II. LITERATURE REVIEW
John et al. (2018) conducted a meta-analysis study on 33 articles from 26 independent studies to determine the relationship between cyberbullying and self-harm or suicidal behaviors in individuals younger than 25 years of age. A major benefit of a meta-analysis is that it provides a quantitative analysis of a large, consolidated body of literature (Haidich, 2010). Using odds ratios as a summary measure of effect size, the findings indicated that, compared with nonvictims, those who experienced cyberbullying were 235% more likely to harm themselves and 257% more likely to attempt suicide (John et al., 2018). In short, victims of cyberbullying were at a greater risk than nonvictims of intentionally hurting themselves.
However, there were several limitations in the John et. al (2018) study. First, because a meta-analysis examines aggregates of data, the relationship between cyberbullying and self-harm for any particular person cannot be determined. Second, a failure to include a majority of existing studies can lead to wrong conclusions (Lee, 2019). Because research studies that do not reject null hypotheses tend to remain unpublished, there may be a bias toward using studies with positive results. This will compromise the validity of a meta-analysis. Third, a meta-analysis relies on shared subjectivity when deciding how similar studies should be combined. Indeed, every form of analysis in a meta-analysis requires some subjective decisions. Finally, quantitative studies do not provide an in-depth understanding of the meanings that the participants have associated with their lived experiences (Berg, 2007).
Hinduja and Patchin (2010) conducted a study and surveyed 1,963 middle school students from one school district in the United States to assess the relationship between cyberbullying and suicide ideation. The students were surveyed on their Internet use and experiences. Respondents were asked about their experiences with bullying and peer harassment, both online and offline, and thoughts about committing suicide. The researchers utilized logistic regression analysis and found that being a victim of cyberbullying is associated with an increase in suicidal ideation.
However, the Hinduja and Patchin (2010) study has several limitations. First, because the data were not collected over time, it is impossible to determine the proper temporal ordering among the variables. Second, the participants may have engaged in acquiescence bias by simply selecting positive responses over negative responses. Third, recall bias may have occurred due to individuals misrepresenting or distorting facts from previous time periods. Fourth, because the students who were surveyed attended only one school district, they may not necessarily be representative of other populations across the country.
Peng et al. (2019) conducted a study to determine if being a victim of cyberbullying is related to self-harm and suicide attempts. The researchers used a sample of 2,647 Chinese students from 10 junior high schools. Data for self-harm and suicide attempts were collected using a selfadministered survey. The psychopathy of each student was assessed using the Strengths and Difficulties Questionnaire. Using multinomial logistic regression, the findings indicated that, compared to nonvictims, victims of cyberbullying were at a greater risk of harming themselves and attempting suicide.
However, the Peng et al. (2019) study has several limitations. First, although the study determined correlational relationships, the study did not indicate causal relationships. Second, because the study was quantitative in nature, it cannot reveal the meanings that participants have given to various phenomena (Berg, 2007). Finally, the students in the study were raised in a social learning environment that is specific to China, which may be different in a meaningful way from other cultures (Peng et al., 2019). In short, the study’s findings cannot be generalized to other populations that do not match the sample’s characteristics.
Peng and Davis (2017) conducted a study to investigate the correlation between the percentage of students who were electronically bullied and the percentage of students who seriously considered suicide. Data were collected in 2011, 2013, and 2015 using a three-stage cluster sample design, which produced a nationally representative sample of students in grades 9–12 who attended public and private schools. The researchers employed linear regression analysis and the findings indicated that once female students were cyberbullied, they seriously considered suicide.
However, the Peng and Davis (2017) study has several limitations. First, the study was not an experimental study and cannot determine causal relationships. Second, because the sample was limited to American students in grades 9-12 in public and private schools, the findings cannot be generalized to other populations that do not match the sample’s characteristics. Finally, linear regression assumes that the observations were independent, which may be a problem because data in the study were collected multiple times from the same states (Norusis, 2008; Su, 2020).
Kyriacou and Issitt (2018) explored the perceptions of student teachers regarding cyberbullying by students against other students. Data were collected from 97 secondary student teachers at a university in England who were mostly in their early 20s and who attended high school when cyberbullying was becoming common. The sample consisted of 41 males, 54 females, and two who did not indicate their sex. The participants were provided a questionnaire that asked them to express their views about the motives behind cyberbullying and about the actions that can be implemented to deal with the problem. Most of the questions on the survey used a Likert-type scale. Based on mean scores and on one open-ended question, the findings indicated that student teachers believe that cyberbullying can be reduced through education on esafety and by implementing heavy sanctions against perpetrators.
However, the Kyriacou and Issitt (2018) study has several limitations. First, because the study was qualitative in nature, it failed to provide patterns of relationships through numerical representations. Second, the participants were student teachers who lived in England, and they may not necessarily reflect the opinions of seasoned teachers in America. Third, because Likerttype scales were used, there is the possibility that the participants a) committed central tendency bias by simply selecting the middle option rather than the best option, b) committed acquiescence bias by simply selecting positive responses over negative responses, and c) were forced to select options that did not accurately represent their realities (Antonovich, 2008). Finally, the sample consisted of student teachers who, compared to seasoned teachers, may not be fully aware of the variety of resources available at educational institutions.
Finally, Paul, Smith, and Blumberg (2012) investigated the students’ perceived effectiveness of coping strategies and school interventions related to traditional bullying and cyberbullying. A sample of 217 students, which consisted of 118 males and 99 females in grades 7-9, completed a worksheet on coping strategies. A sample of 190 students, which consisted of 95 males and 95 females in grades 7-9, evaluated school interventions. The worksheets that were used to collect the data were designed to measure student perception of different coping strategies and school interventions related to traditional bullying and cyberbullying. The analysis of the item ratings was conducted using the Wilcoxon signed ranks test. In addition, comparisons between item ratings and participant role type (i.e., victim, bully, or no role) were conducted using the Kruskall-Wallis test and the Mann-Whitney U post-hoc test. Although the students felt that seeking help and advice were effective coping mechanisms, and that school sanctions, informal approach, and support approach were effective intervention mechanisms, the findings indicated that the most helpful approach requires the support of family members, especially the parents.
However, the Paul et al. (2012) study has several limitations. First, the findings cannot necessarily be generalized to other individuals who do not match the sample’s characteristics. Second, because the application of anti-bullying interventions is unique to each school, it is important to exercise caution in drawing conclusions. Finally, the coping skills adopted by each students might be influenced by the atmosphere of the student’s particular environment.