Predictors of Anonymous Cyber Aggression: The Role of Adolescents' Beliefs About Anonymity,Aggression,and the Permanency of Digital Content

Abstract

Little attention has been given to whether adolescents' beliefs about anonymity and their normative beliefs about cyber aggression jointly increase their perpetration of cyber aggression. To this end, the present longitudinal study examined the moderating influence of these variables on the relationships among adolescents' attitudes toward the permanency of digital content, confidence with not getting caught, and anonymous cyber aggression (ACA) assessed 1 year later (Time 2). These associations were examined among 274 7th and 8th graders and through five technologies, including social networking sites (SNS), e-mail, instant messenger (IM), mobile phones, and chatrooms. Findings indicated that increases in Time 2 ACA and attitudes toward the permanency of digital content were more strongly related when adolescents reported greater confidence with not getting caught and higher normative beliefs concerning cyber aggression through SNS and mobile phones. In addition, higher levels of attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs regarding cyber aggression were related to greater Time 2 ACA through e-mail, IM, and chatrooms. All findings are discussed in the context of adolescents' positive attitudes toward ACA, and an appeal for additional research is made to understand more about anonymity in cyberspace.

Introduction

The use of electronic technology allows many conveniences in adolescents' lives, but it is not without consequences.¹ Cyber aggression is one such consequence, which is gaining increasing attention among parents, educators, and researchers.^2–4 To understand cyber aggression better, researchers^5,6 have examined predictors of this behavior, such as anonymity, but little attention has been given to the role of anonymity in cyber aggression. In one of the few studies to examine this relationship, Wright⁷ found that beliefs about anonymity were indirectly related to cyber aggression through the mediation of two explanatory mechanisms (i.e., attitudes toward the permanency of digital content and confidence with not getting caught), which were both directly associated with cyber aggression.

Although Wright's⁷ study added valuable knowledge to the literature, another fruitful direction is to examine whether beliefs about anonymity strengthen the associations among attitudes toward the permanency of digital content, confidence with not getting caught, and cyber aggression. Furthermore, adolescents' normative beliefs about cyber aggression might also provide insight into these relationships, as such beliefs increase cyber aggression. To this end, the present study examined the longitudinal relationships among attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, normative beliefs about cyber aggression, and anonymous cyber aggression (ACA) through social networking sites (SNS), instant messenger (IM), e-mail, mobile phones, and chatrooms. In addition, the moderating effects of beliefs about anonymity and normative beliefs about cyber aggression were also included to advance our understanding of their joint influence on ACA. Findings from this study will contribute valuable knowledge to the literature regarding the impact of adolescents' social cognitions on their behaviors.

Cyber aggression, anonymity, and normative beliefs

Taking various behavioral forms (e.g., threats, insults), cyber aggression occurs through an assortment of electronic technologies (e.g., cell phones, the Internet).^8–12 As a broader form of cyberbullying (i.e., electronic extension of school bullying), cyber aggression includes cyberbullying behaviors but extends the definition to include behaviors that do not typically happen in a face-to-face setting (e.g., hacking into someone's e-mail account).^4,9–15 Behaviors described as cyber aggression may not necessarily mean they include an imbalance of power or repetition, such as outing, happy slapping, and flaming. An example of cyber aggression that includes these components is a knowledgeable individual (i.e., imbalance of power) repeatedly hacking someone's Facebook account (i.e., repetition). It is not repetition or an imbalance of power, central to the definition of cyberbullying, that define what behaviors constitute cyber aggression. Instead, it is the behavior (i.e., hacking someone's Facebook account), regardless of repetition and an imbalance of power, that makes the behavior cyber aggression instead of cyberbullying. The behavior is cyber aggression when it has no face-to-face equivalent, unlike the definition of cyberbullying. Prevalence rates of adolescents' cyberbullying and cyber aggression involvement vary from 4.1% to 31.5%, depending on the sample and measurement.^9,16 Methodology issues impact frequencies of cyberbullying and cyber aggression, as standardized measures for assessing these behaviors have not been identified in the literature. Therefore, some studies utilize one item to assess these behaviors, others replicate scales on face-to-face aggression and then describe these behaviors as occurring online, and yet other researchers ask respondents how often they experienced cyberbullying. Such measurement differences contribute to adolescents' understanding of the questionnaires, impacting their responses.

With the growing interest in cyber aggression, researchers are concerned with identifying predictors of these behaviors.^9–13 Hypothesized^17–22 as a contributor to cyber aggression, anonymity allows adolescents to control their self-presentation and to avoid the consequences of their actions, influencing their impulsive behaviors. In the literature, around 9–69% of adolescents reported that they did not know the identity of an online perpetrator.^16,21–23 Although many researchers^16,22,23 have focused on frequency rates of ACA, other researchers^7,24 have aimed to understand the variables that might influence anonymity in the cyber context. In particular, Bartlett and Gentile²⁴ found that favorable attitudes toward anonymous cyberbullying predicted these behaviors among young adults. Utilizing a short-term longitudinal design, Wright⁷ found that beliefs about anonymity indirectly related to cyber aggression 6 months later through young adults' attitudes toward the permanency of digital content and their confidence with not getting caught. To follow up these studies, the present study examined anonymity and cyber aggression among early adolescents, who have the highest rates of these behaviors.²⁵ Such an investigation is important, as adolescents' beliefs about anonymity might differ from young adults because cyber aggression may be more common in adolescents' lives. Furthermore, adolescents utilize various electronic technologies, and some may be easier for adolescents to remain anonymous.^26,27 Therefore, the present study investigated all variables through SNS, IM, e-mail, mobile phones, and chatrooms. Adolescents are able to use fake information and user names through these technologies, making it tough to identify which technology allows them to remain the most anonymous.²⁸ Mobile phones are different from the other technologies, as sending text messages can reveal someone's phone number. However, many adolescents utilize mobile phones for the Internet, and through the web browser, they are able to send anonymous e-mails and log into fake SNS.²⁹ Also, many mobile phones allow adolescents to utilize various chat programs in which they can anonymously send harassing IMs.²⁸ In sum, these technologies offer adolescents some degree of anonymity, warranting an investigation of each technology.

In the literature, researchers have found positive links between normative beliefs about cyber aggression and engagement in these behaviors.^30–32 Higher normative beliefs about cyber aggression guide adolescents' acceptance of aggressive behaviors, whereas endorsing higher beliefs about anonymity indicate their perception of how easy it is to remain anonymous through a particular technology. Although distinctive processes, the combined influence of normative beliefs about cyber aggression and beliefs about anonymity might increase cyber aggression, along with adolescents' attitudes toward the permanency of digital content and their confidence with not getting caught. To this end, the present study investigated the moderating influence of normative beliefs about cyber aggression and beliefs about anonymity when examining the relationship of attitudes toward the permanency of digital content and confidence with not getting caught to ACA assessed 1 year later. Three-way and four-way interactions were also included among each of these variables.

People sometimes do and say things online that they would not do or say face to face. This phenomenon occurs often, and it is known as the Online Disinhibition Effect,³³ which refers to the loosening or abandonment of social restrictions and inhibitions in an online environment. These restrictions and inhibitions are usually present in normal face-to-face interactions. Working in tandem with the Online Disinhibition Effect, anonymity in the cyber context allows people to separate their actions from the real world and their identity. Utilizing the Online Disinhibition Effect as a theoretical foundation, it was expected that social inhibitions present in adolescents' face-to-face interactions would be reduced in the cyber context when they endorsed greater normative beliefs about cyber aggression, beliefs about anonymity, attitudes toward the permanency of digital content, and confidence with not getting caught, which would each contribute to greater ACA perpetration. Due to the lack of research, hypotheses regarding whether these associations would vary based on the type of electronic technology were not made.

Methods

Participants

Participants were 274 7th and 8th graders (51% male; M_age=12.62, SD_age=0.68) from one Midwestern middle school. Adolescents primarily identified as white (74.5%), followed by Latino/a (12.8%), Black/African American (7.3%), Asian (5.1%), and other (0.3%).

Measures and procedures

After meeting with a school principal, classroom announcements were made in the 6th and 7th grade classrooms to explain the purpose of the study, how adolescents could participate, assure them that their participation was voluntary, and that their answers were confidential. There were 306 parental permission slips passed out, and 300 were returned, with 286 adolescents having their parents' permission. During the fall, adolescents provided their assent (100% participating) and completed questionnaires on: (a) self-reported cyber aggression, (b) self-reported ACA, (c) beliefs about anonymity, (d) attitudes toward the permanency of digital content, (e) confidence with not getting caught, and (f) normative beliefs about cyber aggression. The following year, in the 7th or 8th grade (Time 2), adolescents completed measures on self-reported ACA. Twelve adolescents did not participate because they were either absent (n=6) during the 3 days of data collection or had moved away (n=6).

Self-reported cyber aggression

This measure asked adolescents how often (1=“never”; 5=“all the time”) they engaged in aggressive behaviors through SNS, IM, e-mail, mobile phones, and chatrooms.²² There were six questions, including: “How often do you gossip about others through [technologies listed]?” and “How often do you insult another peer through [technologies listed]?” All items were averaged to form a final score (α=0.91).

Self-reported ACA

This measure asked adolescents how often (1=“never”; 5=“all the time”) they remained anonymous while engaging in aggressive behaviors through SNS, IM, e-mail, chatrooms and mobile phones. The six questions from the previous measure were asked for the five technologies, resulting in a total of 30 questions. Items were averaged to form final scores for each of the technologies. This measure was administered at Time 1 (α=0.83 for SNS; α=0.91 for mobile phones; α=0.86 for IM; α=0.86 for e-mail; α=−.80 for chatrooms) and Time 2 (α=0.83 for SNS; α=0.88 for mobile phones; α=0.89 for IM; α=0.89 for e-mail; α=0.82 for chatrooms).

Beliefs about anonymity

Adolescents were asked how easy (1=“not easy at all”; 4=“very easy”) it would be for them to hide their identity while engaging in aggressive behaviors through SNS, IM, e-mail, mobile phones, and chatrooms.⁷

Attitudes toward the permanency of digital content

Adolescents rated the statement “I do not believe that anything you say or write about another person through [insert technology] stays in ‘cyberspace’ in some form. That is, if someone does something mean to someone through [insert technology] it ‘goes away’” on a scale of 1 (“strongly disagree”) to 5 (“strongly agree”).⁷ This statement was asked about each technology (SNS, IM, e-mail, chatrooms, and mobile phones).

Confidence with not getting caught

A description of ACA was given through SNS, IM, e-mail, chatrooms, and mobile phones. Next, adolescents rated the statement “I am confident that I will not be caught when engaging in these behaviors through [insert technology]” on a scale of 1 (“not confident at all”) to 7 (“complete confidence”).⁷

Normative beliefs about cyber aggression

This measure assessed the extent to which adolescents believe that cyber aggression is acceptable, tolerable, and typical.^32,34 Adolescents were presented with a description of cyber aggression (i.e., spreading rumors, gossiping, insulting, teasing, threatening to hurt someone physically, sending hurtful messages to someone else). Next, they rated acceptability, tolerability, and typicality of acting this way through SNS, IM, chatrooms, e-mail, and mobiles phones using a scale of 1 (“very unacceptable/completely intolerable/very atypical”) to 5 (“very acceptable/completely tolerable/very typical”). These three aspects were averaged to form final scores for SNS (α=0.88), IM (α=0.88), e-mail (α=0.91), mobile phones (α=0.90), and chatrooms (α=0.83).

Results

To examine how the variables in the study longitudinally relate to ACA 1 year later, correlational and multiple hierarchical regression analyses were conducted. To facilitate reading, Table 1 presents all variables as they relate to Time 2 ACA. Next, hierarchical multiple regressions were conducted with ACA through five technologies (SNS, IM, e-mail, mobile phones, chatrooms) as the dependent variables. Block 1 included control variables, such as gender, Time 1 cyber aggression, and Time 1 ACA. Block 2 included attitudes toward the permanency of digital content. Block 3 included confidence with not getting caught. Block 4 included beliefs about anonymity. Block 5 included normative beliefs about cyber aggression. Block 6 included two-way interactions between attitudes toward the permanency of digital content and beliefs about anonymity, between confidence with not getting caught and beliefs about anonymity, between attitudes toward the permanency of digital content and normative beliefs about cyber aggression, and between confidence with not getting caught and normative beliefs about cyber aggression. Block 7 included three-way interactions among attitudes toward the permanency of digital content, confidence with not getting caught, and beliefs about anonymity, and among attitudes toward the permanency of digital content, confidence with not getting caught, and normative beliefs about cyber aggression. Block 8 included a four-way interaction among attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression. All continuous variables were centered to protect against multicollinearity.³⁵ Multicollinearity was also checked by examining the variance inflation factor (VIF). VIF was not greater than 1.17 for any of the variables, indicating that multicollinearity was not an issue.

Table 1.

Correlations of All Variables to Time 2 ACA

	T2 ACA–SNS	T2 ACA–IM	T2 ACA–e-mail	T2 ACA–mobile phones	T2 ACA–chatrooms
T1 ACA–SNS	0.61^***	0.18^*	0.05	0.16^*	0.08
T1 ACA–IM	0.12^*	0.53^***	0.19^*	0.26^*	0.29^***
T1 ACA–e-mail	0.01	0.19^*	0.67^***	0.31^*	0.33^***
T1 ACA–mobile phones	0.12^*	0.20^*	0.23^*	0.53^***	0.03
T1 ACA–chatrooms	0.08	0.36^***	0.31^***	0.02	0.49^***
T1 CA	0.33^***	0.45^***	0.45^***	0.26^***	0.31^***
T1 BAA–SNS	0.43^***	0.10^****	0.05	0.02	0.01
T1 BAA–IM	0.12^****	0.49^***	0.02	0.09^****	0.37^***
T1 BAA–e-mail	0.05	0.27^***	0.56^***	0.10^****	0.04
T1 BAA–mobile phones	0.02	0.13^*	0.16^**	0.50^***	0.01
T1 BAA–chatrooms	0.12^*	0.04	0.08	0.05	0.51^***
T1 ATPDC–SNS	0.60^***	0.01	0.07	0.18^**	0.18^**
T1 ATPDC–IM	0.06	0.57^***	0.51^***	0.05	0.32^***
T1 ATPDC–e-mail	0.42^***	0.39^***	0.40^***	0.19^***	0.16^*
T1 ATPDC–mobile phones	0.11^****	0.31^***	0.05	0.54^***	0.14^*
T1 ATPDC–chatrooms	0.12^****	0.24^***	0.10	0.26^***	0.53^***
T1 CWNGC–SNS	0.52^***	0.24^***	0.01	0.26^***	0.34^***
T1 CWNGC–IM	0.34^***	0.53^***	0.49^***	0.05	0.37^***
T1 CWNGC–e-mail	0.37^***	0.17^**	0.56^***	0.01	0.14^*
T1 CWNGC–mobile phones	0.01	0.22^***	0.40^***	0.58^***	0.11^****
T1 CWNGC–chatrooms	0.05	0.44^***	0.17^**	0.06	0.52^***
T1 NBCA–SNS	0.65^***	0.16^**	0.10+	0.09^****	0.20^***
T1 NBCA–IM	0.31^***	0.62^***	0.44^***	0.11^****	0.36^***
T1 NBCA–e-mail	0.26^***	0.16^**	0.61^***	0.16^**	0.11^****
T1 NBCA–mobile phones	0.10^****	0.26^***	0.23^***	0.59^***	0.19^**
T1 NBCA–chatrooms	0.09^****	0.43^***	0.19^**	0.54^***	0.54^***

This table provides the relations of all variables to Time 2 anonymous cyber aggression through all technologies. If interested in the full correlation matrix, please e-mail the first author.

****

p<0.1; ^*p<0.05; ^**p<0.01; ^***p<0.001.

T1, Time 1; ACA, anonymous cyber aggression; SNS, social networking site; IM, instant messenger; CA, cyber aggression; BAA, beliefs about anonymity; ATPDC, attitudes toward the permanency of digital content; CWNGC, confidence with not getting caught; NBCA, normative beliefs about cyber aggression.

Gender did not predict Time 2 ACA through the five technologies (see Table 2 for SNS and mobile phones; Table 3 for e-mail, IM, and chatrooms). Time 1 cyber aggression related positively to Time 2 ACA through all technologies. In addition, Time 1 ACA was associated positively with Time 2 ACA for all technologies. Furthermore, attitudes toward the permanency of digital content, confidence with not getting caught, and normative beliefs about cyber aggression were related positively to Time 2 ACA through each of the five technologies. Beliefs about anonymity were not related to Time 2 ACA through SNS or mobile phones.

Table 2.

Regression Analyses for Time 2 ACA Through SNS and Mobile Phones

	Time 2 ACA–SNS			Time 2 ACA–mobile phones
	β	R²	R²	β	R²	R²
Block 1		.05	.05^***		.03	.03^**
Gender	.02			.01
T1 CA	.06^*			.04^*
T1 ACA	.12^*			.13^*
Block 2		.66	.61^***		.62	.61^***
ATPDC	.32^***			.26^***
Block 3		.80	.24^***		.79	.27^***
CWNGC	.21^***			.23^**
Block 4		.80	.01^*		.79	.01^**
BAA	.14			.07
Block 5		.85	.04^***		.81	.02^***
NBCA	.19^**			.20^**
Block 6		.87	.02^**		.83	.02^***
ATPDC×BAA	.05			.03
CWNGC×BAA	.06			.02
ATPDC×NBCA	.26^***			.25^**
CWNGC×NBCA	.20^*			.22^**
Block 7		.89	.02^***		.83	.02^***
ATPDC×CWNGC×BAA	.08			.06
ATPDC×CWNGC×NBCA	.34^***			.29^***
Block 8		.89	.01		.83	.01
ATPDC×CWNGC×BAA×NBCA	.06			.03

p<0.05; ^**p<0.01; ^***p<0.001.

Table 3.

Regression Analyses for Time 2 ACA Through E-Mail, IM, and Chatrooms

	Time 2 ACA–e-mail			Time 2 ACA–IM			Time 2 ACA–chatrooms
	β	R²	ΔR²	β	R²	ΔR²	β	R²	ΔR²
Block 1		0.02	0.02^*		0.03	0.03^*		0.04	0.04^**
Gender	0.01			0.01			0.02
T1 CA	0.03^*			0.03^*			0.04^*
T1 ACA	0.13^**			0.10^*			0.15^**
Block 2		0.66	0.63^***		0.66	0.63^***		0.60	0.55^***
ATPDC	0.27^***			0.27^**			0.30^***
Block 3		0.80	0.24^***		0.70	0.04^***		0.67	0.17^***
CWNGC	0.25^***			0.26^***			0.24^**
Block 4		0.81	0.01^***		0.77	0.07^***		0.70	0.03^***
BAA	0.21^***			0.24^***			0.23^***
Block 5		0.83	0.03^***		0.77	0.01^*		0.73	0.02^***
NBCA	0.17^**			0.15^*			0.19^**
Block 6		0.84	0.01^***		0.83	0.06^***		0.74	0.01^***
ATPDC×BAA	0.22^**			0.25^*			0.27^*
CWNGC×BAA	0.25^***			0.34^***			0.23^*
ATPDC×NBCA	0.15^*			0.26^*			0.26^*
CWNGC×NBCA	0.18^*			0.27^*			0.27^*
Block 7		0.85	0.01^***		0.84	0.01^***		0.76	0.02^***
ATPDC×CWNGC×BAA	0.31^**			0.19^*			0.30^***
ATPDC×CWNGC×NBCA	0.26^*			0.23^**			0.29^***
Block 8		0.86	0.01^***		0.86	0.01^***		0.80	0.04^***
ATPDC×CWNGC×BAA×NBCA	0.22^***			0.33^***			0.35^***

p<0.05; ^**p<0.01; ^***p<0.001.

Significant two-way interactions were found between attitudes toward the permanency of digital content and normative beliefs about cyber aggression as well as confidence with not getting caught and normative beliefs about cyber aggression when predicting Time 2 ACA through all technologies. In addition, two-way interactions between attitudes toward the permanency of digital content and beliefs about anonymity, and confidence with not getting caught and beliefs about anonymity were significant for e-mail, IM, and chatrooms. Furthermore, three-way interactions among attitudes toward the permanency of digital content, confidence with not getting caught, and normative beliefs about cyber aggression were significant for Time 2 ACA via all technologies. On the other hand, the three-way interactions among attitudes toward the permanency of digital content, confidence with not getting caught, and beliefs about anonymity were significant only for Time 2 ACA through e-mail, IM, and chatrooms. The four-way interaction among attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression was significant for Time 2 ACA via e-mail, IM, and chatrooms.

Based on differences in the significant interactions, three-way interactions among attitudes toward the permanency of digital content, confidence with not getting caught, and normative beliefs about cyber aggression were probed further for Time 2 ACA through SNS and mobile phones. The associations between attitudes toward the permanency of digital content and Time 2 ACA through SNS (see Figure 1) and mobiles phones (see Figure 2) were greater for adolescents with high confidence with not getting caught and high normative beliefs about cyber aggression when compared to high confidence with not getting caught and low normative beliefs about cyber aggression (t=2.65, p=0.009 for SNS; t=3.19, p=0.002 for mobile phones), low confidence with not getting caught and high normative beliefs about cyber aggression (t=3.08, p=0.002 for SNS; t=4.09, p<0.001 for mobile phones), and low confidence with not getting caught and low normative beliefs about cyber aggression (t=4.35, p<0.001 for SNS; t=5.53, p<0.001 for mobile phones).

FIG. 1.

Depiction of the three-way interaction among ATPDC, CWNGC, and NBCA when predicting Time 2 anonymous cyber aggression through social networking sites. ATPDC, attitudes toward the permanency of digital content; CWNGC, confidence with not getting caught; NBCA, normative beliefs about cyber aggression.

FIG. 2.

Depiction of the three-way interaction among ATPDC, CWNGC, and NBCA when predicting Time 2 anonymous cyber aggression through mobile phones. ATPDC, attitudes toward the permanency of digital content; CWNGC, confidence with not getting caught; NBCA, normative beliefs about cyber aggression.

For Time 2 ACA through e-mail, IM, and chatrooms, the four-way interactions among attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression were probed. Due to such interactions being difficult to depict graphically, “high” versus “low” subgroups were created using median split analyses.³⁵ The high subgroup (n=132) was created by selecting respondents who were equal to or above the median on the four attitudes toward ACA. Similarly, respondents who had scores below the median on the four attitudes were included in the low subgroup (n=133). The mean for adolescents' engagement in Time 2 ACA through e-mail was 3.01 (SD=0.18) for the low subgroup and 4.25 (SD=0.58) for the high subgroup. For Time 2 ACA via IM, the mean for the low subgroup was 3.47 (SD=0.51) and 3.61 (SD=0.49) for the high subgroup. For Time 2 ACA through chatrooms, the low subgroup had a mean of 2.21 (SD=0.70), while the mean for the high subgroup was 2.45 (SD=0.77). Using independent samples t tests, findings indicated that the high subgroup engaged in greater Time 2 ACA through e-mail, t(263)=−25.85, p<0.001; IM, t(263)=−2.43, p<0.05; and chatrooms, t(263)=−2.71, p<0.01.

Discussion

Although argued as an important mechanism distinguishing face-to-face and cyber aggression, anonymity has been understudied in the literature.^{3,7,8,17,18,21,23,24} Many technologies allow adolescents to remain anonymous, reducing their social accountability and inhibitions toward aggressive acts.^19,20 The present study aimed to understand the combined influence of attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression on adolescents' engagement in ACA 1 year later (Time 2).

Findings indicated that adolescents' perpetration of Time 2 ACA was differentiated based on electronic technology and their attitudes toward such behaviors. Consistent with the literature,^7,36 greater attitudes toward ACA (i.e., attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, normative beliefs about cyber aggression) predicted higher Time 2 ACA. The two exceptions to these patterns were for beliefs about anonymity through SNS and mobile phones. After including all other variables in the analyses, beliefs about anonymity were no longer associated with Time 2 ACA via SNS and mobile phones. A potential explanation may be that attitudes toward the permanency of digital content, confidence with not getting caught, and normative beliefs about cyber aggression accounted for more variance in their association with Time 2 ACA, potentially indicating that these variables were better predictors of such behaviors through SNS and mobile phones.

Although main effects of attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression were found for most of the technologies, we are better able to understand the complexity of the relationship between these variables and Time 2 ACA through the significant three-way and four-way interactions. Three-way interactions were found among attitudes toward the permanency of digital content, confidence with not getting caught, and normative beliefs about cyber aggression for SNS and mobile phones. Such findings indicate that higher levels of normative beliefs about cyber aggression and confidence with not getting caught made the positive association stronger between attitudes toward the permanency of digital content and Time 2 ACA through SNS and mobile phones. These findings are consistent with the literature, as confidence with not getting caught and normative beliefs about cyber aggression each relate to cyber aggression.^7,30–32 Therefore, the contribution of confidence with not getting caught and normative beliefs about cyber aggression to ACA cannot be ignored, as the combination of these variables relate to greater levels of ACA through SNS and mobile phones. Adolescents' normative beliefs about cyber aggression and confidence with not getting caught may lessen their concerns with engaging in negative behaviors, such as ACA.^7,36

The four-way interactions among attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression are better representations of Time 2 ACA through e-mail, IM, and chatrooms. Supported by previous research, higher levels of these attitudes toward ACA related to greater levels of ACA 1 year later. Drawing on the Online Disinhibition Effect,³³ holding higher levels of attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression may help adolescents disregard the social inhibitions present in their face-to-face interactions that usually prevent them from acting aggressively. Therefore, it may be reasonable that a combination of these variables contributes to this disinhibition, resulting in more ACA. In addition, only three-way interactions for SNS and mobile phones were found versus the four-way interactions for IM, e-mail, and chatrooms, indicating that the combined influence of attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression was more predictive of Time 2 ACA through e-mail, IM, and chatrooms. These findings may indicate that e-mail, IM, and chatrooms are perceived by adolescents as more anonymous technologies when compared to SNS and mobile phones.²⁸

Despite contributing to the literature on cyber aggression, there are a few limitations and future directions that should be mentioned to further this literature. First, one item was used to access all constructs, except for ACA and normative beliefs about cyber aggression. Although the reliability and validity of these one-item constructs may be questionable, results from the present study replicated previous research utilizing these constructs.⁷ However, researchers should consider adding additional items to understand more about adolescents' attitudes toward ACA. Such items may include specific behaviors adolescents use to mask their identity through electronic technologies. Another important direction is for researchers to examine the attitudes toward ACA in younger populations. Adolescents begin using electronic technologies in early childhood, which may impact their technological savvy.³⁷ Being technologically savvy may provide children with the skills they need to remain anonymous through these technologies. Future research should also focus on removing self-report biases by including peer nominations of cyberbullying perpetration and victimization. The literature^22,38 indicates that peer nominations of these behaviors provide reliable and valid information concerning cyberbullying.

Contributing to research aimed at understanding predictors of cyber aggression, the present study provided an initial investigation of the conjoint impact of attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression on later ACA. Findings from the present study indicate that attitudes toward the permanency of digital content and Time 2 ACA through SNS and mobiles phones were more strongly connected when adolescents reported higher levels of confidence with not getting caught and normative beliefs about cyber aggression. Similarly, holding greater attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression related positively to higher Time 2 ACA via IM, email, and chatroom. Although Finkelhof et al.³⁹ found that rates of victimization have decreased over the past few years, results of the present study indicate that such behaviors continue to occur among adolescents and that efforts should be made to reduce victimization. This study may inform prevention and intervention efforts by identifying individuals prone to ACA based on their higher endorsement of attitudes toward the permanency of digital content, confidence with not getting caught, beliefs about anonymity, and normative beliefs about cyber aggression.

Footnotes

Acknowledgment

This work was supported by the project “Employment of Best Young Scientists for International Cooperation Empowerment” (cz 1.07/23.00/30.0037) co-financed from European Social Fund and the state budget of the Czech Republic.

Author Disclosure Statement

No competing financial interests exist.

References

Subrahmanyam

, Greenfield

. Online communication and adolescent relationships. The Future of Children, 2008, 18:119–146.

Aricak

, Siyahhan

, Uzunhasanoglu

, et al. Cyberbullying among Turkish adolescents. CyberPsychology, Behavior, & Social Networking, 2008; 11: 253–261.

Heirman

, Walrave

. Assessing concerns and issues about the mediation of technology in cyberbullying. CyberPsychology: Journal of Psychosocial Research on Cyberspace, 2008; 2.

Smith

, Mahdavi

, Carvalho

, et al. Cyberbullying: its nature and impact in secondary school pupils. Journal of Child Psychology & Psychiatry, 2008; 49:376–385.

Mason

. Cyberbullying: a preliminary assessment for school personnel. Psychology in the Schools, 2008; 45:323–348.

Varjas

, Talley

, Meyers

, et al. High school students' perceptions of motivation for cyberbullying: an exploratory study. Western Journal of Emergency Medicine, 2010; 11:269–273.

Wright

. The relationship between young adults' beliefs about anonymity and subsequent cyber aggression. CyberPsychology, Behavior, & Social Networking, 2013; 16:858–862.

Patchin

, Hinduja

. Bullies move beyond the schoolyard: a preliminary look at cyberbullying. Youth Violence & Juvenile Justice, 2006; 4:148–169.

Pornari

, Wood

. Peer and cyber aggression in secondary school students: the role of moral disengagement, hostile attribution bias, and outcome expectancies. Aggressive Behaviors, 2010; 36:81–94.

10.

Slonje

, Smith

. Cyberbullying: another main type of bullying?. Scandinavian Journal of Psychology, 2008; 49:147154.

11.

Wang

, Iannotti

, Nansel

. School bullying among adolescents in the United States: physical, verbal, relational, and cyber. Journal of Adolescent Health, 2009; 45:368–375.

12.

Williams

, Guerra

. Prevalence and predictors of internet bullying. Journal of Adolescent Health, 2007; 41:14–21.

13.

Grigg

. Cyber-aggression: definition and concept of cyberbullying. Australian Journal of Guidance & Counseling, 2010; 20:143–156.

14.

Walker

, Sockman

, Koehns

. An exploratory study of cyberbullying with undergraduate students. TechTrends, 2011; 55:31–38.

15.

König

, Gollwitzer

, Steffgen

. Cyberbullying as an act of revenge?. Australian Journal of Guidance & Counseling, 2010; 20:210–224.

16.

Kowalski

, Limber

. Electronic bullying among middle school students. Journal of Adolescent Health, 2007; 41:22–30.

17.

Joinson

. Knowing me, knowing you: Reciprocal self-disclosure on the Internet. CyberPsychology, Behavior, & Social Networking, 2001; 4:587–591.

18.

Joinson

. Self-disclosure in computer-mediated communication: the role of self-awareness and visual anonymity. European Journal of Social Psychology, 2001; 31:177–192.

19.

Valkenburg

, Peter

, Schouten

. Friend networking sites and their relationship to adolescents' well-being and social self-esteem. CyberPsychology, Behavior, & Social Networking, 2006; 9:84–90.

20.

Valkenburg

, Peter

. Online communication among adolescents: an integrated model of its attraction, opportunities, and risks. Journal of Adolescent Health, 2010; 48:121–127.

21.

Ybarra

, Mitchell

. Online aggressor/targets, aggressors, and targets: a comparison of associated youth characteristics. Journal of Child Psychology & Psychiatry, 2004; 45:308–316.

22.

Wright

, Li

. The association between cyber victimization and subsequent cyber aggression: the moderating effect of peer rejection. Journal of Youth & Adolescence, 2013; 42:662–674.

23.

Moore

, Nakano

, Enomoto

, et al. Anonymity and roles associated with aggressive posts in an online forum. Computers in Human Behavior, 2012; 28:861–867.

24.

Barlett

, Gentile

. Attacking others online: the formation of cyber-bullying in late adolescence. Psychology of Popular Media Culture, 2012; 1:123–135.

25.

Sevcıkova

, Smahel

. Online harassment and cyberbullying in the Czech Republic: comparison across age groups. Journal of Psychology, 2009; 217, 227–229.

26.

Lenhart

, Madden

, Hitlin

. Teens and technology: youth are leading the transition to a fully wired and mobile nation. www.pewinternet.org/∼/media/Files/Reports/2005/PIP_Teens_Tech_July2005web.pdf (accessed February 6, 2013).

27.

Lenhart

, Purcell

, Smith

, et al. Social media and mobile Internet use among teens and young adults. http://web.pewinternet.org/*/media/files/reports/2010/pip_social_media_and_young_adults_report_final_with_toplines.pdf (accessed February 6, 2013).

28.

Willard

. (2006) Cyberbullying and cyberthreats: responding to the challenge of online social cruelty, threats and distress. Eugene: Research Press.

29.

Madden

, Lenhart

, Duggan

, et al. Teens and Technology. 2013. www.pewinternet.org/Reports/2013/Teens-and-Tech/Main-Findings/Teens-and-Technology.aspx (accessed February 6, 2013).

30.

Ang

, Tan

, Mansor

. Normative beliefs about aggression as a mediator of narcissistic exploitativeness and cyberbullying. Journal of Interpersonal Violence, 2011; 26:2619–2634.

31.

Nicol

, Fleming

. “I h8 u”: the influence of normative beliefs and hostile response selection in predicting adolescents' mobile phone aggression—a pilot study. Journal of School Violence, 2010; 9:212–231.

32.

Wright

, Li

. Normative beliefs about aggression and cyber aggression among young adults: a longitudinal investigation. Aggressive Behavior, 2013; 39:161–170.

33.

Suler

. The online disinhibition effect. CyberPsychology & Behavior, 2004; 7:321–326.

34.

Huesmann

, Guerra

. Normative beliefs and the development of aggressive behavior. Journal of Personality & Social Psychology, 1997; 72:1–12.

35.

Aiken

, West

. (1991) Multiple regression: testing and interpreting interactions. Newbury Park, CA: Sage.

36.

Solove

. (2007) The future of reputation: gossip, rumor, and privacy on the Internet. Washington, DC: Yale University Press.

37.

Heim

, Brandtzaeg

, Kaare

, et al. Children's usage of media technologies and psychosocial factors. New Media & Society, 2007; 9:425–454.

38.

Wright

. Longitudinal investigation of the associations between adolescents' popularity and cyber social behaviors. Journal of School Violence, 2013. [Epub ahead of print]; DOI: 10.1080/15388220.2013.849201.

39.

Finkelhor

, Turner

, Ormrod

, et al. Trends in childhood violence and abuse exposure: evidence from 2 national surveys. Archives of Pediatrics & Adolescent Medicine, 2010; 163:238–242.