Prevalence and correlates of bullying perpetration and victimization among school-going adolescents in Chandigarh, North India.

OBJECTIVES: Bullying among adolescents is one of the important but neglected health concerns, especially in low- and middle-income countries. The objective of this study was to estimate the prevalence and correlates of bullying among Indian adolescents. STUDY
DESIGN: This was a cross-sectional study.
MATERIALS AND METHODS: The prevalence of self-reported involvement in any kind of bullying was assessed among sixth to tenth class students (n = 667, mean age 13 years), across government (n = 359) and private (n = 308) schools using Olweus Bully-Victim Questionnaire in Chandigarh, a North Indian union territory. Self-esteem and emotional and behavioral difficulties of the participants were measured by using standard Rosenberg Self Esteem Scale and Strengths and Difficulties Questionnaire, respectively. Multinomial logistic regression was done to determine the predictors of bullying.
RESULTS: Prevalence of any kind of bullying was 25.6% (16% victimization, 5.2% perpetration, and 4.3% being bully-victim). Verbal bullying was the most common (55.1%), followed by physical (32.7%) and relational (25.2%) bullying. The prevalence of cyberbullying was 2.7%. Around 44% of students reported that adults in school never did anything to stop bullying. Bully-victims had the highest mean difficulty score (16.07). Significant predictors of bullying were being male (odds ratio [OR] = 2.5 [1.5-4.2], P < 0.001); studying in government school (OR = 0.63 [0.41-0.99], P = 0.048); having abnormal emotions (OR = 2.24 [1.1-4.7], P = 0.035); and poor peer relations (OR = 2.77 [1.44-5.35], P = 0.002).
CONCLUSIONS: One in four adolescents experience some form of bullying in schools in a North Indian city. Bullying perpetration and victimization is associated with gender, type of school, and abnormal difficulties (emotional and behavioral problems). Copyright:

INTRODUCTION

Bullying is a significant public health problem among adolescents, and it may have harmful consequences such as depression, anxiety, delinquency, and suicides. To achieve the United Nation's sustainable development goal three and four of ensuring good health and well-being at all ages including adolescents, and providing quality education, respectively, it is important to measure the burden and correlates of bullying, which is a form of school violence and has detrimental effects on the health of adolescents and their academic performance.[12] Olweus described bullying as a behavior in which the intention is to inflict injury or discomfort upon another individual who has difficulty in defending himself or herself. Bullying has three main criteria: bully's intention is to harm victim; bullying behavior is repetitive; and there is an imbalance of power between bully and victim.[3]

The prevalence of bullying victimization was assessed in 66 countries across North America, Europe, and Israel in a multicountry study. The prevalence of bullying in that study was reported as low as 7% to as high as 70%, which indicates large variations in its prevalence among students.[4] Modecki et al. conducted a meta-analysis of eighty international studies and reported that the prevalence of bullying perpetration worldwide is 34.5% and of victimization is 36%.[5] A review of bullying studies from South East Asian countries reported that the prevalence of bullying was either comparable or higher than that of Western countries, which can be attributed to factors such as educational or cultural, but methodological issues cannot be ignored such as criteria or definition used to describe the term bullying.[6]

The existing literature on the correlates of bullying have shown that the students involved in bullying have higher risk of suicidal ideation and attempts,[7] carry weapon in school, and indulge in physical fights and getting hurt.[8] Students who are victimized are also more likely to be depressed and suffer from anxiety, low self-esteem, and dropout from school.[910] In spite of the fact that bullying has serious mental health effects that can last up to adulthood,[11] this problem has not been given the required attention, especially, in low- and middle-income countries (LMICs) where focus from several decades is on reproductive and sexual health of adolescents.[12]

There is limited research on bullying in South East Asian region and in particular from India. Studies evaluating bullying in India are more important because the education system comprises of memorizing long lessons and stressful examinations. This puts students with additional mental stress, anxiety, and frustration, which is a precipitator of bullying.[13] In the 21st century, India has witnessed increase in globalization and increase of media in our daily lives.[14] Compared to any age group, adolescents are more exposed to globalization effects, i.e., they are influenced by different types of media such as movies, music videos, and Internet.[15] Many such media promote violence, and it increases the likelihood of adolescents to get involved in bullying as victim, perpetrator, or bystander.[16] Social inequality is another factor whose linkage to bullying is a proven fact. Students from countries with large socioeconomic disparities are more at risk of bullying.[17] Keeping in mind that India has large socioeconomic inequalities, it is important to understand the status of bullying in Indian schools.

With this background, the aim of the present study is to estimate the prevalence and to determine the correlates of bullying behaviors (victimization, perpetration, and bully-victims) among school children studying in grades sixth to tenth.

MATERIALS AND METHODS

Study design and setting

A cross-sectional study was conducted to obtain self-reported data on bullying. The study was conducted in private and government schools across Chandigarh, a North Indian union territory.

Study population

Boys and girls studying in classes from sixth to tenth grade were included in the cross-sectional study. Adolescents' indulgence in problem behaviors usually starts at around 12–15 years of age, hence this is the target group of this study.[18]

Sample size

A sample size of 654 was calculated based on expected proportion of children who were involved in bullying others to be 13%, significance level of 5%, power of 80%, nonresponse rate of 10%, and clustering effect of 1.78. The clustering effect is statistically defined as 1 + (m − 1) p, where m is the number of units in a cluster (students within a section of a class) and p is intraclass correlation coefficient (ICC) and is multiplied to the final sample size.[19] On the basis of prior research, the expected ICC is 0.02 for both victimization and perpetration of bullying.[20] On an average, the number of students in a class was assumed to be approximately 40, thus leading to clustering effect of 1.78. In a school-based study, for sample size calculation, clustering effect has to be taken into account, as the assumption in such studies is that the data within cluster, i.e., classes or schools, are more similar than that between clusters.

Sampling technique

A two-stage cluster sampling technique was used to first select the study schools in government (n = 5) and private school (n = 5) categories by simple random sampling technique using computer-generated random numbers to have equal representation from both the types of the schools. According to the data from the Chandigarh Education Department, one section of a given class comprised of approximately forty students. Hence, to achieve the sample size, two sections per class per school were needed to be selected. As there were five classes from sixth to tenth, hence one class (having at least two sections) per each government and private school was selected. In the second stage, one class and two sections of this class were randomly selected in each selected government and private school, in such a way that all the classes are included in the sample. All the students were included in the study from the selected sections.

Data collection methods and questionnaire

Data were collected anonymously through self- administration of the following study instruments in the class room.

Interview schedule

The information on age, gender (boy or girl), grade (sixth to tenth), type of school (government or private), type of locality (urban, rural, or slum), religion (Hinduism, Islam, Sikhism, and Christianity), category (general or schedule caste/tribe or other backward classes), and living arrangement (nuclear family or joint family) was collected by using a pretested background interview schedule.

Olweus Bully/Victim Questionnaire

Information on bullying was obtained by using the revised version of Olweus Bully/Victim Questionnaire.[21] Previous studies have reported good reliability and validity of this questionnaire.[2223] The bullying questionnaires provided the respondents with a definition of bullying and investigated how often they were involved in bullying in the past 6 months. The participants were later categorized into one of the four bullying subgroups according to their frequency of responses with bullying:

Not involved participants: Defined as who had not been bullied/bullied others or involved “only once or twice”

Victim: Categorized as who had been bullied “2 or 3 times a month” or more

Bully: Categorized who bullied other “2 or 3 times a month” or more

Bully-victim: Involved in bullying others and had been bullied “2 or 3 times a month” or more.

School culture was assessed using three questions from the Olweus Bully-Victim Questionnaire. The first question was “How do you usually react if you see or understand that a student your age is being bullied by other students?”. The second question was “How often do the teachers or other adults at school try to put a stop to it when a student is being bullied at school?”. Another question included to measure the culture of bullying in school was “Overall how much do you think your class teachers have done to counteract bullying since school started?”.

Rosenberg Self-Esteem Scale

This scale was used to assess self-esteem.[24] This scale generally has a high reliability, and has been validated in India.[25] This is a 10-item scale, and each of these items is scored on a 4-point scale. The resultant score can range from 0 to 30. Students having scores >15 are categorized into high (score >15) and those with scores ≤15 into having low self-esteem.

Strengths and Difficulties Questionnaire

The youth self-reported measure (11–17 years) of the Strengths and Difficulties Questionnaire (SDQ),[26] was used to assess emotional and behavioral difficulties. It is a validated tool with good psychometric properties and is available in English as well as Hindi version.[27] There are 25 items in SDQ, which comprises of five subscales including, emotional problems scale, conduct problems scale, hyperactivity scale, peer problems scale, and pro-social scale.

Both Olweus Bully-Victim Questionnaire and Rosenberg Self-Esteem Questionnaire were translated and re-translated in Hindi language, according to WHO methodology,[28] and finalized after discussion with the experts from the field of psychiatry, pediatrics, and community medicine.

Data analysis

Data were entered into Statistical Package for Social Sciences (SPSS) (Version 16.0. Chicago, SPSS Inc), which was used for all statistical purposes. Proportion of adolescents involved in bullying (victims, bullies, and bully-victims) was calculated. Chi-square test was used to compare the differences in the proportions between different groups.

Based on the findings of bivariate analysis, factors found to be significantly associated with the bullying status were analyzed further in multinomial logistic regression. Multinomial logistic regression was carried out to determine the predictors associated with involvement in any form of bullying (victims, bullies, and bully-victims). Adjusted odds ratio was estimated and considered significant at 95% levels. The dependent variable was any bullying including being a victim, bully, or a bully-victim, and not involved students were the reference category. The independent variables or predictors were type of school, gender, liking for school, number of friends, emotional problems, conduct problems, and peer relation problems.

Institute's ethics committee had approved the study. Permission from Director Higher Education and principals of respective schools was obtained. Student's assent and parental consent was obtained prior to participation in the study.

RESULTS

A total of 689 students studying in classes sixth to tenth had filled the Olweus Bully-Victim, Strengths and Difficulties, and Rosenberg Self-Esteem Questionnaire. However, 22 students (3.2%) were excluded from the study because of nonresponse to bullying victimization and perpetration-related questions. Hence, data were analyzed for 667 students (53.8% from government and 46.2% from private schools) students.

The background characteristics of the students in government and private schools are given in Table 1. The mean age of the students was 13.02 years (standard deviation = 1.6). The mean age was significantly higher in government (13.2) as compared to private schools (12.8) (P < 0.001). Significantly higher proportion of government school students belonged to nuclear family (P = 0.016), had four or more friends (P = 0.029), and had liked the school (P = 0.012) as compared to the private school students.

Table 1

Background characteristics of the study participants from government and private schools, Chandigarh

Background variables	Category	Total (n=667), n (%)	Government (n=359), n (%)	Private (n=308), n (%)	P*
Mean age (n=667)	Mean age	13.02 (SD=1.6)	13.18 (SD=1.7)	12.84 (SD=1.3)	<0.001
Class	6	129 (19.3)	69 (19.2)	60 (19.5)	0.969
	7	137 (20.5)	74 (20.6)	63 (20.5)
	8	135 (20.2)	72 (20.1)	63 (20.5)
	9	131 (19.6)	68 (18.9)	63 (20.5)
	10	135 (20.2)	76 (21.2)	59 (19)
Gender	Boy	410 (61.5)	219 (61)	191 (62)	0.789
	Girl	257 (38.5)	140 (39)	117 (38)
Religion	Hinduism	541 (81.2)	302 (84.2)	239 (77.6)	<0.001
	Islam	68 (10.2)	17 (4.7)	51 (16.6)
	Sikhism	51 (7.6)	35 (9.7)	16 (5.2)
	Christianity	7 (1)	5 (1.4)	2 (0.6)
Area of residence	Urban	533 (79.9)	288 (80.2)	245 (79.5)	0.466
	Rural	108 (16.2)	60 (16.7)	48 (15.6)
	Slum	26 (3.9)	11 (3.1)	15 (4.9)
Category	General	550 (82.5)	300 (83.6)	250 (81.2)	0.475
	Schedule caste or tribe/other backward classes	117 (17.5)	59 (16.4)	58 (18.8)
Living arrangement	Nuclear family	456 (68.4)	260 (72.4)	196 (63.6)	0.016
	Joint family	211 (31.6)	99 (27.6)	112 (36.4)
Number of friends	3 or less friends	309 (46.3)	152 (42.3)	157 (51)	0.029
	4 or more friends	358 (53.7)	207 (57.7)	151 (49)
School liking	Like	557 (83.5)	312 (86.9)	245 (79.5)	0.012
	Do not like	120 (16.5)	47 (13.1)	63 (20.5)

*P value indicates level of significance for the difference in proportion between government and private schools. SD – Standard deviation

Prevalence of bullying

Overall, the prevalence of any bullying behavior was 25.6%, victimization was 16%, only bullying was 5.2%, and bullying and victimization was 4.3%. The prevalence of bullying behavior was significantly higher in private schools (any 33%, victims 18.5%, bullies 7.5%, and bully-victims 7.1%) as compared to the government schools (any 19.2%, victims 13.9%, bullies 3.3%, and bully victim 1.9%) (P < 0.001). Victimization was more prevalent in the seventh (19.7%) and ninth grades (19.8%). The prevalence of cyberbullying was 2.7%. Boys (31.5%) were significantly more involved in any form of bullying as compared with girls (16.3%) (P < 0.001) [Table 2].

Table 2

Prevalence of bullying as per sociodemographic variables

Characteristics	Victim, n (%)	Bully, n (%)	Bully-victim, n (%)	Any kind of bullying, n (%)	Not involved, n (%)	χ2	P#
Total (n=667)	107 (16)	35 (5.2)	29 (4.3)	171 (25.6)	496 (74.4)
Class
6th (n=129)	23 (17.8)	7 (5.4)	7 (5.4)	37 (28.7)	92 (71.3)	6.002	0.191
7th (n=137)	27 (19.7)	7 (5.1)	8 (5.8)	42 (30.7)	95 (69.3)
8th (n=135)	18 (13.3)	7 (5.2)	9 (6.7)	34 (25.2)	101 (74.8)
9th (n=131)	26 (19.8)	5 (3.8)	3 (2.3)	34 (26)	97 (74)
10th (n=135)	13 (9.6)	9 (6.7)	2 (1.5)	24 (17.8)	111 (82.2)
Type of school
Government (n=359)	50 (13.9)	12 (3.3)	7 (1.9)	69 (19.2)	290 (80.8)	22.129	<0.001
Private (n=308)	57 (18.5)	23 (7.5)	22 (7.1)	102 (33.1)	206 (66.9)
Gender
Boy (n=410)	80 (19.5)	29 (7.1)	20 (4.9)	129 (31.5)	281 (68.5)	20.293	<0.001
Girl (n=257)	27 (10.5)	6 (2.3)	9 (3.5)	42 (16.3)	215 (83.7)
Area of residence
Urban (n=533)	84 (15.8)	22 (4.1)	22 (4.1)	128 (24)	405 (76)	8.457	0.206
Rural (n=108)	19 (17.6)	10 (9.3)	5 (4.6)	34 (31.5)	74 (68.5)
Slum (n=26)	4 (15.4)	3 (11.5)	2 (7.7)	9 (34.6)	17 (65.4)
Religion
Hinduism (n=541)	82 (15.2)	25 (4.6)	25 (4.6)	132 (24.4)	409 (75.6)	4.621	0.202
Others* (n=126)	25 (19.8)	10 (7.9)	4 (3.2)	39 (31)	87 (69)
Category
General (n=550)	91 (16.5)	30 (5.5)	19 (3.5)	140 (25.5)	410 (74.5)	6.520	0.089
Schedule caste or tribe/other backward classes (n=117)	16 (13.7)	5 (4.3)	10 (8.5)	31 (26.5)	86 (73.5)
Living arrangement
Nuclear family (n=456)	71 (15.6)	24 (5.3)	18 (3.9)	113 (24.8)	343 (75.2)	0.875	0.832
Joint family (n=211)	36 (17.1)	11 (5.2)	11 (5.2)	58 (27.5)	153 (72.5)
School liking
Do not like (n=557)	23 (20.9)	7 (6.4)	9 (8.2)	39 (35.5)	71 (64.5)	8.419	0.038
Like (n=120)	84 (15.1)	28 (5)	20 (3.6)	132 (23.7)	425 (76.3)
Number of friends
3 or less friends (n=309)	64 (20.7)	16 (5.2)	8 (2.6)	88 (28.5)	221 (71.5)	12.553	0.006
4 or more friends (n=358)	43 (12)	19 (5.3)	21 (5.9)	83 (23.2)	275 (76.8)

*Sikhism, Islam, and Christianity categories were merged as the frequencies were <10%. #P value indicates the level of significance for the difference in proportion among bullies, victims, bully-victims, and uninvolved students across various sociodemographic variables

Ways of being bullied

Majority of the students who were victimized were bullied verbally (55.1%), followed by physical bullying (32.7%) and bullying by spreading rumors (25.2%). Bullying by giving comments on caste, bullying by sexual comments or gestures, damaging property or forcibly taking money, and exclusion from group were reported by 24.3%, 22.4%, 18.7%, and 17.8% students, respectively. When the bullying was stratified by gender, verbal bullying was statistically significantly more prevalent in girls (66.7%) than boys (51.3%), whereas physical bullying was more prevalent in boys (37.5%) than girls (18.5%) (P = 0.03) [Figure 1].

Figure 1

Ways of being bullied among boys and girls in school, as per victims (n = 107)

Culture of bullying in school

The culture of bullying in school was assessed by asking whether adults in schools did anything to stop bullying. Around 44% of the students reported that adults in school almost never did anything to stop bullying, followed by 30.6% of students who reported once in a while or sometimes and 25.2% of students reported “often” or “almost always.” Students were also asked whether their class teacher did anything to counteract bullying incident in class, for which around 49% of students reported “nothing or little,” followed by 42.4% of students who reported “good deal or much,” and around 9% of students reported “somewhat.” Another aspect of culture of bullying in school was the bystander behavior, i.e., students' own response to peer's bullying incidents. Majority (59.1%) of the students (n = 394) felt sorry for the student who was bullied and 24% of the students tried to help the victim.

Correlates of bullying

When the association of bullying and strength and difficulties was evaluated, the proportion of students having abnormal total difficulties was statistically significantly higher (P < 0.001) among victims (26.1%) followed by bully-victims (12.5%) and bullied (11.4%) [Table 3].

Table 3

Prevalence of bullying as per mental health variables

Characteristics	Victim (n=107), n (%)	Bully (n=35), n (%)	Bully-victim (n=29), n (%)	Any bullying (n=171), n (%)	Not involved (n=496), n (%)	χ2	P*
Self-esteem
Low (n=221)	34 (15.4)	14 (6.3)	12 (5.4)	60 (27.1)	161 (72.9)	1.826	0.609
High (n=446)	73 (16.4)	21 (4.7)	17 (3.8)	111 (24.9)	335 (75.1)
Total difficulties
Normal (n=469)	58 (12.4)	19 (4.1)	11 (2.3)	88 (18.8)	381 (81.2)	51.289	<0.001
Borderline (n=110)	26 (23.6)	6 (5.5)	7 (6.4)	39 (35.5)	71 (64.5)
Abnormal (n=88)	23 (26.1)	10 (11.4)	11 (12.5)	44 (50)	44 (50)
Emotional problems
Normal (n=559)	77 (13.8)	27 (4.8)	17 (3)	121 (21.6)	438 (78.4)	38.699	<0.001
Borderline (n=45)	14 (31.1)	1 (2.2)	4 (8.9)	19 (42.2)	26 (57.8)
Abnormal (n=63)	16 (25.4)	7 (11.1)	8 (12.7)	31 (49.2)	32 (50.8)
Conduct problems
Normal (n=454)	61 (13.4)	17 (3.7)	15 (3.3)	93 (20.5)	361 (79.5)	28.663	<0.001
Borderline (n=97)	17 (17.5)	6 (6.2)	4 (4.1)	27 (27.8)	70 (72.2)
Abnormal (n=116)	29 (25)	12 (10.3)	10 (8.6)	51 (44)	65 (56)
Hyperactivity
Normal (n=560)	87 (15.5)	29 (5.2)	23 (4.1)	139 (24.8)	421 (75.2)	10.533	0.104
Borderline (n=71)	13 (18.3)	4 (5.6)	1 (1.4)	18 (25.4)	53 (74.6)
Abnormal (n=36)	7 (19.4)	2 (5.6)	5 (13.9)	14 (38.9)	22 (61.1)
Peer problems
Normal (n=341)	33 (9.8)	20 (5.9)	13 (3.8)	66 (19.4)	275 (80.6)	32.92	<0.001
Borderline (n=228)	48 (21.1)	6 (2.6)	9 (3.9)	63 (27.6)	165 (72.4)
Abnormal (n=98)	26 (26.5)	9 (9.2)	7 (7.1)	42 (42.9)	56 (57.1)
Prosocial behavior
Normal (n=472)	75 (15.9)	24 (5.1)	19 (4)	118 (28.6)	354 (71.4)	8.266	0.219
Borderline (n=55)	9 (16.4)	6 (10.9)	5 (9.1)	20 (36.4)	35 (63.6)
Abnormal (n=140)	23 (16.4)	5 (3.6)	5 (3.6)	33 (23.6)	107 (76.4)

*P value indicates level of significance for the difference in proportion among bullies, victims, bully-victims, and uninvolved students across various mental health variables

Male gender (victims: odds ratio [OR] = 2.53, 95% confidence interval [CI]: 1.54–4.2; P < 0.001 and bullies: OR = 4.24; 95% CI: 1.67–10.75; P = 0.002), having emotional problems (victims: OR = 2.24; 95% CI: 1.1–4.7; P = 0.035 and bully victims: OR = 4.36; 95% CI = 1.48–12.86; P = 0.008), peer relation problems (victims: OR = 2.77; 95% CI: 1.44–5.35; P = 0.002), and abnormal conduct problem (bully: OR = 2.58; 95% CI: 1.04–6.42; P = 0.042) were significant predictors of prevalence of bullying (being a victim, bully, or a bully victim, respectively), as per multinomial logistic regression. Studying in government school was a protective factor for being a victim (OR = 0.63; 95% CI: 0.41–0.99; P = 0.048), bully (OR = 0.37; 95% CI = 0.17–0.79, P = 0.01), and bully victim (OR = 0.22, 95% CI: 0.09–0.54; P = 0.001). Having four or more friend was also a significant protective factor for being a bully victim (OR = 0.33; 95% CI: 0.13–0.83; P = 0.019) [Table 4].

Table 4

Multinomial logistic regression showing predictors of bullying (victim, bully, or bully-victim)

Characteristics	OR (95%CI)
	Victim (n=107)	Bully (n=35)	Bully-victim (n=29)
Type of school
Government	0.63* (0.41-0.99)	0.37* (0.17-0.79)	0.22* (0.09-0.54)
Private	1	1	1
Gender
Boy	2.53* (1.54-4.2)	4.24* (1.67-10.75)	1.81 (0.76-4.29)
Girl	1	1	1
Liking for school
Do not like	1.25 (0.71-2.2)	1.29 (0.51-3.22)	2.91* (1.17-7.19)
Like	1	1	1
Number of friends
3 or less friends	1.56 (0.98-2.648)	1.00 (0.47-2.15)	0.33* (0.13-0.83)
4 or more friends	1	1	1
Emotional problems
Abnormal	2.24* (1.1-4.73)	2.36 (0.81-6.91)	4.36* (1.48-12.86)
Borderline	2.44* (1.13-5.23)	0.40 (0.05-3.3)	3.14 (0.87-11.25)
Normal	1	1	1
Conduct problems
Abnormal	1.52 (0.83-2.77)	2.58* (1.04-6.42)	1.76 (0.65-4.76)
Borderline	1.21 (0.65-2.27)	1.73 (0.63-4.72)	0.99 (0.29-3.33)
Normal	1	1	1
Peer relation problems
Abnormal	2.77* (1.44-5.35)	1.91 (0.72-5.07)	2.26 (0.74-6.89)
Borderline	2.14 (1.29-3.55)	0.45 (0.17-1.19)	1.19 (0.48-2.97)
Normal	1	1	1

*P<0.05: Reference category is not involved students. OR – Odds ratio; CI – Confidence interval

The total difficulty score was statistically significantly associated with bullying status when adjusted for gender, number of good friends, liking for school, and type of school. Abnormal difficulty score was statistically significantly associated with being a victim (OR = 3.29, 95% CI: 1.82–5.96; P < 0.001), bully (OR = 4.67; 95% CI: 1.99–10.96; P < 0.001), or bully-victim (OR = 10.25; 95% CI: 3.99–26.36; P < 0.001).

DISCUSSION

A cross-sectional study was used to estimate the prevalence of self-reported involvement in any kind of bullying among sixth to tenth grade students. The prevalence of any kind of bullying was 25.6%, victimization was 16%, perpetration was 5.2%, and bully-victim was 4.3%. Half of the students who were victimized were bullied by mean names, teasing, or passing hurtful comments. Boys had significantly higher risk of being a victim or bully as compared with the girls.

The prevalence of any kind of bullying reported in the present study was 25.6%, which is within the range reported in the European countries (9% in Sweden to 54% in Lithuania).[29] A recent meta-analysis reported a mean prevalence rate of 35% for bullying involvement, across eighty worldwide studies.[5] Prevalence reported in other high-income countries (Australia: victims: 18%; bullies: 3%; and bully-victims: 5%; the USA: victims: 8.4%; bullies: 8.2%, and bully-victims: 3.7%; and North America: victims: 16.7%; bullies: 15.1%, and bully-victims: 6.8%)[28-30] is also comparable to the prevalence observed in the present study. There is a huge variation in the prevalence of bullying in LMICs (from as low as 8% in Tajikistan to as high as 70% in Zambia).[30] The prevalence rates obtained in the present study are much higher than that reported in a study from Guangdong province in China (victims: 4.8%; bullies: 7.1%, and bully-victims: 2.6%) in school-aged youth.[31] Nguyen et al. reported the prevalence and correlates of victimization among low-resource countries; the highest prevalence of victimization was reported from Peru (61.5%) followed by India (56.4%), Vietnam (38.6%), and Ethiopia (27.4%).[32] Skrzypiec et al. reported the following prevalence rates of victimization: the Philippines: 37.1%; Indonesia: 21.2%; and Taiwan: 5.4%.[33]

One of the most important factors for variability in bullying prevalence, highlighted by cross-country studies, is the difference in socioeconomic and cultural factors, across different countries.[323334] In an earlier study from Chandigarh, the prevalence of any type of bullying was reported to be higher (any bullying 53%, victimization 19.2%, perpetration 13%, and bully-victim: 20.7%) as compared to this study.[35] The reason could be attributed to the fact that in the present study, the definition of bullying according to the Olweus was provided to the students, which could also result in the difference of rates in both the studies. Valliancourt et al. (2008) reported that when participants were given the definition of bullying, the prevalence of victimization reported was less, hence accounting for the less prevalence in the present study in comparison to Malhi et al.[3536]

Verbal bullying was the most common form of bullying (55.1%), which is also in line with the existing literature. In a study done among US adolescent students, it was reported that verbal bullying was the most common way through which students were victimized (53.8%) followed by social bullying (51.4%) and physical bullying (20.8%).[37] A similar study done in the Greater Beirut area has also reported that verbal bullying (12.48%) was the most common form of victimization. Khamis reported that boys were bullied more physically (boys vs. girls: 3.3% vs. 0.8%), whereas girls used relational form of victimization (boys vs. girls: 5.9% vs. 6.3%), which are similar to the findings of the present study.[38] Wang et al. reported similar results; boys (9.9%) reported more physical victimization than girls (5.5%), whereas girls (27.6%) reported more of relational victimization than boys (20.7%).[37]

The results of this study show that there is a culture of bullying in Indian schools, as teachers or adults in school did nothing to counteract bullying, which is consistent with the findings of the previous studies.[3940] A positive culture of bullying in school encourages bullies to victimize more as they will think that no action will be taken against them and will also discourage victims to report to adults in the school.

In the present study, being in government (public) school was significantly protective for bullying, which is similar to the results reported from a recent study done in Brazil (2015), which reported that bullying victimization was more likely to be reported in the private schools than public schools (OR = 1.17; 95% CI: 1.04–1.31).[40] A possible explanation could be that in government schools, a homogenous group of students study, i.e., almost all come from socioeconomically weaker sections of the society, so there are fewer chances of bullying, whereas in private schools, there will be a mixed population from both lower and higher socioeconomical groups. This difference of status may contribute to economically weaker students to be a target of victimization. In a meta-analysis, it was also reported that victims (OR = 1.40; 95% CI: 1.36, 1.74) and bully-victims (OR = 1.54; 95% CI = 1.36, 1.74) were more likely to come from the low socioeconomic background.[41]

The results of the present study have shown that significantly more boys (19.5%) were engaged in bullying victimization and perpetration as compared with girls (10.5%). Similar results were obtained in a cross-country study across forty countries (bullying was higher in boys, which ranged from 8.6% to 45.2%, and the range for girls varied from 4.8% to 35.8%,[29] in Beijing, China (victimization was reported to be more in males [23%] than females [17%]);[42] and in India (boys were twice as likely as girls to bully and 2.25 timely to be classified as bully-victims).[35] Smith et al. reviewed gender differences in bullying and reported that due to more risk-taking behavior among boys and increased empathy in girls, boys are more involved in bullying than girls.[43]

The results of a multinomial regression analysis of the present study showed that victimization was significantly associated with abnormal emotional problems and abnormal peer relation problems. These findings are similar to the earlier studies that reported higher emotional problems, i.e., depression and loneliness, among victims.[444546] Previous research has consistently reported significant association of being a bully and externalizing problems such as aggressiveness and conduct problems,[4748] similar to the findings of the present study. In line with the existing literature, the present study also reported the highest mean score for bully-victims and significant association with emotional problems.[354648]

The limitation of the study is that it is a cross-sectional study, and it limits the ability to make causal associations. All the data collected for the study were through self-administered questionnaire. The disadvantage of using self-administered questionnaire is that students may find difficulty in understanding the meaning of some questions. To tackle this limitation, the authors had explained all the questions to the participants one by one before administering the questionnaire.

CONCLUSION

The findings of the present study clearly indicate that bullying perpetration and victimization is widespread among school-going adolescents. One in four adolescents were involved in some kind of bullying, leading to adverse psychosocial outcomes. The findings of this study might be useful for the school administrators, policymakers, public health experts, and mental health professionals to understand the problem of bullying and associated psychosocial problems among school-going adolescents. Bullying interventions which aim to improve self-esteem and provide youth with skills to respond effectively to bullying, will help in reducing psychosocial distress that bullied students suffer in school. Whole school-based intervention programs involving school administrators, teachers, parents, and students are recommended to deal with this issue on a sustainable basis. The present study reinforces the need for promotion of a zero tolerance policy toward bullying in Indian schools.

Financial support and sponsorship

This work was supported by the Indian Council of Medical Research: Junior Research fellowship grant to undergo PhD program of the first author.

Conflicts of interest

There are no conflicts of interest.