The impact of cyberbullying on loneliness and well-being among Malaysian adolescents: The mediation role of psychological distress

Cyberbullying and well-Being

In recent years, as academics have explored the quality of life of individuals, there has been a growing interest in the study of happiness (Neira et al., 2018). The presence of well-being in one's life is a protective factor against the development of internalising symptoms like depression (Llorent et al., 2021) and psychosomatic conditions (Smith et al., 2018). It is common knowledge that happiness is an individual's perception of their own level of well-being (Li et al., 2022b). Several researchers have hypothesised that cybervictimisation has more harmful effects than traditional bullying (Eyuboglu et al., 2021; Ng et al., 2022) due to the anonymity of the attacker and the absence of monitoring (Graham, 2023). It entails personal life pleasure and positive interactions (Ford et al., 2015), which are essential for adolescents attempting to form an identity. Sadly, cyberbullying encounters can have a severe impact on psychological health, safety, and well-being (Lonergan et al., 2021), notably in adolescents, who are going through a period of rapid physical, psychological, and social development, all of which can contribute to feelings of anxiety, confusion, and emotional instability (Fomina et al., 2020). In this paper, we tested the premise that cyberbullying has a detrimental effect on mental health (well-being) in light of the severity of cybervictimisation's impacts on teenagers. The victim is put into a state of depression as a result of being subjected to digital threats, insults, and denigration. This depression prevents the victim from being able to enjoy any form of social engagement, whether it be online or offline, in a tranquil manner, because it produces dissatisfaction.

In addition, given that being a cybervictim is a bad experience that can lead to a decline in well-being and that depression can play a significant part in the dynamic that exists between being a cybervictim and one's overall state of well-being (DeSmet et al., 2019; Santos et al., 2021), it can be inferred that depression can impair well-being in adolescents, as mentioned previously. Some behaviours, such as cyberbullying, appear to increase when stressful circumstances are present. Stressful life events are a wide concept that includes negative social–environmental experiences encompassing multiple domains, such as mental states and social connections (Slavich & Immunity, 2019). Since adolescence is a life stage characterised by multiple physiological, psychosocial, and social–environmental changes (Castellví et al., 2020; Yu et al., 2020), stressful life events are widespread among teenagers (Chau et al., 2022).

Cyberbullying and loneliness

Several researchers believe that loneliness plays a significant role. To put it simply, it is one of the most pressing issues facing contemporary society (Lim et al., 2020). Jeste et al. (2020) believed that loneliness is a subjective feeling of distress induced by a disparity between the desire for and the sense of social ties. There are studies that show links between loneliness and the use of technical devices like mobile phones and computers (Liu et al., 2021b), as is evident when perusing the associated literature. Since adolescent technology use is so pervasive, investigating possible links between the two seems appropriate (O’reilly, 2020). Because of this, it's reasonable to assume that young people turn to online communication as a means of overcoming isolation. Besides that, many physical and behavioural diseases, including heart disease, depression, and cognitive impairment, have been related to loneliness (Freak-Poli et al., 2022; Tan et al., 2019). It is considered as a problem of the elderly. However, Wang et al. (2023) and MacDonald, Willemsen, Boomsma, and Schermer (2020) were able to demonstrate that loneliness is U-shaped across the lifetime, suggesting that it is most prevalent among individuals under the age of 25 and those over the age of 65. While the quality of social contacts protects older generations from loneliness, younger generations place a greater emphasis on their quantity (Schwartz-Mette et al., 2020). Intriguingly, the desire for contact with friends has a bigger impact on loneliness in both groups than actual contact (Verity et al., 2021).

The victims of cybercrime have been found to suffer from feelings of discontentment and disappointment, as well as difficulties in adjusting to their new environments (Şahin, 2012). It has been said that those who exhibit those kinds of actions have a greater propensity to experience feelings of loneliness (Brighi et al., 2019; Yurdakul & Ayhan, 2021). As a result, there is a strong connection between being a cyberbully or cybervictim and experiencing feelings of isolation. In addition, Shin and Kim (2022) and Méndez et al. (2020) believed that adolescents who are victims of cyberbullying spend a significant amount of time on the internet and on their mobile devices. It's possible that this is due to the fact that people who are predi28sposed to feelings of isolation make use of technologies like mobile phones and the internet as a way to cut themselves off from the rest of society. This lends credence to the idea that cybervictims experience feelings of isolation. Cybervictims do not behave aggressively and do not incite others to behave aggressively towards them. This implies that the high levels of loneliness experienced by cybervictims can be attributable to the negative experiences they have with their peers. [Cyber] victims are more likely to be isolated. According to the Zhou (2021) findings, when teenagers are excluded from their peer groups, this may result in a psychological trauma, which in turn leads to feelings of isolation. Cyberbullying includes activities such as labelling people with derogatory names and ostracising others on the internet (Ronis & Slaunwhite, 2019). The combination of being lonely and spending a lot of time online produces an atmosphere that is ripe for the development of cyberbullying practises.

The present study

According to the conversations that came before, there has been an exponential increase in the number of researches on cyberbullying. In spite of the abundant and well-established evidence for researching risk factors that contribute to cyberbullying, relatively few studies have focused on depression, stress, and anxiety when investigating the factors that determine whether or not someone engages in cyberbullying. Yet, while studying the relationship between well-being and loneliness in relation to cyberbullying, the prior studies have not been able to come to a consistent set of conclusions. In addition, the potential mediating effects of internal stimuli (such as depression, stress, and anxiety towards cyberbullying) between well-being, loneliness, and cyberbullying have not been systematically tested within an integrated framework. This is despite the fact that these factors have been shown to have a strong relationship with each other.

A serial multiple mediator model was constructed using the framework of structural equation modelling (SEM) and AMOS 21 in order to fully investigate the mechanism by which cyberbullying influences well-being and loneliness through indirect paths. This was done in order to gain a comprehensive understanding of how cyberbullying can have a negative impact on an individual's emotional state. In light of this, the purpose of the current investigation is to look into the following:

The proposed model is depicted in Figure 1.

Figure 1. Research model.

Ethics approval

The survey was conducted with the University of Malaya Research Ethics Committee approval (Ref. no: UM.TNC2/UMREC_2492). The research methods were performed in accordance with the relevant guidelines and regulations.

Consent to participate

Respondents were provided with an explanation of the research purposed and informed consent was obtained from all respondents. We obtained written informed consent from the parents or legal guardians of all participants below the legal age of consent. This ethical requirement was met to ensure the protection of participants' rights and well-being.

Participants

There were a total of 386 Malaysian adolescents between the ages of 12 and 18 years old (the mean age was 15.73 years, and the standard deviation was 1.08 years). We used two different approaches to determine the most reliable and accurate number of participants based on the structure of our research model when determining the size of our sample. The first method that we used was based on the hypothesis presented by Hair et al. (2014). According to this theory, the required number of samples for the research should be proportional to the number of latent variables included in the investigation. This number should also take into account the number of indicators contained inside the latent variables, as detailed below:

The second way, the minimal sample size is predicted to be 10 times the most complex relationship in the study model by the PLS-SEM rule of thumb, which provides support for the sample size. The G*Power software's power analysis also revealed that a minimum sample size of 85 participants is necessary with an effect size of 0.15, an alpha value of 0.05, and a power of 0.80. As a result, if we use either of the two strategies mentioned above, we will require at least 100 participants for this study.

Purposive sampling was used to recruit participants for the study through various social media platforms like Instagram, Tiktok, Facebook, Telegram, and Twitter. Only Malaysian who were fluent in English and between the ages of 12 and 18 were eligible to participate. Participants contributed to our investigation by filling out an English version of a questionnaire that was hosted on Google Forms. The participants gave their informed written consent, which covers the matters pertaining to anonymity and confidentiality, and it was received from them. There were 386 participants total, with 180 males and the remainder participants being females. It was made up of 150 Malay people, 168 Chinese people, and 68 Indian people. According to the findings, twenty individuals used the Internet for less than one hour per day, 188 used it for two to five hours per day, and 178 used it for more than six hours per day.

Instruments

The level of cyberbullying was evaluated with the help of the Revised Adolescent Peer Relations Instrument (RAPRI), which was established by Griezel et al. (2012). It contains cyberbullying as well as cyber victims' goods, although for this particular study, just the cyberbullying materials were utilised. The RAPRI-Bully comprises of 5 items of bully visual and 8 items of bully text, all of which were scored on a Likert-type scale with 6 points, ranging from 1 (never) to 6 (always) (every day). Bullying through the use of video or photographs is referred to as “bully visual,” while bullying through the use of electronic communications such as emails, text messages, or instant chat messages is referred to as “bullying through the use of bully text.” For example, I used a mobile phone to forward a video to a student that I knew they wouldn't like (e.g., Sent a student an email with a message I knew would hurt their feelings). For measuring depression, anxiety, and stress we involved Depression Anxiety Stress Scales-21 (DASS-21) based on Lovibond and Lovibond (1995) theory. A total score on the DASS-21 can range anywhere from 0 to 63, with a subscale score ranging anywhere from 0 to 21, and an item score ranging anywhere from 0 (did not relate to me at all) to three (applied to me very much). The Warwick-Edinburg Mental Wellbeing Scale (WEMWBS) was used to evaluate an individual's positive mental well-being (positive functioning, happiness, and subjective wellbeing) over the course of the most recent two weeks (Tennant et al., 2007). The 14-item scale comprises five response categories, ranging from “None” (meaning “not at all ”) to “All the time” (meaning “always”). The final score ranges from 14 to 70 and indicates low to high levels of positive mental well-being based on the total of the respondents' ratings. The Pittsburgh Sleep Quality Index (PSQI) (Buysse, Reynolds III, Monk, Berman, & Kupfer, 1989) is a self-administered questionnaire that analyses a person's sleep quality over the course of the previous month. A modified version of the Likert scale developed by Russell et al. (1980) was utilised to assess levels of loneliness. This scale is comprised of four points that range from never to always. This index was used to assess the quality of the respondents' sleep. Subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disruption, usage of sleep medication, and daytime dysfunction are the seven components that make up the 19 different sections.

Reliability and validity analysis

The results of this study show that the average variance extracted (AVE) values are higher than 0.50, which indicates that all of the measures have high-quality content and convergent validity. In addition, the validity is supported by the fact that all of the item loadings are higher than the cutoff value of 0.60. The item loadings range from 0.66 to 0.95. It would appear that the overall composite dependability of all latent variables is above 0.75 and sufficient. The values of the variables' Cronbach's alpha have been determined to be greater than the cutoff value of 0.70. The specifics of the discriminant validity are presented in Table 2. When the square roots of all AVE scores are greater than their corresponding inter-correlations, discriminant validity has been demonstrated. In light of the findings presented above, one can draw the conclusion that the measurement model displayed a degree of validity and reliability that is satisfactory.

Structural equation modelling output

It was discovered that the proposed model provided an excellent fit with the data. The model fit was perfect, including [χ2/df = 1.33, p < 0.05], [CFI = 0.93], [GFI = 0.91], [TLI = 0.96], and [RMSEA = 0.031]. It has been stated that there is a positive and significant association between cyberbullying and stress (beta = 0.39; p < 0.001), anxiety (beta = 0.53; p < 0.001), and depression (beta = 0.46; p < 0.001). It has been discovered that all three psychological indices—depression, anxiety, and stress—have a considerable favourable impact on feelings of loneliness, but a significant negative impact on feelings of overall well-being (see Table 3).

Limitations and future research directions

In this study, it is necessary to address a few constraints. First, this study's conclusions were based on a cross-sectional design, which was incapable of establishing causal links between the research variables. Consequently, the bidirectional or causative relationships must be investigated further utilising longitudinal and experimental approaches. Second, the sample was comprised of junior teens from a particular region of Malaysia, limiting the generalizability of the findings. Hence, offline and online bullying may have larger detrimental consequences on the mental health of Malaysian teenagers. Consequently, the correlations between cyberbullying and depression, anxiety, stress, and the significance of loneliness and well-being may be larger in samples from collectivist cultures than from individualistic ones. Thus, it was suggested that future research duplicate our findings by gathering data from other cultural backgrounds (Malays, Chinese, and Indian). Thirdly, information about the frequency of assistance requests to private/public healthcare services relating to disordered eating behaviours will be useful in corroborating the findings of this study. Due to response bias, the use of self-reported metrics may influence the results to some extent. Future research should employ numerous informants (parents and instructors) and approaches, including experimentation, observation, and interviews.