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

Abstract

Cyberbullying has been subject to a debate about whether it is a subtype of traditional bullying or a distinct deviant behavior from traditional bullying. Applying a longitudinal South Korean youth sample and latent group-based trajectory modeling, the current study examines (a) an overlap of developmental trajectories between cyberbullying and traditional bullying, and (b) effects of predictors on developmental trajectory groups for both cyberbullying and traditional bullying. It is concluded that cyberbullying is close to a variation of bullying rather than a distinct deviant behavior and reported an overlap of developmental trajectories between cyberbullying and traditional bullying, and strong associations between both forms of bullying and peer-related predictors. Policy implications and suggestions for future research are discussed.

Keywords

cyberbullying bullying predictors overlap joint trajectory modeling

Introduction

Within the past few decades, the Internet emerged as a new form of communication and source of information, revolutionizing modern society. Currently, the majority of American adults, approximately 73%, access the Internet at least once a day (Pew Research Center, 2015). For adolescents, the numbers are much greater. According to recent surveys, 95% of American youth report accessing the Internet (Pew Research Center, 2012), and they reportedly spend an average of 4.5 hr daily on the Internet (Common Sense, 2015).In a consequence of the increased time spent involved in Internet-based activity, a new form of deviant behavior emerged, that is, cyberbullying.

Specifically, cyberbullying broadly includes harassment, threats, and other offensive behaviors via the Internet (Tokunaga, 2010). According to the Youth Internet Safety Survey, the proportions of youth who have experienced online harassment increased from 6% in 2000 to 9% in 2005, and to 11% in 2010 (Jones, Mitchell, & Finkelhor, 2013). Furthermore, cyberbullying victimization has been shown to be associated with a number of adverse emotional and behavioral consequences for youth (Patchin & Hinduja, 2010; Tokunaga, 2010). For instance, studies have found that cyberbullying victimization is related to both externalizing delinquent behaviors (Hinduja & Patchin, 2007; Ybarra & Mitchell, 2004a) and internalizing forms of deviance such as depression, anxiety, self-harm, and suicidal thoughts (Hay, Meldrum, & Mann, 2010; Hinduja & Patchin, 2010; Kowalski & Limber, 2013; Wang, Nansel, & Iannotti, 2011; Ybarra & Mitchell, 2004a).

In recognition of these issues, it is essential to gain an understanding of the nature of cyberbullying to set the foundation for prevention strategies and policies. While studies on cyberbullying have been increasingly published in peer-reviewed journals, the developmental nature of cyberbullying has not been thoroughly examined. There has been a debate as to whether cyberbullying is a variation of or a distinct form of bullying (Bauman, 2013; Dooley, Pyżalski, & Cross, 2009; Law, Shapka, Hymel, Olson, & Waterhouse, 2012). However, previous work has been limited in regard to applying a rigorous methodological approach. Cyberbullying is novel in that it occurs specifically on the Internet. Given this unique feature, a comprehensive comparison of cyberbullying with traditional bullying is necessary to understand the developmental nature of this relatively new form of deviance.

To begin, we discuss relationships between traditional bullying and cyberbullying in terms of whether cyberbullying is a variation of bullying behavior or a distinct form of bullying. We introduce literature on definitions of bullying, generality of deviance, and overlaps and shared predictors between traditional bullying and cyberbullying. In addition, probable heterogeneities of age and opportunity variables between cyberbullying and traditional bullying are discussed. Finally, we analyze longitudinal data from the Korean Youth Panel Survey (KYPS) to assess the developmental nature of cyberbullying. Specifically, this study examines whether there are distinguishable longitudinal patterns of cyberbullying, including the exploration of possible risk and protective factors for cyberbullying trajectories, as well as investigates the possible overlap in the trajectories of cyberbullying and traditional bullying.

Literature Overview

Cyberbullying as a Variation of Bullying

Cyberbullying has recently emerged as a type of deviant behavior that shares similar characteristics with traditional bullying, such as an offender perpetrated aggression, although this behavior is unique in that it does not require the physical presence of the offender (Bauman, 2013; Dooley et al., 2009; Kowalski, Giumetti, Schroeder, & Lattanner, 2014). As the definition of bullying contains a broad range of aggressive behaviors including physical harassment as well as non-physical verbal aggression (Dooley et al., 2009; Espelage & Swearer, 2003), some scholars regard cyberbullying as one of the means of bullying rather than a discrete form of deviance (Bauman & Newman, 2013; Salmivalli, Karna, & Poskiparta, 2011; Vandebosch & Van Cleemput, 2009).

In addition to the definitional differences, the generality of deviance characterized both bullying behavior illustrating that they may have similar primary causes. Arguing for the trait low self-control as a general cause of delinquency, Gottfredson and Hirschi (1990) claimed that all forms of deviance could be explained by an individual’s propensity to seek immediate benefits or pleasure rather than consider long-term negative consequences of their behaviors, referred to as low self-control. The colleagues also emphasized that individuals’ variations in self-control could account for many other noncriminal behaviors analogous to crime, such as smoking and alcohol use. That is, various types of deviance and analogous behaviors are manifestations of offenders’ low self-control, and individuals with low self-control are more likely to commit a general variety of deviant behaviors, regardless of variations in purposes, targets, and seriousness of the deviance.

Evidence supporting the notion of the generality of deviance has documented invariant age effects on all forms of crime and analogous acts (Gottfredson & Hirschi, 1990; Hirschi & Gottfredson, 1994). As all types of crime tend to peak among middle to late teens and drop rapidly and steadily throughout life, regardless of social and demographic variations (Hirschi & Gottfredson, 1983), this universality can lead us to extrapolate that there is a common cause of deviance which should be a stable individual trait. That is, age distributions should appear heterogeneous if the social and demographical differences have meaningful effects on crime and if there are specific characteristics across types of crime. Based on the homogeneity of age distribution, Gottfredson and Hirschi (1990) concluded that the stable individual trait of low self-control would lead offenders to all types of crime and analogous behaviors in general.

Overall, the generality hypothesis has been supported by many empirical studies (Dembo et al., 1992; Jennings, Zgoba, Donner, Henderson, & Tewksbury, 2014; Klein, 1984; LeBlanc & Girard, 1997; Osgood, Johnston, O’Malley, & Bachman, 1988; Piquero, 2000; Zhang, Welte, & Wieczorek, 2002). Regarding online contexts, Donner, Jennings, and Banfield (2015) examined the generality of online and offline offenses. Drawing on a college student sample, they reported that low self-control significantly accounted for not only offline offenses, but also online offenses, which is consistent with the generality hypothesis. In the South Korean context, more recently, Yun, Kim, and Kwon (2016) found that Gottfredson and Hirschi’s (1990) generality hypothesis was largely supported for typical delinquency and analogous behaviors including smoking, drinking, and Internet/smartphone addictions among South Korean adolescents.

Studies on cyberbullying have documented two forms of shared similarities between traditional bullying and cyberbullying: (a) an overlap between cyberbullying and traditional bullying, and (b) common predictors of both bullying behaviors. Regarding the overlap, prior research has found that both bullying behaviors were closely related to each other (Hinduja & Patchin, 2008; Smith et al., 2008; Ybarra & Mitchell, 2004a). In addition, a wealth of research has documented that traditional bullying perpetration and victimization were significant predictors for cyberbullying perpetration (Kowalski et al., 2014). A series of longitudinal studies have devoted attention to the dynamic relationship between the two, focusing on whether previous traditional bullying perpetration and cyberbullying victimization influence subsequent cyberbullying perpetration. Findings indicated that previous traditional bullying had significant effects on involvement in cyberbullying perpetration (Espelage, Rao, & Craven, 2013; Fanti, Demetriou, & Hawa, 2012; Hemphill & Heerde, 2014; Higgins, Khey, Dawson-Edwards, & Marcum, 2012; Jang, Song, & Kim, 2014; Pabian & Vandebosch, 2016).

Second, prior research has found that cyberbullying and traditional bullying perpetration share similar individual and contextual predictors (Hemphill et al., 2012; Kowalski & Limber, 2013; Patchin & Hinduja, 2011). As with traditional bullying (Cook, Williams, Guerra, Kim, & Sadek, 2010), predictors of cyberbullying fall into both individual and contextual domains. Regarding individual predictors, low self-control (Bayraktar, Machackova, Dedkova, Cerna, & Sevcikova, 2015; Donner et al., 2015; Donner, Marcum, Jennings, Higgins, & Banfield, 2014; Jang et al., 2014; Marcum, Higgins, Freiburger, & Ricketts, 2014; Vazsonyi, Machackova, Sevcikova, Smahel, & Cerna, 2012), aggression (Ang, Huan, & Florell, 2014; Bayraktar et al., 2015; Calvete, Orue, Estévez, Villardón, & Padilla, 2010; Hemphill et al., 2012; Runions, Shapka, Dooley, & Modecki, 2013), strain (Jang et al., 2014; Patchin & Hinduja, 2011), and low self-esteem (Bayraktar et al., 2015; Patchin & Hinduja, 2010) have been demonstrated to be associated with cyberbullying perpetration. Regarding the contextual predictors including family and peer influences, prior research has found that negative family environments (Bayraktar et al., 2015; Hemphill & Heerde, 2014; Ybarra & Mitchell, 2004b) and negative relationships with peers (Williams & Guerra, 2007) had positive associations with cyberbullying. Associations with delinquent peers also reportedly had positive effects on cyberbullying (Hinduja & Patchin, 2013; Jang et al., 2014). Specifically, recent studies found that those who had associated with cyberbully peers were at an increased likelihood to engage in cyberbullying perpetration (Barlett et al., 2014) and having peers who cyberbully others had been shown to mediate the effects of low self-control on cyberbullying (Li, Holt, Bossler, & May, 2016).

Cyberbullying as a Distinct Form of Bullying

Despite the evidence supporting the notion of cyberbullying as a variation of bullying, there still seems to be some evidence on opposing hypothesis. Indeed, cyberbullying has a unique set of features that may qualify it as being a distinct behavior, especially given that cyberbullying appears to have heterogeneous age distributions and differential opportunistic factors compared with traditional bullying. Due to these differences, it can be hypothesized that there are unique predictors of cyberbullying unrelated to traditional bullying. Although both bullying behaviors have common predictors, these distinct features may lead common predictors to have different effect sizes on cyberbullying and traditional bullying.

There have been mixed findings regarding the age distribution of cyberbullying. As discussed earlier, the generality hypothesis largely relies on the universal age distribution of crimes. For traditional bullying, this seems to be applied as well. Prior research has found evidence of escalation, peak, and decline of traditional bullying patterns during adolescence and early adulthood (Barker, Arseneault, Brendgen, Fontaine, & Maughan, 2008; Nansel et al., 2001; Pepler, Jiang, Craig, & Connolly, 2008). In contrast, only a limited number of studies have examined the longitudinal patterns of cyberbullying, and these studies have generated findings that largely contradicted the longitudinal patterns documented in traditional bullying research. For example, several studies have reported that cyberbullying peaked among middle schoolers and then dropped in the high school years (Calvete et al., 2010; Williams & Guerra, 2007). Other studies found that older youth were more likely than younger youth to engage in cyberbullying and online harassment (Hinduja & Patchin, 2008; Ybarra & Mitchell, 2004b, 2007) or reported no age-varying perpetrations of cyberbullying (Werner, Bumpus, & Rock, 2010).

While the generality hypothesis has been widely supported by many studies, it must also be noted that this has not exclusively and extensively accounted for all forms of deviance, but only a part of them (Lussier, LeBlanc, & Proulx, 2005; Osgood et al., 1988; Tittle & Grasmick, 1997; Yun et al., 2016). Some studies based on a life-course perspective have also found that offenders did not always exhibit versatility of offending, but rather that offenders could specialize in a certain crime during the short term, typically due to the influence of some local life circumstances (McGloin, Sullivan, & Piquero, 2009; McGloin, Sullivan, Piquero, & Pratt, 2007; Sullivan, McGloin, Pratt, & Piquero, 2006). In the context of computer crimes, Khey, Jennings, Lanza-Kaduce, and Frazier (2009) found evidence supporting specialization in computer crime. They reported that college students who were older, non-Whites, and not a member of a Greek organization were more likely to be computer crime specialists than computer crime generalists. Along with the results that age-varying patterns of cyberbullying are inconsistent with traditional bullying, these findings can be thus reasonable grounds to conclude that cyberbullying could be a new form of bullying.

If a tendency to specialize in cyberbullying indeed exists, this may relate to individuals’ variations in situational opportunities derived from differential contextual conditions. As Cloward and Ohlin (1960) argued, forms of delinquency can be dependent on individual’s differential opportunity in certain contexts determining which illegitimate means potential offenders easily access. If cyberbullying is a distinct form of deviance, it is thus expected that cyberbullies have some differential opportunity contexts to allow them to more easily access cyberbullying than other deviant behaviors, which may include (1) online proficiency and (2) parental (or caregiver’s) supervision of online activities.

As cyberbullying is a deviant behavior which takes place in the online setting, understanding online mechanisms and proficiency in computer and other mobile devices may be closely associated with an individual’s tendency of cyberbullying. Highlighting the heterogeneous age distribution between cyberbullying and traditional bullying, Hinduja and Patchin (2014) speculated that there might be intervening opportunity factors such as time spent online and proficiency in online settings moderating the relationship between age and cyberbullying (e.g., Ybarra & Mitchell, 2004a). Specifically, both time spent online and online proficiency are positively correlated with each other (Bossler & Holt, 2009), and prior studies reported that longer time spent computer/online led to more computer crimes and online harassment (J. Kim & Kim, 2015a, 2015b; Moon, McCluskey, & McCluskey, 2010; Ybarra & Mitchell, 2004b, 2007). Accordingly, cyberbullies, who hypothetically have greater proficiency in computer and Internet and spend longer time online, may exhibit their negative emotions online rather than face-to-face contexts because they easily access the virtual society and feel more comfortable being online due to their greater proficiency in the online setting.

Parents or caregivers’ supervision of children’s online activities can be regarded as another opportunity predictor of cyberbullying. While it is expected that both traditional bullying and cyberbullying occur when adult supervision is absent (Farrington, 1993), supervision provided by family contexts may have a greater impact on cyberbullying as cyberbullying is less likely to be supervised by adults due to the lack of agents with an enforcing role. Tokunaga (2010) pointed out the issue, stating that there are no specific individuals or groups in charge of regulating cyberbullying whereas teachers or school administrators are involved in regulating traditional bullying offenses in school as an agent of enforcement. Thus, regulation of cyberbullying heavily depends on adults in family and supervision provided by parents or caregivers of children is one of the few means to control the deviant behavior. That is, cyberbullies may be less likely to be supervised by parents or caregivers at home. Research up to date seems to approve this hypothesis. Ybarra and Mitchell (2004a, 2004b) found that youths who were involved in online harassment reported infrequent parental monitoring of online activities. Additionally, some studies reported that family management, family/parental supports, and parents’ emotional attachment to children were associated with youths being involved in cyberbullying perpetration (Fanti et al., 2012; Hemphill & Heerde, 2014; Kowalski et al., 2014; Wang, Iannotti, & Nansel, 2009).

The Current Study

The mixed findings of previous research make it difficult to conclude whether cyberbullying is a distinct form of behavior or a variation of traditional bullying. To understand the nature of cyberbullying, this study draws on data from a longitudinal sample of South Korean youth to examine similarities and differences between cyberbullying and traditional bullying. Latent classes of cyberbullying and traditional bullying are initially examined to investigate developmental patterns. Next, a set of multinomial logistic regressions are estimated to assess the predictors of trajectory group membership. Finally, a dual-trajectory analysis identifies the overlap between the developmental trajectories of cyberbullying and traditional bullying.

Method

Data

This study uses five waves of the KYPS following South Korean juveniles annually from 2003 to 2007. The first wave of the KYPS data set targeted students in their second year of junior high school, and these respondents were 14 years old at the time of the first survey. The variables of KYPS included diverse aspects of the respondent’s personal traits, peer relations, parental–juvenile relations, and socioeconomic backgrounds. Cyberbullying and traditional bullying were examined through all five waves of the KYPS. The total number of respondents for this study was n = 2,721. A small percentage of missing responses (less than 5% of total responses) were addressed with multiple imputation techniques. Table 1 shows descriptive information of the examined variables.

Table 1.

Descriptive Statistics of Variables.

	M	SD	Minimum	Maximum	Wave
Dependent variables
Cyberbullying	0.43	0.50	0	1	1
	0.23	0.42	0	1	2
	0.12	0.32	0	1	3
	0.10	0.29	0	1	4
	0.07	0.26	0	1	5
Traditional bullying	0.27	0.44	0	1	1
	0.11	0.31	0	1	2
	0.07	0.25	0	1	3
	0.05	0.22	0	1	4
	0.04	0.20	0	1	5
Individual characteristics
Time spent on computer	2.59	1.60	0.1	12.4	1
Self-control	10.06	3.96	0	24	1
Victimization (traditional bullying)	0.12	0.33	0	1	1
Peer relationships
Peer attachment	12.23	2.59	0	16	1
Peer violent delinquency	0.14	0.35	0	1	1
Peer non-violent delinquency	0.33	0.47	0	1	1
Parental–juvenile relations
Parental attachment	4.30	1.95	0	8	1
Parental supervision	8.90	3.42	0	16	1
Controls
Sex	0.50	0.50	0	1	1
Education level (father)	3.68	1.28	0	7	1
Education level (mother)	3.18	1.07	0	7	1
SES	290.06	190.41	5	3,000	1

Note. SES = socioeconomic status.

Measures

Dependent variables

Cyberbullying

This study utilized two questions for the measure of the respondent’s cyberbullying experiences. The first question was, “Have you ever intentionally circulated false information on the Internet message boards about others during the last year?” The second was, “Have you ever cursed/insulted other people through chats/message boards during the last year?” Both questions were measured through the five waves consecutively. When an individual exhibited either or both of those forms of cyberbullying, they were categorized as a 1 (yes) and 0 (no) if otherwise (Jang et al., 2014).

Traditional bullying

For the purpose of measuring traditional bullying, four questions were combined into a dichotomous variable. Each question asked about an individual’s behavior during the last year in regard to “severely beating other people,” “severely teasing or bantering other people,” “threatening other people,” and “treating other people as an outcast.” Similar to cyberbullying, these variables were measured during all five waves of the KYPS.

Independent variables

Individual characteristics

The current study measured three individual characteristics: time spent on a computer, self-control, and victimization via traditional bullying. Time spent on a computer was assessed by the respondent’s average number of hours of computer usage in a day (The average number of computer usage hours was 2.6/day). Portable devices such as smartphone and tablet were not commonly used to access the Internet at the time of the KYPS (between 2003 and 2007), therefore time spent on computer is logical to measure an individual’s involvement in cyber activity. Self-control was measured by the sum of six questions that represented five features of self-control: impulsivity, avoidance of difficult tasks in favor of simple tasks, risk-taking, self-centeredness, and short temper (Grasmick, Tittle, Bursik, & Arneklev, 1993). Each question was measured on a 5-point Likert scale. The higher scores indicate lower levels of self-control (α = .63). Victimization was assessed as whether the individual had experienced traditional types of bullying. There were two questions that were asked to measure an individual’s experience of victimization. These questions measured if the individual was severely teased, bantered, and/or treated as an outcast. When the respondent reported having experienced at least one form of traditional bullying, they categorized into 1 (yes) and 0 (no) if otherwise (Jang et al., 2014).

Peer relationships

Three peer-related predictors were included for this study. Peer attachment was measured by the sum of four questions: (a) I want to maintain a friendship with my close friends, (b) I enjoy hanging out with my close friends, (c) I try to have the same feelings and thoughts as my friends, and (d) I can share my thoughts and feelings with my close friends in an honest way (α = .75; E. Kim, Kwak, & Yun, 2010). Peer delinquency was separately obtained for violent and non-violent delinquency, and these were both dichotomous variables. The peer violent delinquency variable indicated an intimate association with peers who conducted serious beatings and/or forcefully took money from other people. The peer non-violent delinquency variable measured the peer’s misconduct in the areas of smoking, drinking, and skipping school.

Parental–juvenile relationships

Parental attachment and parental supervision were included in this study. Parental attachment was examined by two questions, “I am comfortable sharing my thoughts and feelings with my parents,” and “I often talk about what happens to me outside home.” (α = .72; E. Kim et al., 2010). Parental supervision was measured by four questions: (a) When I go out, my parents usually know where I am, (b) When I go out, my parents usually know whom I am with, (c) When I go out, my parents usually know what I am doing, and (d) When I go out, my parents usually know when I will return. (α = .84; Han & Grogan-Kaylor, 2012).

Controls

Four variables were included in this study as statistical controls. The respondent’s sex was measured by 1 for male, and 0 for female. Education level was separately measured for the father’s and mother’s level of education. This was categorized into seven levels ranging from 0 = elementary school, 1 = middle school . . . 6 = master’s degree, and 7 = doctoral degree. Socioeconomic status (SES) was measured by the monthly income of the family.

Analytic Strategy

The current study uses diverse analytic stages to investigate the developmental patterns and predictors of cyberbullying and traditional bullying, and their overlap. The descriptive statistics of the sample demonstrate the general summary of variables. Second, group-based trajectory analysis examines the latent trajectory groups for cyberbullying and traditional bullying. The best fitting models of developmental patterns is selected based on the Bayesian Information Criterion (BIC) and group mean posterior probabilities of assignment (Nagin, 2005, 2010). We use the logistic model for both cyberbullying and traditional bullying. Third, a series of multinomial logistic regressions will compare the risk and protective factors related to trajectory group membership. Last, the relationship between cyberbullying and traditional bullying will be examined to investigate mutual interactions between trajectory groups.

Results

Figure 1 shows the developmental changes of the pattern of cyberbullying. Half of juveniles exhibited a sharp-decreasing pattern, with almost no participation after the age of 16. Of the respondents, 33.8% of juveniles did not commit cyberbullying, while 15.5% indicated chronic participation between the ages of 14 and 18. The cyberbullying trajectory groups were created using five different quadratic specification models. Based on changes in the BIC, the model with three latent groups of cyberbullying was selected. The minimum BIC score was −5,491.62. The current study conducted several model selections with a different order of the polynomials to ensure posterior probabilities were over .70 threshold offered by Nagin (2005, 2010). Figure 2 examines the trajectories of traditional bullying following the same procedures. The BIC value was minimized (−3,941.93), when three latent groups of traditional bullying were chosen for the model. The general patterns between the groups are similar; however, differences exist in the ratio of each trajectory group. The most common trajectory group of traditional bullying was the non-participation group, which occupies almost half of the total sample (51.9%). Comparatively, the sharp-decreasing group was the second largest group of traditional bullying (35.5%).

Figure 1.

Group-based trajectories of cyberbullying.

Figure 2.

Group-based trajectories of traditional bullying.

Table 2 summarizes the results of the multinomial logistic regression models investigating the risk and protective factors of the cyberbullying trajectories. When compared with the non-participation group, the sharp-decreasing and chronic groups indicate general similarities in predictors, while the peer-related variables showed opposite results. For instance, victimization from traditional bullying was associated with both the sharp-decreasing (relative risk ratio [RRR] = 1.36, SE = .19, p < .05) and the chronic cyberbullying groups (RRR = 1.61, SE = .31, p < .05); however, there were not any significant factors to distinguish between those two groups. Both the sharp-decreasing (RRR = 0.95, SE = .01, p < .01) and chronic groups (RRR = .93, SE = .02, p < .01) indicated lower levels of parental supervision when compared with non-participating group. A lower level of parental attachment was related to chronic cyberbullying group (RRR = .91, SE = .03, p < .05) participation, but not for the sharp-decreasing group. The peer-related variables were the key factors that distinguished between the sharp-decreasing and chronic group membership. Peer attachment (RRR = .91, SE = .02, p < .001) and peer violence (RRR = 2.08, SE = .42, p < .001) was strongly related to the chronic group participation. Having a lower level of peer attachment (RRR = .94, SE = .02, p < .01) and violent peer association (RRR = 1.74, SE = .32, p < .01) was associated and chronic cyberbullying participation even compared with sharp-decreasing group. Among the individual demographic control variables, sex was the only variable that exhibited a statistically significant relationship, such that males were more likely to commit cyberbullying than females.

Table 2.

Bivariate RRR Between Factors and Each Pair of Trajectory Group Comparisons: Cyberbullying.

Cyberbullying	1 vs. 2		1 vs. 3		2 vs. 3
Cyberbullying	RRR	SE	RRR	SE	RRR	SE
Individual characteristics
Time spent on computer	1.06	0.03*	1.07	0.04	1.01	0.04
Self-control	1.05	0.01***	1.04	0.02**	0.99	0.02
Traditional bullying	1.90	0.20***	2.20	0.34***	1.16	0.17
Victimization (traditional bullying)	1.36	0.19*	1.61	0.31*	1.19	0.21
Peer relationships
Peer attachment	0.98	0.01	0.91	0.02***	0.94	0.02**
Peer violent delinquency	1.19	0.18	2.08	0.42***	1.74	0.32**
Peer non-violent delinquency	1.49	0.16***	1.19	0.20	0.80	0.13
Parental–juvenile relations
Parental attachment	0.97	0.02	0.91	0.03*	0.94	0.03
Parental supervision	0.95	0.01***	0.93	0.02**	0.98	0.02
Controls
Sex	0.95	0.08	1.93	0.27***	2.03	0.28***
Education level (father)	0.98	0.05	0.98	0.07	1.00	0.07
Education level (mother)	1.03	0.06	0.98	0.09	0.94	0.08
SES	1.00	0.00	1.00	0.00	1.00	0.00

Note. 1 = non-involved; 2 = sharp-decreasing; 3 = chronic; RRR = relative risk ratio; SES = socioeconomic status.

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

Table 3 displays same models in which traditional bullying served as the dependent variable. Similarities were found for the prediction of each developmental trajectory, with the main differences between cyberbullying and traditional bullying being the role of peers and the importance of parental supervision. The individual characteristics were related to sharp-decreasing and chronic traditional bullying participation, especially in regard to victimization from traditional bullying (RRR = 3.11, SE = .43, p < .001 and RRR = 3.26, SE = .61, p < .001, respectively). The peer relations variables were unable to distinguish between the sharp-decreasing and chronic group memberships as opposed to the cyberbullying models. With the parental–juvenile relations, there was no meaningful relationship for parental attachment and group membership. In contrast, lower levels of parental supervision did allow for distinction between the chronic traditional bullying group and the non-participation group (RRR = .93, SE = .02, p < .01), as well as from the sharp-decreasing group (RRR = .95, SE = .02, p < .05). Parental supervision was also a meaningful predictor of cyberbullying group membership; however, it was not able to distinguish between the sharp-decreasing and the chronic group membership. An individual’s sex exhibited similar predictions as the cyberbullying model. In addition, paternal education level was also associated with traditional bullying.

Table 3.

Bivariate RRR Between Factors and Each Pair of Trajectory Group Comparisons: Traditional Bullying.

Traditional bullying	1 vs. 2		1 vs. 3		2 vs. 3
Traditional bullying	RRR	SE	RRR	SE	RRR	SE
Individual characteristics
Time spent on computer	1.02	0.03	1.01	0.04	0.99	0.04
Self-control	1.04	0.01**	1.06	0.02***	1.03	0.02
Cyberbullying	1.65	0.17***	1.79	0.27***	1.09	0.17
Victimization (traditional bullying)	3.11	0.43***	3.26	0.61***	1.05	0.19
Peer relationships
Peer attachment	0.90	0.01***	0.86	0.02***	0.96	0.02
Peer violence delinquency	3.04	0.45***	4.08	0.78***	1.34	0.26
Peer non-violent delinquency	1.38	0.16**	1.83	0.31***	1.33	0.24
Parental–juvenile relations
Parental attachment	0.96	0.03	0.93	0.04	0.97	0.04
Parental supervision	0.97	0.02	0.93	0.02**	0.95	0.02*
Controls
Sex	1.11	0.11	1.57	0.23**	1.41	0.22*
Education level (father)	1.01	0.05	0.83	0.07*	0.82	0.07*
Education level (mother)	0.92	0.06	1.08	0.11	1.18	0.12
SES	1.00	0.00	1.00	0.00	1.00	0.00

Note. 1 = non-involved; 2 = sharp-decreasing; 3 = chronic; RRR = relative risk ratio; SES = socioeconomic status.

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

Table 4 represents three different sets of cross-tabulations estimated from a dual-trajectory analysis: (a) the probability of the cyberbullying group conditional on the traditional bullying group, (b) the probability of the traditional bullying group conditional on the cyberbullying group, and (c) the joint probability of the cyberbullying group and the traditional bullying group. The first model indicates the general similarities in the patterns of the traditional bullying group membership with the cyberbullying group membership. When examining the probability of the traditional bullying group conditional on the cyberbullying group, the general similarities were found for all groups except for the chronic cyberbullying group. Even though the chronic traditional bullying group was the most prevalent group (47.8%), the association with the other two trajectory groups was relatively mixed. The third model shows the joint probability of both types of bullying. Almost half of the juveniles were categorized as being assigned to both the non-participation groups (28.8%) and the sharp-decreasing groups (27.5%), and 18.6% of juveniles did not participate in traditional bullying, but they showed a sharp-decreasing pattern of cyberbullying. The prevalence of being assigned to both of the chronic groups was 7.4% of total sample.

Table 4.

The Overlap Between Cyberbullying and Traditional Bullying Trajectory Group Membership.

Cyberbullying group	Traditional bullying group
Cyberbullying group	Non-involved	Sharp-decreasing	Chronic
(a) Probability of cyberbullying group conditional on traditional bullying group
Non-involved	55.6	12.4	4.8	100.0
Sharp-decreasing	35.9	77.3	36.4	100.0
Chronic	8.5	10.3	58.8	100.0
(b) Probability of traditional bullying group conditional on cyberbullying group
Non-involved	85.2	13.0	1.8	100.0
Sharp-decreasing	36.7	54.2	9.1	100.0
Chronic	28.5	23.7	47.8	100.0
(c) Joint probability of cyberbullying group and traditional bullying group
Non-involved	28.8	4.4	0.6
Sharp-decreasing	18.6	27.5	4.6
Chronic	4.4	3.7	7.4

Discussion

Initially, this study examined the developmental patterns of cyberbullying and traditional bullying, particularly focusing on identifying latent trajectory groups. We suggested that if cyberbullying exhibited unique patterns from traditional bullying, this could be evidence for the need to consider cyberbullying as a new form of deviant behavior. Our results found similar developmental patterns for both, which provided little evidence to conclude that cyberbullying is distinct from traditional bullying. Additionally, the overlap between cyberbullying and traditional bullying supported such findings. First, the results from a set of multinomial regressions identified that committing traditional bullying increases the risk to commit cyberbullying and vice versa. Additionally, results from a dual-trajectory analysis also found that more than 60% of juveniles indicated similar developmental patterns for both outcomes. Among those compositions of trajectories, the overlap patterns were dominant except for sharp-decreasing cyberbullying and non-traditional bullying patterns.

Regarding predictors of cyberbullying and traditional bullying, we expected that cyberbullying would be mainly explained by different factors, or that cyberbullying would have distinctive effect sizes from these common factors, when cyberbullying is a distinct form of bullying. Initially, differential opportunity for online access opportunity, measured by time spent on computer, was associated with the sharp-decreasing pattern of cyberbullying, but not for the chronic group. In addition, parental supervision was an important factor in facilitating different opportunity contexts for ease of cyberbullying (Ybarra & Mitchell, 2004a, 2004b), but the effect size was largely similar to that of traditional bullying. Such results provided partial support for the notion that cyberbullying is a distinct form of deviant behavior, although we encountered limitations of marginal explanatory power, as other predictors exhibited dominant explanations for cyberbullying.

Among other predictors, significant effects were found from individual’s low self-control, while peer-related factors, especially differential association with violent peers, suggested strongest explanations for both cyberbullying and traditional bullying. Initially, along with previous work examining the causal links between low self-control and online/offline offenses (Donner et al., 2015) our results showed that self-control was related to membership in the sharp-decreasing and chronic group for cyberbullying, when compared with the non-involved group. Yet, it was not dominant, and similar findings emerged for predictors of traditional bullying. Such results are also opposed to the expectation that delayed rewards of cyberbullying makes the distinction from traditional bullying for its motives (Kowalski et al., 2014).

Second, the most dominant and insightful effects were found among peer-related factors for both cyberbullying and traditional bullying. It has been hypothesized that the desire to obtain or maintain dominant status in a peer group may be an important motive for traditional bullying (Espelage & Swearer, 2003; Pellegrini, 2002), and this notion has been empirically supported (Pellegrini, Bartini, & Brooks, 1999; Reijntjes et al., 2013). However, peer influence on cyberbullying has been questioned for the given features of the online space, including accessibility, anonymity, lack of physical contacts, and disinhibition. Due to such nature, cyberbullies may not have the same need to pursue physical and social dominance as traditional bullies. However, results for cyberbullying predictions were largely similar with traditional bullying, and two different theoretical domains of peer relations, attachment and differential association, had significant effects. In detail, peer violent delinquency was an important factor in distinguishing chronic group membership from other groups, while peer non-violent delinquency was an important factor in distinguishing membership in the sharp-decreasing group from non-involved group. These results are suggesting the possibility of causal mechanism of delinquent peer influence is mostly rooted on the hypotheses of similar levels of delinquency resemblance (Burt & Rees, 2015; Young, Rebellon, Barnes, & Weerman, 2014). The results of the current study are also supportive of the taxonomic separation between adolescent-limited and chronic offenders in regard to seriousness of deviancy (Moffitt, 1993, 2003). That is, adolescent-limited cyberbullying offenders were likely to associate with peers who engaged in non-violent delinquency, and not with those who engaged in violent delinquency.

The results of the current study have practical implications for setting the foundation of prevention strategies and policies. As cyberbullying is a newly emerged social problem, practitioners and policy makers are less familiar with reality of and coping strategies for cyberbullying. Prior strategies based on technological backgrounds, including strict privacy settings and online identities regulations, have indicated no or marginal effectiveness (Slonje, Smith, & Frisén, 2013; Tokunaga, 2010). Those technical coping strategies were mostly focused on blocking access for cyberbullies, rather than influencing these individuals’ motivations for cyberbullying. This study found general similarity between cyberbullying and traditional bullying, which may suggest that establishing strategies based on conventional bullying prevention programs may help control for both behaviors. Prior research also supports such an expectation, finding that a conventional bullying intervention program decreased cyberbullying, though the program did not include cyberbullying prevention (Salmivalli et al., 2011). Bauman and Newman (2013) also suggested similar clinical and policy implications, arguing that bullying intervention programs need to focus on the context and severity of bullying itself, rather than on whether the method of bullying is traditional or online.

This study is not without limitations. We adopted a dichotomous prevalence measure (yes/no) for estimating the developmental patterns of traditional bullying and cyberbullying. The original survey questions consisted of two separate steps, with respondents initially asked whether they perpetrated each type of bullying during last year, and were then asked “If yes, how many times have you done each of them?” This led to the development of two separate variables: (a) yes/no indication and (b) the number of times the respondent engaged in each behavior. The original data set (KYPS) has numerous missing answers for the frequency measure, especially for the first wave. At Wave 1, individuals who answered that they had perpetrated cyberbullying comprised of 43% of the total sample, while the sum of the frequency question indicated that only 22.6% answered that they had perpetrated these behaviors more than once. We also found many individuals who reported perpetration of cyberbullying, but also skipped reporting their frequency of involvement. Similar issues emerged for the traditional bullying assessment; prevalence at Wave 1 was 27%, while frequency of perpetrating behavior more than once was 10.5%. Although such problems were minimized after Wave 1, we decided to use prevalence as dependent variable, rather than the frequency measure given these concerns.

This study comprehensively examined risk and protective factors of cyberbullying when compared with traditional bullying, and investigated the overlap between the two outcomes. Yet, there remains room for future research to devote further attention to investigating cyberbullying. Initially, this study examined peer influence as three factors as peer attachment, and two types of differential association. However, when considering the nature of cyberbullying, interactions inside the cyber network are likely to also be important factors. Although accessibility and time spent on the Internet were not significantly related to chronic participation or for distinguishing between the chronic and sharp-decreasing group, this does not necessarily indicate that cyberbullying can be fully understood by conventional risk and protective factors. Earlier explanations of interpersonal interactions by Akers (1998) focused on the importance of a comprehensive examination of the learning process occurring via the Internet. For instance, the Internet allows individuals to interact with a diverse group of people without the limitations of physical boundaries, also making it easier to reinforce their thinking through the Internet community activity. Future research could extend the conventional understanding for cyberbullying, and consider such interactions that happen on the Internet. In addition, future research is encouraged to further explore the role of gender and gender orientation when examining traditional bullying and/or cyberbullying. For instance, LGBTQ youth are at a heightened risk for traditional bullying in general (Russell, Kosciw, Horn, & Saewyc, 2010) and cyberbullying specifically (Robinson & Espelage, 2011), and as such future research should make an effort to include measures on LGBTQ status when data are available. Also, future research considering gendered pathways of cyberbullying may have a great potential to reveal important insights. This study paid little attention to gender issues, given our focus on illustrating the general nature of cyberbullying and its comparison with traditional bullying. However, gender showed significant predictability for cyberbullying. Prior research has also supported the importance of gender on cyberbullying (Bauman, 2013; Dooley et al., 2009); however, research is largely limited in regard to examining the pathways perspective. Last but not least, it is also important to note that there is some research that has demonstrated a theoretical overlap between important mechanisms and constructs such as parental and peer modeling that exists in social learning and social control theories. In fact, this research has argued for the theoretical integration of these theoretical perspectives (see Jennings, Higgins, Akers, Khey, & Dobrow, 2013). As such, future research should investigate the degree to which this theoretical and conceptual overlap exists and its implications for traditional bullying and cyberbullying theorizing.

Ultimately, the current study is one of the first studies to date to examine the overlap between traditional bullying and cyberbullying, and to evaluate the role of risk and protective factors for distinguishing unique developmental patterns of involvement in these two behaviors. More importantly, this study is the first study to focus on these research questions in a South Korean sample. Future research is encouraged to explore the degree to which these findings may replicate in other cross-cultural and international investigations to further contribute to the evidence regarding the global applicability of these relationships.

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.

Author Biographies

Jihoon Kim is a PhD student in the Department of Sociology and Criminology & Law at the University of Florida. His research interests lie in the area of life course/developmental theory, peer influence, measurement, and quantitative analysis. His recent or forthcoming publications appear in Crime & Delinquency and Deviant Behavior.

Hyojong Song is a doctoral candidate in the Department of Criminology at the University of South Florida. His research interests include cyber-crime, environmental crime, and macro-level theories of crime.

Wesley G. Jennings, PhD, is associate professor, associate chair, and undergraduate director in the Department of Criminology, has a courtesy appointment in the Department of Mental Health Law and Policy, and is a faculty affiliate of the Florida Mental Health Institute in the College of Behavioral and Community Sciences at the University of South Florida. In addition, 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.

References

Akers

R. L.

(1998). Social learning and social structure: A general theory of crime and deviance. Boston, MA: Northeastern University Press.

Ang

R. P.

Huan

V. S.

Florell

(2014). Understanding the relationship between proactive and reactive aggression, and cyberbullying across United States and Singapore adolescent samples. Journal of Interpersonal Violence, 29, 237-254. doi:10.1177/0886260513505149

Barker

E. D.

Arseneault

Brendgen

Fontaine

Maughan

(2008). Joint development of bullying and victimization in adolescence: Relations to delinquency and self-harm. Journal of the American Academy of Child & Adolescent Psychiatry, 47, 1030-1038. doi:10.1097/CHI.ObO13e31817eec98

Barlett

C. P.

Gentile

D. A.

Anderson

C. A.

Suzuki

Sakamoto

Yamaoka

Katsura

(2014). Cross-cultural differences in cyberbullying behavior: A short-term longitudinal study. Journal of Cross-Cultural Psychology, 45, 300-313. doi:10.1177/0022022113504622

Bauman

(2013). Cyberbullying: What does research tell us? Theory Into Practice, 52, 249-256. doi:10.1080/00405841.2013.829727

Bauman

Newman

M. L.

(2013). Testing assumptions about cyberbullying: Perceived distress associated with acts of conventional and cyber bullying. Psychology of Violence, 3, 27-38. doi:10.1037/a0029867

Bayraktar

Machackova

Dedkova

Cerna

Sevcikova

(2015). Cyberbullying: The discriminant factors among cyberbullies, cybervictims, and cyberbully-victims in a Czech adolescent sample. Journal of Interpersonal Violence, 30, 3192-3216. doi:10.1177/0886260514555006

Bossler

A. M.

Holt

T. J.

(2009). On-line activities, guardianship, and malware infection: An examination of routine activities theory. International Journal of Cyber Criminology, 3, 400-420.

Burt

C. H.

Rees

(2015). Behavioral heterogeneity in adolescent friendship networks. Justice Quarterly, 32, 872-899. doi:10.1080/07418825.2013.856932

10.

Calvete

Orue

Estévez

Villardón

Padilla

(2010). Cyberbullying in adolescents: Modalities and aggressors’ profile. Computers in Human Behavior, 26, 1128-1135. doi:10.1016/j.chb.2010.03.017

11.

Cloward

R. A.

Ohlin

L. E.

(1960). Delinquency and opportunity: A theory of delinquent gangs. New York, NY: The Free Press of Glencoe.

12.

Cook

C. R.

Williams

K. R.

Guerra

N. G.

Kim

T. E.

Sadek

(2010). Predictors of bullying and victimization in childhood and adolescence: A meta-analytic investigation. School Psychology Quarterly, 25, 65-83. doi:10.1037/a0020149

13.

Common Sense. (2015). The Common Sense Census: Media Use by Tweens and Teens. Retrieved Retrieved January 25, 2016 from https://www.commonsensemedia.org/research/the-common-sense-census-media-use-by-tweens-and-teens

14.

Dembo

Williams

Wothke

Schmeidler

Getreu

Berry

Wish

E. D.

(1992). The generality of deviance: Replication of a structural model among high-risk youths. Journal of Research in Crime & Delinquency, 29, 200-216. doi:10.1177/0022427892029002006

15.

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, 663-679. doi:10.1177/0894439314555949

16.

Donner

C. M.

Marcum

C. D.

Jennings

W. G.

Higgins

G. E.

Banfield

(2014). Low self-control and cybercrime: Exploring the utility of the general theory of crime beyond digital piracy. Computers in Human Behavior, 34, 165-172. doi:10.1016/j.chb.2014.01.040

17.

Dooley

J. J.

Pyżalski

Cross

(2009). Cyberbullying versus face-to-face bullying. Zeitschrift Für Psychologie / Journal of Psychology, 217, 182-188. doi:10.1027/0044-3409.217.4.182

18.

Espelage

D. L.

Rao

M. A.

Craven

R. G.

(2013). Theories of cyberbullying. In Bauman

Cross

Walker

(Eds.), Principles of cyberbullying research: Definitions, measures, and methodology (pp. 49-67). New York, NY: Routledge.

19.

Espelage

D. L.

Swearer

S. M.

(2003). Research on school bullying and victimization: What have we learned and where do we go from here? School Psychology Review, 32(3), 365-384.

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, 168-181. doi:10.1080/17405629.2011.643169

21.

Farrington

D. P.

(1993). Understanding and preventing bullying. In Crime and justice: An annual review of research (Vol. 17, pp. 381-458). Chicago, IL: The University of Chicago Press. doi:10.1177/1362361306068508

22.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. San Francisco, CA: Stanford University Press.

23.

Grasmick

H. G.

Tittle

C. R.

Bursik

R. J.

Arneklev

B. J.

(1993). Testing the core empirical implications of Gottfredson and Hirschi’s general theory of crime. Journal of Research in Crime & Delinquency, 30, 5-29. doi:10.1177/0022427893030001002

24.

Han

Grogan-Kaylor

(2012). Parenting and youth psychosocial well-being in South Korea using fixed-effects models. Journal of Family Issues, 34, 689-715. doi:10.1177/0192513X12440091

25.

Hay

Meldrum

Mann

(2010). Traditional bullying, cyber bullying, and deviance: A general strain theory approach. Journal of Contemporary Criminal Justice, 26, 130-147. doi:10.1177/1043986209359557

26.

Hemphill

S. A.

Heerde

J. A.

(2014). Adolescent predictors of young adult cyberbullying perpetration and victimization among Australian youth. Journal of Adolescent Health, 55, 580-587. doi:10.1016/j.jadohealth.2014.04.014

27.

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, 59-65. doi:10.1016/j.jadohealth.2011.11.019

28.

Higgins

G. E.

Khey

D. N.

Dawson-Edwards

B. C.

Marcum

C. D.

(2012). Examining the link between being a victim of bullying and delinquency trajectories among an African American sample. International Criminal Justice Review, 22, 110-122. doi:10.1177/1057567712443965

29.

Hinduja

Patchin

J. W.

(2007). Offline consequences of online victimization. Journal of School Violence, 6, 89-112. doi:10.1300/J202v06n03_06

30.

Hinduja

Patchin

J. W.

(2008). Cyberbullying: An exploratory analysis of factors related to offending and victimization. Deviant Behavior, 29, 129-156. doi:10.1080/01639620701457816

31.

Hinduja

Patchin

J. W.

(2010). Bullying, cyberbullying, and suicide. Archives of Suicide Research, 14, 206-221. doi:10.1080/13811118.2010.494133

32.

Hinduja

Patchin

J. W.

(2013). Social influences on cyberbullying behaviors among middle and high school students. Journal of Youth & Adolescence, 42, 711-722. doi:10.1007/s10964-012-9902-4

33.

Hinduja

Patchin

J. W.

(2014). Bullying beyond the schoolyard: Preventing and responding to cyberbullying. Thousand Oaks, CA: Corwin Press.

34.

Hirschi

T. W.

Gottfredson

M. R.

(1983). Age and the explanation of crime. American Journal of Sociology, 89, 552-584. Retrieved from https://www-jstor-org.web.bisu.edu.cn/stable/2779005

35.

Hirschi

T. W.

Gottfredson

M. R.

(1994). The generality of deviance. New Brunswick, NJ: Transaction.

36.

Jang

Song

Kim

(2014). Does the offline bully-victimization influence cyberbullying behavior among youths? Application of general strain theory. Computers in Human Behavior, 31, 85-93. doi:10.1016/j.chb.2013.10.007

37.

Jennings

W. G.

Higgins

G. E.

Akers

R. L.

Khey

D. N.

Dobrow

(2013). Examining the influence of delinquent peer association on the stability of self-control in late childhood and early adolescence: Toward an integrated theoretical model. Deviant Behavior, 34, 407-422. doi:10.1080/01639625.2012.735903

38.

Jennings

W. G.

Zgoba

K. M.

Donner

C. M.

Henderson

B. B.

Tewksbury

(2014). Considering specialization/versatility as an unintended collateral consequence of SORN. Journal of Criminal Justice, 42, 184-192. doi:10.1016/j.jcrimjus.2013.06.010

39.

Jones

L. M.

Mitchell

K. J.

Finkelhor

(2013). Online harassment in context: Trends from three Youth Internet Safety Surveys (2000, 2005, 2010). Psychology of Violence, 3(1), 53-69.

40.

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, 421-434. doi:10.1080/14786010903358166

41.

Kim

Kwak

D.-H.

Yun

(2010). Investigating the effects of peer association and parental influence on adolescent substance use: A study of adolescents in South Korea. Journal of Criminal Justice, 38, 17-24. doi:10.1016/j.jcrimjus.2009.11.003

42.

Kim

(2015a). International note: Teen users’ problematic online behavior: Using panel data from South Korea. Journal of Adolescence, 40, 48-53. doi:10.1016/j.adolescence.2015.01.001

43.

Kim

(2015b). Software piracy among Korean adolescents: Lessons from panel data. Deviant Behavior, 36, 705-724. doi:10.1080/01639625.2014.977111

44.

Klein

M. W.

(1984). Offence specialisation and versatility among juveniles. The British Journal of Criminology, 24, 185-194.

45.

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, 1073-1137. doi:10.1037/a0035618

46.

Kowalski

R. M.

Limber

S. P.

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

47.

Law

D. M.

Shapka

J. D.

Hymel

Olson

B. F.

Waterhouse

(2012). The changing face of bullying: An empirical comparison between traditional and internet bullying and victimization. Computers in Human Behavior, 28, 226-232. doi:10.1016/j.chb.2011.09.004

48.

LeBlanc

Girard

(1997). The generality of deviance: Replication over two decades with a Canadian sample of adjudicated boys. Canadian Journal of Criminology, 39, 171-183.

49.

C. K. W.

Holt

T. J.

Bossler

A. M.

May

D. C

. (2016). Examining the mediating effects of social learning on the low self-control—Cyberbullying relationship in a youth sample. Deviant Behavior, 37, 126-138. doi:10.1080/01639625.2014.1004023

50.

Lussier

LeBlanc

Proulx

(2005). The generality of criminal behavior: A confirmatory factor analysis of the criminal activity of sex offenders in adulthood. Journal of Criminal Justice, 33, 177-189. doi:10.1016/j.jcrimjus.2004.12.009

51.

Marcum

C. D.

Higgins

G. E.

Freiburger

T. L.

Ricketts

M. L.

(2014). Exploration of the cyberbullying victim/offender overlap by sex. American Journal of Criminal Justice, 39, 538-548. doi:10.1007/s12103-013-9217-3

52.

McGloin

J. M.

Sullivan

C. J.

Piquero

A. R.

(2009). Aggregating to versatility? Transitions among offender types in the short term. British Journal of Criminology, 49, 243-264. doi:10.1093/bjc/azn072

53.

McGloin

J. M.

Sullivan

C. J.

Piquero

A. R.

Pratt

T. C.

(2007). Local life circumstances and offending specialization/versatility: Comparing opportunity and propensity models. Journal of Research in Crime & Delinquency, 44, 321-346. doi:10.1177/0022427807302664

54.

Moffitt

T. E.

(1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674-701. doi:10.1037/0033-295X.100.4.674

55.

Moffitt

T. E.

(2003). Life-course-persistent and adolescence -limited antisocial behavior: A 10-year research review and a research agenda. In General and integrative causal models (pp. 49-75). New York, NY: Guildford Press.

56.

Moon

McCluskey

J. D.

McCluskey

C. P.

(2010). A general theory of crime and computer crime: An empirical test. Journal of Criminal Justice, 38, 767-772. doi:10.1016/j.jcrimjus.2010.05.003

57.

Nagin

D. S.

(2005). Group-based modeling of development. Cambridge, MA: Harvard University Press.

58.

Nagin

D. S.

(2010). Group-based trajectory modeling: An overview. In In Piquero

A. R.

Weisburd

(Eds.), Handbook of quantitative criminology (pp. 53-67). New York, NY: Springer.

59.

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, 2094-2100. doi:10.1001/jama.285.16.2094

60.

Osgood

D. W.

Johnston

L. D.

O’Malley

P. M.

Bachman

J. G.

(1988). The generality of deviance in late adolescence and early adulthood. American Sociological Review, 53, 81-93. doi:10.2307/2095734

61.

Pabian

Vandebosch

(2016). Developmental trajectories of (cyber)bullying perpetration and social intelligence during early adolescence. Journal of Early Adolescence, 36, 145-170. doi:10.1177/0272431614556891

62.

Patchin

J. W.

Hinduja

(2010). Cyberbullying and self-esteem. Journal of School Health, 80, 614-621. doi:10.1111/j.1746-1561.2010.00548.x

63.

Patchin

J. W.

Hinduja

(2011). Traditional and nontraditional bullying among youth: A test of general strain theory. Youth & Society, 43, 727-751. doi:10.1177/0044118X10366951

64.

Pellegrini

A. D.

(2002). Affiliative and aggressive dimensions of dominance and possible functions during early adolescence. Aggression and Violent Behavior, 7, 21-31. doi:10.1016/S1359-1789(00)00033-1

65.

Pellegrini

A. D.

Bartini

Brooks

(1999). School bullies, victims, and aggressive victims: Factors relating to group affiliation and victimization in early adolescence. Journal of Educational Psychology, 91, 216-224. doi:10.1037/0022-0663.91.2.216

66.

Pepler

Jiang

Craig

Connolly

(2008). Developmental trajectories of bullying and associated factors. Child Development, 79, 325-338. doi:10.1111/j.1467-8624.2007.01128.x

67.

Pew Research Center. (2012). Internet User Demographics. Retrieved January 25, 2016 from http://www.pewinternet.org/data-trend/teens/internet-user-demographics/

68.

Pew Research Center. (2015). One-fifth of Americans Report Going Online ‘Almost Constantly’. Retrieved January 25, 2016 from http://pewrsr.ch/1NIurjL

69.

Piquero

A. R.

(2000). Frequency, specialization, and violence in offending careers. Journal of Research in Crime & Delinquency, 37, 392-418. doi:10.1177/0022427800037004003

70.

Reijntjes

Vermande

Goossens

F. A.

Olthof

van

Schoot

Aleva

van der Meulen

(2013). Developmental trajectories of bullying and social dominance in youth. Child Abuse & Neglect, 37, 224-234. doi:10.1016/j.chiabu.2012.12.004

71.

Robinson

J. P.

Espelage

D. L.

(2011). Inequities in educational and psychological outcomes between LGBTQ and straight students in middle and high school. Educational Researcher, 40, 315-330. doi:10.3102/0013189X11422112

72.

Runions

Shapka

J. D.

Dooley

Modecki

(2013). Cyber-aggression and victimization and social information processing: Integrating the medium and the message. Psychology of Violence, 3, 9-26. doi:10.1037/a0030511

73.

Russell

S. T.

Kosciw

Horn

Saewyc

(2010). Safe schools policy for LGBTQ students. Sharing Child and Youth Development Knowledge, 24(4), 1-25.

74.

Salmivalli

Karna

Poskiparta

(2011). Counteracting bullying in Finland: The KiVa program and its effects on different forms of being bullied. International Journal of Behavioral Development, 35, 405-411. doi:10.1177/0165025411407457

75.

Slonje

Smith

P. K.

Frisén

(2013). Computers in human behavior the nature of cyberbullying, and strategies for prevention. Computers in Human Behavior, 29, 26-32. doi:10.1016/j.chb.2012.05.024

76.

Smith

P. K.

Mahdavi

Carvalho

Fisher

Russell

Tippett

(2008). Cyberbullying: Its nature and impact in secondary school pupils. Journal of Child Psychology and Psychiatry, 49, 376-385. doi:10.1111/j.1469-7610.2007.01846.x

77.

Sullivan

C. J.

McGloin

J. M.

Pratt

T. C.

Piquero

A. R.

(2006). Rethinking the “norm” of offender generality: Investigating specialization in the short-term. Criminology, 44, 199-233. doi:10.1111/j.1745-9125.2006.00047.x

78.

Tittle

C. R.

Grasmick

H. G.

(1997). Criminal behavior and age: A test of three provocative hypotheses. The Journal of Criminal Law and Criminology, 88, 309-342. doi:10.2307/1144079

79.

Tokunaga

R. S.

(2010). Following you home from school: A critical review and synthesis of research on cyberbullying victimization. Computers in Human Behavior, 26, 277-287. doi:10.1016/j.chb.2009.11.014

80.

Vandebosch

Van Cleemput

(2009). Cyberbullying among youngsters: Profiles of bullies and victims. New Media & Society, 11, 1349-1371. doi:10.1177/1461444809341263

81.

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, 210-227. doi:10.1080/17405629.2011.644919

82.

Wang

Iannotti

R. J.

Nansel

T. R.

(2009). School bullying among adolescents in the United States: Physical, verbal, relational, and cyber. Journal of Adolescent Health, 45, 368-375. doi:10.1016/j.jadohealth.2009.03.021

83.

Wang

Nansel

T. R.

Iannotti

R. J.

(2011). Cyber and traditional bullying: Differential association with depression. Journal of Adolescent Health, 48, 415-417. doi:10.1016/j.jadohealth.2010.07.012

84.

Werner

N. E.

Bumpus

M. F.

Rock

(2010). Involvement in internet aggression during early adolescence. Journal of Youth and Adolescence, 39, 607-619. doi:10.1007/s10964-009-9419-7

85.

Williams

K. R.

Guerra

N. G.

(2007). Prevalence and predictors of internet bullying. Journal of Adolescent Health, 41(6), S14-S21. doi:10.1016/j.jadohealth.2007.08.018

86.

Ybarra

M. L.

Mitchell

K. J.

(2004a). Online aggressor/targets, aggressors, and targets: A comparison of associated youth characteristics. Journal of Child Psychology and Psychiatry, 45, 1308-1316. doi:10.1111/j.1469-7610.2004.00328.x

87.

Ybarra

M. L.

Mitchell

K. J.

(2004b). Youth engaging in online harassment: Associations with caregiver-child relationships, internet use, and personal characteristics. Journal of Adolescence, 27, 319-336. doi:10.1016/j.adolescence.2004.03.007

88.

Ybarra

M. L.

Mitchell

K. J.

(2007). Prevalence and Frequency of Internet Harassment Instigation: Implications for Adolescent Health. Journal of Adolescent Health, 41, 189-195. doi:10.1016/j.jadohealth.2007.03.005

89.

Young

J. T. N.

Rebellon

C. J.

Barnes

J. C.

Weerman

F. M.

(2014). Unpacking the black box of peer similarity in deviance: Understanding the mechanisms linking personal behavior, peer behavior, and perceptions. Criminology, 52, 60-86. doi:10.1111/1745-9125.12029

90.

Yun

Kim

S.-G.

Kwon

(2016). Low self-control among South Korean adolescents: A test of Gottfredson and Hirschi’s generality hypothesis. International Journal of Offender Therapy and Comparative Criminology, 60, 1185-1208. doi:10.1177/0306624X15574683

91.

Zhang

Welte

J. W.

Wieczorek

W. F.

(2002). Underlying common factors of adolescent problem behaviors. Criminal Justice and Behavior, 29, 161-182. doi:10.1177/0093854802029002003