Scandinavian http://sjp.sagepub.com/ Journal of Public Health
Bullying, cyberbullying, and mental health in young people Evelina Landstedt and Susanne Persson Scand J Public Health published online 7 March 2014 DOI: 10.1177/1403494814525004 The online version of this article can be found at: http://sjp.sagepub.com/content/early/2014/03/06/1403494814525004
Published by: http://www.sagepublications.com
Additional services and information for Scandinavian Journal of Public Health can be found at: Email Alerts: http://sjp.sagepub.com/cgi/alerts Subscriptions: http://sjp.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Mar 7, 2014 What is This?
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
525004 research-article2014
SJP0010.1177/1403494814525004Bullying, cyberbullying, and mental health in young peopleE. Landstedt and S. Persson
Scandinavian Journal of Public Health, 1–7
ORIGINAL ARTICLE
Bullying, cyberbullying, and mental health in young people
EVELINA LANDSTEDT1 & SUSANNE PERSSON2 1Department
of Health Sciences, Mid Sweden University, Sundsvall, Sweden, and 2Department of Public Health and Community Medicine, University of Gothenburg, Gothenburg, Sweden
Abstract Objective: To investigate the factors associated with exposure to in-real-life (IRL) bullying, cyberbullying, and both IRL and cyberbullying and to explore the relationship between these types of bullying and mental health among 13–16-year-old Swedish boys and girls. Methods: Data was derived from a cross-sectional web-based study of 13–16-year-old students in northern Sweden (n=1214, response rate 81.9%). Results: The combination of IRL- and cyberbullying was the most common type of bullying. A non-supportive school environment and poor body image were related to exposure to bullying for both genders but the relationship was more distinct in girls. All types of bullying were associated with depressive symptoms in both boys and girls and all forms of bullying increased the likelihood of psychosomatic problems in girls. Conclusions: Cyberbullying can be seen as an extension of IRL bullying. A combination of IRL- and cyberbullying seems to be particularly negative for mental health. Interventions should focus on improved school environment and body image as well as anti-violence programmes. Gender aspects of bullying need to be acknowledged. Key Words: Adolescents, body image, bullying, depressive symptoms, online harassment, psychosocial school environment, psychosomatic problems
Background Bullying is a major health problem in young people and individuals involved are at a risk of psychological problems such as psychosomatic symptoms, depression, and attempted or actual suicide [1–3]. Traditional (in-reallife, IRL) bullying can include physical acts, verbal abuse, and more subtle acts like social exclusion and spreading rumours [4]. According to Olweus’ (1994) [4] definition, bullying is when a person is exposed, repeatedly and over time, to negative actions by one or several people in their surroundings. For young people, school is the main arena for bullying [3, 4]. Increased risk of bully victimisation has been suggested to be associated with poor body image [5] and ethnicity [6]. As digital technology has created new ways to interact, new arenas for harassment have emerged. Young people use electronic-based communication
extensively and are therefore at risk of cyberbullying [7, 8]. Cyberbullying has been conceptualised as harassments where an electronic medium is used to threaten or harm and to inflict humiliation and fear [9, 10]. Many aspects of cyberbullying are similar to those of IRL bullying [9, 10] although there are some important differences [11]. Victims of cyberbullying are, for example, never left alone since harassments can reach them almost everywhere. Unlike IRL bullying, cyberbullying does not have to be repeated since a single episode can be reposted and forwarded [12]. Electronic communication also allows bullies to remain anonymous and, because of absence of direct reactions from the victim, there is risk of increased cruelty [8, 9]. Furthermore, unwanted contacts and harassments online are often of sexual nature [13].
Correspondence: Evelina Landstedt, Department of Health Sciences, Mid Sweden University, Holmgatan 10, 85170 Sundsvall, Sweden. E-mail: evelina. [email protected] (Accepted 25 January 2014) © 2014 the Nordic Societies of Public Health DOI: 10.1177/1403494814525004
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
2 E. Landstedt and S. Persson According to existing literature, 25–72% of adolescents report life-time exposure to cyberbullying [14–16]. Some studies suggest no or few gender differences [9, 12], while others found that girls are more exposed than boys [17]. Importantly, victims of cyberbullying are at risk for poor mental health [9, 11, 17]. There is also evidence of overlap between IRL bullying and cyberbullying [9, 13] and that exposure to several types of bullying increases the risk of mental health problems [11, 13, 18]. Most existing research on cyberbullying focuses on prevalence and not on potential differences in risk factors for IRL and cyberbullying victimisation, as well as relationships between cyberbullying and mental health. To our knowledge, there is no earlier study with a sample larger than 1000 adolescents that has examined mental health associations (especially several aspects of mental health) of cyberbullying and IRL victimisation. Given the existing evidence of links between poor school environment and rates of bullying as well as young people’s poor mental health [19], the next step is to evaluate whether the school environment also matters to cyberbullying. Also, it is unclear if the overlap of different types of harassment and its associations with mental health is the same in boys and girls. The objective of this study was to explore what factors are associated with exposure to IRL bullying, cyberbullying, and both IRL and cyberbullying and to explore the relationship between these types of bullying and mental health among 13–16-year-old Swedish boys and girls. Research questions were: What is the prevalence of exposure to IRL bullying, cyberbullying, and the combination of IRL and cyberbullying? How is poor body image, foreign background, age, and poor school climate associated with exposure to the types of bullying? How is exposure to IRL bullying, cyberbullying, and both IRL and cyberbullying associated with psychosomatic problems and depressive symptoms? Methods Context The current study is a part of a larger project undertaken in a municipality in northern Sweden (the main town in the region, population 59,000). In the region, 20% of the population has at least 3 years post-secondary education, which is the average for northern Sweden but slightly less than for the country in total (23%)[20]. The proportion of people born abroad is lower than the Swedish average (6 and 15% respectively), and approximately 20% of those are born in Norway (compared to 4% for whole
Sweden) [21]. The employment rate is about the same as the Swedish average of 80% [20]. Participants and procedure All high schools (grades 6–9, age 12–16 years) in the municipality were invited to participate in the study. All nine public schools and one of four independent schools agreed to participate. With the help of school staff, data was collected using a self-administered web-based questionnaire which the students accessed via their school e-mail. The questionnaires were completed in class with an adult present. The current study focuses on grades 7–9 (n=1214; 52.7% girls). The overall response rate was 81.9% (for grades 7, 8, and 9: 84.5, 81.8, and 79.9%, respectively). An information letter was sent to students and parents to gain informed consent. Seven students actively refrained participation (no background data available). Identities of respondents were confidential. The study was approved by Umeå Regional Ethical Review Board (Dnr 09-179M). Measures Psychosomatic problems were measured by the Psychosomatic Problem Scale (PSP) [23] which includes the following eight items: “How often during the past six months have you had difficulty in concentrating; had difficulty in sleeping; suffered from headache; suffered from stomach ache; felt tense; had little appetite; felt sad, and felt giddy” (5-point scale: never, seldom, sometimes, often, and always). A sum score was computed (range 0–32). Due to skewed distribution, the scale was dichotomised by upper quartile indicating poor psychosomatic health (cut off 14) [22]. Depressive symptoms were measured by the Centre for Epidemiologic Studies 20-item Depression Scale (CES-D), which has been found acceptable and reliable to assess depressive symptoms in high school students as well as in adults [23]. Respondents are asked how many times they experienced various symptoms during the past week and response categories for all items were “rarely or none of the time”, “some or little of the time”, “occasionally or a moderate amount of time”, and “most or all the time”. The range was 0–60 with a cut off at 16 for depressive symptoms, as suggested by Radloff Sawyer [23]. IRL bullying was indicated by the question “It happens that one or several students tease, pick fights or exclude somebody. Has any of that happened to you during the past six months?” In accordance with Olweus’ definition (1994) [4], responses were
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
Bullying, cyberbullying, and mental health in young people 3 dichotomised into: (0) no, never; Yes, once; and (1) Yes, almost all the time; Yes, several times; Yes, a few times. Exposure to cyberbullying was measured by three items: “Has something of the following happened to you when you have been by the computer during the past six months?” “Somebody called you bad words or was mean to you in some other way?”; “Somebody spread false rumours about you?”; “Somebody threatened or been aggressive towards you?” This variable was dichotomised into: (0) never; and (1) one time; a few times; several times. Given that a single act of cyberbullying can be repeated, exposure to cyberbullying was differently categorised to IRL bullying. A composite bullying measure, both IRL and cyberbullying, was constructed as follows: (0) no bullying; (1) IRL; (2) cyber; and (3) both. Foreign background was measured by two items regarding in which country their parents were born. Due to the low number of students having parents born outside Sweden, the variable was dichotomised into: (0) both parents born in Sweden; and (1) one/ both parents born in another country. Body image was measured by an index of four questions (Cronbach’s alpha=0.87, range 5–25). Students were asked: “Are you satisfied with your: looks, body, weight, height?” Response alternatives were: always, often, sometimes, seldom, never. In factor analysis all four items loaded on one component that explained 73% of the total variance (loadings 0.50–0.85). The variable was dichotomised by median split (9) into: (0) good/fairly good; and (1) poor. Experience of discrimination in school was indicated by the question “Has discrimination based on the following occurred at your school during the past six months: sex; ethnicity; disability; religion; sexual orientation; other?” Hence, the measure is based on the participants’ understandings of discrimination. Responses were coded into: (0) no discrimination; and (1) any type of discrimination. Students were asked if they had ever been bullied by a teacher (single-item question). Responses were coded into: (0) never; and (1) yes (seldom, sometimes, often, and always). Teacher support was measured by six items (Cronbach’s alpha=0.91): “Do your teachers support you when you need it?”; “Do you think your teachers would notice if you didn’t like it at school?”; “Do your teachers treat you fairly?”; “Do you receive praise and encouragement from your teachers?”; “Do you ask your teachers about things you don’t understand?”; and “Do your teachers acknowledge your thoughts and opinions?” A 5-point scale was used (never, seldom, sometimes, often and always).
In factor analysis all six items loaded on one component that explained 67.0% of the variance (loading 0.73–0.86). The index had a range from six to 30 and variables were dichotomised by median split (14) into: (0) good/fairly good support; and (1) low support. Statistical analyses Statistical analyses were performed in SPSS PASW Statistics 20 for Windows. Between-group differences were assessed by chi-squared statistics. Factor analysis was used to construct indices. Multinomial and binary logistic regressions were used to estimate odds ratios (OR) for associations. Regressions were run separately for boys and girls. Level of significance was set at 5%. Results According to Table I, 15.1% of the adolescents had been bullied IRL only, 17.2% reported cyberbullying only, and 19.6% were subjected to both IRL and cyberbullying. Approximately every second student had experienced some type bullying. No significant gender differences in exposure to bullying were found. Thirty per cent of the girls and 13.7% of the boys described psychosomatic problems, whereas 55.5% of the girls and 35.1% of the boys had depressive symptoms. Experience of being bullied by a teacher was the only variable significantly associated with all types of bullying in both boys and girls (Table II). Poor body image and discrimination in school were associated with experience of both IRL and cyberbullying in both boys and girls, whereas discrimination in school was significantly related to all forms of bullying in girls. In girls, poor body image and low teacher support were associated with exposure to cyberbullying and both IRL and cyberbullying. Higher age (school grade) was found to be a protective factor for IRL bullying and both IRL and cyberbullying among girls only. Foreign background was not related to any form of bullying in either sex. As further displayed in Table II, boys and girls reporting poor body image, discrimination in school, and teaching bullying were at increased risk of experiencing psychosomatic problems (PSP) and depressive symptoms. Low teacher support was associated with psychosomatic problems in girls only, and foreign background as well as age was significantly related with psychosomatic problems in boys. Based on the associations, these variables were included in the analyses exploring associations between bullying exposure and mental health.
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
4 E. Landstedt and S. Persson Table I. Descriptive statistics for dependent and independent variables. Variable
Total % (n)
Boys % (n)
Girls % (n)
p
Psychosomatic problems Depressive symptoms Bullying No IRL Online Both Foreign background Yes Grade 7th 8th 9th Body image Poor Discrimination in school Any Teacher bullying Yes Teacher support Low
22.4 (267) 46.0 (541)
13.7 (77) 35.1 (191)
30.2 (190) 55.5 (350)
48.1 (503) 15.1 (158) 17.2 (180) 19.6 (205)
51.4 (238) 12.3 (57) 17.3 (80) 19.0 (88)
45.4 (264) 17.3 (101) 17.2 (100) 20.1 (117)
14.4 (173)
12.8 (72)
15.9 (101)
31.8 (386) 32.6 (396) 35.6 (432)
30.8 (177) 30.5 (175) 38.7 (222)
32.7 (209) 34.5 (221) 32.8 (210)
46.6 (514)
26.2 (170)
59.4 (444)
26.8 (324)
22.6 (138)
25.7 (186)
22.9 (276)
24.9 (151)
16.9 (125)
50.7 (614)
44.6 (261)
48.8 (353)
Bullying, cyberbullying, and mental health in young people Evelina Landstedt and Susanne Persson Scand J Public Health published online 7 March 2014 DOI: 10.1177/1403494814525004 The online version of this article can be found at: http://sjp.sagepub.com/content/early/2014/03/06/1403494814525004
Published by: http://www.sagepublications.com
Additional services and information for Scandinavian Journal of Public Health can be found at: Email Alerts: http://sjp.sagepub.com/cgi/alerts Subscriptions: http://sjp.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Mar 7, 2014 What is This?
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
525004 research-article2014
SJP0010.1177/1403494814525004Bullying, cyberbullying, and mental health in young peopleE. Landstedt and S. Persson
Scandinavian Journal of Public Health, 1–7
ORIGINAL ARTICLE
Bullying, cyberbullying, and mental health in young people
EVELINA LANDSTEDT1 & SUSANNE PERSSON2 1Department
of Health Sciences, Mid Sweden University, Sundsvall, Sweden, and 2Department of Public Health and Community Medicine, University of Gothenburg, Gothenburg, Sweden
Abstract Objective: To investigate the factors associated with exposure to in-real-life (IRL) bullying, cyberbullying, and both IRL and cyberbullying and to explore the relationship between these types of bullying and mental health among 13–16-year-old Swedish boys and girls. Methods: Data was derived from a cross-sectional web-based study of 13–16-year-old students in northern Sweden (n=1214, response rate 81.9%). Results: The combination of IRL- and cyberbullying was the most common type of bullying. A non-supportive school environment and poor body image were related to exposure to bullying for both genders but the relationship was more distinct in girls. All types of bullying were associated with depressive symptoms in both boys and girls and all forms of bullying increased the likelihood of psychosomatic problems in girls. Conclusions: Cyberbullying can be seen as an extension of IRL bullying. A combination of IRL- and cyberbullying seems to be particularly negative for mental health. Interventions should focus on improved school environment and body image as well as anti-violence programmes. Gender aspects of bullying need to be acknowledged. Key Words: Adolescents, body image, bullying, depressive symptoms, online harassment, psychosocial school environment, psychosomatic problems
Background Bullying is a major health problem in young people and individuals involved are at a risk of psychological problems such as psychosomatic symptoms, depression, and attempted or actual suicide [1–3]. Traditional (in-reallife, IRL) bullying can include physical acts, verbal abuse, and more subtle acts like social exclusion and spreading rumours [4]. According to Olweus’ (1994) [4] definition, bullying is when a person is exposed, repeatedly and over time, to negative actions by one or several people in their surroundings. For young people, school is the main arena for bullying [3, 4]. Increased risk of bully victimisation has been suggested to be associated with poor body image [5] and ethnicity [6]. As digital technology has created new ways to interact, new arenas for harassment have emerged. Young people use electronic-based communication
extensively and are therefore at risk of cyberbullying [7, 8]. Cyberbullying has been conceptualised as harassments where an electronic medium is used to threaten or harm and to inflict humiliation and fear [9, 10]. Many aspects of cyberbullying are similar to those of IRL bullying [9, 10] although there are some important differences [11]. Victims of cyberbullying are, for example, never left alone since harassments can reach them almost everywhere. Unlike IRL bullying, cyberbullying does not have to be repeated since a single episode can be reposted and forwarded [12]. Electronic communication also allows bullies to remain anonymous and, because of absence of direct reactions from the victim, there is risk of increased cruelty [8, 9]. Furthermore, unwanted contacts and harassments online are often of sexual nature [13].
Correspondence: Evelina Landstedt, Department of Health Sciences, Mid Sweden University, Holmgatan 10, 85170 Sundsvall, Sweden. E-mail: evelina. [email protected] (Accepted 25 January 2014) © 2014 the Nordic Societies of Public Health DOI: 10.1177/1403494814525004
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
2 E. Landstedt and S. Persson According to existing literature, 25–72% of adolescents report life-time exposure to cyberbullying [14–16]. Some studies suggest no or few gender differences [9, 12], while others found that girls are more exposed than boys [17]. Importantly, victims of cyberbullying are at risk for poor mental health [9, 11, 17]. There is also evidence of overlap between IRL bullying and cyberbullying [9, 13] and that exposure to several types of bullying increases the risk of mental health problems [11, 13, 18]. Most existing research on cyberbullying focuses on prevalence and not on potential differences in risk factors for IRL and cyberbullying victimisation, as well as relationships between cyberbullying and mental health. To our knowledge, there is no earlier study with a sample larger than 1000 adolescents that has examined mental health associations (especially several aspects of mental health) of cyberbullying and IRL victimisation. Given the existing evidence of links between poor school environment and rates of bullying as well as young people’s poor mental health [19], the next step is to evaluate whether the school environment also matters to cyberbullying. Also, it is unclear if the overlap of different types of harassment and its associations with mental health is the same in boys and girls. The objective of this study was to explore what factors are associated with exposure to IRL bullying, cyberbullying, and both IRL and cyberbullying and to explore the relationship between these types of bullying and mental health among 13–16-year-old Swedish boys and girls. Research questions were: What is the prevalence of exposure to IRL bullying, cyberbullying, and the combination of IRL and cyberbullying? How is poor body image, foreign background, age, and poor school climate associated with exposure to the types of bullying? How is exposure to IRL bullying, cyberbullying, and both IRL and cyberbullying associated with psychosomatic problems and depressive symptoms? Methods Context The current study is a part of a larger project undertaken in a municipality in northern Sweden (the main town in the region, population 59,000). In the region, 20% of the population has at least 3 years post-secondary education, which is the average for northern Sweden but slightly less than for the country in total (23%)[20]. The proportion of people born abroad is lower than the Swedish average (6 and 15% respectively), and approximately 20% of those are born in Norway (compared to 4% for whole
Sweden) [21]. The employment rate is about the same as the Swedish average of 80% [20]. Participants and procedure All high schools (grades 6–9, age 12–16 years) in the municipality were invited to participate in the study. All nine public schools and one of four independent schools agreed to participate. With the help of school staff, data was collected using a self-administered web-based questionnaire which the students accessed via their school e-mail. The questionnaires were completed in class with an adult present. The current study focuses on grades 7–9 (n=1214; 52.7% girls). The overall response rate was 81.9% (for grades 7, 8, and 9: 84.5, 81.8, and 79.9%, respectively). An information letter was sent to students and parents to gain informed consent. Seven students actively refrained participation (no background data available). Identities of respondents were confidential. The study was approved by Umeå Regional Ethical Review Board (Dnr 09-179M). Measures Psychosomatic problems were measured by the Psychosomatic Problem Scale (PSP) [23] which includes the following eight items: “How often during the past six months have you had difficulty in concentrating; had difficulty in sleeping; suffered from headache; suffered from stomach ache; felt tense; had little appetite; felt sad, and felt giddy” (5-point scale: never, seldom, sometimes, often, and always). A sum score was computed (range 0–32). Due to skewed distribution, the scale was dichotomised by upper quartile indicating poor psychosomatic health (cut off 14) [22]. Depressive symptoms were measured by the Centre for Epidemiologic Studies 20-item Depression Scale (CES-D), which has been found acceptable and reliable to assess depressive symptoms in high school students as well as in adults [23]. Respondents are asked how many times they experienced various symptoms during the past week and response categories for all items were “rarely or none of the time”, “some or little of the time”, “occasionally or a moderate amount of time”, and “most or all the time”. The range was 0–60 with a cut off at 16 for depressive symptoms, as suggested by Radloff Sawyer [23]. IRL bullying was indicated by the question “It happens that one or several students tease, pick fights or exclude somebody. Has any of that happened to you during the past six months?” In accordance with Olweus’ definition (1994) [4], responses were
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
Bullying, cyberbullying, and mental health in young people 3 dichotomised into: (0) no, never; Yes, once; and (1) Yes, almost all the time; Yes, several times; Yes, a few times. Exposure to cyberbullying was measured by three items: “Has something of the following happened to you when you have been by the computer during the past six months?” “Somebody called you bad words or was mean to you in some other way?”; “Somebody spread false rumours about you?”; “Somebody threatened or been aggressive towards you?” This variable was dichotomised into: (0) never; and (1) one time; a few times; several times. Given that a single act of cyberbullying can be repeated, exposure to cyberbullying was differently categorised to IRL bullying. A composite bullying measure, both IRL and cyberbullying, was constructed as follows: (0) no bullying; (1) IRL; (2) cyber; and (3) both. Foreign background was measured by two items regarding in which country their parents were born. Due to the low number of students having parents born outside Sweden, the variable was dichotomised into: (0) both parents born in Sweden; and (1) one/ both parents born in another country. Body image was measured by an index of four questions (Cronbach’s alpha=0.87, range 5–25). Students were asked: “Are you satisfied with your: looks, body, weight, height?” Response alternatives were: always, often, sometimes, seldom, never. In factor analysis all four items loaded on one component that explained 73% of the total variance (loadings 0.50–0.85). The variable was dichotomised by median split (9) into: (0) good/fairly good; and (1) poor. Experience of discrimination in school was indicated by the question “Has discrimination based on the following occurred at your school during the past six months: sex; ethnicity; disability; religion; sexual orientation; other?” Hence, the measure is based on the participants’ understandings of discrimination. Responses were coded into: (0) no discrimination; and (1) any type of discrimination. Students were asked if they had ever been bullied by a teacher (single-item question). Responses were coded into: (0) never; and (1) yes (seldom, sometimes, often, and always). Teacher support was measured by six items (Cronbach’s alpha=0.91): “Do your teachers support you when you need it?”; “Do you think your teachers would notice if you didn’t like it at school?”; “Do your teachers treat you fairly?”; “Do you receive praise and encouragement from your teachers?”; “Do you ask your teachers about things you don’t understand?”; and “Do your teachers acknowledge your thoughts and opinions?” A 5-point scale was used (never, seldom, sometimes, often and always).
In factor analysis all six items loaded on one component that explained 67.0% of the variance (loading 0.73–0.86). The index had a range from six to 30 and variables were dichotomised by median split (14) into: (0) good/fairly good support; and (1) low support. Statistical analyses Statistical analyses were performed in SPSS PASW Statistics 20 for Windows. Between-group differences were assessed by chi-squared statistics. Factor analysis was used to construct indices. Multinomial and binary logistic regressions were used to estimate odds ratios (OR) for associations. Regressions were run separately for boys and girls. Level of significance was set at 5%. Results According to Table I, 15.1% of the adolescents had been bullied IRL only, 17.2% reported cyberbullying only, and 19.6% were subjected to both IRL and cyberbullying. Approximately every second student had experienced some type bullying. No significant gender differences in exposure to bullying were found. Thirty per cent of the girls and 13.7% of the boys described psychosomatic problems, whereas 55.5% of the girls and 35.1% of the boys had depressive symptoms. Experience of being bullied by a teacher was the only variable significantly associated with all types of bullying in both boys and girls (Table II). Poor body image and discrimination in school were associated with experience of both IRL and cyberbullying in both boys and girls, whereas discrimination in school was significantly related to all forms of bullying in girls. In girls, poor body image and low teacher support were associated with exposure to cyberbullying and both IRL and cyberbullying. Higher age (school grade) was found to be a protective factor for IRL bullying and both IRL and cyberbullying among girls only. Foreign background was not related to any form of bullying in either sex. As further displayed in Table II, boys and girls reporting poor body image, discrimination in school, and teaching bullying were at increased risk of experiencing psychosomatic problems (PSP) and depressive symptoms. Low teacher support was associated with psychosomatic problems in girls only, and foreign background as well as age was significantly related with psychosomatic problems in boys. Based on the associations, these variables were included in the analyses exploring associations between bullying exposure and mental health.
Downloaded from sjp.sagepub.com at NORTHEASTERN UNIV LIBRARY on April 5, 2014
4 E. Landstedt and S. Persson Table I. Descriptive statistics for dependent and independent variables. Variable
Total % (n)
Boys % (n)
Girls % (n)
p
Psychosomatic problems Depressive symptoms Bullying No IRL Online Both Foreign background Yes Grade 7th 8th 9th Body image Poor Discrimination in school Any Teacher bullying Yes Teacher support Low
22.4 (267) 46.0 (541)
13.7 (77) 35.1 (191)
30.2 (190) 55.5 (350)
48.1 (503) 15.1 (158) 17.2 (180) 19.6 (205)
51.4 (238) 12.3 (57) 17.3 (80) 19.0 (88)
45.4 (264) 17.3 (101) 17.2 (100) 20.1 (117)
14.4 (173)
12.8 (72)
15.9 (101)
31.8 (386) 32.6 (396) 35.6 (432)
30.8 (177) 30.5 (175) 38.7 (222)
32.7 (209) 34.5 (221) 32.8 (210)
46.6 (514)
26.2 (170)
59.4 (444)
26.8 (324)
22.6 (138)
25.7 (186)
22.9 (276)
24.9 (151)
16.9 (125)
50.7 (614)
44.6 (261)
48.8 (353)
