A Path from Traditional Bullying to Cyberbullying in South Korea: Examining the Roles of Self-Control and Deviant Peer Association in the Different Forms of Bullying

Abstract

Despite a recent growth in studies on cyberbullying, extant knowledge on the underlying mechanisms of cyberbullying remain limited. The objective of the present study is to explore the dynamics of cyberbullying via traditional bullying, self-control, and delinquent peer association. Specifically, the following hypotheses guide the present study: (1) traditional bullying, low self-control, and delinquent peer association are predictive of cyberbullying, respectively, (2) the interaction between traditional bullying and low self-control has a significant impact on cyberbullying, and (3) the interaction between traditional bullying and delinquent peer association has a significant impact on cyberbullying. The present study relies on five waves of the Korean Youth Panel Survey (KYPS), a representative sample of South Korean adolescents. Data collection occurred annually and respondents were 14 years old at the first wave in 2003. KYPS is an almost gender-equal and racially/ethnically homogenous sample. Results of cross-lagged dynamic panel models show (1) significant effects of traditional bullying on cyberbullying with and without low self-control and delinquent peer affiliation, (2) the respective roles of self-control and delinquent peer association in the prediction of cyberbullying, and (3) an interaction effect between low self-control and traditional bullying on cyberbullying. These findings demonstrate the theoretical validity of self-control theory and social learning theory in online delinquent behavior as well as confirm their cross-cultural generalizability in a non-Western sample. The findings also highlight the importance of investing in early life-course prevention/intervention programs and policies to prevent and/or reduce the occurrence of bullying, regardless of whether it is being perpetrated face-to-face or online, and these programs and policies should also target components to improve self-control and reduce delinquent peer associations.

Keywords

cyberbullying traditional bullying self-control deviant peer association

Bullying is often defined as intentional repeated aggression that occurred in an asymmetric power relationship (Olweus, 1997), which has been regarded as a major social concern around the world. Traditional bullying that usually takes place in school has had adverse impacts on youth in many ways. Negative consequences of traditional bullying include mental health problems (Fekkes et al., 2004; Gini & Pozzoli, 2009), physical health issues (Fekkes et al., 2006; Jennings et al., 2019), poor academic performance (Eisenberg et al., 2003), problematic behaviors at school (Smith et al., 2004), and externalizing behavior (Gini & Pozzoli, 2009). To make matters worse, recent innovations in information and communication technologies form a new type of bullying, that is, cyberbullying.

Although a relatively new phenomenon, cyberbullying quickly has become a global issue. South Korea is certainly not an exception. For example, a Korean national survey has documented that almost 28% of adolescents reported cyberbullying victimization, and almost 20% of teenagers engaged in cyberbullying perpetration (Lee, 2014). Analogous to traditional bullying, cyberbullying is linked to a wide array of adverse outcomes including: mental health/psychological symptoms including depression, social anxiety, poor mental well-being, low self-esteem, social difficulties, substance use, and suicidal attempts (Bottino et al., 2015; Chang et al., 2013; Fahy et al., 2016), physical health issues (Kowalski & Limber, 2013), problems at school such as poor performance, poor concentration, absences, and low grades (Beran & Li, 2008; Kowalski et al., 2012; Vazsonyi et al., 2012) as well as detentions, suspensions, and weapon carrying (Ybarra et al., 2007).

The appearance of cyberbullying, “willful and repeated harm inflicted through the medium of electronic text” (Patchin & Hinduja, 2006, p.152), resulted in scholars inquiring about the nature of this new type of bullying, and there is now a growing body of research in this literature exploring the relationship between cyberbullying and traditional bullying. In this vein, prior studies generally suggest that cyberbullying shares similarities with traditional bullying, and then, the next question arises as to whether traditional bullies are likely to also be (or become future) cyberbullies. Exploring answers to such a question is important to make effective and comprehensive prevention/intervention programs and policies, as well as aiding in the understanding of cyberbullying itself. Besides, the extant knowledge in this area is largely limited about factors that accelerate or buffer such linkages that are derived from a criminological theoretical perspective. To answer these questions, the present study seeks to unravel the dynamic mechanism of cyberbullying with consideration of the role of traditional bullying. Through the use of five waves of the Korean Youth Panel Survey (KYPS), the present study integrates the criminological theoretical frameworks of self-control and delinquent peer association to examine these linkages.

The Relationship Between Cyberbullying and Traditional Bullying

In cyberbullying literature, there is an ongoing discussion on its connection to traditional bullying: that is, whether cyberbullying is a newly invented deviance or a subtype of bullying. Some view cyberbullying as distinct due to the unique features of cyberbullying (see, Kowalski et al., 2012; Thomas et al., 2015). In the digital world, cyberbullies feel that he/she can hide behind a computer monitor since anonymity offers a feeling of safety or an inability of detection/getting caught. While verbal and nonverbal cues are observable in face-to-face communication, electronic communications lack instant emotional reactivity, which could worsen the situations (Kowalski et al., 2014; Kowalski et al., 2012). Moreover, it is quick and simple to copy and paste any rumor or gossip in electronic communication environment, allowing them to spread instantly. Although traditional bullying brings damages in terms of the victimization experience, it is often limited in terms of time and place (Nansel et al., 2001). In contrast, cyberbullying has no boundaries due to technology development and the worldwide accessibility of such technology. Another critical difference between these two forms of bullying is that victims of traditional bullying were relatively safe outside the school while cyberbullying can victimize individuals no matter where they are, what day/time it is, etc. As such, this 24/7 accessibility makes victims more vulnerable and leads to perception (and oftentimes the reality) that they cannot escape the bullying (Kowalski et al., 2012). Another contributing factor that increases online bullying and victimization is the parents’ and teachers’ lower familiarity with the digital world compared to their children. This digital divide reduces the ability for parents and teachers to prevent, monitor, or intervene in cyberbullying experiences as they emerge.

Despite the conceptual differences, empirical studies have shown a link between traditional bullying and cyberbullying. For example, employing a dual-trajectory analysis, Kim et al. (2017) reported general similarities in the developmental trajectories between the two forms of bullying among Korean youth. More specifically, they identified three latent groups for both bullying behaviors. Furthermore, the joint probability showed that almost 28.8% of adolescents were assigned to the non-involved group for both traditional bullying and cyberbullying, 27.5% were assigned to both sharp-decreasing groups, and 7.4% were assigned to both chronic groups. Or, in other words, an overlap in trajectory group assignment for the two forms of bullying was observed for greater than 60% of the sample. Comparatively, based on the criminal career approach, Donner et al. (2015) tested the offending generality/specialization hypothesis in cybercrime, relying on college students in the U.S. To be specific, the study examined whether online offenders engage in various types of crime or specialize in specific types of crime from a life-course perspective. Their results also demonstrated that offline offending and online offending were significantly related, thus supporting the generality hypothesis (for an exception, Khey et al., 2009).

The close association between traditional bullying and cyberbullying has also been observed in studies with different cultures. For instance, relying on a sample of Australian adolescents, Hemphill et al. (2012) found relational aggression as “a covert form of bullying” (p.59) (e.g., exclusion and spreading rumors) was predictive of cyberbullying perpetration. Similarly, engaging in traditional bullying predicted cyberbullying among Greek students aged 12–14 (Athanasiades et al., 2016). This was also the case for adolescents living in Cyprus (Fanti et al., 2012) as school bullying predicted cyberbullying 1 year later and school victimization predicted cyber victimization 1 year later. Jose et al. (2012) also documented the bidirectional relationship between traditional bullying and cyberbullying among teenagers in New Zealand. In addition, both traditional bullying perpetration and victimization have been determined to be significant predictors of cyberbullying perpetration among Korean youth (You & Lim, 2016).

The Role of Self-Control and Delinquent Peers on Cyberbullying

Considering prior literature has identified the shared characteristics between traditional bullying and cyberbullying, it is plausible to apply extant criminological theories explaining traditional delinquent behavior to cyberbullying. For example, Gottfredson and Hirschi (1990) highlighted the importance of self-control in their general theory of crime. Specifically, those who have low self-control are described as being impulsive, shortsighted, insensitive, lacking empathy, and seek immediate gratification. Therefore, they are vulnerable to the pleasures of the moment and do not take into account the long-term consequences of their behavior. Claiming its general effects, Gottfredson and Hirschi have argued that self-control plays a crucial role in various types of delinquency, which includes bullying. Individuals with low self-control are insensitive and have poor empathy, therefore, it is unlikely for them to realize victims’ pain while enjoying using their power over victims, which may pleasure them.

Low self-control as a key factor for deviance has been considerably documented in the bullying research (Benda, 2005; Chapple, 2005; Lee & Kim, 2017; Vazsonyi et al., 2017). For example, Kim et al. (2017) revealed the salience of self-control in identifying heterogenous developmental pathways in both traditional bullying and cyberbullying among Korean adolescents. Donner et al. (2015) has also reported that low self-control is significantly related to online and offline offending. Finally, a cross-cultural study drawing on samples from 25 European countries has shown both direct and indirect effects of low self-control on cyberbullying perpetration (Vazsonyi et al., 2012).

Comparatively, Akers’ social learning theory (Akers, 2011) is another major criminological theory that has been widely applied to different types of delinquency and crime in the field of criminology. Social learning theory posits that individuals learn criminal behavior through social interaction with others and reveals the complex mechanisms by which behavior is learned. In this regard, the underlying assumption is that a dynamic learning process takes place through differential association, definitions, imitation, and differential reinforcement.

Differential association with deviant peers (or commonly referred to as delinquent peer association) is particularly important for adolescents who are heavily influenced by their peers (Warr, 2002). Deviant friends play several roles in the broader context of social learning mechanisms (Akers et al., 2020). Youth are exposed to definitions favorable to deviance through interactions with delinquent friends. Also, deviant kids may serve as a good model for delinquency through imitation and provide rewards via differential association. The influence of deviant peer association has been frequently tested and its importance has been widely supported by empirical evidence (Hoeben et al., 2016; McGloin & Thomas, 2019; Pratt et al., 2010). As is the case with previous research on traditional delinquency, a growing number of studies examining the delinquent peer effect on cybercrime has been emerging in the literature. For instance, Sasson and Mesch (2017) reported the significant association between peer norms and cyberbullying. Specifically, peer delinquency was identified as a significant factor for distinguishing high levels of cyberbullying and high levels of traditional bullying perpetration (Kim et al., 2017). Holt et al. (2012) have also found peer deviance to be significantly related to an increase in different types of cyber deviance.

Moreover, the moderation effects of self-control and deviant peers were suggested in theoretical frameworks and tested in prior studies on general delinquency and bullying literature. Although Gottfredson and Hirschi (1990) remained in their position about self-control as a stable trait, self-control is often thought of as a dynamic entity, which is open to change (Burt et al., 2014; Hay & Forrest, 2006; Na & Paternoster, 2012; Pratt, 2016). For instance, an emerging body of research found that prior delinquency influences the level of self-control (Clinkinbeard et al., 2018; de Kemp et al., 2009; Vazsonyi & Jiskrova, 2018). This is also the case for delinquent peer affiliation. According to Hirschi (1969), it is hard for deviant youth to build prosocial relationships with conventional kids; therefore, they are likely to hang out with other deviant ones. Recent bullying studies found support for the thesis. Bullying is positively linked to negative peer influence (Espelage et al., 2000) and the relationship between bullying and deviance was moderated by deviant friends (J. Lee et al., 2018). Against this background, we would expect that low self-control and delinquent peers increase the likelihood of offline bullies engaging in online bullying.

The Current Study

Although an emerging body of literature has made initial forays into examining cyberbullying and its relationship with traditional bullying, the majority of this prior research has been cross-sectional and focused on Western samples. As such utilizing a nationally representative sample of Korean adolescents, the objective of the present study is to explore the dynamics of cyberbullying via traditional bullying, self-control, and delinquent peer association. Specifically, the following hypotheses guide the present study: (1) traditional bullying, low self-control, and delinquent peer association are predictive of cyberbullying, respectively, (2) the interaction between traditional bullying and low self-control has a significant impact on cyberbullying, and (3) the interaction between traditional bullying and delinquent peer association has a significant impact on cyberbullying.

Method

Data

The present study relies on five waves of the KYPS, a representative sample of South Korean adolescents. A stratified multistage cluster sampling was used to ensure the representativeness of the sample. The Korean youth population was divided into three groups: (1) those who live in the Seoul metropolitan city (the largest city in South Korea), (2) those who live in a compilation of six other metropolitan cities, and (3) those who live in other cities and counties equally drawn from all nine Korean provinces. A total of 104 schools were randomly chosen, and then, one class was randomly selected from a school.

The KYPS encompasses a wide variety of domains relevant to adolescent development. Data collection occurred annually and respondents were 14 years old at the first wave in 2003. Although there was some attrition, approximately 80% of the respondents remained in the sample by the end of the five waves. Respondents took a paper-and-pencil survey and parents had a phone survey. KYPS is an almost gender-equal and racially/ethnically homogenous sample (Table 1).

Table 1.

Sample Descriptives.

Variable	Mean	SD	Min	Max	SD (between)	SD (within)
Dependent variable
Cyberbullying	0.265	0.835	0	7.168	0.543	0.635
Independent variables
Traditional bullying	0.087	0.473	0	6.905	0.297	0.368
Low self-control	15.956	3.968	6	30	3.109	2.467
Delinquent peer association	0.088	0.141	0	1	0.107	0.093
Time variant control measures
Lagged cyberbullying	0.409	0.958	0	6.905	0.615	0.740
Parental abuse	3.255	1.635	2	10	1.207	1.103
Parental monitoring	13.639	3.314	4	20	2.584	2.075
Parental attachment	6.677	1.849	2	10	1.467	1.127
Peer attachment	16.935	2.538	4	20	1.836	1.756
Time invariant control measures
Gender	0.503	0.500	0	1
Education level: Father	3.677	1.282	0	7
Education level: Mother	3.180	1.066	0	7

SD = standard deviation.

Measures

Dependent variable (waves 2–5)

Cyberbullying was measured by two items (Choi & Kruis, 2019; Kim et al., 2017; Song et al., 2020): (1) How many times have you intentionally circulated false information on the Internet message boards about others during the last year? and (2) How many times have you ever cursed/insulted other people through chats/message boards during the last year? As the responses were open-ended, frequencies of these two items were summed and the natural logarithmic transformation was used to deal with skewed responses. Pearson’s correlation coefficient between two items of cyberbullying was .048, which meets the criterion (r = .30–.70) suggested by Nunnally & Bernstein (1994).

Independent variables (waves 1–4)

Traditional bullying was measured by three items (Cho et al., 2017; Choi & Kruis, 2019; Kim et al., 2017): (1) How many times have you severely teased or bantered other people? (2) How many times have you threatened other people, and (3) How many times have you treated other people as an outcast? As the responses were open-ended, frequencies of these three items were summed and the natural logarithmic transformation was used to deal with skewed responses.

Self-control was composed of six items measured on a 5-point Likert scale ranging from 1 (strongly disagree) through 5 (strongly agree) (Kim & Lee, 2019). Items tapped into five of the dimensions of self-control identified by Gottfredson and Hirschi (1990): impulsivity, avoidance of difficult tasks for simple tasks, risk-taking, self-centeredness, and short temper (e.g., “I enjoy risky activities” and “I lose my temper whenever I get angry”). Higher scores indicated lower self-control (α = .629).

Delinquent peer association was estimated as the ratio of delinquent peers among their best friends, which took three steps. The first question asked to report their total number of best friends. Second, respondents were then asked to report the number of best friends who drank alcohol, smoked, were truant, or had engaged in assault, robbery, and/or larceny, separately. Third, based on the two items, we divided the number of delinquent friends by the total number of friends (α = .69). That is, the response reflects the proportion of deviant friends among their best friends.

Time variant control measures (waves 1–4)

Lagged cyberbullying refers to previous cyberbullying. For example, when estimating cyberbullying at time t, cyberbullying at t-1 was included in the estimating model as a control variable. This procedure was repeated across waves. Lagged cyberbullying was measured in the same manner as the dependent variable as described above.

Parental abuse was measured by two items based on a 5-point Likert scale ranging from 1 (strongly disagree) through 5 (strongly agree): (1) I often receive verbal abuse from my parents and (2) I am often badly hit by my parents. Higher values presented greater parental abuse (α = .752).

Parental monitoring was composed of four items measured on a 5-point Likert scale ranging from 1 (strongly disagree) through 5 (strongly agree): When I go out, my parents usually know (1) where I am, (2) whom I am with, (3) what I am doing, and (4) when I will return. Higher values indicated greater parental monitoring (α = .844).

Parental attachment was composed of two items measured on a 5-point Likert scale ranging from 1 (strongly disagree) through 5 (strongly agree): (1) I am comfortable sharing my thoughts and feelings with my parents and (2) I often talk about what happens to me outside the home. Higher values presented greater parental attachment (α = .723).

Peer attachment was measured by four items based on a 5-point Likert scale ranging from 1 (strongly disagree) through 5 (strongly agree): (1) I want to maintain the friendships I have with my close friends, (2) I am happy with my friends, (3) I try to have same thoughts and feelings as my friends, and (4) I have candid conversations with my friends. Higher values indicated greater peer attachment (α = .753).

Time-invariant control measures (wave 1)

Gender was coded 0 for females and 1 for males. Parental education level was measured for father and mother, respectively. The response was categorized into seven levels ranging from 0 = elementary school, 1 = middle school … 6 = master’s degree, and 7 = doctoral degree.

Analytic Strategy

Cross-lagged dynamic panel models are employed to examine the effects of traditional bullying, low self-control, and delinquent peer association on cyberbullying. Cross-lagged dynamic panel models are suitable for the purpose of this study since the models permit the ability to consider time-variant and time-invariant covariates over time. In other words, cross-lagged dynamic panel models enable to estimate how the independent variables influence the outcome behavior over time while controlling various correlates. Another key advantage is to include lagged measures of the dependent variable as a control variable, which distinguishes cross-lagged dynamic panel models from fixed- or mixed-effects models that are also used in the estimation of time-varying and time-invariant factors (Allison et al., 2017; Williams et al., 2018). Considering “delinquent behavior is determined largely by previous delinquent behavior” (Gottfredson & Hirschi, 1987; Matsueda, 1986), the inclusion of previous dependent variables in the models enables a more rigorous analysis.

The analytic strategy for the present study is conducted in several steps. First, the descriptive statistics are reported for all dependent, independent, and control variables. Second, we estimate the effects of traditional bullying on cyberbullying with the time-variant and time-invariant measures including lagged cyberbullying. Third, low self-control and delinquent peer association are included in the analyses with traditional bullying and the control variables, respectively. Fourth, we estimate the effects of all independent variables (i.e., traditional bullying, low self-control, and delinquent peer association) on cyberbullying net of time-variant and time-invariant measures. Fifth, two interaction effects (i.e., traditional bullying x low self-control and traditional bullying x delinquent peer association) are examined, respectively. Missing values were handled via full-information maximum likelihood estimation. All analyses were performed in Stata 15.

Results

Table 1 displays descriptive statistics including mean and standard deviation (SD) across five waves. We also provided SD (between) and SD (within) since the estimations were repeated over time. The overall mean of the dependent variable, cyberbullying is .265 and a SD is .835 with a range from 0 to 7168. Its SD (between) and SD (within) are .543 and .635, respectively, which reflects changes in cyberbullying across waves. Regarding traditional bullying, the mean is .087 with a SD of .473 (range from 0 to 6.905). SD (between) of traditional bullying is .297 and SD (within) is .368, indicating the varying levels of traditional bullying over time.

Results of the cross-lagged dynamic panel model assessing cyberbullying are shown in Table 2. Overall, the fit indices represent a good model fit (RMSEA below .05 and CFI above .95).¹ Following Allison and colleagues’ advice (2017), we compared the estimated models with saturated models, where higher p values are indicative of being better fitting models. Model 1 includes only one independent variable, traditional bullying with time-variant and time-invariant control variables, and shows that engaging in traditional bullying is related to a .304 point increase in cyberbullying across time. Models 2 and 3 demonstrate that low self-control and delinquent peer association are associated with a .019 point increase and .343 point increase in cyberbullying scale across time, respectively. When the three independent variables are included in Model 4, traditional bullying is associated with a .290 point increase, low self-control is related to a .019 point increase, and delinquent peer association is associated with a .310 point increase in cyberbullying across time. Among the control variables, lagged cyberbullying is associated with .095, .096, .094, and .095 point increase in cyberbullying from Models 1 to 4. Parental abuse and gender are significantly associated with higher levels of cyberbullying.

Table 2.

Cross-Lagged Dynamic Panel Data Models Assessing Cyberbullying.

	Model 1			Model 2			Model 3			Model 4
	Coef	SE	p	Coef	SE	p	Coef	SE	p	Coef	SE	p
Independent variables
Traditional bullying	0.304	0.025	***	0.297	0.025	***	0.296	0.024	***	0.289	0.024	***
Low self-control				0.019	0.005	***				0.019	0.005	***
Delinquent peer association							0.343	0.130	**	0.310	0.127	*
Time variant control measures
Lagged cyberbullying	0.095	0.012	***	0.096	0.012	***	0.094	0.012	***	0.095	0.012	***
Parental abuse	0.026	0.010	**	0.020	0.010	*	0.025	0.010	*	0.018	0.010
Parental monitoring	−0.010	0.006		−0.009	0.006		−0.009	0.006		−0.008	0.006
Parental attachment	−0.007	0.011		−0.007	0.011		−0.008	0.011		−0.007	0.011
Peer attachment	0.003	0.006		0.002	0.006		0.004	0.006		0.003	0.006
Time invariant control measures
Gender	0.163	0.020	***	0.157	0.020	***	0.153	0.020	***	0.148	0.020	***
Education level: Father	0.008	0.009		0.010	0.009		0.010	0.009		0.011	0.009
Education level: Mother	−0.011	0.011		−0.011	0.011		−0.011	0.011		−0.010	0.011
Fit indices
χ2 (model vs. Saturated)	61.500*			65.311*			74.231**			80.201**
RMSEA	0.014			0.012			0.015			0.015
CFI	0.989			0.990			0.985			0.985

Coef. = coefficient; SE = standard error; RMSEA = root mean square error of approximation; CFI = comparative fit index.

*p < .05; **p < .01; ***p < .001.

Table 3 presents the results of the cross-lagged dynamic panel model assessing potential low self-control and/or delinquent peer association and traditional bullying interaction effects on cyberbullying. The fit indices suggest that the model fits the data well (RMSEA below .05 and CFI above .95). Model 1 illustrates the positive and significant interaction effect of low self-control and traditional bullying on cyberbullying, which suggests that the two factors together lead to significant increases in cyberbullying. In contrast, Model 2 fails to identify a significant interaction effect for delinquent peer association and traditional bullying on cyberbullying. Similar to the results displayed in Table 2, lagged cyberbullying is related to a .094 and .093 increase in cyberbullying in Models 1 and 2, respectively. Males report .150 points higher for cyberbullying than females in both Models 1 and 2.

Table 3.

Cross-Lagged Dynamic Panel Data Models Assessing Interactions on Cyberbullying.

	Model 1			Model 2
	Coef	SE	p	Coef	SE	p
Independent variables
Traditional bullying	0.033	0.105		0.264	0.032	***
Low self-control	0.018	0.005	***	0.019	0.005	***
Delinquent peer association	0.304	0.127	*	0.287	0.131	*
Interactions
Traditional bullying × low self-control	0.013	0.005	*
Traditional bullying × delinquent peer association				0.086	0.089
Time variant control measures
Lagged cyberbullying	0.094	0.012	***	0.093	0.012	***
Parental abuse	0.017	0.010		0.019	0.010
Parental monitoring	−0.008	0.006		−0.008	0.006
Parental attachment	−0.008	0.011		−0.007	0.011
Peer attachment	0.003	0.006		0.003	0.006
Time invariant control measures
Gender	0.150	0.020	***	0.150	0.020	***
Education level: Father	0.011	0.009		0.011	0.009
Education level: Mother	−0.010	0.011		−0.011	0.011
Fit indices
χ2 (model vs. saturated)	91.562**			106.017***
RMSEA	0.015			0.018
CFI	0.982			0.975

Coef. = coefficient; SE = standard error; RMSEA = root mean square error of approximation; CFI = comparative fit index.

*p < .05; **p < .01; ***p < .001.

Discussion

The current study provided an examination of the linkages between traditional bullying and cyberbullying informed by low self-control and social learning theoretical perspectives. Relying on cross-lagged dynamic panel models estimated with a representative longitudinal sample of Korean youth, several key findings emerged that are briefly highlighted below.

Consistent with prior research (Athanasiades et al., 2016; Donner et al., 2015; Fanti et al., 2012; Jose et al., 2012; Kim et al., 2017; Hemphill et al., 2012), the analysis documented a significant association between the two different forms of bullying, that is, traditional bullying was predictive of cyberbullying. Furthermore, both low self-control and delinquent peer association were predictive of cyberbullying alongside traditional bullying. Subsequent analysis upheld these robust linkages, as these initial effects still maintained their significance once estimated simultaneously and net of the effects of time variant and invariant control variables. In addition, a significant interaction effect between traditional bullying and self-control was observed, but this same interactive effect was not found for traditional bullying and delinquent peer association. The latter was not an expected finding, and may suggest that self-control exerts a greater influence on cyberbullying in particular contexts, that is, in the presence of traditional bullying behaviors. In contrast, it appears that delinquent peer association operates only directly in terms of its effect on cyberbullying and its effect is not significantly amplified in the presence of traditional bullying behaviors. These findings overall fit within the larger literature about cyberbullying as prior systematic reviews have documented that cyberbullying can be mediated or moderated by demographic and social factors, and that the interrelationship between theoretically informed risk factors and cyberbullying can be used to inform intervention components (Kwan et al., 2020).

These results have several theoretical and policy implications. Regarding the theoretical implications, the findings suggest that both the low self-control (Gottfredson & Hirschi, 1990) and delinquent peer association (i.e., social learning) (Akers et al., 2020) theoretical perspectives have validity as explanations of cyberbullying behaviors. Therefore, future cyberbullying research that fails to incorporate these theoretical perspectives into their empirical investigations run the risk of being mis-specified. Also, the results contribute to extending the cross-cultural generalizability of both theoretical perspectives for predicting cyberbullying in non-Western samples of youth. The robust linkage that was observed between traditional bullying and cyberbullying yields additional evidence suggesting that some bullies may just be bullies, and they often do not discriminate between or specialize in one particular mode of bullying. Essentially, both forms of bullying are forms of the same underlying problem behavior syndrome for some, but certainly not for all. This has implications for future bullying research as it indicates that perhaps these constructs (i.e., traditional bullying and cyberbullying) are not as statistically or conceptually distinct constructs as previously thought. As such, future traditional bullying research should incorporate cyberbullying items into their measurement and vice versa, and this research should rely on similar theoretical perspectives and risk factors as predictors of both of these outcomes. These findings also highlight the importance of investing in early life-course prevention/intervention programs and policies to prevent and/or reduce the occurrence of bullying, regardless of whether it is being perpetrated face-to-face or online, and these programs and policies should also target components to improve self-control and reduce delinquent peer associations (Forgatch et al., 2009; Forgatch & Kjøbli, 2016; Piquero et al., 2016a, 2016b). In this same vein, these findings combined with prior evidence that the risk factors for involvement in crime and cyberbullying or online and offline/traditional offending are fairly similar (see Donner et al., 2015) indicate that prevention programs that target delinquency in general may also be sufficient and appropriate to reduce the occurrence of cyberbullying specifically rather than having a standalone cyberbullying prevention program. Nevertheless, the extant literature has also identified promising components for bullying intervention in South Korea youth specifically including: peer/teacher support (Lee et al., 2020), school interest in the issue of bullying, a reporting and monitoring system in the schools, and education and help-seeking (Chun et al., 2021).

Notwithstanding these implications, several limitations are worth noting. First, these data are from Korean youth, and therefore the extent to which these findings may generalize to other Asian and non-Western samples is subject to future empirical investigations. Second, as bullying perpetration is just one specific form of deviance, future research is encouraged to account for involvement in other delinquent behaviors when examining traditional bullying and cyberbullying when data permit. Third, given the well documented overlap between victimization and offending, future research should attempt to capture the bullying victimization experiences alongside the bullying perpetration experiences when further testing the low self-control and social learning theoretical perspectives as factors in the etiology of bullying and cyberbullying. Ultimately, these criminological perspectives exhibit promise in generalizing to explain cyberbullying and do so in an international, non-Western (Korean) context. Future studies should continue this effort to take these theories “global” as bullying in general and cyberbullying specifically are indeed global issues. Fourth, although we employ a rigorous analytical approach, our data presented several limitations in regards to measurement. Our bullying measures were unable to reflect the power imbalance that is often regarded as one aspect of bullying. While the bullying measures we used have been widely used in bullying literature, future work might wish to consider expanded measures. In addition, a few measures (e.g., self-control) had a reliability of less than .7. Further research relying on measures with stronger reliability could help enhance the generalizability of this study. Despite the study limitations, the current study enhances not only our understanding of how cyberbullying is evolved but also how key criminological factors influence the link between different types of bullying. We also provided empirical evidence of traditional criminological theories in a new type of delinquent behavior in a digital world. Ultimately, the unique contribution of the current study lies in its incorporation of multiple theoretical perspectives and time variant and invariant factors, application of advanced statistical methodologies, examination of two overlapping forms of bullying, and doing so in an international context, that is, South Korea.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Jihoon Kim

Notes

Author Biographies

Jihoon Kim is an assistant professor in the Department of Criminology and Criminal Justice at the University of Alabama. His research interests lie in the area of juvenile delinquency, life-course/developmental theory, quantitative methods, and criminal justice policies. His recent publications appear in Journal of Criminal Justice, Crime & Delinquency, and Journal of Research in Crime and Delinquency.

Yeungjeom Lee is an assistant professor in the School of Economic, Political and Policy Sciences at the University of Texas at Dallas. Her primary interests include life-course/developmental criminology, victimization, psychopathy, and juvenile delinquency.

Wesley Jennings, PhD, is Chair and Professor in the Department of Legal Studies at the University of Mississippi. He also has a Courtesy Appointment in the Department of Health Outcomes & Policy and is a Faculty Affiliate of the Institute for Child Health Policy in the College of Medicine at the University of Florida. He received his doctorate degree in Criminology from the University of Florida.

References

Akers

R. L

. (2011). Social learning and social structure: A general theory of crime and deviance. Transaction Publishers.

Akers

R. L.

Sellers

Jennings

W. G.

(2020). Criminological theories: Introduction, evaluation, and application (8th ed.). Oxford University Press.

Allison

P. D.

Williams

Moral-Benito

(2017). Maximum likelihood for cross-lagged panel models with fixed effects. Socius: Sociological Research for a Dynamic World, 3(1), 1–17. https://doi.org/10.1177/2378023117710578

Athanasiades

Baldry

A. C.

Kamariotis

Kostouli

Psalti

(2016). The “net” of the Internet: Risk factors for cyberbullying among secondary-school students in Greece. European Journal on Criminal Policy and Research, 22(2), 301–317. https://doi.org/10.1007/s10610-016-9303-4

Benda

B. B.

(2005). The robustness of self-control in relation to form of delinquency. Youth & Society, 36(4), 418–444. https://doi.org/10.1177/0044118X04268071

Beran

(2008). The relationship between cyberbullying and school bullying. Journal of Student Wellbeing, 1(2), 15–33. https://doi.org/10.21913/jsw.v1i2.172

Bottino

S. M. B.

Bottino

C. M. C.

Regina

C. G.

Correia

A. V. L.

Ribeiro

W. S.

(2015). Cyberbullying and adolescent mental health: Systematic review. Cadernos de Saude Publica, 31(3), 463–475. https://doi.org/10.1590/0102-311X00036114

Burt

C. H.

Sweeten

Simons

R. L.

(2014). Self-control through emerging adulthood: Instability, multidimensionality, and criminological significance. Criminology, 52(3), 450–487. https://doi.org/10.1111/1745-9125.12045

Chang

F.-C.

Lee

C.-M.

Chiu

C.-H.

Hsi

W.-Y.

Huang

T.-F.

Pan

Y.-C.

(2013). Relationships among cyberbullying, school bullying, and mental health in Taiwanese adolescents. Journal of School Health, 83(6), 454–462. https://doi.org/10.1111/josh.12050

10.

Chapple

C. L.

(2005). Self-control, peer relations, and delinquency. Justice Quarterly, 22(1), 89–106. https://doi.org/10.1080/0741882042000333654

11.

Cho

Hong

J. S.

Sterzing

P. R.

Woo

(2017). Parental attachment and bullying in South Korean adolescents: Mediating effects of low self-control, deviant peer associations, and delinquency. Crime and Delinquency, 63(9), 1168–1188. https://doi.org/10.1177/0011128717714968

12.

Choi

Kruis

N. E.

(2019). The effects of life domains on cyberbullying and bullying: Testing the generalizability of Agnew’s integrated general theory. Crime and Delinquency, 65(6), 772–800. https://doi.org/10.1177/0011128718814860

13.

Chun

Lee

Kim

(2021). Conceptualizing the protective factors of cyberbullying victimization in Korean adolescents. School Mental Health, 13(3), 473–486. http://doi.org/10.1007/s12310-021-09422-0

14.

Clinkinbeard

S. S.

Barnum

T. C.

Rhodes

T. N.

(2018). The other side of the coin: Exploring the effects of adolescent delinquency on young adult self-control. Journal of Criminal Justice, 56, 86-97. https://doi.org/10.1016/j.jcrimjus.2017.08.001.

15.

de Kemp

R. A. T.

Vermulst

A. A.

Finkenauer

Scholte

R. H. J.

Overbeek

Rommes

E. W. M.

Engels

R. C. M. E.

(2009). Self-control and early adolescent antisocial behavior: A longitudinal analysis. Journal of Early Adolescence, 29(4), 497–517. https://doi.org/10.1177/0272431608324474

16.

Donner

C. M.

Jennings

W. G.

Banfield

(2015). The general nature of online and off-line offending among college students. Social Science Computer Review, 33(6), 663–679. https://doi.org/10.1177/0894439314555949

17.

Eisenberg

M. E.

Neumark-Sztainer

Perry

C. L.

(2003). Peer harassment, school connectedness, and academic achievement. Journal of School Health, 73(8), 311–316. https://doi.org/10.1107/S0907444999000839

18.

Espelage

D. L.

Bosworth

Simon

T. R.

(2000). Examining the social context of bullying behaviors in early adolescence. Journal of Counseling and Development, 78(3), 78–85. https://doi.org/10.1002/j.1556-6676.2000.tb01914.x

19.

Fahy

A. E.

Stansfeld

S. A.

Smuk

Smith

N. R.

Cummins

Clark

(2016). Longitudinal associations between cyberbullying involvement and adolescent mental health. Journal of Adolescent Health, 59(5), 502–509. https://doi.org/10.1016/j.jadohealth.2016.06.006

20.

Fanti

K. A.

Demetriou

A. G.

Hawa

V. V.

(2012). A longitudinal study of cyberbullying: Examining risk and protective factors. European Journal of Developmental Psychology, 9(2), 168–181. https://doi.org/10.1080/17405629.2011.643169

21.

Fekkes

Pijpers

F. I. M.

Fredriks

A. M.

Vogels

Verloove-Vanhorick

S. P.

(2006). Do bullied children get ill, or do ill children get bullied? A prospective cohort study on the relationship between bullying and health-related symptoms. Pediatrics, 117(5), 1568–1574. http://doi.org/10.1542/peds.2005-0187

22.

Fekkes

Pijpers

F. I. M.

Verloove-Vanhorick

S. P.

(2004). Bullying behavior and associations with psychosomatic complaints and depression in victims. Journal of Pediatrics, 144(1), 17–22. https://doi.org/10.1016/j.jpeds.2003.09.025

23.

Forgatch

M. S.

Kjøbli

(2016). Parent management training—Oregon model: Adapting intervention with rigorous research. Family Process, 55(3), 500–513. https://doi.org/10.1111/famp.12224

24.

Forgatch

M. S.

Patterson

G. R.

Degarmo

D. S.

Beldavs

Z. G.

(2009). Testing the Oregon delinquency model with 9-year follow-up of the Oregon Divorce Study. Development and Psychopathology, 21(2), 637–660. https://doi.org/10.1017/S0954579409000340

25.

Gini

Pozzoli

(2009). Association between bullying and psychosomatic problems: A meta-analysis. Pediatrics, 123(3), 1059–1065. https://doi.org/10.1542/peds.2008-1215

26.

Gottfredson

M. R.

Hirschi

(1987). The methodological adequacy of longitudinal research on crime. Criminology, 25(3), 581–614. https://doi.org/10.1111/j.1745-9125.1987.tb00812.x

27.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. Stanford University Press.

28.

Hay

Forrest

(2006). The development of self-control: Examining self-control theory’s stability thesis. Criminology, 44(4), 739–774. http://doi.org/10.1111/j.1745-9125.2006.00062.x

29.

Hemphill

S. A.

Kotevski

Tollit

Smith

Herrenkohl

T. I.

Toumbourou

J. W.

Catalano

R. F.

(2012). Longitudinal predictors of cyber and traditional bullying perpetration in Australian secondary school students. Journal of Adolescent Health, 51(1), 59–65. https://doi.org/10.1016/j.jadohealth.2011.11.019

30.

Hirschi

(1969). Causes of delinquency. University of California.

31.

Hoeben

E. M.

Meldrum

R. C.

Walker

Young

J. T. N.

(2016). The role of peer delinquency and unstructured socializing in explaining delinquency and substance use: A state-of-the-art review. Journal of Criminal Justice, 47, 108-122. https://doi.org/10.1016/j.jcrimjus.2016.08.001

32.

Holt

T. J.

Bossler

A. M.

May

D. C.

(2012). Low self-control, deviant peer associations, and juvenile cyberdeviance. American Journal of Criminal Justice, 37(3), 378–395. https://doi.org/10.1007/s12103-011-9117-3

33.

Jennings

W. G.

Song

Kim

Fenimore

D. M.

Piquero

A. R.

(2019). An examination of bullying and physical health problems in adolescence among South Korean youth. Journal of Child and Family Studies, 28(9), 2510–2521. https://doi.org/10.1007/s10826-017-0885-3

34.

Jose

P. E.

Kljakovic

Scheib

Notter

(2012). The joint development of traditional bullying and victimization with cyber bullying and bictimization in adolescence. Journal of Research on Adolescence, 22(2), 301–309. https://doi.org/10.1111/j.1532-7795.2011.00764.x

35.

Khey

D. N.

Jennings

W. G.

Lanza‐Kaduce

Frazier

C. E.

(2009). An exploration into the factors associated with specialization among college student computer criminals. Criminal Justice Studies, 22(4), 421–434. https://doi.org/10.1080/14786010903358166

36.

Kim

Lee

(2019). Does it take a school? Revisiting the tnfluence of first arrest on subsequent delinquency and educational attainment in a tolerant educational background. Journal of Research in Crime and Delinquency, 56(2), 254–302. https://doi.org/10.1177/0022427818801053

37.

Kim

Song

Jennings

W. G.

(2017). A distinct form of deviance or a variation of bullying? Examining the developmental pathways and motives of cyberbullying compared with traditional bullying in South Korea. Crime and Delinquency, 63(12), 1600–1625. https://doi.org/10.1177/0011128716675358

38.

Kowalski

R. M.

Giumetti

G. W.

Schroeder

A. N.

Lattanner

M. R.

(2014). Bullying in the digital age: A critical review and meta-analysis of cyberbullying research among youth. Psychological Bulletin, 140(4), 1073. http://doi.org/10.1037/a0035618

39.

Kowalski

R. M.

Limber

S. P.

(2013). Psychological, physical, and academic correlates of cyberbullying and traditional bullying. Journal of Adolescent Health, 53(Suppl 1), S13–S20. https://doi.org/10.1016/j.jadohealth.2012.09.018

40.

Kowalski

R. M.

Limber

S. P.

Agatston

P. W.

(2012). Cyberbullying: Bullying in the digital age (2nd ed.). Wiley-Blackwell.

41.

Kwan

Dickson

Richardson

MacDowall

Burchett

Stansfield

Thomas

(2020). Cyberbullying and children and young people’s mental health: A systematic map of systematic reviews. Cyberpsychology, Behavior, and Social Networking, 23(2), 72–82. http://doi.org/10.1089/cyber.2019.0370

42.

Lee

(2014). Youth research brief: Cyberbullying among adolescents. National Youth Policy Institute.

43.

Lee

Chun

Kim

Lee

(2020). Cyberbullying victimisation and school dropout intention among South Korean adolescents: The moderating role of peer/teacher support. Asia Pacific Journal of Social Work and Development, 30(3), 195–211. http://doi.org/10.1080/02185385.2020.1774409

44.

Lee

Hong

J. S.

Resko

S. M.

Tripodi

S. J.

(2018). Face-to-face bullying, cyberbullying, and multiple forms of substance use among school-age adolescents in the USA. School Mental Health, 10(1), 12–25. https://doi.org/10.1007/s12310-017-9231-6

45.

Lee

Kim

(2017). Developmental patterns of substance use by gender and their relation to individual, parental, and peer factors. Criminal Justice and Behavior, 44(11), 1413–1443. https://doi.org/10.1177/0093854817717291

46.

Matsueda

(1986). The dynamics of belief and delinquency. Presented at the annual meetings of the American Sociological AssociationStates, September 1st, New York City, NY, United States, September 1st, 1986.

47.

McGloin

J. M.

Thomas

K. J.

(2019). Peer influence and delinquency. Annual Review of Criminology, 2(1), 241–264. https://doi.org/10.1146/annurev-criminol-011518-024551

48.

Paternoster

(2012). Can self-control change substantially over time? Rethinking the relationship between self- and social control. Criminology, 50(2), 427–462. https://doi.org/10.1111/j.1745-9125.2011.00269.x

49.

Nansel

T. R.

Overpeck

Pilla

R. S.

Ruan

W. J.

Simons-Morton

Scheidt

(2001). Bullying behaviors among US youth: Prevalence and association with psychosocial adjustment. Journal of the American Medical Association, 285(16), 2094–2100. https://doi.org/10.1001/jama.285.16.2094

50.

Nunnally

J. C.

Bernstein

I. H.

(1994). Psychometric theory (3rd ed.). New York: McGraw-Hill.

51.

Olweus

(1997). Bully/victim problems in school: Facts and intervention. European Journal of Psychology of Education, XlI(4), 495–510. https://doi.org/10.1201/9780203164662.ch3

52.

Patchin

J. W.

Hinduja

(2006). Bullies move beyond the schoolyard: A preliminary look at cyberbullying. Youth Violence and Juvenile Justice, 4(2), 148–169. https://doi.org/10.1177/1541204006286288

53.

Piquero

A. R.

Jennings

W. G.

Diamond

Farrington

D. P.

Tremblay

R. E.

Welsh

B. C.

Gonzalez

J. M. R.

(2016a). A meta-analysis update on the effects of early family/parent training programs on antisocial behavior and delinquency. Journal of Experimental Criminology, 12(2), 229–248. https://doi.org/10.1007/s11292-016-9256-0

54.

Piquero

A. R.

Jennings

W. G.

Farrington

D. P.

Diamond

Gonzalez

J. M. R.

(2016b). A meta-analysis update on the effectiveness of early self-control improvement programs to improve self-control and reduce delinquency. Journal of Experimental Criminology, 12(2), 249–264. https://doi.org/10.1007/s11292-016-9257-z

55.

Pratt

T. C.

(2016). A self-control/life-course theory of criminal behavior. European Journal of Criminology, 13(1), 129–146. https://doi.org/10.1177/1477370815587771

56.

Pratt

T. C.

Cullen

F. T.

Sellers

C. S.

Thomas Winfree

Madensen

T. D.

Daigle

L. E.

Fearn

N. E.

Gau

J. M.

(2010). The empirical status of social learning theory: A meta-analysis. Justice Quarterly, 27(6), 765–802. https://doi.org/10.1080/07418820903379610

57.

Sasson

Mesch

(2017). The role of parental mediation and peer norms on the likelihood of cyberbullying. Journal of Genetic Psychology, 178(1), 15–27. https://doi.org/10.1080/00221325.2016.1195330

58.

Smith

P. K.

Talamelli

Cowie

Naylor

Chauhan

(2004). Profiles of non-victims, escaped victims, continuing victims and new victims of school bullying. British Journal of Educational Psychology, 74(4), 565–581. https://doi.org/10.1348/0007099042376427

59.

Song

Lee

Kim

(2020). Gender differences in the link between cyberbullying and parental supervision trajectories. Crime and Delinquency, 66(13–14), 1914–1936. https://doi.org/10.1177/0011128720912371

60.

Thomas

H. J.

Connor

J. P.

Scott

J. G.

(2015). Integrating traditional bullying and cyberbullying: Challenges of definition and measurement in adolescents – a Review. Educational Psychology Review, 27(1), 135–152. https://doi.org/10.1007/s10648-014-9261-7

61.

Vazsonyi

A. T.

Jiskrova

G. K.

(2018). On the development of self-control and deviance from preschool to middle. Journal of Criminal Justice, 56, 60-69. https://doi.org/10.1016/j.jcrimjus.2017.08.005

62.

Vazsonyi

A. T.

Machackova

Sevcikova

Smahel

Cerna

(2012). Cyberbullying in context: Direct and indirect effects by low self-control across 25 European countries. European Journal of Developmental Psychology, 9(2), 210–227. https://doi.org/10.1080/17405629.2011.644919

63.

Vazsonyi

A. T.

Mikuška

Kelley

E. L.

(2017). It’s time: A meta-analysis on the self-control-deviance link. Journal of Criminal Justice, 48, 48-63. https://doi.org/10.1016/j.jcrimjus.2016.10.001

64.

Warr

(2002). Companions in crime: The social aspects of criminal conduct. Cambridge University Press.

65.

Williams

Allison

P. D.

Moral-Benito

(2018). Linear dynamic panel-data estimation using maximum likelihood and structural equation modeling. Stata Journal, 18(2), 293–326. https://doi.org/10.1177/1536867x1801800201

66.

Ybarra

M. L.

Diener-West

Leaf

P. J.

(2007). Examining the overlap in internet harassment and school bullying: Implications for school intervention. Journal of Adolescent Health, 41(6), 42–50. https://doi.org/10.1016/j.jadohealth.2007.09.004

67.

You

Lim

S. A.

(2016). Longitudinal predictors of cyberbullying perpetration: Evidence from Korean middle school students. Personality and Individual Differences, 89, 172-176. https://doi.org/10.1016/j.paid.2015.10.019