An international systematic review of cyberbullying measurements

By Prof. JongSerl Chun (jschun@ewha.ac.kr)
Department of Social Welfare

Cyberbullying is a growing concern in many countries worldwide. In particular, adolescents who used to be bullied in school settings are now experiencing bullying by electronic means. For example, the EU Kids Online Project (2010–2014) indicated that cyberbullying victimization among 9- to 16-year-old adolescents rose from 8% to 12% across seven countries in the European Union (Livingstone et al., 2014). In South Korea, the Korea Internet and Security Agency (2013) reported that 30.3% of Korean children and adolescents experienced cyberbullying victimization. Numerous studies have reported the severity of cyberbullying issues in their respective countries; however, there is no established definition of cyberbullying (Langos, 2012). Given the lack of consensus on the conceptualization of the cyberbullying construct, the development of a measurement tool faces several issues and limitations. Despite the extensive research conducted on cyberbullying, the results of cyberbullying perpetration and victimization remain inconsistent across studies.

As such, Dr. Chun’s recent publication “An international systematic review of cyberbullying measurements” in Computers in Human Behavior, which systematically reviewed all existing cyberbullying measurements internationally, provides limitations of the current cyberbullying scales and suggestions that future researchers attempting to develop cyberbullying scale should consider.

Following the PRISMA guidelines (Moher et al., 2009), the number of scholarly manuscripts has been identified, screened, reviewed for eligibility, and finally included in the analysis. A total of 62 articles from 17 countries were included in the review. To provide an overview of the existing scales and suggest ways to standardize cyberbullying measurement, these studies were coded using the following categories: general characteristics, definition of cyberbullying, study sample characteristics, sample size, type of device or social media, time frame, survey type, item-pooling method, subscales, reliability, and validity.

Although 46 of the 64 studies explained the concept of “cyberbullying,” there was still a lack of consensus on the criteria and definition, and half of the studies used the word other than cyberbullying such as cyber aggression and electronic bullying. Taking all different kinds of definitions used in these 46 studies, the authors concluded that cyberbullying is a repeated and intentional act to threaten/harass/embarrass others through electronic means or devices.

The age of the sample population ranged from 8 to 25 years, and all the studies recruited both male and female participants. However, there were no gender-specific cyberbullying scales. Of the 64 studies, 27 reported a significant gender difference in the prevalence of cyberbullying victimization and perpetration. In terms of the type of devices or social media platforms, mobile phones were the most commonly included device, followed by e-mails and messages. The time frame varied across studies, but the most frequently used time periods were the “the last three months” and “the last six months.” The survey type was largely based on the self-reported, paper-and-pencil questionnaire, but some also used the online survey.

The item-pooling method is crucial in scale development; however, only 15 studies followed the recommended guideline for scale development by DeVellis (2003), either fully or partially. This implies that many existing scales on cyberbullying simply extracted or modified the questionnaires from other scales. Of the 30 instruments that had subscales, eight consisted of two subscales, and the other 22 had more than three subscales. Furthermore, 33 instruments performed either confirmatory or exploratory factor analyses. Most cyberbullying instruments revealed moderate to high reliability, and only half of the studies assessed the validity of cyberbullying measurements, with a high portion testing the construct validity.

Our findings first address the need for a consistent and standardized definition of cyberbullying to use worldwide, which may be the most important factor in measuring cyberbullying behaviors. As most cyberbullying measurement studies were conducted in the United States and Western countries, we propose the development and validation of scales that fully reflect the sociocultural components of cyberbullying in respective countries. Furthermore, noting the gender difference in cyberbullying victimization and perpetration, future researchers must develop a gender-sensitive cyberbullying measurement scale. Finally, the level of sophistication in item generation seemed to vary greatly across studies. We highly recommend future studies to provide a concrete, theoretical background of the item-pooling method using deductive, inductive, or both methods to yield more accurate results.

* Related Article
An international systematic review of cyberbullying measurements, December 2020, Computers in Human Behavior 113, 106485