Has the COVID-19 Pandemic Affected Cyberbullying and Cybervictimization Prevalence among Children and Adolescents? A Systematic Review

In light of the alarming results emerging from some studies and reports on the significant increase in aggressive online behaviors among children and adolescents during the COVID-19 pandemic, the current research aimed at providing a more detailed evaluation of the investigations focusing on the cyberbullying prevalence rates published between 2020 and 2023. To this purpose, systematic searches were conducted on four databases (Web of Science, APA PsycInfo, Scopus and Google Scholar), and following PRISMA guidelines, 16 studies were included and qualitatively reviewed. Although studies were characterized by a large variety in cyberbullying operationalization and measurement, and by different methodologies used for data collection, the prevalence rates of the involvement in cyberbullying and/or cybervictimization generally revealed opposite trends: an increase in many Asian countries and Australia and a decrease in Western countries. The findings were also discussed by considering the effects of the COVID-19 pandemic. Finally, some suggestions were provided to policy makers for promoting prevention and intervention anti-cyberbullying programs in school contexts.

The picture of such a phenomenon becomes sadder when considering the contextual factor of the COVID-19 pandemic that negatively affected children's psychological health due to the prolonged school closures and the consequent social isolation [45]. Indeed, adolescents and young people's increased use of the Internet and information and communication technologies (ICT) during the COVID-19 pandemic could have posed a greater risk of being involved in cyberbullying and cybervictimization behaviors, thus exacerbating the likelihood of displaying mental health problems. This is in line with the previous empirical evidence showing that spending much time online and engaging in online activities are crucial risk factors for cyberbullying and cybervictimization [15,16,[46][47][48][49][50].
The hypothesized close connection between time spent on digital devices during the pandemic and the increase in cyberbullying has led some researchers to provide empirical evidence [51][52][53]. In this vein, inspired by the general strain theory (GST), that posits strains or stressors increase negative affect due to exposure to negative stimuli and consequently predict broader aggressive behaviors, Barlett et al. [51] examined the mediating role of stress in the relationship between COVID-19 experiences and antisocial behavior (cyberaggression). The results provided tentative support for the theory among the US adult population by collecting data on public online discussions. Han et al. [52] showed a general increase in cyberbullying prevalence among Chinese children and adolescents, indicating that about 11.0% reported being cybervictimized, whereas 5.0% of participants said they had cyberbullied others during the COVID-19 outbreak in 2020.
A similar and considerable increase was also documented by Karmakar et al. [53], who assessed the impact of COVID-19 on the reported cyberbullying incidents through user-gendered tweets. In the same data-driven direction, the study by L1GHT [54], focused on the analysis of the increase in online toxicity as a result of the pandemic, reported a 900% increase in hate speech on Twitter toward China/Chinese people, a 70.0% increase in cyberbullying among children in chat forums and a 40.0% increase in toxicity on gaming platforms.
Moreover, a European technical report [55] aimed at providing a snapshot of how children across 11 countries experienced different online risks during the COVID-19 lockdown by comparing changes that occurred before and during the crisis underlined that almost 49.0% of the children declared having been cyberbullied at some point and that, at the national level, the highest percentages were recorded in Italy and Ireland (59%) followed by Germany (58.0%) and other countries. A further interesting note is that by comparing the two periods (pre-and during the pandemic), nearly half (44.0%) of children who had already been cybervictimized declared an increase in the phenomenon during the first wave of the pandemic. In contrast, more than a fifth of the children (22.0%) perceived the phenomenon as lesser than before. These results seem to reveal an evident contradiction among European children in estimating their experience with upsetting situations online. Conversely, the UNICEF Canadian report [56] indicated an opposite trend of its prevalence, reporting a 17.0% reduction in cyberbullying during the pandemic.
Given the social alarm raised by some empirical evidence and, to the best of the authors' knowledge, the lack of a detailed evaluation of cyberbullying prevalence rates across studies carried out during the COVID-19 pandemic, the present work aims to systematically review studies that investigated the prevalence of this phenomenon. In fact, restrictions due to the COVID-19 pandemic have limited children and adolescents' face-toface interactions, potentially increasing the chances of being involved in cyberbullying and cybervictimization rather than in school bullying. Although school bullying prevalence rates were, for a long time, much higher than cyberbullying [57,58], the lockdown restriction measures made engagement in school bullying "unfeasible" [59,60].
Understanding cyberbullying and cybervictimization trends during the COVID-19 pandemic would be crucial to designing and implementing primary and secondary prevention activities.

Methods
The systematic review included studies published from 2020 to 2023 (also considering any in-press publications on the topic). The Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA) guidelines [61] were adopted for searching, assessing study eligibility and reporting the results.

Search Strategy and Selection Criteria
A systematic search was conducted by four researchers in December 2022 on four databases: Web of Science, APA PsycInfo, Scopus and Google Scholar. To extract significant data for our research purposes, specific keywords were included in the search string to be launched on the selected databases: "cyberbullying", "cybervictim*", "prevalence", "diffusion", "minor*", "adolescent*", "teenager*", "COVID-19", "SARS-CoV-2", "pandemic" and "lockdown". These keywords were combined utilizing Boolean operators to create a search string. The search string used in all four research databases was (cyberbully* OR cybervictim*) AND (prevalence OR diffusion) AND (minor* OR adolescent* OR teenager*) AND ("COVID-19" OR "SARS-CoV-2" OR pandemic OR lock-down). Researchers (M.S., M.d.F.) individually checked all the titles and abstracts collected from the databases. They afterward discussed limits, inconsistencies and standards until a consensus was reached.
Articles identified were screened as abstracts. After excluding those that did not meet our inclusion criteria, the full texts of the remaining articles were assessed for eligibility, and decisions were made regarding their final inclusion in the review.
The inclusion criteria were as follows: (1) studies written in English and peerreviewed; (2) studies published between 2020 and 2023 (including any publications in press); (3) focusing on a survey population aged 6-18 years; (4) aimed at assessing the incidence of cyberbullying and/or cybervictimization among school-aged students during the pandemic.
On the other hand, exclusion criteria for this study are as follows: (i) reviews, case studies and case series; (ii) articles not published in English; (iii) focusing on general features of cyberbullying not during COVID-19 pandemic; (iv) studies focusing on cyberbullying and/or cybervictimization against minorities (i.e., ethnic/racial, LGBTQ+, SEN, disability, etc.).

Data Extraction
Two independent researchers (M.d.F. and M.S.) separately carried out data extraction, reading full-text articles to include only relevant objects. In case of disagreement, a third researcher (either A.S. or F.S.) was consulted to reach a consensus. Literature was only rejected and excluded if at least two researchers agreed it was not relevant.
To manage the records, Rayyan review-management software was utilized [62]. The variables extracted included lead author/year, country, study design (cross-sectional and longitudinal), sample size, sex (% males), age (mean ± SD, range), period of data collection (presence or absence of lockdown measures), instruments, quality appraisal (see below) and key findings.

Quality Appraisal
To check the quality appraisal, the Critical Appraisal tools in JBI Systematic Reviews for prevalence studies were used [63] (supplementary material, S1). The tool aims to assess a study's methodological quality and determine the extent to which a study has addressed the possibility of bias in its design, implementation and analyses by evaluating 9 domains: appropriateness of the sample frame, participants' recruitment, simple size adequateness, participants and setting described, sample coverage of data analyses conducted, method validity, suitability of statistical analyses, response rate and missing management. Two researchers (A.S. and F.S.) independently appraised the 16 studies included.
All the studies assessed were considered to be eligible as a result of the overall appraisal.

Results
The initial search identified a total of 4191 records. Before the screening phase, a total of 4129 records were removed since 4123 were remarked as ineligible, and 6 were duplicates. Thus, a total of 62 papers were scrutinized based on the search criteria, title and abstract. Twenty reports were excluded. Therefore, the remaining 42 full texts were considered potentially eligible for the study.
The results of the search are shown in a PRISMA 2020 flow diagram ( Figure 1).
validity, suitability of statistical analyses, response rate and missing man researchers (A.S. and F.S.) independently appraised the 16 studies include All the studies assessed were considered to be eligible as a result of praisal.

Results
The initial search identified a total of 4191 records. Before the screenin of 4129 records were removed since 4123 were remarked as ineligible, and cates. Thus, a total of 62 papers were scrutinized based on the search cr abstract. Twenty reports were excluded. Therefore, the remaining 42 full t sidered potentially eligible for the study.
The results of the search are shown in a PRISMA 2020 flow diagram (
The number of school-aged students involved ranged from 107 in Germany [99] to 34.771 in Finland [60]. Fifteen studies involved middle and/or high school students with four also including primary school pupils [52,60,92,101]; one study involved only elementary school students [91].
The time ranges of the data collection varied across the studies reflecting the spread of COVID-19 that followed different rates for each geographic area, based on random and epidemiological factors, as underlined by the Johns Hopkins University research team (https://coronavirus.jhu.edu/data/animated-world-map, accessed on 10 May 2023). Two studies [94,96] reported that data collection took place in 2021, five studies [60,95,97,99,100] were conducted during the early lockdown phase of 2020 and/or shortly after the pandemic outbreak in the spring-summer 2020 (March-August), and six studies were conducted between autumn 2020 and winter 2021 [89,90,92,93,101,102]. Across the two Chinese studies, the data collection period did not overlap, ranging from September-October 2020 [90] to November 2020-January 2021 [89]. In the longitudinal study involving South Korean elementary school students, the authors referred to the first data collection (baseline) in 2019 and a follow-up in 2020 [91].
Finally, two investigations [52,98] did not clearly report the timeframe even if explicit references to the pandemic were present in the title; thus, such information was inferred by consulting indirect sources or textual and/or metadata elements. Indeed, for the former [52], involving participants from the rural region of Shandong in China, since the article was submitted to the journal on February 5th, 2021, we can hypothesize that data collection occurred in 2020. In addition, considering that the questionnaire administered to participants included the item "Have your classmates or peers implemented these actions to you since January 2020?", we can infer that the data collection took place during or just after the Shandong "strictest lockdown" phase, which occurred roughly from late January 2020 to early March 2021, as indicated in other similar Chinese studies [103,104]. For the latter study [98], although the range period was not explicitly indicated, it was reported that pupils were invited to complete an online survey sent to 15 schools in December 2020.
Concerning the data collection procedure, most studies utilized online surveys, and three [89,96,97] collected data during school hours.
In particular, Eroglu et al. [96] reported that data were collected by psychological counselors during school hours, even if the rate of students attending school was low because they worried about both contracting COVID-19 and transmitting it to their families. Xiang et al. [89] collected data by administering a paper-and-pencil questionnaire to Chinese adolescents during school hours. Thumronglaohapun et al. [95] sent envelopes via mail to each institute participating in the survey to let the students decide when and where to fill in the questionnaires. In the study conducted by Thai et al. [97], data were collected while schools were open, but the procedure was not specified.
In the study carried out in Australia [100], data on CB/CV prevalence were collected adopting the Cyberbullying Questionnaire-Revised (CBQ-R) [40] from 2015 to 2019, while the Cyber Bullying Participant Roles Scale [105] was administered online in 2020. Moreover, Shin and Choi [92] and Choi, Shin and Lee [91] assessed CB/CV in person through teachers in 2019, while in 2020, data were collected using an online version of the same questionnaire because of the COVID-19 pandemic. One single question asking "Have other children used any of the following items to bully you?" Students rated their involvement in CB by answering 14 items scored on a 5-point Likert scale (from "0 = never" to "4 = always") One single question asking "Have you used any of the following items to bully other children?" Students rated their involvement in CV by answering 14 items on a 5-point Likert scale (from "0 = never" to "4 = always") items on a 6-point scale (from "1 = not at all" to "6 = many times a week") 9 items on a 6-point scale (from "1 = not at all" to "6 = many times a week") -Wiguna et al. [93] Indonesia August-October 2020 Online survey -3-item scale adapted from other questionnaires One question asking "During the past 6 months, how often have you been cyber-bullied?" Students answered using a 4-point Likert scale (from "1 = never" to "4 = almost daily") One question asking "During the past 6 months, how often have you cyber-bullied others?" Students answered using a 4-point Likert scale (from "1 = never" to "4 = almost daily") One question asking "During the past 6 months, how often have you been cyber-bullied and being cyber-bullied others?" Students answered using a 4-point Likert scale (from "1 = never" to "4 = almost daily") Xiang et al. [89] China November 2020-January 2021 School-based survey Using information and communication technologies to repeatedly and intentionally harm, harass, hurt and/or embarrass a target Chinese version of the E-Bullying and E-Victimization Scale (E-BVS) 6 items on a 7-point Likert scale (from "0 = never" to "6 = 6 times or more") 6 items on a 7-point Likert scale (from "0 = never" to "6 = 6 times or more") Cyberbullying refers to activities committed in cyberspace or utilizing information communication technology with the intent of causing harm to those who are unable to defend themselves Ad hoc questionnaire 6 dichotomous items (0 = "never" and 1 = "rarely", "sometimes" and "often") 6 dichotomous items (0 = "never" and 1 = "rarely", "sometimes" and "often") - (from "0 = never" to "4 = more than once a week") 11 items on a 5-point Likert scale (from "0 = never" to "4 = more than once a week") Notes: CB = Cyberbullying; CV = Cybervictimization. Another important methodological issue is related to the definitions adopted to operationalize CB and CV, which, as summarized in Table 1, differed across the selected studies. Such definitions were analyzed referring to Chun et al.'s categorization [2]: (1) the use of electronic means or devices, (2) vulnerability (e.g., those who cannot easily defend themselves), (3) repeated harm or behavior, (4) deliberate or intentional act, (5) unwanted information of others and (6) the purpose of threatening/harassing/embarrassing others. Thirteen studies included the use of electronic means or devices in the definition for CB and/or CV [52,60,[89][90][91][92]94,[96][97][98][99][100][101]; five definitions referred to victims' vulnerability and incapacity to defend themselves [52,90,94,95,100]; six studies defined it as repeated harm caused to victims [52,89,90,[98][99][100]; six studies also included the perpetrator's deliberate intention to harm in the definition [89,94,96,98,100,101]; and five definitions referred to cyberbullies' purpose of threatening/harassing/embarrassing others [89][90][91][92]97]. None of the sixteen studies used "unwanted information of others" in their CB definition.
Only five studies [60,90,91,94,100] provided a definition of CB to students along with the self-report questionnaires utilized to assess their involvement. In contrast, the remaining 11 studies did not explicitly report if a description of CB/CV was provided before data collection. However, multi-item scales were used to overcome such limits by asking participants to rate the frequency of specific CB/CV behaviors [106,107].
Two studies [93,94] did not explicitly report if a CB/CV definition was provided to participants before collecting data and adopted a single item to measure students' engagement in CB and/or CV.

Data on Prevalence of Cyberbullying/Victimization during the Pandemic
At first glance, almost half of the studies [60,89,90,93,96,100,102] did not report CB and/or CV prevalence rates, and the presented data varied considerably in terms of CB and/or CV percentages (Table 1). In addition, a comparison among them turned out to be difficult, since different methods were used to operationalize and measure such constructs.
Indeed, concerning the psychometric tools assessing CB/CV, more than two-thirds of the studies used validated measures of cyberbullying and cybervictimization, three studies developed ad hoc questionnaires [52,91,92], and one study adopted a single item for involvement in cyberbullying, cybervictimization and cyberbullying/victimization [93], whereas in another one [60], two single items were adopted for measuring cybervictimization only.
Such heterogeneous measurement also implied that not all studies reported prevalence rates concerning all four possible categories of involvement in the phenomenon (i.e., not involved; cyberbully; cybervictim; cyberbully/victim). As shown in Table 2, only two studies reported participants' involvement in cyberbullying, referring to the four involvement categories [52,98].
Three studies reported percentages of participants not involved in any cyberbullying episodes during the pandemic, with the highest rate found in China [52] (86.68%), followed by the Croatian study (73.08%) [98], while the lowest percentage was found among Thai students (50.8%) [95].
Seven studies reported the percentage of cyberbullying. The two longitudinal studies conducted in South Korea found the highest rates, reporting that 12.4% of elementary [92] and 9.5% of school-aged students [91] were involved in cyberbullying during the pandemic. The lowest CB involvement percentage rate was found in China [52], with 1.89% of 1111 participants declaring being cyberbullies.
Furthermore, systematic comparisons between CB and/or CV prevalence rates before and after the pandemic across these 11 cross-sectional studies are not possible. Therefore, the emerging results are compared with those observed in previous investigations if they are explicitly discussed and reported.
The investigation carried out by Thai et al. [97] in May 2020, when all schools were open, found that (i) cybervictimization was identified in 36.5% of 1492 Vietnamese adolescents and almost one-fourth of students experienced multiple types of cybervictimization; (ii) cybervictims had 1.81 times (95% CI [1.42-2.30]) higher odds of reporting symptoms of depression; and (iii) a higher likelihood of having depression was also found among female students who had Internet addiction. By comparing findings with others collected before the pandemic period, it emerged that the prevalence rate resulted higher than those collected in Vietnam, i.e., in Hue [108] (9.0%; of 648) and in northern provinces (24.0% of 763 adolescents) [109]. The higher prevalence rate was attributed to the higher level of depression found among adolescents, whereas the lower rate was attributed to the different Internet development and usage patterns.
In the Chinese context, 13.32% of 1111 adolescents reported being involved in the phenomenon, with 8.19% being involved in cybervictimization, 1.89% being involved in cyberbullying and 3.24% being cyberbullies/victims [52]. The reported prevalence rates were generally higher compared to prior studies carried out in China, highlighting a victimization prevalence rate ranging from 5.51% [110] to 7.49% and a perpetration rate of 2.05% [13]. According to the researchers, the higher prevalence rates in these phenomena were due to intensive use of the Internet service under the age of 10 and greater loneliness due to social distancing.
A similar conclusion was drawn in the Malaysian study involving 1290 high school students [94]. The researchers found that 3.8% of the participants were cyberbullies and 13.7% were cybervictims. Prevalence rates were lower compared to previous investigations conducted in the same country. In particular, a study carried out in the state of Negeri Sembilan found a prevalence rate of 52.2% in cybervictimization [111], while a study conducted in the state of Penang reported a prevalence rate of 20.9% in cyberbullying and 31.6% in cybervictimization [112]. According to the authors [94], cyberbullying prevalence varies across the states as some states, particularly the more urban ones, may have greater access to technology, thus leading to a higher prevalence of cyberbullying. Moreover, in this investigation [94], a higher rate of suicide prevalence was found compared to previous studies conducted in the same country, highlighting that adolescents who had been cybervictimized showed an increased risk of depression and anxiety symptoms and were more prone to engage in suicidal behaviors than those who had not been cybervictimized.
In the study carried out in Germany [99], 107 adolescents reported their cybervictimization frequency, with 25.7% indicating experiencing cybervictimization less frequently, 54.3% indicating experiencing cybervictimization equally frequently and 20.0% indicating more frequent cybervictimization during the pandemic as compared to before. It should be noted that the percentage indicating less cybervictimization was inferred from participants' self-report and might be an underestimation of the true decrease in cybervictimization. In the same study, greater levels of cybervictimization were found to be related to lower emotional self-efficacy, and lower self-efficacy was related to lower self-esteem. Moreover, cybervictimization was related to lower well-being in this sample.
Similarly, in a retrospective randomized study [101], the results showed a significant decrease in CB and CV involvement among Canadian participants during the COVID-19 pandemic compared to the pre-COVID-19 situation: indeed, with regard to the cyberbullying involvement, 13.8% of participants reported to be involved as perpetrators before the pandemic compared to the 2.3% during the pandemic. The same reduction emerged for cybervictimization (13.8% vs. 11.5%). These results are consistent with the recent Canadian UNICEF report [56].

Trends in the Prevalence of Cyberbullying/Cybervictimization before and during the Pandemic
To track the trends in the prevalence of cyberbullying/cybervictimization before and during the pandemic, four longitudinal investigations are scrutinized, although the diversity in sample size and methodological and measurement strategies make it difficult to compare the observed data. Indeed, the study carried out by Repo et al. [60] assessed the prevalence rate of cybervictimization only, the two Korean studies [91,92] were part of a broader national project on cyberbullying and cybervictimization started in 2019, and the last one reported CB and CV prevalence at four time points from 2015 to 2020 [100]. Concerning the first investigation [60] carried out on a sample of 34,771 students attending the Kiva anti-bullying program, the findings showed a decrease in cybervictimization with the average rate decreasing from 2.0% before the pandemic to 1.0% during the lockdown. An opposite trend was reported in the two Korean studies [91,92], where the rates of the prevalence of cybervictimization increased from 19.0% in 2019 to 19.7% in 2020 [92] with the highest prevalence increasing from 25.8% in 2019 to 32.7% in 2020 among elementary school students [91]. Conversely, a general decrease was recorded in cyberbullying when looking at different age groups from 18.0% in 2019 to 9.5% in 2020 and in cyberbullying/victimization involvement from 10.1% in 2019 to 6.4% in 2020 [92]. These results seem to confirm the negative impact of COVID-19 of increasing cybervictimization diffusion among primary school pupils.
The study carried out in Australia from 2015 to 2020 [100] showed an increase in the trend of cybervictimization over time, with higher rates in 2017 and 2019 compared to 2015; CV rates in 2020 were also higher than those reported in 2015 and 2019. By contrast, a different pattern emerged in the prevalence rates of cyberbullying, highlighting a significant increase only in 2020 compared to the timeframe 2015-2019.
To sum up, three studies found an increasing trend with two studies reporting a significant increase in cybervictimization behaviors [91,92] and one study reporting an increase in the diffusion of both cyberbullying and cybervictimization [100]. On the other hand, only one study showed a decreasing trend in the involvement in cybervictimization among children and youth [60]. An overview of all studies is provided in Table 2.

Gender Differences in Cyberbullying and Cybervictimization Involvement
Half of the analyzed studies indicated gender differences in the involvement in cyberbullying and cybervictimization during the COVID-19 pandemic, although mixed results emerged: three studies found that males were more likely to be involved in cyberbullying [89] both before and during the pandemic [92,100]. A different pattern was found in the study carried out by Choi et al. [91] involving only elementary school pupils, indicating that girls were more likely to be involved in cyberbullying before and during the COVID-19 pandemic. Two studies reported no significant gender differences in cyberbullying involvement [98,101].
Concerning cybervictimization, three studies found no significant gender differences [99,101,103], and one study found a significant association between being male and cybervictimization [89], whereas Vaillancourt et al. [101] found that, both before and during the COVID-19 pandemic, girls scored higher in cybervictimization.

Discussion
The COVID-19 pandemic has brought about a rapid and drastic change in human relationships, especially among children and young people, due to the prolonged closure of schools and the long-lasting social isolation. In this regard, several studies highlighted the negative impact of COVID-19 on children and youth mental health and an increase in youth psychological distress, worry, loneliness, anxiety, depression and traumatic symptoms [45].
It is conceivable that these adverse psychological outcomes could have been affected or worsened by the youth's increased involvement in cyberbullying and cybervictimization, as underlined by documents released in the initial stages of the pandemic [55,113]. Indeed, especially during lockdown periods, children and adolescents increased their screen time, thus increasing their risk of being involved in cyberbullying and/or cybervictimization in more severe ways [10,114]. Furthermore, the significant overlap and inter-correlation between cyberbullies and cybervictims' roles [43] potentially create a recursive negative circle that may cause several negative and deviant behaviors that could in turn result in internalizing and externalizing symptoms [11].
Given the lack of a review concerning the incidence of this phenomenon during the pandemic period, the current qualitative synthesis attempted to fill this gap by analyzing the prevalence rates of cyberbullying and/or cybervictimization among children and adolescents across countries, bearing in mind the different variations of the SARS-CoV-2 transmission in each geographic area. Sixteen studies resulted eligible for our investigation. Ten studies were conducted in Asian countries, three in Europe and the remaining three in Australia, Canada and Chile. Although the selected studies could provide a global overview for a deeper understanding of the prevalence rates of the phenomenon, the large variety of the operationalization and measurement of the construct and different methodologies used for data collection limited the evaluation, thus adding further complexity to the phenomenon.
Indeed, despite prior recommendations suggesting the need to follow shared criteria or categorization in defining cyberbullying before collecting data [2,106,107,115], from the current qualitative synthesis, it emerges that only five studies were in line with this methodological orientation.
Concerning the prevalence rates, the review generally indicated two opposite trends: an increase, for example, among Australian teenagers in both CB and CV [100], and a decrease among South Korean primary schoolers in CV [91] in the context of a general decrease in CB/CV among South Korean elementary, middle and high school students [92]. A decreasing trend was also reported among Canadian children and adolescents in both CB and CV [101] and among Finnish adolescents in CV involvement [60].
These findings might reflect the cross-cultural differences as well as the different impact and severity of the lockdown measures during the pandemic. Studies underlining a decrease in CB and/or CV during the pandemic seem to corroborate the arguments proposed by Olweus and Limber [106], who rejected the general claim "Screen time is up -and so is cyberbullying" [113] thus stressing that cyberbullying involvement is not a mere consequence of the increased time on digital devices but a long-established problem. Moreover, such a reduction in CB and CV involvement might reflect the fact that, as a consequence of the lockdown measures, parents and teachers were more involved in closely monitoring young people's virtual activities [100,116,117]. By contrast, studies that found an increasing trend in CB and CV prevalence rates hypothesized the contributing role of factors such as higher levels of loneliness due to social distancing, increased time spent online and poor levels of psychological well-being [52,94,97,99]. Such findings could foster the general assumption that cyberbullying is compounded by the interplay of different individual, relational and contextual factors [118]. In this vein, investigations carried out by Shin and Choi [92] and Choi, Shin and Lee [91] highlighted that the COVID-19 pandemic exposed especially primary school pupils to a greater risk of CB and CV, shedding light on the importance of including children in anti-cyberbullying prevention and intervention programs. The development and implementation of such programs with this population could help to pinpoint effective primary prevention strategies that may foster an increase in protective factors toward future youth and adolescents' involvement in CB and/or CV. In this direction, primary schools could represent the place of choice for the implementation of effective primary prevention actions, able to inform and sensitize children to the "safe" use of electronic devices and social media and to avoid the psychosocial consequences (e.g., depression, anxiety) resulting from the involvement in cyberbullying and/or cybervictim-ization [119]. In this direction, Repo et al.'s [60] results remarked the significant role played by the KiVa anti-bullying program [120] in promoting greater awareness of risks related to online settings and, therefore, in lowering cybervictimization rates.

Conclusions
Considering the importance of developing and delivering anti-cyberbullying programs, primary prevention and awareness-raising activities about online risks for children and adolescents need to be addressed at governmental and community levels. In this regard, considering the increasing use of social media and the Internet by younger users or "digital natives", particular attention should be paid by social media platforms in terms of defining and setting clear anti-cyberbullying policies. From an educational perspective, social media allow users to easily connect with knowledge, share it and transform and rework content-internalizing it and making the learning flow smoother [121]. However, even if some popular social media platforms set restrictions for young children, their use can still become potentially dangerous for children's psychological and physical well-being. Indeed, a study based on a qualitative analysis of 14 social media companies' policies and interviews with social media company representatives found that cyberbullying continues to take place regardless of social media policies [122]. Although some social media platforms have improved their policies over time, by implementing methods to report cyberbullying behaviors and encouraging non-anonymous experiences by making users use their real names [123], specific curricula focusing on media education and awareness of online risks should be included in cyberbullying prevention and intervention programs [124].
Furthermore, future research should therefore focus on developing, implementing and assessing the long-term efficacy and sustainability of multi-componential CB and CV programs targeting (also) primary school students.
Although the aim of the current systematic review was not fully achieved due to the above-mentioned limitations arising from the qualitative evaluation of the selected studies, it could contribute to providing a partial snapshot of the prevalence rates of involvement in cyberbullying and/or cybervictimization across the countries during the pandemic period and to remarking the need for a shared agreement on how to define and measure CB and CV [2]. Furthermore, our results stress the need for future research assessing retrospectively the possible long-term impact of the pandemic period on CB and CV prevalence.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijerph20105825/s1, Table S1: Quality assessment results based on Critical Appraisal tools in JBI Systematic Reviews for prevalence studies, Table S2: Description of studies excluded from the systematic review.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.