innovating their practice. Tom also works at the Netherlands Youth institute, where he aims to improve the quality of child and family services.

Augustinus Jacobus Petrus Schrijvers, PhD, is an Emeritus Professor of Public Health. He has performed and guided research on health care innovations for over 25 years at the UMC Utrecht. He recently published the book Health Care Innovation with the Cappuccinomodel. He is Chairman and Founder of the International Foundation of integrated Care.

Johannes Rob Josephus De Leeuw, PhD, is a Senior Researcher at the Julius Center, Department of Public Health at the University Medical Center Utrecht. His main professional experience consists of the acquisition of public health research grants, actively performing performing research and guiding (PhD) students in their work in scientific research. Recently this entailed studies on disease management of mental and physical disease, on the evaluation of a Health Promoting School intervention, and on the effects of mental and physical chronic disease on quality of life and mood.

References

Arseneault

Bowes

Shakoor

(2010) Bullying victimization in youths and mental health problems: ‘Much ado about nothing’? Psychological Medicine 40(5): 717–729. doi: 10.1017/S0033291709991383.

Baldry

A. C.

(2004) The impact of direct and indirect bullying on the mental and physical health of Italian youngsters. Aggressive Behavior 30(5): 343–355. doi: 10.1002/ab.20043.

Bollen

K. A.

(1989) Structral equations with latent variables, Hoboken, NJ: John Wiley & Sons, Inc.

Bond

Carlin

J. B.

Thomas

Kerryn

Patton

(2001) Does bullying cause emotional problems? A prospective study of young teenagers. British Medical Journal 323(7311): 480–484. doi: 10.1136/bmj.323.7311.480.

Byrne

M. D.

(1998) Structural equation modeling with lisrel, prelis, and simplis: Basic concepts, applications, and programming, Mahway, NJ: Erlbaum Associates.

Camodeca

Goossens

F. A.

Terwogt

M. M.

Schuengel

(2002) Bullying and victimization among school-age children: Stability and links to proactive and reactive aggression. Social Development 11(3): 332–345. doi: 10.1111/1467-9507.00203.

Carbone-Lopez

Esbensen

F.-A.

Brick

B. T.

(2010) Correlates and consequences of peer victimization: Gender differences in direct and indirect forms of bullying. Youth Violence and Juvenile Justice 8(4): 332–350. doi: 10.1177/1541204010362954.

Copeland

W. E.

Wolke

Angold

Costello

(2013) Adult psychiatric outcomes of bullying and being bullied by peers in childhood and adolescence. The Journal of the American Medical Association: Psychiatry 70(4): 419–426. doi: 10.1001/jamapsychiatry.2013.504.

Craig

(1998) The relationship among bullying, victimization, depression, anxiety and aggression in elementary school children. Personality and Individual Differences 24(1): 123–130. doi: 10.1016/S0191-8869(97)00145-1.

10.

Craig

W. M.

Pepler

D. J.

(2003) Identifying and targeting risk for involvement in bullying and victimization. Canadian Journal of Psychiatry 48(9): 577–582.

11.

Currie

Molcho

Boyce

Holstein

Torsheim

Richter

(2008) Researching health inequalities in adolescents: The development of the Health Behaviour in School-Aged Children (HBSC) family affluence scale. Social Science and Medicine 66(6): 1429–1436. doi: 10.1016/j.socscimed.2007.

12.

Dao

T. K.

Kerbs

J. J.

Rollin

S. A.

Potts

Gutierrez

Choi

(2006) The association between bullying dynamics and psychological distress. Journal of Adolescent Health 39(2): 277–282. doi:10.1016/j.jadohealth.2005.11.001.

13.

Esbensen

F. A.

Carson

D. C.

(2009) Consequences of being bullied: Results from a longitudinal assessment of bullying victimization in a multisite sample of American students. Youth and Society 41(2): 209–233. doi: 10.1177/0044118X09351067.

14.

Fekkes

Pijpers

F. I. M.

Fredriks

A. M.

Vogels

Verloove-Vanhorick

S. P.

(2006) Do bullied children get ill, or do ill children get bullied? A prospective cohort study on the relationship between bullying and health-related symptoms. Pediatrics 117(5): 1568–1574. doi: 10.1542/peds.2005-0187.

15.

Goldweber

Waasdorp

T. E.

Bradshaw

C. P.

(2013) Examining the link between forms of bullying behaviors and perceptions of safety and belonging among secondary school students. Journal of School Psychology 51(4): 469–485. doi: 10.1016/j.jsp.2013.04.004.

16.

Goodman

(1997) The Strengths and Difficulties Questionnaire: A research note. Journal of Child Psychology and Psychiatry 38(5): 581–586. doi: 10.1111/j.1469-7610.1997.tb01545.x.

17.

Goodman

Meltzer

Bailey

(1998) The strengths and difficulties questionnaire: A pilot study on the validity of the self-report version. European Child and Adolescent Psychiatry 7(3): 125–130. http://link.springer.com.proxy.library.uu.nl/journal/787/7/3/page/1.

18.

Hawker

D. S.

Boulton

M. J.

(2000) Twenty years’ research on peer victimization and psychosocial maladjustment: A meta-analytic review of cross-sectional studies. Journal of Child Psychology and Psychiatry 41(4): 441–455. doi: 10.1111/1469-7610.00629.

19.

Hodges

E. V.

Perry

D. G.

(1999) Personal and interpersonal antecedents and consequences of victimization by peers. Journal of Personality and Social Psychology 76(4): 677–685. doi: 10.1037/0022-3514.76.4.677.

20.

Huitsing

Veenstra

(2012) Bullying in classrooms: Participants roles from a social network perspective. Aggressive Behaviors 38(6): 494–509. doi: 10.1002/ab.21438.

21.

Juvonen

Graham

Schuster

M. A.

(2003) Bullying among young adolescents: The strong, the weak, and the troubled. Pediatrics 112(6): 1231–1237. http://pediatrics.aappublications.org/content/112/6/1231.full.html.

22.

Kaltiala-Heino

Rimpela

Rantanen

Rimpela

(2000) Bullying at school: An indicator of adolescents at risk for mental disorders. Journal of Adolescence 23(6): 661–674. doi: 10.1006/jado.2000.0351.

23.

Kyriakides

Kaloyirou

Lindsay

(2006) An analysis of the Revised Olweus Bully/Victim Questionnaire using the Rasch measurement model. British Journal of Educational Psychology 76(4): 781–801. doi: 10.1348/000709905X53499.

24.

Muris

Meesters

van den Berg

(2003) The Strengths and Difficulties Questionnaire (SDQ) Further evidence for its reliability and validity in a community sample of Dutch children and adolescents. European Child and Adolescent Psychiatry 12(1): 1–8. doi: 10.1007/s00787-003-0298-2.

25.

Muthén

L. K.

Muthén

B. O.

(2007) Mplus User’s Guide, fifth edition, Los Angeles, CA: Muthén & Muthén.

26.

Nansel

T. R.

Overpeck

Pilla

R. S.

Ruan

Simons-Morton

Scheidt

(2001) Bullying behaviors among US youth: Prevalence and association with psychosocial adjustment. The Journal of the American Medical Association 285(16): 2094–2100. doi: 10.1001/jama.285.16.2094.

27.

Olweus

(1993) Bullying at school: What we know and what we can do, Hoboken, NJ: John Wiley & Sons Ltd.

28.

Olweus, D., Limber, S., & Mihalic, S.(1999). Bullying prevention program. In D. S. Elliott (Series Ed.), Blueprints for violence prevention: Book nine. Boulder, CO: Center for the Study and Prevention of Violence, University of Colorado.

29.

Pepler

Craig

(2008) Understanding and adressing bullying: An international perspective, Bloomington, IN: AuthorHouse.

30.

Pepler

Jiang

Craig

Connolly

(2008) Developmental trajectories of bullying and associated factors. Child Development 79(2): 325–338. http://ww1.cpa-apc.org:8080/Publications/Archives/CJP/2003/october/craig.asp.

31.

Reijntjes

Kamphuis

J. H.

Prinzie

Boelen

P. A.

van der Schoot

Telch

M. J.

(2011) Prospective linkages between peer victimization and externalizing problems in children: A meta-analysis. Aggressive Behavior 37(3): 215–222. doi: 10.1002/ab.20374.

32.

Reijntjes

Kamphuis

J. H.

Prinzie

Telch

M. J.

(2010) Peer victimization and internalizing problems in children: A meta-analysis of longitudinal studies. Child Abuse and Neglect 34(4): 244–252. doi: 10.1016/j.chiabu.2009.07.009.

33.

Richard

J. F.

Schneider

B. H.

Mallet

(2012) Revisiting the whole-school approach to bullying: Really looking at the whole school. School Psychology International 33(3): 263–284. doi: 10.1177/0143034311415906.

34.

Schwartz

(2000) Subtypes of victims and aggressors in children’s peer groups. Journal of Abnormal Child Psychology 28(2): 181–192. doi: 10.1023/A:1005174831561.

35.

Siegel

R. S.

Greca

A. M.

Harrison

H. M.

(2009) Peer victimization and social anxiety in adolescents: Prospective and reciprocal relationships. Journal of Youth and Adolescence 38(8): 1096–1109. doi: 10.1007/s10964-009-9392-1.

36.

Solberg

M. E.

Olweus

(2003) Prevalence estimation of school bullying with the Olweus Bully/Victim Questionnaire. Aggressive Behavior 29(3): 239–268. doi: 10.1002/ab.10047.

37.

Stavrinides

Georgiou

Nikiforou

Kiteri

(2011) Longitudinal investigation of the relationship between bullying and psychosocial adjustment. European Journal of Developmental Psychology 8(6): 730–743. doi: 10.1080/17405629.2011.628545.

38.

Stavrinides

Georgiou

Theofanous

(2010) Bullying and empathy: A short-term longitudinal investigation. Educational Psychology 30(7): 793–802. doi: 10.1080/01443410.2010.506004.

39.

van Dorsselaer, S., de Looze, M., Vermeulen-Smit, E., de Roos, S., Verdurmen, J., ter Bogt, T., & Vollebergh, W. (2009). Health, well-being and education of Dutch adolescents, HBSC 2009. Trimbos Instituut; Universiteit Utrecht; Sociaal en Cultureel Planbureau, Utrecht, The Netherlands.

40.

Veenstra

Lindenberg

Munniksma

Dijkstra

J. K.

(2010) The complex relation between bullying, victimization, acceptance, and rejection: Giving special attention to status, affection, and sex differences. Child Development 81(2): 480–486. doi: 10.1111/j.1467-8624.2009.01411.x.

41.

Veenstra

Lindenberg

Oldehinkel

A. J.

de Winter

A. F.

Verhulst

F. C.

Ormel

(2005) Bullying and victimization in elementary schools: A comparison of bullies, victims, bully/victims, and uninvolved preadolescents. Developmental Psychology 41(4): 672–682. doi: 10.1037/0012-1649.41.4.672.

42.

Wang

Ianotti

R. J.

Nansel

T. R.

(2009) School bullying among adolescents in the United States: Physical, verbal, relational, and cyber. Journal of Adolescent Health 45(4): 368–375. doi: 10.1016/j.jadohealth.2009.03.021.

43.

Wang

Y.-C.

Lee

C.-M.

Lew-Ting

C.-Y.

Hsiao

C. K.

Chen

D.-R.

Chen

W. J.

(2005) Survey of substance use among high school students in Taipei: Web-based questionnaire versus paper-and-pencil questionnaire. Journal of Adolescent Health 37(4): 289–295. doi:10.1016/j.jadohealth.2005.03.017.

44.

Wienke Totura

C. M.

Green

A. E.

Karver

M. S.

Gesten

E. L.

(2009) Multiple informants in the assessment of psychological, behavioral, and academic correlates of bullying and victimization in middle school. Journal of Adolescence 32(2): 193–211. doi: 10.1016/j.adolescence.2008.04.005.

45.

Wissmann, M., Toutenburg, H., & Shalabh, H. (2008). Role of categorical variables in multicollinearity in the linear regression model. Technical Report 008, Department of Statistics, University of Munich.

46.

Wolke

Woods

Bloomfield

Karstadt

(2000) The association between direct and relational bullying and behaviour problems among primary school children. Journal of Child Psychology and Psychiatry 41(8): 989–1002. doi: 10.1111/1469-7610.00687.

47.

Wolke

Woods

Bloomfield

Karstadt

(2001) Bullying involvement in primary school and common health problems. Archives of Diseases in Childhood 85(3): 197–201. doi: 10.1136/adc.85.3.197.

48.

Woods

Done

Kalsi

(2009) Peer victimisation and internalising difficulties: The moderating role of friendship quality. Journal of Adolescence 32(2): 293–308. doi: 10.1016/j.adolescence.2008.03.005.

49.

Young

Sweeting

(2004) Adolescent bullying, relationships, psychological well-being, and gender-atypical behavior: A gender diagnosticity approach. Sex Roles 50(7–8): 525–537. doi: 10.1023/B:SERS.0000023072.53886.86.

Longitudinal associations from baseline (T0) to time 1 (T1)	Odds Ratio^a (95% CI)
Longitudinal associations from baseline (T0) to time 1 (T1)	Boys	Girls
Being bullied at T0 > Being bullied at T1	5.03* (3.00–8.44)	5.26* (3.42–8.08)
Being bullied at T0 > Psychosocial Problems at T1	1.49 (0.84–2.66)	2.00* (1.29–3.12)
Bullying at T0 > Psychosocial Problems at T1	1.72* (1.01–2.91)	1.12 (0.62–2.04)
Bullying at T0 > Bullying at T1	2.86* (1.66–4.93)	3.90* (2.13–8.04)
Psychosocial Problems at T0 > Psychosocial Problems at T1	10.61* (6.40–17.60)	5.26* (3.42–8.08)
Psychosocial Problems at T0> Being bullied at T1	1.23 (0.62–2.46)	1.64 (0.88–3.08)
Psychosocial Problems at T0 > Bullying at T1	1.31 (0.59–2.18)	0.79 (0.37–1.68)

Longitudinal associations		Odds Ratio^a (95% CI)
Baseline (T0)	Follow-up (T1)	Boys	Girls
Being bullied	Emotional problems	0.87 (0.36–2.07)	1.57 (0.88–2.79)
Bullying	Emotional problems	0.83 (0.36–1.87)	1.02 (0.49–2.14)
Being bullied	Conduct problems	1.75* (1.03–3.00)	2.31* (1.41–3.78)
Bullying	Conduct problems	2.00* (1.23–3.27)	1.54 (0.83–2.88)
Being bullied	Peer problems	1.09 (0.64–1.86)	1.46 (0.88–2.42)
Bullying	Peer problems	1.84* (1.12–3.02)	2.26* (1.27–3.99)
Being bullied	Inattention-Hyperactivity	1.20 (0.70–2.06)	0.70 (0.43–1.12)
Bullying	Inattention-Hyperactivity	0.88 (0.51–1.51)	1.29 (0.74–2.24)
Being bullied	Pro-social problems	1.02 (0.62–1.70)	1.12 (0.64–1.97)
Bullying	Pro-social problems	1.98* (1.24–3.15)	3.84* (1.95–6.21)
Emotional problems	Emotional problems	5.67* (2.22–14.44)	16.60* (10.45–23.40)
	Being bullied	0.36 (0.12–1.10)	2.18* (1.17–4.06)
	Bullying	0.37 (0.11–1.33)	0.78 (0.33–1.88)
Conduct problems	Conduct problems	4.59* (2.94–7.19)	1.91* (1.00–3.64)
	Being bullied	1.21 (0.61–2.38)	1.05 (0.46–2.37)
	Bullying	1.59 (0.86–2.92)	1.58 (0.62–4.02)
Peer problems	Peer problems	6.34* (4.07–9.88)	3.79* (2.38–6.04)
	Being bullied	1.10 (0.60–1.99)	0.98 (0.49–1.97)
	Bullying	1.10 (0.57–2.11)	0.84 (0.29–1.03)
Inattention-Hyperactivity	Inattention-hyperactivity	5.82* (3.91–8.66)	9.91* (6.99–14.05)
	Being bullied	2.98* (1.78–4.97)	1.28 (0.72–2.28)
	Bullying	1.27 (0.77–2.08)	2.15* (1.24–3.72)
Pro-social problems	Pro-social behavior	3.67* (2.47–5.50)	6.67* (3.85–11.57)
	Being bullied	1.21 (0.70–2.09)	0.34 (0.10–1.21)
	Bullying	1.98* (1.20–3.26)	1.97* (1.00–3.99)