Description

profilereddygs17
ResearchArticle4.pdf

Cyberbullying on Social Networking Sites: The Crime Opportunity and Affordance Perspectives

TOMMY K. H. CHAN, CHRISTY M. K. CHEUNG, AND RANDY Y. M. WONG

Tommy K. H. Chan ([email protected]; corresponding author) is a Lecturer in Business Information Management at Northumbria University, United Kingdom. He earned a Ph.D. in Information Systems and e-Business Management from Hong Kong Baptist University. Dr. Chan’s research interests include societal implications of information technology use, such as cyberbullying and game addiction, and online consumer behaviors, such as customer engagement and social media firestorm. His work has been published in such journals as Information & Management, Industrial Marketing Management, Electronic Commerce Research and Applications, Internet Research, and others.

Christy M. K. Cheung ([email protected]) is an Associate Professor at Hong Kong Baptist University. She earned a Ph.D. in Information Systems from the College of Business at City University of Hong Kong. Her research interests include technology use and well-being, IT adoption and use, societal implications of IT use, and social media. She has published over one hundred refereed articles in scholarly journals and conference proceedings, including, Journal of Management Information Systems, Journal of Information Technology, Journal of the Association for Information Science and Technology, and MIS Quarterly, among others. Dr. Cheung is President of the Association for Information Systems - Hong Kong Chapter. She also serves as Editor-in-Chief of Internet Research.

Randy Y. M. Wong ([email protected]) is a Ph.D. candidate in the Department of Finance and Decision Sciences at Hong Kong Baptist University. Her research interests include social media and social networking, and societal implications of technology use. Her work has appeared in Computers in Human Behavior as well as in the proceedings of the International Conference on Information Systems, European Conference on Information Systems, Pacific Asia Conference on Information Systems, and Hawaii International Conference on System Sciences.

ABSTRACT: Cyberbullying on social networking sites (SNS bullying) is an emerging societal challenge related to the deviant use of technologies. To address the research gaps identified in the literature, we draw on crime opportunity theory and the affordance perspective to propose a meta-framework that guides our investigation into SNS bullying. The meta-framework explains how SNS affordances give rise to the evaluation of favorable SNS environmental conditions for SNS bullying, which,

Journal of Management Information Systems / 2019, Vol. 36, No. 2, pp. 574–609.

Copyright © Taylor & Francis Group, LLC

ISSN 0742–1222 (print) / ISSN 1557–928X (online)

DOI: https://doi.org/10.1080/07421222.2019.1599500

in turn, promote SNS bullying. The research model was empirically tested using a longitudinal online survey of 223 SNS users. The results suggest that the evalua- tion of SNS environmental conditions predict SNS bullying, and SNS affordances influence the evaluation of these environmental conditions. This work offers a new theoretical perspective to study SNS bullying, highlighting the critical impacts of environmental conditions in shaping such behavior. It also provides actionable insights into measures that combat SNS bullying.

KEY WORDS AND PHRASES: cyberbullying, SNS bullying, crime opportunity, affor- dance, social networking sites, meta-framework, societal impacts of technology use, IT deviant use.

Introduction

Social networking sites (SNSs) have become increasingly popular vehicles for individuals to communicate with their friends and family, anytime and any- where. Despite their promising potential for online social interactions, SNSs are also ripe for abuse because they provide perpetrators with an ideal venue for cyberbullying—in other words, for harassing, threatening, and exploiting poten- tial targets. Cyberbullying on social networking sites (SNS bullying) refers to any form of aggressive behavior on SNSs conducted by a group or an indivi- dual, repeatedly and over time, against targets who cannot easily defend them- selves [88]. SNS bullying is a relatively recent phenomenon; however, researchers have

already devoted much attention to reporting and documenting its prevalence and the adverse consequences associated with it. The Pew Research Center [74] found that 40 percent of Internet users had experienced cyberbullying. Facebook has been found to be the most common venue for SNS bullying: 54 percent of Facebook users reported that they have experienced cyberbully- ing on Facebook [37]. Previous research has demonstrated that SNS bullying incidents have adverse consequences for victims (e.g., Sticca et al. [91]), such as depression, anxiety, low self-esteem, substance abuse, and in extreme cases, self-harming behaviors and suicide attempts. Frequent news headlines report- ing suicide cases linked to SNS bullying document the severity of this pro- blem, including, for example, the recent case of an eighteen-year-old girl who shot herself dead in front of her family after being relentlessly bullied for her weight on Facebook [40]. Given its adverse consequences on individuals and society, SNS bullying has not

surprisingly become an important and emerging research topic across disciplines. With roots in psychology, education, and public health research, most studies have focused on individual traits and characteristics that lead to SNS bullying (or to cyberbullying in general) [see 39, 47, for a review]. However, the research into SNS bullying is still emerging in the information systems (IS) discipline. Only recently Lowry et al. [56] drew on social learning theory to examine how social media

CYBERBULLYING ON SOCIAL NETWORKING SITES 575

anonymity affects adults’ engagement in SNS bullying. In general, there have been few investigations into the phenomenon within the IS discipline. How SNS, as a form of new information technology, shapes and fosters cyberbullying remains relatively unexplored from a technological perspective. Understanding SNS bullying from a technological perspective is vital in order

to shed light onto new measures that may effectively combat this emerging societal challenge, given that existing research has mostly focused on identifying individual characteristics associated with SNS bullying. Indeed, numerous social science theories, such as social cognitive theory and crime opportunity theory, have stressed the importance of the environment in shaping human behaviors. Neglecting the environmental component in SNS bullying research could be potentially dangerous because this produces a lopsided view into the causes of the phenomenon. Accordingly, our study aims to advance the scientific understanding of cyberbullying

by developing a meta-framework that explains how SNS affordances and the evalua- tion of favorable SNS environmental conditions influence SNS bullying. We use crime opportunity theory [30] to explain SNS bullying, considering both the perpetrator characteristic and SNS environmental conditions that offer the criminogenic opportu- nities. We further adopt the affordance perspective [63] to delineate how SNS affor- dances give rise to such a favorable evaluation of the environmental conditions for SNS bullying. We endeavor to answer two primary research questions:

Research Question 1: What are the key factors driving SNS bullying?

Research Question 2: How do SNS affordances influence the evaluation of SNS environmental conditions for SNS bullying?

This work responds to calls for research on the societal impacts of technology use (e.g., Majchrzak et al. [61] and Tarafdar et al. [94]) and contributes to theory and practice in three distinct ways. First, this work advances the scientific knowledge of cyberbullying by investigating how the SNS environment drives SNS bullying from the crime opportunity perspective. We test how presence of suitable targets and absence of capable guardianships affected SNS bullying and explore how the favor- able evaluation of such environmental conditions intensified the relationship between inclination to bully and SNS bullying. Second, this work enriches the IS literature by examining how users interpret SNSs

and the resultant deviant behaviors from the affordance perspective. We test four SNS affordances (i.e., accessibility, information retrieval, editability, and association) that influence perpetrators’ evaluation of SNS environmental conditions for SNS bullying. Although prior research has focused on the positive connotation of SNS affordances, our work breaks new ground for the study of unintended and negative acts afforded by the SNSs. Finally, for practitioners, the findings of this work could provide insights into how to

effectively combat SNS bullying. Based on the empirical results, SNS developers could prioritize resources to rectify the criminogenic environmental conditions that

576 CHAN ET AL.

exacerbate SNS bullying. Meanwhile, government agencies could launch campaigns to educate users on the appropriate use of SNSs. Together, the findings of this work offer a more proactive approach to tackle cyberbullying and maintain a healthy social networking environment.

Theoretical Background

Definition of Cyberbullying

Cyberbullying is a new form of bullying that involves the use of information technology. Different terminologies have been used to describe the phenomenon, such as electronic bullying [79], Internet bullying [106], and cyberbullying [98], with the last term being the most popular and widely adopted. Most cyberbullying studies have derived definitions from traditional bullying literature. For instance, cyberbullyingwas defined aswillful and repeated harm inflicted through the medium of electronic text [72]. Later, a more refined definition, proposing that cyberbullying is an aggressive online behavior that encom- passes three characteristics: (1) it is performed by individuals or groups using electronic or digital media; (2) hostile or aggressivemessages are repeatedly communicated; and (3) the behavior is conducted with the intent to cause discomfort or inflict harm on the target, was advocated [95]. Research also suggested that there are different types of cyberbully- ing behavior, such as flaming, harassment, cyberstalking, denigration, masquerade, out- ing and trickery, exclusion, and impersonation [48, 105]. At present, there is no exhaustive list of the types of bullying behavior perpetrated on SNSs.

Nature of SNS Bullying

SNS bullying is a form of aggressive behavior on SNSs conducted by individuals or groups, repeatedly and over time, against targets who cannot easily defend themselves. It shares three definitional criteria with the related concepts of bullying and cyberbullying: intentionality, repetition, and power imbalance [24]. SNS bullying is distinguished from other forms of online deviant behavior, such as Internet trolling and flaming, because it is deliberate, repeated, and involves exploitation of a power imbalance to intentionally harm a target by leveraging the functionalities and capabilities of social networking platforms. SNS bullying is often viewed as a form of deviant behavior fostered by the emergence

of information technologies [29, 47]. Specifically, the widespread deployment of perso- nal communication devices (such as smartphone, tablet, and laptop) and the ease of connectivity to online platforms have led to individuals spending more time with technologies. This shift in social activities, moving from offline venues to social net- working platforms, creates criminogenic opportunities for SNS bullying. In particular, the rapid growth in SNS users has created a wealth of online profiles that make it easy for perpetrators to identify vulnerable individuals. Guardianships of SNS bullying behaviors (e.g., SNS self-reporting functions, laws, and regulations prohibiting bullying) become ineffective because there are thousands to millions of social interactions happen on SNSs

CYBERBULLYING ON SOCIAL NETWORKING SITES 577

every day. It is virtually impossible to monitor, moderate, and control all the uses that have violated the community standards. Such a view is consistent with crime opportunity theory [30], which asserts that social and technological changes produce new opportu- nities for crime and deviance. In some countries, individuals face criminal charges and prison time if found guilty of

SNS bullying. For instance, in the United Kingdom, Section 127 of the Communications Act of 2003 makes SNS bullying a criminal offense for anyone sending something grossly offensive, indecent, obscene, or menacing character via a public electronic communications network. The law states that a perpetrator can face up to six months in jail, a fine, or both if found guilty [52]. Similarly, nearly half of the states in America include cyberbullying as part of their broader bullying laws. The nationwide trend is toward greater accountability for cyberbullying, in general, includ- ing criminal statutes [44]. For example, a bill recently passed in West Virginia, making cyberbullying a misdemeanor offense with a maximum punishment of one year in prison, a $500 fine, or both [14].

Toward a Meta-Framework of SNS Bullying

We use crime opportunity theory [30] and the affordance perspective [63] to develop a meta-framework that guides our investigation into SNS bullying. Specifically, crime opportunity theory posits two primary components contribute to a crime being com- mitted: (1) a likely perpetrator, and (2) environmental conditions that offer criminogenic opportunities. These are the building blocks of our meta-framework explaining SNS bullying.We further incorporate the affordance perspective into crime opportunity theory to explain how an SNS allows a perpetrator to evaluatewhether environmental conditions would facilitate an SNS bullying act. By integrating the affordance perspective into well- established theoretical frameworks, prior research has demonstrated the viability to obtain contextualized insights into a wide spectrum phenomenon related to information technol- ogy uses (e.g., Chatterjee et al. [15], Seidel et al. [85], andSuh et al. [93]). For instance, the affordance perspective has been integrated into the notion of virtue ethics to explain the effects of organizational ITaffordances on organizational virtues and innovation improve- ment [15]. Hence, we expect that integrating crime opportunity theory and the affordance perspectivewould provide a useful theoretical foundation for developing a contextualized understanding of SNS bullying. Figure 1 depicts the meta-framework of SNS bullying.

Crime Opportunity Theory

Crime opportunity theory [30] asserts that social and technological changes produce new opportunities for crime and deviance. Opportunities play a central role in every category of offense, regardless of its nature and severity. Subscribing to this perspective, we stipulate that the shifts in social activities from offline venues to SNS platforms provide opportunities for likely perpetrators to engage in SNS bullying. We argue that the rapid growth of user populations creates ample opportunities for SNS bullying. Specifically,

578 CHAN ET AL.

perpetrators can easily identify vulnerable individuals through browsing their online profiles on SNSs. The massive amount of information flow and social interactions also makes it difficult to monitor and identify acts of SNS bullying, which, in turns, weaken the capabilities of authorities and detection mechanisms in regulating such acts. Crime opportunity theory further emphasizes that the occurrence of crime and

deviance is influenced not just by the perpetrators’ characteristics but also by the environmental conditions that offer criminogenic opportunities. Our review of past studies suggest that SNS bullying research has mainly investigated the “likely perpe- trator” component, and have included aspects such as the perpetrators’ demographic characteristics (e.g., Cao and Lin [10] and Sengupta and Chaudhuri [87]), their intensity of SNS usage (e.g., Kwan and Skoric [49]), their cyberbullying victimization experience (e.g., Marcum et al. [62]), and their personality traits (e.g., Kokkinos et al. [46]) (see Appendix A for a review). The potential impacts of the “environment” have only recently attracted attention in the literature. For instance, the anonymous SNS environment has been found to be exploited by heavy SNS users to perpetrate others on the platform [56]. As Lowry et al. [56] noted, most cyberbullying studies “have glossed over the central issue: the role of information technology or social media artifacts themselves in promoting cyberbullying” (p. 3). Over the last two decades, researchers have been increasingly using opportunity

theories to investigate technology-related crime and deviance, such as data breaches [86] and computer crimes [107]. Empirical studies have also illustrated the applic- ability of crime opportunity theory for understanding bullying behaviors (e.g., Cho et al. [17]). Hence, considering both the theoretical assumptions and empirical applica- tions, together with the criminogenic nature of SNS bullying discussed in the previous section, we believe that crime opportunity theory is a viable theoretical perspective for explaining SNS bullying. Specifically, our study continues to advance the literature by focusing on the “environment” component and by examining how the SNS environ- ment fosters the development of SNS bullying. Building on prior criminology literature [30, 100], we propose two SNS environmental conditions that offer the criminogenic

Figure 1. Meta-Framework of SNS Bullying

CYBERBULLYING ON SOCIAL NETWORKING SITES 579

opportunities for a likely offender to engage in SNS bullying: (1) presence of suitable targets and (2) absence of capable guardianships.

Affordance Perspective

An affordance refers to “the potential for behaviors associated with achieving an immediate concrete outcome and arising from the relationship between an artifact and a goal-oriented actor or actors” [92, p. 69]. Technological affor- dance refers to “the mutuality of actor intentions and technology capabilities that provide the potential for a particular action” [60, p. 39]. It arises when one interprets a technology through his or her goals for action. The relational view of affordance is advantageous for understanding technology use because it allows researchers to consider the symbiotic relationship between the cap- abilities of the technology and the actor’s goal and action [36], treating the entanglement between them as a unit of analysis [60]. Research has further shown that one technology can support different goal-oriented actions for members of different social groups [20, 53]. In other words, it is individuals’ goals that shape what they come to believe the technology can afford them [96], which in turn leads to a wide spectrum of desirable or undesirable—or intended or unintended—behaviors [60]. For instance, Majchrzak et al. [60] identified four affordances of social media that affect employees’ engagement in group online workplace conversations. They suggest that some workers believed metavoicing affordance (i.e., the action possibility enabled by social media for users to engage in the ongoing online knowledge conversation by reacting online to others’ presence, profiles, content, and activities) fostered productive knowledge conversations, whereas some thought it inhibited pro- ductivity by promoting potentially biased and inaccurate information. Acting on this perspective, we argue that one could interpret an SNS differ-

ently depending on his or her goal [53]. The actualization of affordances occurs when an actor takes advantage of one or more affordances of the SNS to achieve immediate concrete outcomes that support their goals. In this study, the artifact is an SNS, and the goal-oriented actor is a user who purposefully uses an SNS to bully a target (i.e., a perpetrator). For general users, the actualization of SNS affordances occurs when they make use of the SNS to, perhaps, engage in self-disclosure and read their friends’ newsfeed in support of their relationship maintenance and socialization [16]. However, for a likely offender whose goal is to leverage the SNS to bully someone, the actualization of affordances could be completely different. For instance, they might see the SNS as affording them the ability to access information about the background and activities of other users, which would help them to identify suitable targets, giving rise to a favorable evaluation of SNS environmental conditions for SNS bullying.

580 CHAN ET AL.

Based on the review of the literature on technological affordances [60, 96] and social network research [45], we propose four types of SNS affordances and suggest that they have the potential to influence how one evaluates the SNS environmental condition for SNS bullying. These affordances include accessibility, information retrieval, editability, and association. Table 1 sum- marizes the definitions and illustrations of these affordances.

Table 1. SNS Bullying Affordances

SNS affordance Definition

How the affordance relates to SNS bullying

Related SNS affordances/SNS

features

Accessibility The extent to which a user believes that an SNS offers the opportunity to connect to another user on the platform.

This affordance allows a perpetrator to transcend time and spatial constraints in identifying a target for SNS bullying.

Network-informed associating [60]; network transparency [45]

Information retrieval

The extent to which a user believes that an SNS offers the opportunity to obtain information about a user on the platform.

This affordance allows a perpetrator to obtain contents created by a target to understand his/her background, preferences, and daily activities for the purpose of SNS bullying.

Persistence [96]; search and privacy [45]

Editability The extent to which a user believes that an SNS offers the opportunity to manipulate the content that he/she posted, commented on, and shared on the platform.

This affordance allows a perpetrator to deny his SNS bullying acts by erasing, editing, or hiding bullying related contents and identification cues.

Editability [96]; digital profile [45]

Association The extent to which a user believes that an SNS offers the opportunity to associate the responsibility for his/ her post with other users who interacted with the post on the platform.

This affordance allows a perpetrator to elude sole accountability for creating the bullying contents by attributing the contents with other users.

Association [96]; relational ties [45]

CYBERBULLYING ON SOCIAL NETWORKING SITES 581

Research Model and Hypotheses

Our meta-framework provides a theoretical basis to construct a research model explaining SNS bullying. First, drawing on crime opportunity theory [30], we propose that SNS bullying is driven by two primary components: (1) a likely offender, which is conceptualized as one’s inclination to bully and (2) the evalua- tion of SNS environmental conditions that offer the criminogenic opportunity, which include presence of suitable targets and absence of capable guardianships. Second, subscribing to the affordance perspective [63], we examine how SNS affordances (i.e., accessibility, information retrieval, editability, and association) influence the evaluation of environmental conditions for SNS bullying. Figure 2 depicts the research model.

Likely Offender and SNS Bullying

According to crime opportunity theory, a likely offender refers to a person who might commit a crime or engage in deviant behavior for any reason [30]. Crime opportunity theory presumes that crimes would not happen without an offender, therefore the presence of a likely offender is a prerequisite for any crime or deviance [30]. In this study, we conceptualize a likely offender as someone who has an inclina-

tion to bully on an SNS, which refers to one’s tendency to engage in SNS bullying for any reason [42]. Past studies have shown that positive inclinations toward

Figure 2. Proposed Research Model

582 CHAN ET AL.

bullying (e.g., probullying beliefs and favorable attitudes toward cyberbullying) predicted perpetrators’ engagement in cyberbullying behaviors (e.g., Lazuras et al. [50] and Wiklund et al. [104]). For instance, adolescents’ inclination to cyberbully was found to positively predict self-reported cyberbullying behaviors among teen- agers [42] and secondary students [71]. Therefore, we hypothesize that:

Hypothesis 1: Inclination to bully positively influences SNS bullying.

Evaluation of SNS Environmental Conditions and SNS Bullying

Crime opportunity theory presumes that favorable environmental conditions play a critical role in the occurrence of any crime or deviance [30]. In this study, we propose two SNS environmental conditions that offer the criminogenic opportu- nities for a likely offender to engage in SNS bullying: (1) presence of suitable targets and (2) absence of capable guardianships [30, 100].

Presence of Suitable Targets

Crime opportunity theory [30] states that “targets of crime can be a person or an object, whose position in space or time puts it at more or less risk of criminal attack” (p. 5). The theory asserts that certain characteristics of a target will be of greater interest to a likely offender, such as being visible (e.g., a valuable good is placed near windows) and accessible (e.g., a house with doors left unlocked). In this work, we define presence of suitable targets as the extent to which a perpetrator

believes there are suitable targets in the SNS environment available for SNS bullying. As discussed earlier, the prevalence and popularity of SNSs create newopportunities for SNS bullying [47]. In recent years, not only have the number of SNS users dramatically increased but also the amount of personal information that users posted and shared online. In 2017, 71 percent of Internet users had an SNS profile on one of the major SNS platforms [90]. Of these, 92 percent used their real names on their profiles, 91 percent had a picture of themselves on their profiles, and 82 percent had posted other personal information on their profiles—such as birth date, gender, education background, occupa- tion, or country of residence [59]. A large number of users and an ample amount of sensitive personal information available provide a wealth of opportunity to identify suitable targets for SNS bullying. Hence, the perception that the SNS environment is a source of suitable targets is likely to attract more SNS bullying behaviors. This prediction is also evident in the bullying research, which supports a link between suitable targets and bullying behaviors. For instance, students who were perceived to be suitable targets among the perpetrators were more likely to be victimized [76]. Therefore, we hypothesize that:

Hypothesis 2: Presence of suitable targets positively influences SNS bullying.

CYBERBULLYING ON SOCIAL NETWORKING SITES 583

Absence of Capable Guardianships

Crime opportunity theory suggests that in the absence of capable guardianships, crime and deviance are more likely to occur [30]. According to the theory, guardianships are not confined to government officials alone, but rather include “anybody whose pre- sence or proximity would discourage a crime from happening” [30, p. 4]. In this work, we define absence of capable guardianships as the extent to which

a perpetrator evaluates that guardianships are incapable of fortifying SNS environ- ments against SNS bullying. Guardianships here represent both offline authorities (e.g., laws and regulations) and online mechanisms (e.g., reporting systems and detection algorithms) that aim to protect users from being victimized on SNSs. For instance, Facebook has implemented a built-in reporting system that permits users to report any content that is not commensurate with its community standards (such as nudity, hate speech, or violence). The Facebook team regularly reviews the reported materials and removes them if they are deemed inappropriate. These functions serve as a guardianship, protecting general users against SNS bullying. However, with the growing number of posts uploaded and shared on SNSs daily, it has become increas- ingly challenging for these protective measures to effectively tackle bullying activities on SNSs [5]. Though there have been initiatives to use more advanced techniques— such as machine learning and natural language processing to detect SNS bullying— their effectiveness is restricted by computers’ ability to interpret meanings, variations, and metaphors in human language [11]. It remains difficult for guardianships to fortify SNS environments against SNS bullying effectively. Past studies have found support for the link between a lack of guardianships and bullying behaviors. For instance, social guardianships was found to decrease victimization among young people [57]. Therefore, we hypothesize that:

Hypothesis 3: Absence of capable guardianships positively influences SNS bullying.

SNS Affordances and the Evaluation of SNS Environmental Conditions

Drawing on the affordance perspective [63], we further examine how the SNS affor- dances outlined above (accessibility, information retrieval, editability, and association) affect the evaluation of SNS environmental conditions (i.e., presence of suitable targets and absence of capable guardianships), in which criminogenic opportunities for SNS bullying are perceived.

Accessibility Affordance

Accessibility affordance refers to the extent to which a user believes that an SNS offers the opportunity to connect with a user on the platform. In SNS bullying, accessibility

584 CHAN ET AL.

affordance allows a perpetrator to transcend time and spatial constraints to reach potential targets. Kane et al. [45] suggested that network transparency is one of the essential features of a social network; it allows users to view their connections within a network and offers the opportunity to connect each other. In SNSs, users are given various opportunities to contact and connect with an unlimited number of users, including friends, family members, acquaintances, and even strangers. For perpetra- tors, however, accessibility affordance facilitates overcoming barriers of time and space to connect with potentially suitable targets. In a recent SNS bullying case, for example, a perpetrator used the hashtag (i.e., #hashtag) and handle (i.e., @username) on Instagram to repeatedly bully a group of young people [66]. The unconstrained and boundless accessibility afforded by SNSs may lead a perpetrator to evaluate that the SNS provides an environment where suitable targets can be easily identified and accessed. Therefore, we hypothesize that:

Hypothesis 4: Accessibility affordance positively influences presence of suita- ble targets.

Information Retrieval Affordance

Information retrieval affordance refers to the extent to which a user believes that an SNS offers the opportunity to obtain information about a user on the platform. In SNS bullying, information retrieval affordance allows a perpetrator to access material created by a potential target, which provides information about the background, preferences, and daily activities of the potential target. SNS updates often include new features that aim to entice users to continuously create and share information on the platforms. For instance, Facebook’s “On This Day” feature shows old photos and newsfeeds to a user and encourages the user to forward these posts and stories with their friends. Instagram, Twitter, and other SNSs often ask users to provide precise information when uploading a photo. Such updates are part of an oversharing phenom- enon, with a recent survey estimating that about 40 percent of users overshare sensitive information on SNSs [64]. Such abundance of unrestricted information puts users at risk for SNS bullying victimization. For instance, the Facebook timeline provides an easy interface for quickly reading others’ activity logs. It is like a scrapbook, providing snapshots of information that can be used to understand a particular user. It allows a perpetrator to trawl back through a target’s history, gleaning information from shared photos and statuses and eventually using them to create harassing materials or even to impersonate the person identified as a suitable target [13]. Past studies have also shown that individuals who did not restrict access to their online profiles or who disclosed too much sensitive personal information online were considered more attractive and vulnerable by perpetrators [65, 73]. Therefore, we hypothesize that:

Hypothesis 5: Information retrieval affordance positively influences presence of suitable targets.

CYBERBULLYING ON SOCIAL NETWORKING SITES 585

Editability Affordance

Editability affordance refers to the extent to which a user believes that an SNS offers the opportunity to manipulate a content that he or she posted, commented on, and/or shared on the platform. In SNS bullying, editability affordance allows a perpetrator to deny his SNS bullying acts by erasing, editing, or otherwise hiding bullying related contents and identification cues. In offline bullying, it is difficult for a perpetrator to conceal his or her identity because the victim can at least recognize the physical appearance of the perpetrator. Physical damages inflicted on the target are also difficult to hide. In contrast, in SNSs, it is fairly easy for a perpetrator to modify, erase, or hide identification cues in relation to the bullying and his or her identity. For instance, Facebook allows users to edit descriptions of their posts or even delete contents published on their walls. One can also register a new email domain and create an alternative SNS account to engage in SNS bullying. As a result, this affordance weakens the effect of guardianships on SNS because it is difficult for authorities to track and punish SNS bullying behaviors. Therefore, we hypothesize that:

Hypothesis 6: Editability affordance positively influences absence of capable guardianships.

Association Affordance

Association affordance refers to the extent to which a user believes that an SNS offers the opportunity to share responsibility for his or her post with other users who interact with the post on the platform. In SNS bullying, association affordance allows a perpetrator to avoid accountability for the bullying act by inviting other SNS members; that is, the perpetrator can deny sole responsibility for carrying out the action. User engagement and cocreation are core values on most social networking platforms. SNS providers not only entice users to share more information but also encourage others to interact with these posts. For instance, Facebook now offers more nuanced reactions to posts beyond the “like” reaction (i.e., “love,” “ha-ha,” “wow,” “sad,” and “angry”) to encourage users to express themselves after reading a post. The long-standing tag feature (@user name) allows users to invite others to respond to a post and jointly develop the conversation. Recent statistics show that 44 percent of Facebook users “Liked” content posted by their friends at least once a day, and 31 percent made comments on posts daily [89]. On the one hand, association affor- dance fosters meaningful exchange among ordinary users. On the other hand, it allows perpetrators to invite other users to view and participate in bullying posts, making it difficult to designate responsibility for the hurtful contents [82], mitigating the effect of guardianships. Therefore, we hypothesize that:

Hypothesis 7: Association affordance positively influences absence of capable guardianships.

586 CHAN ET AL.

Control Variables

Past studies have demonstrated that demographic characteristics, computer usage, and cyberbullying self-efficacy can influence cyberbullying [47]. Accordingly, we include age, gender, education, SNS usage, SNS experience, SNS real name registration, and self-efficacy in SNS bullying, as the control variables.

Research Method

Research Design

We used an anonymous, self-reported, longitudinal online survey design with Facebook users to test the proposed research model. The survey method has been used to examine a broad range of deviant behaviors related to technology use, such as online software piracy [43], information system misuse [19], and cyberbullying [56]. The self-report questionnaire technique has been used to test crime opportu- nity theory and the affordance perspective in both offline and online contexts, such as bullying victimization [17], workplace sexual harassment [22], online hate on SNSs [78], and gamification [93]. Using a longitudinal setting can also reduce the threat of common method bias and enhance causal inference [75, 81]. We selected Facebook as the research context because it is the leading SNS worldwide [28]. A recent survey also revealed that cyberbullying is most likely to take place on this platform [23]. Therefore, we believed that Facebook represents a suitable context for testing our proposed research model. To participate in the study, individuals had to: (1) be users of Facebook; (2) live in the United States (this requirement ensured a standardized perception of laws and norms regarding SNS bullying on Facebook [56]).

Measure

The measurement items were adapted from the literature where possible (e.g., SNS bullying). Minor modifications were made to measurement items to fit the current research context. When measurement items were unavailable (e.g., SNS affordances and crime opportunity components), we followed the guidelines set out in the instrument development literature [68] to develop new instruments to measure the constructs. The instrument development process and the complete list of measurement items for the focal constructs are shown in the online supplement – section A. As the research context examines a socially undesirable behavior, the social desirability scale was also included to detect for potential response bias [80].

CYBERBULLYING ON SOCIAL NETWORKING SITES 587

Data Collection and Procedures

Respondents for the online survey were recruited from the Amazon Mechanical Turk (MTurk). MTurk is an online crowdsourcing platform that allows people to participate in Human Intelligence Task (HIT) for remuneration. The use of MTurk is appropriate for the current research purpose, as suggested in recent cyberbullying research [e.g., 83] and advocated in senior IS literature [e.g., 56]. Specifically, cyberbullying is a sensitive issue and is socially unacceptable in most cultures. Hence, using MTurk as a portal to reach the target sample helped ensure respondents’ anonymity, thereby eliciting responses that are more honest and reducing social desirability bias. Furthermore, since cyberbullying is a general topic that requires minimal expertise, using MTurk to collect data is a good fit. It allows researchers to reach a huge pool of potential respondents with SNS bullying experiences, which is virtually impossible using other data collection methods. To ensure data quality, we followed guidelines as described in the latest methodological literature onMTurk in designing and distributing the survey study [34, 54]. For instance, we checked the workers’ location based on their IP address to ensure they reside in the United States.We detected “super workers,”who generally put less time and effort into a task, using their completion time and number of tasks completed. We also included randomly appearing attention-check questions and reverse-coded questions to affirm the accuracy of the responses. The data collection consisted of two waves. At time t (Wave 1), HIT requests

were posted on MTurk. At this stage, responses related to independent variables (i.e., SNS affordances and crime opportunity components) were collected. The respondents in Wave 1 were then invited to answer another online questionnaire at time t+1 (Wave 2), in which responses related to the dependent variables (i.e., SNS bullying behaviors) were collected. A unique code was used to match respon- dents’ responses across the two waves of data collection. At the beginning of the survey, respondents were asked to answer screening

questions to determine their eligibility to participate. In particular, they were asked to indicate the three social networking platforms they had visited most frequently during the past three months and asked to report their country of residence. We filtered out respondents who did not pass these screening ques- tions. Following the screening questions, respondents were asked to complete a questionnaire that included measures of the variables of interest in each wave. Finally, they were asked to answer the social desirability items. We collected their demographic information at the end of the survey. We provided a monetary incentive upon successful completion of the questionnaire. Ten randomly pre- sented attention-check questions were included to detect any careless, random, or haphazard responses that may have occurred as a result of the online survey method. Responses from individuals who attempted to participate multiple times (as identified through respondents’ MTurk ID and IP address), failed to pass the attention-check questions, and from those who completed the survey in an exceptionally short time (i.e., less than 15 minutes) were filtered out of the sample to ensure data quality.

588 CHAN ET AL.

Respondent Profile

We launched the online surveys in June 2018 (time t, Wave 1) and September 2018 (time t + 1, Wave 2). 1,023 respondents attempted the survey in Wave 1, with 530 indicating Facebook as their most visited SNS and theUnited States as their country of residence. 32 respondents failed to pass the attention-check questions or provided haphazard responses, leaving 498 complete and valid responses. For Wave 2, we sent an invitation to respon- dents who participated in Wave 1. 262 attempted the survey, and 39 respondents did not pass the attention-check questions or provided haphazard responses, leaving 223 com- plete and valid responses for subsequent analyses. Of the remaining respondents, 98 (43.9 percent) were male, and 125 (56.1 percent) were female. Most were young adults, between the ages of 25 and 34 (45.3 percent). Themajority visited Facebook at least once a day (91.0 percent) and had more than five years of experience using Facebook (85.2 percent). Table 2 presents the respondent profile.

Data Analysis and Results

Survey methodologies may be plagued by common method bias (CMB) and social desirability bias (SDB), we applied several procedural and statistical remedies to minimize these threats. The results suggest that both CMB and SDB were

Table 2. Respondent Profile

No. Percent No. Percent

Gender SNS usage Male 98 43.9 Once a week 4 1.8 Female 125 56.1 2–4 times a week 12 5.4

5–6 times a week 4 1.8 Age Once a day 52 23.3

18–24 15 6.7 2–3 times a day 42 18.8 25–34 101 45.3 4–5 times a day 25 11.2 35–44 51 22.9 More than 5 times a day 84 37.7 45–54 24 10.8 55–64 17 7.6 SNS experience 65 or above 15 6.7 Less than a year 3 1.3

1–2 year(s) 7 3.1 Education 3–4 years 23 10.3

Less than high school 3 1.3 5–6 years 48 21.5 High school 49 22.0 7–8 years 43 19.3 College degree 51 22.9 9–10 years 36 16.1 Bachelor’s degree 79 35.4 More than 10 years 63 28.3 Master’s degree 31 13.9 Doctoral degree 3 1.3 Professional degree 7 3.1

CYBERBULLYING ON SOCIAL NETWORKING SITES 589

negligible in this study [75, 84]. Detailed procedures are reported in the online supplement – section B. We assessed the reliability of the measurement items using Cronbach’s alpha

and examined the convergent and discriminant validity of the constructs using factor analysis and pairwise chi-square tests. Specifically, all of the constructs demonstrate internal consistency with Cronbach’s alpha values exceeding the threshold [38]. Factor analysis showed that items load strongly on their corresponding constructs with low cross-loadings with other constructs. Furthermore, the chi-square tests showed that all chi-square differences for each pair of constructs in the research model are statistically significant. An examination into the variance inflation factors also suggested that the model does not suffer from multicollinearity issue. Taken together, the measurement model demonstrates sufficient convergent validity and discriminant validity [38, 99]. Details of the assessment of the reliability, validity, and multicolli- nearity can be found in the online supplement – section C and D. We performed hierarchical regression analyses to test the hypotheses. To test

the direct effects of the crime opportunity components on SNS bullying, we ran a control effect model and then a main effect model. Table 3 shows the results of these analyses. We first tested the control variables. The control-only model explains 29.5 percent of the variance for SNS bullying. After that, we tested the effects of inclination to bully, presence of suitable targets, and absence of capable guardianships on SNS bullying. The main effect model explains

Table 3. Results of Regression Analysis on Crime Opportunity Components

SNS Bullying

Dependent variable Control-only Main effect

Control variables Gender −.165** −.095 Age −.237*** −.102* Education .111 .051 SNS usage −.037 −.026 SNS experience −.396*** −.216*** SNS real name registration .008 .002 Self-efficacy in SNS bullying .173** .077

Main effects Inclination to bully .443*** Presence of suitable targets .173*** Absence of capable guardianships .118**

R2 .295 .547 Δ R2 .252***

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

590 CHAN ET AL.

54.7 percent of the variance for SNS bullying. Specifically, inclination to bully (β = .443, p < .001), presence of suitable targets (β = .173, p < .001), and absence of capable guardianships (β = .118, p < .01), predict SNS bullying, supporting Hypothesis 1, Hypothesis 2, and Hypothesis 3. To test the effects of SNS affordances on the evaluation of SNS environmental

conditions, we ran a control effect model and then a main effect model. Table 4 shows the results of the analyses. The results indicate that information retrieval affordance (β = .265, p < .001) predicts presence of suitable targets, supporting Hypothesis 5. The model explains 13.3 percent of the variance for presence of suitable targets. Furthermore, the analysis shows that editability affordance (β = .233, p < .01) and association affordance (β = .182, p < .05) predict absence of capable guardianships, supporting Hypothesis 6 and Hypothesis 7. The model explains 13.4 percent of the variance for absence of capable guardianships. However, accessibility affordance has no influence on presence of suitable targets (β = -.098, p > .05), failing to support Hypothesis 4. Table 5 summarizes the hypotheses test results.

Table 4. Results of Regression Analysis on SNS Affordances

Presence of suitable targets

Absence of capable guardianships

Dependent variable Control- only

Main effect

Control- only

Main effect

Control variables Gender −.064 −.034 .095 .098 Age −.146* −.098 −.038 .007 Education .027 −.022 .049 .036 SNS usage −.053 −.060 .125 .135 SNS experience −.168* −.112 −.029 −.036 SNS real name registration −.011 .018 −.104 −.067 Self-efficacy in SNS bullying .103 .086 .125 .075

Main effects Accessibility affordance −.098 Information retrieval affordance

.265***

Editability affordance .233** Association affordance .182*

R2 .070 .133 .134 Δ R2 .063** .094***

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

CYBERBULLYING ON SOCIAL NETWORKING SITES 591

Post Hoc Analyses

Comparison of Alternative Models

We performed a pseudo-F test to assess the effects of excluding the components inclination to bully or evaluation of SNS environmental conditions from the model, along with the resulting change in variance explained for SNS bullying. As shown in Table 6, the exclusion of either of these components leads to a significant drop in variance for SNS bullying. This result indicates that SNS bullying is better explained by examining the likely offender and the environmental condition com- ponents together, providing further support to crime opportunity theory.

Assessment of the Mediation Effects

We conducted bootstrapping analyses to examine the mediating effects using PROCESS [41, 58]. We bootstrapped the effects of SNS affordances (i.e., accessibility,

Table 5. Summary of Hypotheses Test Results

Hypothesis Result

Hypothesis 1: Inclination to bully positively influences SNS bullying. Supported Hypothesis 2: Presence of suitable targets positively influences SNS

bullying. Supported

Hypothesis 3: Absence of capable guardianships positively influences SNS bullying.

Supported

Hypothesis 4: Accessibility affordance positively influences presence of suitable targets.

Not Supported

Hypothesis 5: Information retrieval affordance positively influences presence of suitable targets.

Supported

Hypothesis 6: Editability affordance positively influences absence of capable guardianships.

Supported

Hypothesis 7: Association affordance positively influences absence of capable guardianships.

Supported

Table 6. Results of the Pseudo-F Test

Comparison R2

excluded R2

full ΔR2 ΔF Cohen’s

f2 Effect size

Inclination to bully excluded .411 .547 .135 63.270*** .156 Medium Evaluation of SNS environmental

conditions excluded .495 .547 .052 12.137*** .055 Small

Note: f2 ≥ .02, f2 ≥ .15, and f2 ≥ .35 represent small, medium, and large effect sizes, respectively [18].

592 CHAN ET AL.

T ab le

7. R es ul ts of

th e M ed ia ti on

Te st s

S N S

af fo rd an ce s

(I V )

T he

ev al ua ti on

of S N S

en vi ro nm

en ta l co nd

it io ns

(M )

M ed ia ti on

te st (a b)

F ul l/ P ar ti al

m ed ia ti on

te st (c ’)

In di re ct

ef fe ct

B ia s-

co rr ec te d

95 pe rc en t

co nf id en ce

in te rv al s fo r

in di re ct

ef fe ct

Z er o?

M ed ia ti on

? D ir ec t

ef fe ct

B ia s-

co rr ec te d

95 pe rc en t

co nf id en ce

in te rv al s fo r

di re ct

ef fe ct

Z er o?

T yp

es of

m ed ia ti on

E ff ec t (S E ) L ow

er U pp

er E ff ec t (S E ) L ow

er U pp

er

A cc es

si bi lit y

P re se

nc e of

su ita

bl e ta rg et s

− .0 51

(. 03

0) − .1 16

.0 04

Y es

N o

− .1 54

(. 12

0) − .3 90

.0 83

Y es

N on

e In fo rm

at io n

re tr ie va

l .1 10

(. 03

7) .0 51

.1 98

N o

Y es

− .0 71

(. 06

2) − .1 93

.0 52

Y es

F ul l

E di ta bi lit y

A bs

en ce

of ca

pa bl e

gu ar di an

sh ip s

.0 99

(. 03

4) .0 47

.1 80

N o

Y es

− .1 67

(. 09

7) − .3 59

.0 25

Y es

F ul l

A ss oc

ia tio

n .0 48

(. 02

1) .0 16

.1 03

N o

Y es

.2 04

(. 06

1) .0 84

.3 24

N o

P ar tia

l

CYBERBULLYING ON SOCIAL NETWORKING SITES 593

information retrieval, editability, and association) on the evaluation of SNS environ- mental conditions (i.e., presence of suitable targets, and absence of capable guardian- ships) (a1-4), the effects of the evaluation of SNS environmental conditions on SNS bullying (b1-2), and the effects of SNS affordances on SNS bullying (c’1-4). Table 7 summarizes the mediation tests. Full mediation is observed when the confidence intervals (CIs) of the indirect

effect (i.e., ab) does not involve zero but the direct effect (i.e., c’) does. In our model, presence of suitable targets fully mediates the relationship between information retrieval affordance and SNS bullying; and absence of capable guardianships fully mediates the relationships between editability affordance and SNS bullying. Furthermore, absence of capable guardianships partially mediates the relationships between association affordance and SNS bullying. However, there is no mediation effect found between accessibility affordance and SNS bullying. The results indicate that whereas the effects of information retrieval affordance and editability affordance are explained wholly by presence of suitable targets and absence of capable guardianships, respectively, associa- tion affordance has a direct positive effect on SNS bullying beyond the effect that is mediated by absence of capable guardianships. In other words, being able to associate one’s act with other SNS users may have psychological effects, such as diffusion of responsibility, beyond simply perceiving an absence of capable guardianships [97].

Assessment of the Interaction Effects

Crime opportunity theory holds that offenders behave rationally and engage in crime and deviance when the environment is favorable [30]. Accordingly, we expect that the evaluation of SNS environmental conditions will not only have a direct effect on SNS bullying but also exacerbate perpetrators’ inclination to actually engage in SNS bullying behaviors.

Inclination to Bully × The Evaluation of SNS Environmental Conditions

We expect two two-way interaction effects between the inclination to bully and the evaluation of SNS environmental conditions (i.e., presence of suitable targets, and absence of capable guardianships). In traditional bullying, most bullying takes place among primary and secondary students. In these popula- tions, there is always a large pool of peers from which a perpetrator can easily select a suitable target. Also, bullying often takes places after school, when a vulnerable target is away from teachers’ supervision [24]. Based on this logic, it is plausible that in SNS bullying, when one with an inclination to bully evaluates the SNS environment as favorable, he or she would believe that the effort involved in finding suitable targets or the chances of being caught would be low. As a rational perpetrator, he or she would be more likely to translate the

594 CHAN ET AL.

inclination into action. Therefore, the relationship between inclination to bully and SNS bullying will be stronger when the evaluation of the SNS environ- mental conditions is favorable (i.e., high in terms of presence of suitable targets or absence of capable guardianships).

Presence of Suitable Targets × Absence of Capable Guardianships

We expect a two-way interaction effect between these two environmental conditions. Prior research report that bullying incidents are less likely when teachers are attentive to students at school [17] and that high levels of parental support reduce the risk of cyberbullying victimization among adolescents [101]. These findings suggested that the attractiveness of a target (i.e., the perception of suitability) could be greatly reduced by the presence of capable guardianships. Based on this logic, it is plausible that when the perpetrator perceives a high absence of capable guardianships, he or she would likely estimate a higher number of suitable targets present in the SNS environment. For instance, if a perpetrator perceives the detection mechanism of SNS bullying to be ineffective, he or she would tend to believe that users aremore vulnerable because there is no one to protect them from being bullied. Conversely, if a perpetrator perceives that guardianships are effectively filtering and removing bullying content quickly and therefore safeguarding the potential targets, they may evaluate users on the SNS platform as less suitable for bullying. Therefore, the relationship between presence of suitable targets and SNS bullying is stronger when the perpetrator perceives a higher degree of absence of capable guardianships.

Inclination to Bully × Presence of Suitable Targets × Absence of Capable Guardianships

We expect a three-way interaction effect on SNS bullying between the inclina- tion to bully, presence of suitable targets, and absence of capable guardianships. Crime opportunity theory assumes that crime components (i.e., offender, target, and guardians) are interrelated [35]. Crime and deviance are most likely to occur when an offender is situated in favorable environmental conditions [30]. Therefore, when one with an inclination to bully perceives two favorable SNS environmental conditions existing in time and space (i.e., a high degree of presence of suitable targets and a high degree of absence of capable guardian- ships), he or she expects minimal effort and risk when engaging in SNS bullying. As a result, the perpetrator is more likely to act opportunistically and translate the inclination into actual behavior.

We conducted bootstrapping analyses to examine the interaction effects using PROCESS [41]. Table 8 summarizes the moderation tests. The results show two significant two-way interactions among the crime opportunity components.

CYBERBULLYING ON SOCIAL NETWORKING SITES 595

Specifically, presence of suitable targets (β = .185, p < .05) positively moderates the relationship between inclination to bully and SNS bullying, whereas absence of capable guardianships (β =.197, p <.001) positively moderates the relation- ship between presence of suitable targets and SNS bullying. SNS bullying is more likely to occur when a likely offender who is inclined to bully perceives a higher number of suitable targets. Targets are also more prone to being perceived as vulnerable and suitable for an attack when there is a higher degree of absence of capable guardianships. The significant moderating effects provide additional support for the salience of environmental conditions in exacerbating SNS bullying behaviors, supporting crime opportunity theory. We conducted simple slope analyses to further understand the conditional

effects of the interaction among inclination to bully, presence of suitable targets, absence of capable guardianships, and SNS bullying. We plotted the significant interactions at one standard deviation above and below the mean of the variables [1]. Figure 3a and b show the interaction plots. For the two-way interaction of inclination to bully × presence of suitable targets, we observe a stronger and significant positive relationship between inclination to bully and SNS bullying when presence of suitable targets is high. Furthermore, we observe a stronger and significant positive relationship between presence of suitable targets and SNS bullying when there is a high degree of absence of capable guardianships. Details of the conditional effects at values of the moderators can be found in the online supplement – section E. These results imply that SNS bullying is more likely to occur when there are favorable environmental conditions on SNSs. The results, therefore, support crime opportunity theory, which posits that easy and tempting environmental conditions attract more crime and deviance.

Table 8. Results of the Interaction Effects of the Crime Opportunity Components

Dependent variable SNS bullying

Interaction effects Coeff. (β)

(SE) t-value (sig)

Inclination to bully × Presence of suitable targets .185 (.084) 2.207* Inclination to bully × Absence of capable guardianships .171 (.098) 1.750(n.s.)

Presence of suitable targets × Absence of capable guardianships

.197 (.052) 3.822***

Inclination to bully × Presence of suitable targets × Absence of capable guardianships

.079 (.074) 1.067(n.s.)

*p < .05; ***p < .001. Note: n.s. Not significant.

596 CHAN ET AL.

Discussion

The objectives of this work are to (1) understand the key factors driving SNS bullying, and (2) examine how SNS affordances influence the evaluation of SNS environmental conditions. We build on crime opportunity theory and the affordance perspective to develop a meta-framework that explains the occurrence of SNS bullying and delineates the role of technology affordance. The research model was tested using a longitudinal survey with 223 Facebook users. Empirical results provide strong evidence in support of the research model, and the overall model explains a substantial amount of variance for SNS bullying. In the following sections, we discuss implications for research and practice, limitations, and avenues for future research.

Implications for Research

This work has significant implications for research. First, we offer a comprehensive theoretical explanation and empirical investigation into SNS bullying that considers factors associated with both individual characteristic and SNS environmental condi- tions. We further identify and test the effects of SNS affordances that influence perpetrators’ evaluation of SNS environmental conditions for SNS bullying. The empirical results demonstrate strong support of the integration of the two theoretical perspectives, which offer rich insights into the occurrence of SNS bullying. The meta- framework also serves as a solid basis for future studies aiming to examine the effects of technology affordance on technology-related crime and deviance. Second, our empirical results enrich our scientific understanding of SNS bullying and

add to the knowledge accumulation of the cyberbullying literature. Crime opportunity theory and its predictive power have been validated previously in offline and

(a) (b)

Figure 3. a) Two-way Interaction between Inclination to Bully and Presence of Suitable Targets;. b) Two-way Interaction between Presence of Suitable Targets and Absence of Capable Guardianships

CYBERBULLYING ON SOCIAL NETWORKING SITES 597

organizational contexts. This work extends the generalizability of the theory to the SNS bullying context, contributing to the cumulative tradition of scientific research and the ongoing assessment of the theory. Specifically, our results show that crime opportunity theory is a plausible theoretical lens for investigating technology-related crime and deviance at an individual level. We further explore the interaction effects between the components of crime opportunity theory and identify the combinations that exacerbate SNS bullying. Third, we enrich the IS literature by introducing the affordance perspective into the

study of SNS bullying research. Based on past research on technological affordances and social network research, we identify four SNS affordances and examine their effects on the environmental conditions conducive to SNS bullying. Our empirical results demon- strate the salience of affordance in giving rise to the favorable evaluation of criminogenic opportunities. Technological affordances have long been recognized as a useful concept to explain the action possibilities perceived by users interacting with technologies. However, previous work has tended to associate affordances with positive behaviors, such as maintaining friendships and sharing useful content on social networks, with little understanding of how technological affordances can enable deviant behaviors. Our results offer a novel perspective on the far-reaching and unintended effects of technolo- gical affordances as a potential enabler of technology-related crime and deviance.

Implications for Practice

A large body of research on SNS bullying has shown that online users with certain characteristics are more vulnerable to both SNS bullying perpetration and victimi- zation (e.g., Peluchette et al. [73]). Although these insights are valuable, we contend that actionable and proactive measures can be better developed by focusing on the recertification of the SNS features and environmental conditions. First, our work observes that SNS bullying could be enabled by SNS affordances. We

found that the information retrieval affordance significantly drives the perception of suitable targets on SNSs. Educating SNS users to limit the amount of private and sensitive information that they share on online platforms could help reduce their attractiveness to potential perpetrators. For instance, educational videos that alert users about the potential risks of “friending” strangers and disclosing sensitive personal information could be developed and auto-played on social networking sites themselves. To mitigate unintended uses of personal information, SNS developers should also introduce more sophisticated options for users to control their preferences for informa- tion disclosure. Such measures could help to reduce the attractiveness of users on social networking platforms and keep them safe from SNS bullying. Another potential means of reducing SNS bullying would be introducing and reform-

ing legislation that regulates deviant online behaviors. Recently, national governments have started to engage in legislative action and other measures to protect users from SNS bullying. For instance, the Prime Minister of the United Kingdom has urged social networking giants Facebook and Twitter to tighten their rules to prevent cyberbullying

598 CHAN ET AL.

[21]. Such actions might align SNS bullying with higher potential costs, intensifying the perception of capable guardianships presents on the platform. As editability affordance and association affordance are important drivers for evaluating the absence of capable guardianships in SNS environment, new legislation imposing heavier legal consequences of SNS bullying could be useful in discouraging such deviant behavior. To complement these legislative initiatives, SNS developers should establish zero-tolerance policies toward SNS bullying behaviors and indicate clearly the punishment of deviant behavior to site users. For instance, platforms should give warnings to users if any inappropriate site use is detected, and temporary account suspension should be imposed if a user is found guilty of violating the terms of use. It is also essential for SNS developers to be cautious about their core design principles, which obviously favor maximizing social interaction. Such design principles have constantly been abused by perpetrators who seek to involve more accomplices in the incident, thereby allowing them to deny sole culpability. Finally, SNS platforms should inform users that any information uploaded onto the site will be stored and subject to investigation upon request by the proper authorities.

Limitations and Future Research Directions

Our work does have some limitations that should be acknowledged—which, however, also gesture toward several avenues for future research. First, care must be taken when extrapolating the findings of this study to bullying on other SNSs and in other countries. Specifically, we tested the research model using a single SNS platform with American adult users. The homogeneity of the respondent profile may have affected the general- izability of our conclusions. However, the sample did consist of respondents with heterogeneous demographic characteristics—such as SNS usage experience, educational background, and age—whichmay have helped to overcome sampling limitations. Future research should replicate our research model and test whether users’ evaluation of SNS environmental conditions can be generalized to different user groups (e.g., children), other cultural contexts (e.g., Asia), or social networking platforms (e.g., Twitter). Second, since we used an online survey to collect the data, our findings may be

influenced by response bias. To address these concerns, we used a third-party platform and an anonymous survey setting to minimize the threat of response bias and used the social desirability scale to detect biased responses. We also applied both procedural remedies and statistical remedies to detect and mitigate concerns related to common method bias. Nevertheless, our study may have been influenced by self-selection bias, which is difficult to estimate when using an online survey design. It is also possible that some respondents with SNS bullying experience left the survey after being exposed to sensitive questions. Third, we consolidated four general SNS affordances from the literature and tested

their effects in our research model explaining SNS bullying. Although our study breaks new ground by investigating the unintended effects of SNS affordances on giving rise to favorable environmental conditions for SNS bullying, future research should explore

CYBERBULLYING ON SOCIAL NETWORKING SITES 599

other SNS affordances associated with specific social networking platforms. For instance, Snapchat allows photos to be viewable for a maximum of only 10 seconds. Such design can be further examined by introducing an “erasability” affordance, which may affect the evaluation of capable guardianships on Snapchat and alter SNS bullying behaviors and dynamics. Future research should also examine the technical objects that giving rise to an affordance. In this study, we broadly considered the technical object to be the “SNS” (i.e., Facebook). An experimental setup would, therefore, be beneficial for future studies to better understand and test the exact technical features and char- acteristics that give rise to these affordances. Finally, because we used a typical variance model based on longitudinal online

survey design, we were only able to infer causation from the theoretical foundation and research design. Despite this limitation, we prefer the survey method over other alternatives. It allows us to maximize the predicted frequency of SNS bullying by providing a snapshot of the relative effects and interaction effects among the various crime opportunity components. Future research should use experiments, interviews, and case studies to validate the research findings. However, the use of these alternative research designs may inevitably induce undesirable cyberbullying experiences to the participants, and conflict with participants’ ability to remain anonymous due to the requirement for identification. This may lead to new challenges in eliciting honest responses while maintaining confidentiality.

Conclusion

Drawing on crime opportunity theory and the affordance perspective, we develop and empirically test a research model to explain SNS bullying. The research model explains a substantial amount of the variance for SNS bullying and highlights the imperative role of technology affordance and SNS environment in shaping SNS bullying. We believe that the results have significant implications for research on IT deviant use and provide practical guidance for formulating preventive measures and educational pro- grams to combat SNS bullying.

Acknowledgement: The authors wish to thank the Editor-in-Chief, Professor Zwass, and the reviewers for their support and guidance throughout the review process.

Funding

The work described in this article was partially supported by a grant from the Research Grant Council of the Hong Kong Special Administrative Region, China (Project No. HKBU 12511016).

SUPPLEMENTAL MATERIAL

Supplemental data for this article can be accessed on the publisher’s website.

600 CHAN ET AL.

ORCID

Tommy K. H. Chan http://orcid.org/0000-0001-9930-8897 Christy M. K. Cheung http://orcid.org/0000-0003-4411-0570 Randy Y. M. Wong http://orcid.org/0000-0001-6585-9973

REFERENCES 1. Aiken, L.S.; West, S.G.; and Reno, R.R. Multiple regression: Testing and interpreting

interactions. Thousand Oaks: SAGE, 1991. 2. Alhabash, S.; McAlister, A.R.; Hagerstrom, A.; Quilliam, E.T.; Rifon, N.J.; and

Richards, J.I. Between likes and shares: Effects of emotional appeal and virality on the persuasiveness of anticyberbullying messages on Facebook. Cyberpsychology, Behavior, and Social Networking, 16, 3 (2013), 175–182.

3. Anderson, J.; Bresnahan, M.; and Musatics, C. Combating weight-based cyberbullying on Facebook with the dissenter effect. Cyberpsychology, Behavior, and Social Networking, 17, 5 (2014), 281–286.

4. Bastiaensens, S.; Vandebosch, H.; Poels, K.; Van Cleemput, K.; DeSmet, A.; and De Bourdeaudhuij, I. Cyberbullying on social network sites. An experimental study into bystan- ders’ behavioural intentions to help the victim or reinforce the bully. Computers in Human Behavior, 31, February 2014 (2014), 259–271.

5. Bayern, M. How AI became Instagram‘s weapon of choice in the war on cyberbully- ing, 2017. https://www.techrepublic.com/article/how-ai-became-instagrams-weapon-of- choice-in-the-war-on-cyberbullying/(accessed on July 7, 2017).

6. Bellmore, A.; Calvin, A.J.; Xu, J.-M.; and Zhu, X. The five W‘s of ‘bullying‘ on Twitter: Who, What, Why, Where, and When. Computers in Human Behavior, 44, March 2015 (2015), 305–314.

7. Bowler, L.; Knobel, C.; and Mattern, E. From cyberbullying to well-being: A narrative-based participatory approach to values-oriented design for social media. Journal of the Association for Information Science and Technology, 66, 6 (2015), 1274–1293.

8. Brody, N.; and Vangelisti, A.L. Bystander intervention in cyberbullying. Communication Monographs, 83, 1 (2015), 1–26.

9. Calvin, A.J.; Bellmore, A.; Xu, J.-M.; and Zhu, X. #bully: Uses of hashtags in posts about bullying on Twitter. Journal of School Violence, 14, 1 (2015), 133–153. 10. Cao, B.; and Lin, W.-Y. How do victims react to cyberbullying on social networking

sites? The influence of previous cyberbullying victimization experiences. Computers in Human Behavior, 52, November 2015 (2015), 458–465. 11. Cassidy, A. Are Facebook and Twitter doing enough to protect users?, The Guardian, 2016. 12. Chapin, J. Adolescents and cyber bullying: The precaution adoption process model.

Education and Information Technologies, 21, 4 (2016), 719–728. 13. Charles, C. 5 reasons why accepting strangers on Facebook is a bad idea, 2014. http://

www.thatsnonsense.com/5-reasons-why-accepting-strangers-on-facebook-is-a-bad-idea /(accessed on July 7, 2017). 14. Charleston, W. UPDATE: ‘Cyberbullying‘ bill to make online harassment a crime in

W.Va. passes Senate. WSAZ, 2017. 15. Chatterjee, S; Moody, G; Lowry, P.B.; Chakraborty, S; and Hardin, A. Strategic

Relevance of Organizational Virtues Enabled by Information Technology in Organizational Innovation, Journal of Management Information Systems, 32, 3 (2015), 158–196. 16. Cheung, C.; Lee, Z.W.; and Chan, T.K. Self-disclosure in social networking sites: the role of

perceived cost, perceived benefits and social influence. Internet Research, 25, 2 (2015), 279–299. 17. Cho, S.; Wooldredge, J.; and Park, C.S. Lifestyles/routine activities and bullying

among South Korean youths. Victims & Offenders, 11(2016), 285–314. 18. Cohen, J. Statistical Power Analysis for the Behavioral Sciences Hillsdale, NJ:

Lawrence Erlbaum, 1988.

CYBERBULLYING ON SOCIAL NETWORKING SITES 601

19. D‘Arcy, J.; Hovav, A.; and Galletta, D. User awareness of security countermeasures and its impact on information systems misuse: A deterrence approach. Information Systems Research, 20, 1 (2009), 79–98. 20. Davern, M.; Shaft, T.; and Te‘eni, D. Cognition matters: Enduring questions in cognitive

IS research. Journal of the Association for Information Systems, 13, 4 (2012), 273–314. 21. Davidson, L. STOP THE TROLLS! Theresa May warns social networking giants

Facebook and Twitter to tighten up their rules on preventing cyber-bullying, 2016. https:// www.thesun.co.uk/news/2202601/theresa-may-warns-social-networking-giants-facebook-and- twitter-to-tighten-up-their-rules-on-preventing-cyber-bullying/(accessed on July 7, 2017). 22. de Coster, S.; Estes, S.B.; and Mueller, C.W. Routine activities and sexual harassment

in the workplace. Work and Occupations, 26, 1 (1999), 21–49. 23. Ditch the Label. The cyberbullying report 2013. http://www.ditchthelabel.org/research-

papers/the-cyberbullying-survey-2013/(accessed on July 8, 2017). 24. Dooley, J.J.; Pyżalsk, J.; and Cross, D. Cyberbullying versus face-to-face bullying:

A theoretical and conceptual review. Journal of Psychology, 217, 4 (2009), 182–188. 25. Dredge, R.; Gleeson, J.; and de la PiedadGarcia, X. Cyberbullying in social networking sites:

An adolescent victim’s perspective. Computers in Human Behavior, 36, July 2014 (2014), 13–20. 26. Dredge, R.; Gleeson, J.; and Garcia, X.d.l.P. Presentation on Facebook and risk of

cyberbullying victimisation.Computers in Human Behavior, 40, November 2014 (2014), 16–22. 27. Dredge, R.; Gleeson, J.F.M.; and de la Piedad Garcia, X. Risk factors associated with

impact of severity of cyberbullying victimization: A qualitative study of adolescent online social networking. Cyberpsychology, Behavior, and Social Networking, 17, 5 (2014), 287–291. 28. ebizmba.com. Top 15 most popular social networking sites | December 2017, 2017.

http://www.ebizmba.com/articles/social-networking-websites (accessed on July 7, 2017). 29. Fay, L. New teen survey reveals cyberbullying moving beyond social media to email,

messaging apps, YouTube, 2017. https://www.the74million.org/new-teen-survey-reveals- cyberbullying-moving-beyond-social-media-to-email-messaging-apps-youtube/(accessed on August 29, 2017). 30. Felson, M.; and Clarke, R., “Opportunity makes the thief: Practical theory for crime

prevention,” The Policing and Reducing Crime Unit, London, 1998. 31. Freis, S.D.; and Gurung, R.A.R. A Facebook analysis of helping behavior in online

bullying. Psychology of Popular Media Culture, 2, 1 (2013), 11–19. 32. Gahagan, K.; Vaterlaus, J.M.; and Frost, L.R. College student cyberbullying on social

networking sites: Conceptualization, prevalence, and perceived bystander responsibility. Computers in Human Behavior, 55, Part B (2016), 1097–1105. 33. Ging, D.; and Norman, O.H.J. Cyberbullying, conflict management or just messing?

Teenage girls‘ understandings and experiences of gender, friendship, and conflict on Facebook in an Irish second-level school. Feminist Media Studies, 16, 5 (2016), 805–821. 34. Goodman, J.K.; Cryder, C.E.; and Cheema, A. Data collection in a flat world: The

strengths and weaknesses of Mechanical Turk samples. Journal of Behavioral Decision Making, 26, 3 (2013), 213–224. 35. Gottfredson, M.; and Hirschi, T. A General Theory of Crime. Stanford, CA: Stanford

University Press, 1990. 36. Grgecic, D.; Holten, R.; and Rosenkranz, C. The impact of functional affordances and

symbolic expressions on the formation of beliefs. Journal of the Association for Information Systems, 16, 7 (2015), 580–607. 37. GuardChild.com. Cyber bullying statistics, 2016. http://www.guardchild.com/cyber-

bullying-statistics/(accessed on July 7, 2017). 38. Hair, J.F.; Black, W.C.; Babin, B.J.; and Anderson, R.E. Multivariate Data Analysis,

7th Ed. Upper Saddle River: NJ: Prentice-Hall International, 2009. 39. Hamm, M.P.P.; Newton, A.S.P.; Chisholm, A.B.; Shulhan, J.B.; Milne, A.M.;

Sundar, P.P.; Ennis, H.M.A.; Scott, S.D.P.; and Hartling, L.P. Prevalence and effect of cyberbullying on children and young people: A scoping review of social media studies. JAMA Pediatrics, 169, 8 (2015), 770–777. 40. Hassan, C. Teen who was relentlessly bullied kills herself in front of her family. CNN, 2016.

602 CHAN ET AL.

41. Hayes, A.F. Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-based Approach. New York: Guilford, 2018. 42. Heirman, W.; and Walrave, M. Predicting adolescent perpetration in cyberbullying: An

application of the theory of planned behavior. Psicothema, 24, 4 (2012), 614–620. 43. Hinduja, S. Neutralization theory and online software piracy: An empirical analysis.

Ethics and Information Technology, 9, 3 (2007), 187–204. 44. Hinduja, S.; and Patchin, J.W. State cyberbullying laws: A brief review of state

cyberbullying laws and policies, 2015. https://cyberbullying.org/Bullying-and- Cyberbullying-Laws.pdf (accessed on July 7, 2017). 45. Kane, G.C.; Alavi, M.; Labianca, G.; and Borgatti, S.P. What‘s different about social

media networks? A framework and research agenda. MIS Quarterly, 38, 1 (2014), 275–304. 46. Kokkinos, C.M.; Baltzidis, E.; and Xynogala, D. Prevalence and personality correlates

of Facebook bullying among university undergraduates. Computers in Human Behavior, 55, Part B (2016), 840–850. 47. Kowalski, R.M.; Giumetti, G.W.; Schroeder, A.N.; and Lattanner, M.R. Bullying in the

digital age: A critical review and meta-analysis of cyberbullying research among youth. Psychological Bulletin, 140, 4 (2014), 1073–1137. 48. Kowalski, R.M.; Limber, S.P.; and Agatston, P.W. Cyber Bullying: Bullying in the

Digital Age. Oxford: Blackwell Publishing, 2008. 49. Kwan, G.C.E.; and Skoric, M.M. Facebook bullying: An extension of battles in school.

Computers in Human Behavior, 29, 1 (2013), 16–25. 50. Lazuras, L.; Barkoukis, V.; Ourda, D.; and Tsorbatzoudis, H. A process model of

cyberbullying in adolescence. Computers in Human Behavior, 29, 3 (2013), 881–887. 51. Lee, J.Y.; Kwon, Y.; Yang, S.; Park, S.; Kim, E.-M.; and Na, E.-Y. Differences in

friendship networks and experiences of cyberbullying among Korean and Australian adolescents. The Journal of Genetic Psychology: Research and Theory on Human Development, 178, 1 (2017), 44–57. 52. Legislation.gov.uk. Communications act 2003, 2018. https://www.legislation.gov.uk/

ukpga/2003/21/section/127 (accessed on August 8, 2017). 53. Leonardi, P.M. When does technology use enable network change in organizations?

A comparative study of feature use and shared affordances.MIS Quarterly, 37, 3 (2013), 749–776. 54. Lowry, P.B.; D’Arcy, J.; Hammer, B.; and Moody, G.D. “Cargo Cult” science in

traditional organization and information systems survey research: A case for using nontradi- tional methods of data collection, including Mechanical Turk and online panels. The Journal of Strategic Information Systems, 25, 3 (2016), 232–240. 55. Lowry, P.B.; Moody, G.D.; and Chatterjee, S. Using IT design to prevent

cyberbullying. Journal of Management Information Systems, 34, 3 (2017), 863–901. 56. Lowry, P.B.; Zhang, J.; Wang, C.; and Siponen, M.Why do adults engage in cyberbullying

on social media? An integration of online disinhibition and deindividuation effects with the social structure and social learningmodel. Information Systems Research, 27, 4 (2016), 962–986. 57. Lwin, M.; Stanaland, A.; and Miyazaki, A. Protecting children’s privacy online: How

parental mediation strategies affect website safeguard effectiveness. Journal of Retailing, 84, 2 (2008), 205–217. 58. MacKinnon, D.P.; and Fairchild, A.J. Current directions in mediation analysis. Current

Directions in Psychological Science, 18, 1 (2009), 16–20. 59. Madden, M.; Lenhart, A.; Cortesi, S.; Gasser, U.; Duggan, M.; Smith, A.; and

Beaton, M., “Teens, Social Media, and Privacy,” Pew Research Center, 2013. 60. Majchrzak, A.; Faraj, S.; Kane, G.C.; and Azad, B. The contradictory influence of

social media affordances on online communal knowledge sharing. Journal of Computer- Mediated Communication, 19, 1 (2013), 38–55. 61. Majchrzak, A.; Markus, M.L.; and Wareham, J. ICT and societal challenges. MIS

Quarterly, 37, 1 (2013), 1–3. 62. Marcum, C.D.; Higgins, G.E.; Freiburger, T.L.; and Ricketts, M.L. Exploration of the

cyberbullying victim/offender overlap by sex. American Journal of Criminal Justice, 39, 3 (2014), 538–548.

CYBERBULLYING ON SOCIAL NETWORKING SITES 603

63. Markus, M.L.; and Silver, M.S. A foundation for the study of IT effects: A new look at DeSanctis and Poole‘s concepts of structural features and spirit. Journal of the Association for Information Systems, 9, 10/11 (2008), 609–632. 64. McAfee, “2014 Teens and the screen study: Exploring online privacy, social network-

ing and cyberbullying,” 2014. 65. McHugh, B.C.; Wisniewski, P.; Rosson, M.B.; and Carroll, J.M. When social media

traumatizes teens: The roles of online risk exposure, coping, and post-traumatic stress. Internet Research, 28, 5 (2018), 1169–1188. 66. Mcneel, B. Latest local cyberbullying case contains valuable lessons. The Rivard Report,

2017. 67. Meter, D.J.; and Bauman, S. When sharing is a bad idea: The effects of online social

network engagement and sharing passwords with friends on cyberbullying involvement. Cyberpsychology, Behavior, and Social Networking, 18, 8 (2015), 437–442. 68. Moore, G.C.; and Benbasat, I. Development of an instrument to measure the percep-

tions of adopting an information technology innovation. Information Systems Research, 2, 3 (1991), 192–222. 69. Obermaier, M.; Fawzi, N.; and Koch, T. Bystanding or standing by? How the number of

bystanders affects the intention to intervene in cyberbullying. NewMedia & Society (2014), 1–7. 70. Pabian, S.; De Backer, C.J.S.; and Vandebosch, H. Dark Triad personality traits and

adolescent cyber-aggression. Personality and Individual Differences, 75(2015), 41–46. 71. Pabian, S.; and Vandebosch, H. Using the theory of planned behaviour to understand

cyberbullying: The importance of beliefs for developing interventions. European Journal of Developmental Psychology, 11, 4 (2014), 463–477. 72. Patchin, J.W.; and Hinduja, S. Bullies move beyond the schoolyard. Youth Violence

and Juvenile Justice, 4, 2 (2006), 148–169. 73. Peluchette, J.V.; Karl, K.; Wood, C.; and Williams, J. Cyberbullying victimization: Do

victims’ personality and risky social network behaviors contribute to the problem? Computers in Human Behavior, 52, November 2015 (2015), 424–435. 74. Pew Research Center, “Online Harassment 2017,” July 11 2017. 75. Podsakoff, P.M.; MacKenzie, S.B.; Lee, J.Y.; and Podsakoff, N.P. Common method

biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 5 (2003), 879–903. 76. Popp, A.M. The difficulty in measuring suitable targets when modeling victimization.

Violence and Victims, 27, 5 (2012), 689–709. 77. Rachoene, M.; and Oyedemi, T. From self-expression to social aggression:

Cyberbullying culture among South African youth on Facebook. Communicatio, 41, 3 (2015), 302–319. 78. Räsänen, P.; Hawdon, J.; Holkeri, E.; Keipi, T.; Näsi, M.; and Oksanen, A. Targets of

online hate: Examining determinants of victimization among young Finnish Facebook users. Violence and Victims, 31, 4 (2016), 708–725. 79. Raskauskas, J.; and Stoltz, A.D. Involvement in traditional and electronic bullying

among adolescents. Developmental Psychology, 43, 3 (2007), 564–575. 80. Reynolds, W.M. Development of reliable and valid short forms of the Marlowe-Crowne

Social Desirability Scale. Journal of Clinical Psychology, 38, 1 (1982), 119–125. 81. Rindfleisch, A.; Malter, A.J.; Ganesan, S.; and Moorman, C. Cross-sectional versus

longitudinal survey research: Concepts, findings, and guidelines. Journal of Marketing Research, 45, 3 (2008), 261–279. 82. Runions, K.C.; and Bak, M. Online moral disengagement, cyberbullying, and

cyber-aggression. Cyberpsychology Behavior and Social Networking, 18, 7 (2015), 400–405. 83. Schacter, H.L.; Greenberg, S.; and Juvonen, J. Who‘s to blame?: The effects of victim

disclosure on bystander reactions to cyberbullying. Computers in Human Behavior, 57, April 2016 (2016), 115–121. 84. Schwarz, A.; Rizzuto, T.; Carraher-Wolverton, C.; Roldán, J.L.; and Barrera-Barrera, R.

Examining the impact and detection of the urban legend of common method bias. ACM SIGMIS Database: the DATABASE for Advances in Information Systems, 48, 1 (2017), 93–119.

604 CHAN ET AL.

85. Seidel, S.; Recker, J.C.; and Vom Brocke, J. Sensemaking and sustainable practicing: functional affordances of information systems in green transformations. MIS Quarterly, 37, 4 (2013), 1275–1299. 86. Sen, R.; and Borle, S. Estimating the contextual risk of data breach: An empirical

approach. Journal of Management Information Systems, 32, 2 (2015), 314–341. 87. Sengupta, A.; and Chaudhuri, A. Are social networking sites a source of online

harassment for teens? Evidence from survey data. Children and Youth Services Review, 33, 2 (2011), 284–290. 88. Slonje, R.; Smith, P.K.; and Frisén, A. The nature of cyberbullying, and strategies for

prevention. Computers in Human Behavior, 29, 1 (2013), 26–32. 89. Smith, A., “6 New Facts about Facebook,” Pew Research Center, 2014. 90. Statista.com, “Number of Social Media Users Worldwide from 2010 to 2021 (in billions),”

2017. 91. Sticca, F.; Ruggieri, S.; Alsaker, F.; and Perren, S. Longitudinal risk factors for cyberbul-

lying in adolescence. Journal of Community & Applied Social Psychology, 23, 1 (2013), 52–67. 92. Strong, D.M.; Johnson, S.A.; Tulu, B.; and Trudel, J. A theory of organization-EHR

affordance actualization. Journal of the Association for Information Systems, 15, 2 (2014), 53–85. 93. Suh, A.; Cheung, C.M.; Ahuja, M.; and Wagner, C. Gamification in the workplace:

The central role of the aesthetic experience. Journal of Management Information Systems, 34, 1 (2017), 268–305. 94. Tarafdar, M.; Gupta, A.; and Turel, O. The dark side of information technology use.

Information Systems Journal, 23, 3 (2013), 269–275. 95. Tokunaga, R.S. Following you home from school: A critical review and synthesis of

research on cyberbullying victimization. Computers in Human Behavior, 26, 3 (2010), 277–287. 96. Treem, J.W.; and Leonardi, P.M. Social media use in organizations: Exploring the

affordances of visibility, editability, persistence, and association. Annals of the International Communication Association, 36, 1 (2013), 143–189. 97. Vance, A.; Lowry, P.B.; and Eggett, D. Increasing accountability through the user

interface design artifacts: A new approach to addressing the problem of access-policy violations. MIS Quarterly, 39, 2 (2015), 345–366. 98. Vandebosch, H.; Van Cleemput, K.; and Cleemput, K.V. Cyberbullying among young-

sters: Profiles of bullies and victims. New Media & Society, 11, 8 (2009), 1349–1371. 99. Venkatraman, N. Strategic orientation of business enterprises: The construct, dimen-

sionality, and measurement. Management Science, 35, 8 (1989), 942–962. 100. Vold, G.B.; Bernard, T.J.; and Snipes, J.B. Theoretical Criminology. New York: Oxford University Press, 1998. 101. Wang, J.; Iannotti, R.J.; and Nansel, T.R. School bullying among adolescents in the United States: Physical, verbal, relational, and cyber. Journal of Adolescent Health, 45, 4 (2009), 368–375. 102. Wegge, D.; Vandebosch, H.; Eggermont, S.; and Walrave, M. The strong, the weak, and the unbalanced: The link between tie strength and cyberaggression on a social network site. Social Science Computer Review, 33, 3 (2015), 315–342. 103. Whittaker, E.; and Kowalski, R.M. Cyberbullying via social media. Journal of School Violence, 14, 1 (2015), 11–29. 104. Wiklund, G.; Ruchkin, V.V.; Koposov, R.A.; and af Klinteberg, B. Pro-bullying attitudes among incarcerated juvenile delinquents: Antisocial behavior, psychopathic tenden- cies and violent crime. International Journal of Law and Psychiatry, 37, 3 (2014), 281–288. 105. Willard, N.E. An educator‘s guide to cyberbullying and cyberthreats, 2004. http:// cyberbully.org/(accessed on July 7, 2017). 106. Williams, K.R.; and Guerra, N.G. Prevalence and predictors of Internet bullying. Journal of Adolescent Health, 41, 6 (2007), 14–21. 107. Willison, R.; and Backhouse, J. Opportunities for computer crime: considering systems risk from a criminological perspective. European Journal of Information Systems, 15, 4 (2006), 403–414.

CYBERBULLYING ON SOCIAL NETWORKING SITES 605

A pp en di x

T ab le

A . S um

m ar y of

P ri or

S tu di es

on S N S B ul ly in g

S tu dy

O bj ec ti ve

T he or et ic al

fo un

da ti on

M et ho

d S am

pl e

A lh ab

as h

et al . [2 ]

T o ex

pl or e th e pe

rs ua

si ve

ef fe ct s of

th e us

e of

em ot io na

la pp

ea l

an d m es

sa ge

vi ra lit y of

F ac

eb oo

k st at us

up da

te s as

a co

rr ec

tiv e

to ol

fo r cy be

rb ul ly in g

D id

no t sp

ec ify

E xp

er im

en t

U ni ve

rs ity

st ud

en t

(n = 36

5)

A nd

er so

n et

al . [3 ]

T o te st

ho w

so ci al

su pp

or t fo r th e vi ct im

, vi a di ss en

tin g co

m m en

ts ,

m ay

af fe ct

by st an

de rs ’ be

ha vi or s in

a cy be

rb ul ly in g ep

is od

e D id

no t sp

ec ify

E xp

er im

en t

U ni ve

rs ity

st ud

en t( n = 18

1)

B as

tia en

se ns

et al . [4 ]

T o ex

am in e th e in flu

en ce

of co

nt ex

tu al

fa ct or s on

by st an

de rs ’

be ha

vi or al

in te nt io ns

to he

lp th e vi ct im

or re in fo rc e th e bu

lly in

ca se

s of

ha ra ss m en

t on

F ac

eb oo

k

D id

no t sp

ec ify

E xp

er im

en t

H ig h sc ho

ol st ud

en ts

(n = 45

3)

B el lm

or e et

al .

[6 ]

T o un

de rs ta nd

cy be

rb ul ly in g us

in g so

ci al

m ed

ia da

ta D id

no t sp

ec ify

M ac

hi ne

le ar ni ng

m et ho

ds

P ub

lic tw ee

ts be

tw ee

n S ep

te m be

r 20

11 , an

d A ug

us t 20

13 (n

= 97

64 58

3) B ow

le r et

al .

[7 ]

T o ge

ne ra te

a va

lu es

-o rie

nt ed

, us

er -g en

er at ed

co nc

ep tu al

fr am

ew or k fo r un

de rs ta nd

in g an

d gu

id in g th e de

si gn

of so

ci al

m ed

ia th at

m ig ht

co un

te ra ct

or pr ev

en t cy be

rb ul ly in g

C he

ng an

d F le is ch

m an

’s va

lu es

fr am

ew or k

V is ua

l na

rr at iv e

in qu

iry

U ni ve

rs ity

st ud

en ts

an d

T ee

ns (n

= 9)

B ro dy

an d

V an

ge lis ti

[8 ]

T o ex

am in e va

ria bl es

th at

w er e ex

pe ct ed

to in flu

en ce

th e

pr op

en si ty

of a by

st an

de r to

ac t in

cy be

rb ul ly in g in ci de

nt s

B ys ta nd

er ef fe ct

S ur ve

y; E xp

er im

en t

U ni ve

rs ity

st ud

en ts

(n = 26

5; n = 37

9)

606 CHAN ET AL.

C al vi n et

al .

[9 ]

T o un

de rs ta nd

th e bu

lly in g to pi cs

th at

T w itt er

us er s po

st ed

ab ou

t ac

ro ss

20 12

by st ud

yi ng

w hi ch

ha sh

ta gs

w er e em

pl oy

ed an

d ho

w th ey

w er e ut ili ze

d.

D id

no t sp

ec ify

D at a m in in g

H as

ht ag

s be

tw ee

n Ja

nu ar y 1,

20 12

an d

D ec

em be

r 31

, 20

12 (n

= 55

28 31

) C ao

an d Li n

[1 0]

T o in ve

st ig at e ho

w vi ct im

iz at io n ex

pe rie

nc es

, in flu

en ce

te en

ag er s’

re ac

tio n st ra te gi es

w he

n w itn

es si ng

cy be

rb ul ly in g on

S N S s

D id

no t sp

ec ify

S ur ve

y T ee

ns (n

= 62

2)

C ha

pi n [1 2]

T o do

cu m en

t ad

ol es

ce nt s us

e of

F ac

eb oo

k an

d ex

pe rie

nc e w ith

cy be

rb ul ly in g

T he

pr ec

au tio

n ad

op tio

n pr oc

es s

m od

el

S ur ve

y A do

le sc en

ts (n

= 14

88 )

D re dg

e et

al .

[2 5]

T o ex

am in e ad

ol es

ce nt

vi ct im

s’ un

de rs ta nd

in g of

cy be

rb ul ly in g,

th e

sp ec

ifi c cy be

rb ul ly in g ev

en ts

ex pe

rie nc

ed in

S N S an

d im

pa ct s

D id

no t sp

ec ify

In te rv ie w

H ig h sc ho

ol st ud

en ts

(n = 25

) D re dg

e et

al .

[2 6]

T o id en

tif y th e fa ct or s th at

af fe ct

th e im

pa ct

of cy be

rb ul ly in g up

on ad

ol es

ce nt

vi ct im

s w ho

us e S N S

D id

no t sp

ec ify

In te rv ie w

H ig h sc ho

ol st ud

en ts

(n = 25

) D re dg

e et

al .

[2 7]

T o in ve

st ig at e th e as

so ci at io ns

be tw ee

n se

lf- pr es

en ta tio

n be

ha vi or s in

F ac

eb oo

k an

d cy be

rb ul ly in g vi ct im

iz at io n

T he

vi ct im

pr ec

ip ita

tio n

m od

el

C on

te nt

an al ys is

F ac

eb oo

k pr of ile

pa ge

s (n

= 14

7)

F re is

an d

G ur un

g [3 1]

T o de

te rm

in e w ha

t w ill m ak

e a pa

rt ic ip an

t in te rv en

e in

an on

lin e

bu lly in g si tu at io n,

an d to

m ea

su re

th e ty pe

s of

te ch

ni qu

es pa

rt ic ip an

ts us

e to

in te rv en

e

D id

no t sp

ec ify

E xp

er im

en t

U ni ve

rs ity

st ud

en t

(n = 37

)

G ah

ag an

et al . [3 2]

T o in cr ea

se un

de rs ta nd

in g re ga

rd in g cy be

rb ul ly in g ex

pe rie

nc e on

so ci al

ne tw or ki ng

si te s am

on g co

lle ge

st ud

en ts

D id

no t sp

ec ify

S ur ve

y U ni ve

rs ity

st ud

en t

(n = 19

6) G in g an

d N or m an

[3 3]

T o ex

pl or e ho

w fr ie nd

sh ip , co

nf lic t, an

d bu

lly in g ar e ex

pe rie

nc ed

an d un

de rs to od

by Ir is h te en

ag e gi rls

in re la tio

n to

F ac

eb oo

k D id

no t sp

ec ify

S ur ve

y H ig h sc ho

ol st ud

en t

(n = 11

6)

H am

m et

al .

[3 9]

T o re vi ew

ex is tin

g pu

bl ic at io ns

th at

ex am

in e th e he

al th -r el at ed

ef fe ct s of

cy be

rb ul ly in g vi a so

ci al

m ed

ia am

on g ch

ild re n an

d ad

ol es

ce nt s

D id

no t sp

ec ify

Li te ra tu re

re vi ew

P ee

r- re vi ew

ed jo ur na

l ar tic le s (n

= 34

) (c on ti nu es )

CYBERBULLYING ON SOCIAL NETWORKING SITES 607

T ab le

A . C on

ti nu

ed

S tu dy

O bj ec ti ve

T he or et ic al

fo un

da ti on

M et ho

d S am

pl e

K ok

ki no

s et

al . [4 6]

T o ex

am in e th e pr ev

al en

ce of

cy be

rb ul ly in g on

F ac

eb oo

k an

d its

as so

ci at io ns

w ith

in di vi du

al ch

ar ac

te ris

tic s

D id

no t sp

ec ify

S ur ve

y U ni ve

rs ity

st ud

en ts

(n = 22

6) K w an

an d

S ko

ric [4 9]

T o ex

am in e th e ph

en om

en on

of cy be

rb ul ly in g on

F ac

eb oo

k an

d ho

w it is

re la te d to

sc ho

ol bu

lly in g am

on g se

co nd

ar y sc ho

ol st ud

en ts

D id

no t sp

ec ify

S ur ve

y H ig h sc ho

ol st ud

en t

(n = 16

76 )

Le e et

al . [5 1]

T o in ve

st ig at e th e re la tio

ns hi ps

be tw ee

n fr ie nd

sh ip

ne tw or ks

w ith

th e ex

pe rie

nc es

as vi ct im

s, pe

rp et ra to rs , an

d by

st an

de rs

of cy be

rb ul ly in g am

on g yo

un g ad

ol es

ce nt s

D id

no t sp

ec ify

S ur ve

y A do

le sc en

ts (n

= 92

1)

Lo w ry

et al .

[5 6]

T o st ud

y ho

w th e in fo rm

at io n te ch

no lo gy

ar tif ac

t in flu

en ce

an d w hy

pe op

le ar e so

ci al iz ed

to en

ga ge

in cy be

rb ul ly in g

S oc

ia ll ea

rn in g

th eo

ry of

cr im

e S ur ve

y A du

lt (n

= 10

03 )

Lo w ry

et al .

[5 5]

T o ex

pl or e sy st em

ch ar ac

te ris

tic s th at

pr ev

en t cy be

rb ul ly in g

C on

tr ol

ba la nc

e th eo

ry F ac

to ria

l su

rv ey

A du

lt (n

= 50

7) M ar cu

m et

al .

[6 2]

T o ex

pl or e th e di ffe

re nc

es in

m al e an

d fe m al e cy be

rb ul ly in g,

as w el l

as th e vi ct im

-o ffe

nd er

re la tio

ns hi p ex

pe rie

nc ed

by ea

ch se

x D id

no t sp

ec ify

S ur ve

y U ni ve

rs ity

st ud

en ts

(n = 11

39 )

M et er

an d

B au

m an

[6 7]

T o st ud

y th e re la tio

ns hi ps

be tw ee

n so

ci al

ne tw or k en

ga ge

m en

ta nd

cy be

rb ul ly in g in vo

lv em

en t ov

er tim

e T he

so ci al -

ec ol og

ic al

m od

el S ur ve

y S tu de

nt s

(n = 12

72 )

P ab

ia n et

al .

[7 0]

T o em

pi ric

al ly

in ve

st ig at e th e re la tio

ns hi ps

be tw ee

n th e da

rk tr ia d

pe rs on

al ity

tr ai ts

an d cy be

r- ag

gr es

si on

on F ac

eb oo

k D id

no t sp

ec ify

S ur ve

y A do

le sc en

ts (n

= 32

4) P el uc

he tte

et al . [7 3]

T o ex

am in e th e im

pa ct s of

ris ky

so ci al

ne tw or k si te

pr ac

tic es

an d

in di vi du

al di ffe

re nc

es in

se lf- di sc lo su

re an

d pe

rs on

al ity

on cy be

rb ul ly in g vi ct im

iz at io n on

F ac

eb oo

k us

er s

D id

no t sp

ec ify

S ur ve

y Y ou

ng ad

ul ts

(n = 57

2)

O be

rm ai er

et al . [6 9]

T o ex

am in e th e by

st an

de r ef fe ct

in cy be

rb ul ly in g

B ys ta nd

er ef fe ct

E xp

er im

en t

U ni ve

rs ity

st ud

en t

(n = 85

; n = 44

0)

608 CHAN ET AL.

R ac

ho en

e an

d O ye

de m i

[7 7]

T o ex

am in e on

lin e bu

lly in g am

on g S ou

th A fr ic an

yo ut h on

F ac

eb oo

k D id

no t sp

ec ify

D ig ita

l et hn

og ra ph

y F ac

eb oo

k pa

ge (n

= 6)

R äs

än en

et al . [7 8]

T o ex

am in e th e de

te rm

in an

t on

lin e ha

te vi ct im

iz at io n on

F ac

eb oo

k D id

no t sp

ec ify

S ur ve

y F in ni sh

F ac

eb oo

k us

er s (n

= 72

3) S ch

ac te r

et al . [8 3]

T o un

de rs ta nd

th e co

nd iti on

s un

de r w hi ch

by st an

de rs

w ill sh

ow in cr ea

se d su

pp or t fo r vi ct im

s of

cy be

rb ul ly in g

A ttr ib ut io n th eo

ry E xp

er im

en t

A du

lt (n

= 11

8)

S en

gu pt a an

d C ha

ud hu

ri [8 7]

T o id en

tif y th e ke

y fa ct or s as

so ci at ed

w ith

cy be

r- bu

lly in g an

d on

lin e

ha ra ss m en

t of

te en

ag er s in

th e U ni te d S ta te s

D id

no t sp

ec ify

P an

el da

ta fr om

P E W

T ee

n (n

= 93

5)

W eg

ge et

al .

[1 02

] T o ex

am in e ho

w yo

un g pe

op le ’s

co nn

ec tio

ns on

S N S s ar e re la te d

to th ei r ris

k of

be in g in vo

lv ed

in cy be

r- ha

ra ss m en

t an

d cy be

rb ul ly in g

D id

no t sp

ec ify

S ur ve

y H ig h sc ho

ol st ud

en t

(n = 14

58 )

W hi tta

ke r an

d K ow

al sk i

[1 03

]

T o ex

am in e th e pr ev

al en

ce ra te s of

cy be

rb ul ly in g am

on g co

lle ge

- ag

e st ud

en ts

D id

no t sp

ec ify

S ur ve

y da

ta m in in g

U ni ve

rs ity

st ud

en t

(n = 24

4; n = 19

7) F ac

eb oo

k po

st (n

= 29

61 )

CYBERBULLYING ON SOCIAL NETWORKING SITES 609

Copyright of Journal of Management Information Systems is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

  • Abstract
  • Introduction
  • Theoretical Background
    • Definition of Cyberbullying
    • Nature of SNS Bullying
  • Toward a Meta-Framework of SNS Bullying
    • Crime Opportunity Theory
    • Affordance Perspective
  • Research Model and Hypotheses
    • Likely Offender and SNS Bullying
    • Evaluation of SNS Environmental Conditions and SNS Bullying
      • Presence of Suitable Targets
      • Absence of Capable Guardianships
    • SNS Affordances and the Evaluation of SNS Environmental Conditions
      • Accessibility Affordance
      • Information Retrieval Affordance
      • Editability Affordance
      • Association Affordance
    • Control Variables
  • Research Method
    • Research Design
    • Measure
    • Data Collection and Procedures
    • Respondent Profile
  • Data Analysis and Results
  • Post Hoc Analyses
    • Comparison of Alternative Models
    • Assessment of the Mediation Effects
    • Assessment of the Interaction Effects
      • Inclination to Bully × The Evaluation of SNS Environmental Conditions
      • Presence of Suitable Targets × Absence of Capable Guardianships
      • Inclination to Bully × Presence of Suitable Targets × Absence of Capable Guardianships
  • Discussion
    • Implications for Research
    • Implications for Practice
    • Limitations and Future Research Directions
  • Conclusion
  • The authors wish to thank the Editor-in-Chief, Professor Zwass, and the reviewers for their support and guidance throughout the review process.
  • Funding
  • SUPPLEMENTAL MATERIAL
  • References
  • Appendix