Bullying and victimization among Turkish children and adolescents: examining prevalence and associated health symptoms
Over the past decade, concerns about bullying and its effects on school health have grown. However, few studies in Turkey have examined the prevalence of bullying in childhood and adolescence and its association with health problems. The current study aimed to examine the prevalence and manifestation of bullying and victimization among male and female students aged 11–15 years. A second goal was to examine the physical and psychological symptoms associated with being a bully, victim and both a bully and a victim (‘bully–victim’). Participants were 1,315 students from grades 5, 7, and 9, selected from three schools in Western Turkey. Twenty percent of the students were found to be involved in the cycle of bullying (5 % as a bully, 8 % as a victim, and 7 % as bully–victims). Bullies (although not victims) were found to show decreased levels of school satisfaction and school attendance. Being a victim or a bully–victim was associated with a significantly increased risk of experiencing a wide range of physical and psychological health symptoms (victims OR, 1.67–3.38; p < 0.01; bully–victims OR, 2.13–3.15; p < 0.01). Being a bully, in contrast, was associated with high levels of irritability (OR, 2.82; p < 0.01), but no other health concerns. Children that were bullies and victims were almost as vulnerable to health symptoms as children that were purely victims. Conclusion: These findings contribute to a better understanding of bullying in Turkish schools, emphasizing the negative effects of bullying involvement on health and well-being.
Bullying can be defined as a repetitive physical or psychological pressure by a stronger person or group on a weaker person [31, 35, 38], creating a power imbalance between the bully and victim . Bullying is reported to be most common in early adolescence . Although bullying is a global phenomenon [10, 29, 32, 34], previous studies show great variation in prevalence rates across countries. For example, bullying prevalence has been shown to range from 6.3 % in girls from Sweden to 41.4 % in Lithuanian boys . Despite a wealth of research in more developed countries across Europe and the USA, there have been relatively few studies of bullying in less developed countries such as Turkey. The need for improved research in this area has been emphasized by recent media attention along with recommendations from the Turkish Ministry of Education, UNICEF and other related institutions. In particular, few studies have examined the physical and psychological consequences of bullying; an important theme within the current work [1, 2, 22].
Previous studies have highlighted the serious impact that bullying can have on the health of children and adolescents [11, 15, 16, 19, 25, 48]. Physical and psychological harassment can result in internal problems such as headaches or stomachaches coupled with external concerns such as decreased school attendance or satisfaction. Depression, low self-confidence, school anxieties, and somatic symptoms of worry, sadness, and loneliness are also very common among the victims of bullying [1, 28, 44, 46]. In a large prospective study of 9- to 11-year-olds (N = 1,118), Fekkes and colleagues  found that victims of bullying had significantly higher chances of developing new psychosomatic and psychosocial problems compared with children and adolescents who were not bullied. Their longitudinal design led them to conclude that bullying caused the health symptoms, rather than vice versa. Children and adolescents who are bullied may also avoid coming to class or, in severe circumstances, may protect themselves by bringing weapons to school [2, 23, 33]. At worst, victimization can also lead to suicidal ideation  and attempted suicide among children and adolescents . Rigby determined that these negative consequences of bullying can be placed within four overarching categories; (1) low mood; general unhappiness and frustration (2) peer relationship difficulties; frequent absence from school, low school satisfaction with school (3) psychological harassment; depression, suicidal thoughts (4) physical problems; diagnosed psychosomatic and/or medical illnesses .
Although most studies of bullying have focused on victims, it is also important to consider the psychological and physical symptoms associated with being a bully or a bully–victim (both a victim and a bully) [2, 10, 13, 15, 17, 22]. One previous study of 819 children and adolescents (aged 13–16 years) showed that boys who were bullies were significantly more likely to experience somatic symptoms, anxiety, depression and social dysfunction . Being a bully has also been associated with significantly increased rates of suicidal ideation in adolescents . Children and adolescents who are both bullies and victims, may also be at increased risk of somatic and emotional difficulties, although few studies have addressed this question. One exception was a study by Forero and colleagues , who found that bully–victims (constituting 22 % of their sample of 3,918 children and adolescents) were twice as likely to experience high levels of psychosomatic symptoms compared to participants who were not involved in bullying problems.
The effects of bullying victimization may not be limited to short term physical and psychological symptoms. Longitudinal investigations have shown that exposure to bullying during childhood can result in difficulties with low self-esteem, poor peer relationships and high levels of depression later in adult life [3, 41, 46]. Meltzer and colleagues  found that adults who had been bullied earlier in life were more than twice as likely to attempt suicide than controls. This long-term impact can be particularly pronounced in individuals with low levels of social support .
There have been few studies to date investigating the physiological and psychological impact of bullying and victimization in Turkish schools [1, 2, 22]. The most recent study by Karatas and colleagues , investigated the health effects of bullying exposure in 6th grade students (N = 92); finding increased rates of headaches, crying, nervousness, sleep problems and appetite changes in victims. Most previous Turkish studies of the have focused on the health-related symptoms attributed to the victims of bullying [2, 22] with limited investigation of bullies and bully–victims. One study by Alikasifoglu and colleagues  showed that bullies were more likely to be involved in physical fights; while bully–victims were more likely to partake in smoking, drinking and sexual activity. However, no previous studies have explored the physical or psychological symptoms associated with being a bully and a bully–victim.
The current study had three main goals. First, it aimed to assess the prevalence of traditional bullying in the largest sample of 11–15-year-olds in Turkey carried out to date. Second, it explored the factors associated with ‘the bullying cycle’ (including victims, bullies and bully–victims) in terms of home and school characteristics. Finally, it aimed to assess the physical and psychological consequences associated with being a victim, a bully and a bully–victim in both males and females.
Materials and method
Participants included 5th, 7th, and 9th grade children and adolescents (age 11–19) from primary, secondary and high schools in Duzce city, a rural area in the western region of Turkey. Nine schools were chosen (3 elementary (N = 430), 3 secondary (N = 450), 3 high schools (N = 435)) at random from all those affiliated with the National Education Ministry. Questionnaires were distributed to 1,350 children and adolescents to ensure sufficient power. Of these, complete data were available for 1,315 participants. Fifteen children and adolescents did not consent to participate and 20 questionnaires were only partially completed. All classes were co-educational, including both girls and boys.
Questionnaire design: structure and content
A 16-item self-report questionnaire was created to characterize the backgrounds of the participants in terms of home and school characteristics. This measure included self-reported items regarding demographic characteristics (e.g., social economic status (SES) and parental employment), household variables (e.g., number of siblings, one or two-parent family) and aspects of schooling (e.g., school achievement, satisfaction and attendance). SES was operationalized by asking the participants ‘How would you describe your family’s income?’ scored either low, average, or high. Participants reported their level of academic achievement according to the five levels within the Turkish system (an ‘appreciation certificate’, a ‘thanks certificate’, a direct pass, passing with responsibility or repeating the year). For the current analyses, these were divided into three achievement categories: low, average, and high. Participants also reported on whom they lived with (one parents, both parents, other relatives), school satisfaction (either likes or dislikes school) and their school attendance (frequently absent or rarely absent).
Determination of Peer Victims and Bullies Scale (DPVBS)
This 56-item self-report questionnaire assessed the bullying behaviors and victimization of children and adolescents over the past six months. It constitutes an adaptation of the ‘Multidimensional Peer Victimization Questionnaire’ developed by Maynard and Joseph , which included 27- items concerning victimization only. This scale was translated into Turkish by Gultekin and Sayil . Pekel-Uludagli and Ucanok  later added 27 additional items to measure bullying behaviors (e.g. rewording ‘The other children hit me’ to ‘I hit other children’) . Items are scored on a three-point scale: not at all (0) once (1) or more than once (2). To assess frequency of bullying involvement, participants were also asked ‘How often have you bullied other children?’ and ‘how often have you been bullied by other children?’ scored using the same three-point scale. Children and adolescents scoring over one standard deviation above the mean on the bullying items of the Determination of Peer Victims and Bullies Scale (DPVBS) were classified as bullies for these analyses. Participants scoring over one standard deviation above the mean on the victimization items were classed as victims. Finally, if participants scored over one standard deviation above the mean on both the bullying and victimization items, they were categorized as a bully–victim [6, 20, 27, 34]. The scale showed a high level of internal consistency in the current sample across the 27 bullying items (Cronbach alpha 0.90) and the 27 victimization items (Cronbach alpha 0.92). Both the bullying items and the victimization items fall into five categories: terrorizing behaviors (e.g., threats using a weapon or knife, five items), overt bullying (e.g., kicking or punching other children, eight items), teasing (e.g., name calling, six items), relational bullying (e.g., causing harm to others by manipulating their social relationships and reputation, four items), and attacks on property (four items; stealing or damaging property). The Cronbach alpha scores were acceptable to good for each of the bullying subscales (terrorizing, 0.78; overt, 0.85; teasing, 0.77; relational, 0.80; attacks on property, 0.75) and victimization subscales (terrorizing, 0.70; overt, 0.79; teasing, 0.80; relational, 0.86; attacks on property, 0.78)
Measure of health symptomatology
The current study used a self-reported measure of psychological and physical symptoms, taken from the Health Behavior in School-Aged Children study; a standardized, international collaborative investigation by the World Health Organization. This study incorporated repeated cross-sectional surveys among 11, 13, and 15 year-olds in representative samples of schools in participating countries . The measure used 12 items to assess health status. In the current study, each symptom was dichotomized prior to analysis, according to the method described by Due and colleagues . As the health symptoms were scored slightly differently, dichotomizing in this way helped to standardize scoring and ease interpretation of the regressions. For eight symptoms (headache, stomachache, backache, feeling low, bad temper, nervousness, difficulties in getting to sleep, and dizziness) the frequency was dichotomized into every day/more than once a week/about every week versus about every month/rarely/never. ‘Loneliness’ was dichotomized into very often/rather often versus sometimes/never. ‘Tired in the morning’ was dichotomized into once a week or more versus less. The two remaining symptoms, ‘feeling left out of things’ and ‘feeling helpless’ were divided into always/often versus sometimes/rarely/never.
The study was approved by the University of Duzce. Written permission to collect the data was obtained from the Duzce Provincial Education Ministry. The schools were chosen from a list provided by the Department of Education and informed about the study by letter. Potential participants were informed about the purpose of the study, that consent was strictly voluntary and that they could withdraw at any time. Data collection forms were distributed in a class setting and participants were instructed not to interact with each other when answering. It took participants an average of 40 min to answer the data collection forms. All responses were anonymous, with no identifying information collected.
Exploratory analyses of the demographic data were carried out using the statistical software program SPSS 19. Chi-squared analyses were used to determine whether the student’s status (as a victim, bully, bully–victim or neither) was associated with their grade (5th, 7th, 9th), gender, self-reported SES (low, average, high), school satisfaction, school achievement (below average, average, above average) and school attendance (frequently/rarely absent). Pairwise chi-squared tests were used to explore any significant associations (e.g. comparing bullies to non-bullies, victims to non-victims, bully–victims to non-bully–victims). Owing to positively skewed distributions, each of the DPVBS subscales (for both the bullying and victimization scales) was dichotomized into high (above the median for the whole sample) and low (below the median) scores. Chi-squared analyses were then used to determine whether there were differences in these scales by gender. Finally, logistic regression was used to assess the associations between the involvement in the bullying cycle (as bully, victim, bully–victim) and each dichotomized health symptom, covarying for class, self-reported SES and gender.
Self-reported demographic and household/school characteristics of the full sample (N = 1,315)
5 (age 11–14)
7 (age 14–16)
9 (age 16–19)
Social Economic Status
Whom does the student live with?
Lives with both parents
Lives with one parent
Live with other relatives
Self-reported prevalence of bullying
The majority of children and adolescents in the current sample 1,056 (80 %) had not been involved in bullying, as either an aggressor or a victim. Of the participants involved in the bullying cycle, 66 (5 %) were classified as bullies according to the DPVBS, 105 (8 %) were victims and 88 (7 %) were both bullies and victims (bully–victims). A significant sex difference was observed in terms of the proportion of children and adolescents classed as bullies, victims, bully–victims, or uninvolved (χ2(3) = 15.20; p < 0.01). Chi-squared analyses revealed this difference was driven by a greater proportion of boys (7 %) than girls (3 %) that were classified as bullies (χ2 (1) = 11.44, p < 0.01).
There was a significant association between a student’s grade (5, 7, 9) and their involvement in the bullying cycle (victim, bully, bully–victim, uninvolved; χ2(3) = 22.81, p < 0.01). Chi-squared analyses showed that this was driven by a higher number of bully–victims in grade 5 (47 %) than in grades 7 (22 %) and 9 (21 %) (χ2(2) = 11.16, p < 0.01). There was also a low proportion of bullies (15 %) in grade 7 compared to grades 5 (41 %) and 9 (44 %) (χ2 (2) = 9.01, p = 0.01). Involvement in the bullying cycle was significantly associated with school attendance (χ2(3) =13.48, p < 0.01), driven by a higher proportion of bullies (20 %) and bully–victims (21 %) being frequently absent compared to victims (10 %) or participants not involved in bullying (9 %). School satisfaction was also associated with involvement in the bullying cycle (χ2(3) = 20.53,p < 0.01). Chi-squared analyses showed that bullies had a lower level of school satisfaction (29 % were dissatisfied) compared to the rest of the sample (χ2 (1) =17.20, p < 0.01), including victims (17 %), bully–victims (11 %) and participants uninvolved in bullying (11 %). There was no association between involvement in the bullying cycle and a child’s level of school achievement (p > 0.05). According to student self-report, 699 (53.2 %) of the children and adolescents were from a family of average SES, 480 (36.5 %) were from a family of high SES and 36 (10.3 %) were from a family of average SES. There was no association between involvement in the bullying cycle and SES (p > 0.05).
Bullying and health symptoms
Associations between bullying, victimization and self-reported health symptoms
Victim (N = 105)
Bully (N = 66)
Bully–Victim (N = 88)
N with frequent symptoms (%)
Odds Ratio (95 % CI)
N with frequent symptoms (%)
Odds ratio (95 % CI)
N with frequent symptoms (%)
Odds ratio (95 % CI)
Tired in the morning
Feeling left out of things
Being a victim of bullying was associated with increased likelihood of experiencing all of the physiological and psychological symptoms that were assessed in the current study. In particular, victims were over three times as likely as other participants to often feel left out of things (OR, 3.12; p < 0.01) and to feel helpless (OR, 3.38; p < 0.01). They were also over twice as likely to experience frequent loneliness (OR, 2.51; p < 0.01), low mood (OR, 2.11; p < 0.01), tiredness in the morning (OR, 2.69; p < 0.01) and sleep difficulties (OR, 2.17; p < 0.01).
Children and adolescents that were classified as bullies in the current sample were significantly more likely to experience frequent symptoms of irritability compared to the rest of the sample (OR, 2.82; p < 0.01). Being a bully was not significantly associated with any of the other psychological or physiological symptoms assessed here.
Compared to the rest of the sample, participants who were both bullies and victims were more likely to experience 10 out of 12 of the health symptoms that were assessed. In particular, they were over three times as likely as other children and adolescents to feel left out of things (OR, 3.15; p < 0.01). They were also over twice as likely to experience frequent backache, sleeping difficulties, tiredness in the morning, dizziness, low mood, irritability, nervousness, loneliness, helplessness (for odds ratios see Table 2). The only symptoms that were not significantly associated with being a bully–victim were frequent headaches and stomachaches.
Two hundred fifty-nine (20 %) children and adolescents in the current sample were involved in the bullying cycle as a bully, 66 (5 %), a victim, 105 (8 %), or a bully–victim, 88 (7 %). These findings were in keeping with a prior Turkish study, which revealed similar rates of bullies (5 %), victims (6 %), and bully–victims (6 %) in a large sample (N = 1,670) of 9th and 10th grade students . A study by Alikasifoglu placed estimated slightly higher, finding that 22 % of Turkish high school students (N = 3,519) were the victims of bullying, 9 % were bullies and 9 % were bully–victims . In a cross-country bullying study, Nansel and colleagues  revealed a high degree of variability in the prevalence of bullying involvement (as victim, bully or both) across Europe and the USA. Estimates of bullying involvement ranged from 9 to 54 %, with an average of 11 % across 25 countries. This suggests that bullying behaviors in Turkey are of a similar frequency to some of the more developed European countries and the USA.
In the current study, a greater proportion of boys than girls reported that they carried out bullying, in keeping with previous findings [7, 26, 29, 41, 45, 47]. In particular, boys reported significantly higher levels of overt bullying behaviors than girls. This was in line with previous studies showing that boys prefer to bully others in more direct and aggressive ways than girls [5, 32, 35]. In their previous study of Turkish children and adolescents, Alikasifoglu and colleagues  also found that many more boys (n = 1,301, 61 %) than girls (n = 419, 22 %) participated in physical violence. This result may reflect, in part, the social and cultural norms of Turkish society, where males typically display more outwardly aggressive behavior than females. In contrast, girls in the current sample were more likely than boys to be the victims of teasing and relational bullying, in keeping with some previous findings [4, 53]. One exception was a study by Rigby, which found that boys and girls were similarly likely to be bullied through name calling, teasing and deliberate exclusion  .
A student’s involvement in the bullying cycle was also associated with a number of school-related characteristics. First, bullies in the current sample reported more frequent absences than other participants and also lower levels of school satisfaction. Bully–victims (n = 88,) reported more frequent absences but showed a similar level of school satisfaction as participants who were not involved in the bullying cycle. Victims did not report significantly reduced attendance or school satisfaction. These mixed findings are somewhat in keeping with previous studies showing negative impacts of bullying behaviors on attendance and school enjoyment . For example, the cross-national survey by Nansel and colleagues  showed that bullies reported poorer school adjustment than other students (in their sample of 113,200 children and adolescents aged 11 to 15). The current findings do, however, stand in contrast to previous work that revealed decreased school satisfaction in the victims of bullying [15, 18].
There were mixed findings in the current sample concerning school class (grades 5, 7, 9) and involvement in the bullying cycle. There was no association between a student’s grade and being a victim of bullying. However, compared to the other grades, a significantly higher proportion of participants in grade 5 were bully–victims and a significantly lower proportion of participants in grade 7 were bullies. These mixed findings require further investigation using age as a continuous variable. In previous work, bullying behavior has been found to decrease with age, potentially owing to improved coping mechanisms over time [8, 44, 52]. There were no associations between school achievement and bullying involvement in the current sample, in contrast to some previous work [42, 51].
Children and adolescents are vulnerable to a variety of physical, psychological, and social changes. Friendships and family relationships can become increasingly strained; with social acceptance and positive feedback becoming particularly important. Becoming involved in bullying at this age, as either a bully or victim, has been shown to have long-term impacts on the mental and physical well-being of individuals in later life [16, 23, 42]. The current study was the first to examine the physical and psychological health symptoms associated with both bullying and victimization in Turkish children and adolescents.
Being a victim of bullying in the current sample was associated with more frequent occurrences of all the health symptoms examined. These included physical symptoms (e.g., headaches, stomachaches, backaches and dizziness), and psychological symptoms (e.g., nervousness, loneliness and irritability). In particular, victims were over three times as likely as other participants to feel helpless and left out of things. A similar association between victimization and health problems has been described in previous studies across childhood and adolescence [13, 14, 17, 22, 30, 49, 50] and across different countries in Europe and the USA [8, 27]. One previous study also showed increased physical and psychological symptoms in Turkish 9th and 11th grade students (N = 4,153) who had been bullied . Of note, it is not possible within the current cross-sectional design to determine the direction of this association. For example, it may be that the experience of bullying leaves young people vulnerable to developing physical symptoms and psychological difficulties. However, conversely, it may be that children and adolescents with existing health difficulties may be targeted more frequently by bullies. One previous study addressed these alternatives using a longitudinal design, finding that health symptoms appeared to be a consequence of prior victimization at school . It is important that these consequences are monitored, particularly as victimization has also been linked with increased risks anxiety , depression [21, 43], and suicidal ideation .
Previous studies of the health effects of bullying have focused largely on the outcomes for victims [1, 10, 22, 30, 54]. Only a limited number of studies examined the health problems of the bullies themselves [13, 16, 25], reporting mixed findings. The current results showed that bullies were almost three times as likely to experience frequent symptoms of irritability compared to the other participants in the sample. However, being a bully was not associated significantly with other health symptoms. This was in keeping with a previous study showing that, of children and adolescents involved in the bullying cycle, pure bullies (who were never victimized) had the least physical and psychosomatic health problems . In contrast, other studies have shown that bullies experience higher levels of headaches , bedwetting , hyperactivity, sleep problems and feeling tense . Of note, the small number of bullies in the current sample (N = 66) may offer limited the power to detect significant associations using logistic regression. Again, the direction of the association remains unclear in the present study. While being a bully may heighten irritability levels, it is also likely that irritable children and adolescents are more prone to become bullies.
In the current sample, children and adolescents that were both bullies and victims were at increased risk for 10 out of the 12 health problems assessed, including physical symptoms (e.g., backaches, dizziness, tiredness) and psychological concerns (e.g., feeling nervous, lonely and helpless). This suggested that they were greater at risk of health symptoms than the pure bullies, and were almost as vulnerable as pure victims. These findings are in keeping with previous work in other countries, showing that bully–victims are particularly prone to concerns such as externalizing behaviors , poor emotional adjustment  and physical symptoms (e.g., repeated sore throats, colds, and coughs) . In the only previous Turkish study to address this issue, Alikasifoglu and colleagues  found that bully–victims were even more vulnerable to health problems than either bullies or victims alone. Further research is now required to determine what places young people at risk of becoming both a bully and a victim and how the profiles of these individuals differ from others in the bullying cycle.
The current study included a large sample, from a country that has received limited research attention in the field of bullying. However, a number of limitations must be addressed. The study used solely self-report data, including a student rating of family SES and parental education. Caution is needed when interpreting these data, as children and adolescents may be somewhat limited in their judgments of household characteristics. Similarly, individuals may differ significantly in their subjective perception of bullying and victimization. Replication of these findings using corroboration from multiple informants (parents, teachers and peers) would be beneficial. It should be noted that the categorization of bullies, victims and bully–victims was based on the mean and standard deviation of DPVBS scores in the current sample. Consequently, caution is needed when comparing these findings directly to other studies that have used this measure. As mentioned previously, the direction of causation between bullying exposure and health concerns cannot be examined using this cross-sectional sample. Further longitudinal investigation will be vital to disentangle any causal relationships between bullying and health in Turkey and to explore mediating influences such as family SES. Finally, the current sample was drawn from a rural region in Western Turkey. Consequently, caution is needed when generalizing these findings to children and adolescents from more urban areas across the country.
The current findings help to further elucidate the difficulties with bullying that are observed within Turkish primary, secondary and high school secondary schools. Here, the effects of bullying exposure were found to depend both on a student’s gender and their role in the cycle of bullying (as bully, victim or bully–victim). Both victims and bully–victims showed higher levels of health concerns, including physical and psychological difficulties. This information is vital for schools and health professionals, who must be aware of the risk factors and secondary consequences of bullying exposure for young people. Vigilance is vital at home and school, to prevent the negative impact of bullying on school satisfaction, achievement, and psychological and physical well-being.