Bystander Apathy and Intervention in the Era of Social Media

Abstract

Abstract Bystander apathy has been a source of debate for decades. In the past half-century, psychologists developed theoretical frameworks to understand bystander activity, commonly referred to as bystander intervention models (BIMs). More recently, BIMs have been modified to facilitate initiatives to prevent various forms of online victimization. This chapter begins with a review of BIMs and recent applications of bystander intervention research to online environments. We also present several future directions for research along with applications for reducing technology-facilitated violence, including programming recommendations and theoretical development. Keywords Bystander apathy Bystander intervention model Prevention Online communities Social media Online victimization Citation Lytle, R.D., Bratton, T.M. and Hudson, H.K. (2021), "Bystander Apathy and Intervention in the Era of Social Media", Bailey, J., Flynn, A. and Henry, N. (Ed.) The Emerald International Handbook of Technology-Facilitated Violence and Abuse (Emerald Studies In Digital Crime, Technology and Social Harms), Emerald Publishing Limited, Bingley, pp. 711-728. https://doi.org/10.1108/978-1-83982-848-520211052 Publisher: Emerald Publishing Limited Copyright © 2021 Robert D. Lytle, Tabrina M. Bratton, and Heather K. Hudson. Published by Emerald Publishing Limited. This chapter is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of these chapters (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode. License This chapter is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of these chapters (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode. Introduction If you have visited a metropolitan area, you have likely been present for an incident that could be considered an emergency. This might include walking past an unconscious person on the street one night, a shriek from a child as you walk past a park, or perhaps a heated argument between two adults in the neighboring room. When you perceive these types of stimuli in your environment, you engage in a cognitive process that helps to decide whether or not to act. This process includes a range of cognitive mechanisms wherein we interpret otherwise ambiguous information (e.g., Is the argument between the two adults becoming dangerous?), assess the situation (e.g., Can someone else assist the child in the park?), and consider how best to intervene (e.g., Do you wake the unconscious person and call for help?). Though we complete this process quickly, we remain partially paralyzed while resolving the conflict of witnessing a potential emergency. Social scientists have generated a wealth of literature to explain the process driving bystander action in the presence of a possible emergency (Batson, van Lange, Ahmad, & Lishner, 2007). Much of this research has been built upon the bystander intervention model (BIM) proposed by Latané and Darley (1970) which situates the cognitive processes driving decisions to intervene in ambiguous emergencies within an unpredictable social environment. BIMs have informed applications to promote bystander intervention in emergencies, including violent situations, with some effect (Del Rey, Casas, & Ortega, 2016; Polanin, Espelage, & Pigott, 2012). Despite this success, society has experienced significant changes in both cognitive processing and environments following the rise of social media. This chapter begins by briefly reviewing the literature on bystander apathy and BIMs, with a special emphasis on developments in bystander theory in light of recent technological advancements. The remainder of the chapter is dedicated to proposed directions for developing theory, methodological considerations for future research, and applying BIMs to understanding and reducing technology-facilitated violence and abuse (TFVA). The Bystander Intervention Model In 1968, John Darley and Bibb Latané published a seminal work on an emerging psychological concept referred to as bystander apathy. Bystander apathy refers to the phenomenon wherein otherwise moral persons can observe an emergency (e.g., accident, crime) without coming to the aid of those affected by it. Although incidents of bystander apathy were not novel at the time, social commentary grew in the wake of the murder of Kitty Genovese in 1964 (Manning, Levins, & Collins, 2007). Genovese was reportedly murdered while bystanders in the nearby apartments, all aware of the attack, did little to intervene on her behalf. Explanations centered around the moral ambivalence or decay of society at the time. Latané and Darley (1970), however, hypothesized that bystander apathy might emerge from the internal conflict one experiences when confronted with an emergency. In effect, they argued that bystander apathy did not necessarily imply a moral deficiency in the passive bystander; rather, they argued nonintervention could be expected when the situation surrounding the emergency favored nonintervention (Darley & Latané, 1968). After several classic experiments, Latané and Darley (1970) presented their BIM to explain bystander decisions about whether or not to intervene in emergencies. Latané and Darley's (1970) model holds that people experience internal conflict when observing an emergency, regardless of their desire to assist or intervene in the danger. The BIM consists of five steps that influence whether or not a person will intervene. The first step is to witness the incident. Once a potential emergency has been witnessed, the second step requires the bystander to define the situation as an emergency. If the danger to the bystander or victim is unclear, the bystander may follow others' lead in recognizing whether the ambiguous situation (e.g., smoke, loud sounds) demands a response (Clark & Word, 1972; Darley & Latané, 1968; Latané & Darley, 1970). This is referred to as audience inhibition or pluralistic ignorance (Latané & Nida, 1981). This psychological process is so impactful that, even when the situation becomes more clearly threatening, the bystander will remain calm if others witnessing the situation do not react to the threat (Darley & Latané, 1968). If the bystander defines the situation as an emergency, they then look to the environment for justification to avoid the risk often present in an intervention (e.g., injury, embarrassment) (Latané & Darley, 1970). If a bystander is alone with the victim of an emergency, the responsibility of intervening on the victim's behalf is placed entirely on the bystander. As such, the cost of intervention may not outweigh the moral obligation they feel. However, as the number of other persons capable of intervening increases, the bystander can mentally share the responsibility for providing aid to a victim, which is known as diffusion of responsibility (Darley & Latané, 1968; Garcia, Weaver, Moskowitz, & Darley, 2002). As a result, the moral obligation felt by the bystander is attenuated and may not overcome the costs associated with intervention (Darley & Latané, 1968). Scholars often refer to nonintervention facilitated by bystander group size as the “bystander effect” (Levine, Philpot, & Kovalenko, 2020). Although the BIM has enjoyed the theoretical spotlight for understanding helping behaviors, recent developments have revealed conditions to the initial framework (Fischer et al., 2011). One such development in bystander intervention research has centered around the nature of the emergency. Whereas classic bystander effect studies observed inaction in response to emergencies that posed little to no immediate danger to the participant (Levine & Darley, 1968), more recent research has proposed that dangerous or violent emergencies may be less likely to result in bystander inaction; a phenomenon commonly referred to as positive bystander effects (Fischer et al., 2011; Fischer, Greitemeyer, Pollozek, & Frey, 2006). Another relevant development in the bystander intervention literature relevant to understanding online environments involves the group composition of persons surrounding the bystander during the emergency (referred to here as witnesses). Based on meta-analyses of bystander intervention research, helping behavior is more likely to occur when a bystander is surrounded by witnesses who are known to one another (Rutkowski, Gruder, & Romer, 1983), share membership in a common social or demographic group (Levine & Crowther, 2008; Levine, Prosser, Evans, & Reicher, 2005), or are simply active observers (Fischer & Greitemeyer, 2013). Relatedly, bystanders who know the victim are also more likely to intervene in an emergency regardless of the size (or composition) of fellow witnesses to the emergency (Palmer, Nicksa, & McMahon, 2018). In this regard, bystander intervention may be similar in online environments. That is, helping behavior is likely dependent on the online relationship between the witness and victim. Nonetheless, the BIM has demonstrated a moderately weak effect in explaining bystander inaction (Fischer et al., 2011). The utility of this theory can also be observed through its application to various forms of violence reduction and intervention programs, including bullying (Polanin et al., 2012) and campus sexual assault (Banyard, Moynihan, Cares, & Warner, 2014). 1 It has, however, emerged as a helpful tool to combat non-reporting and inaction from persons who observe an emergency, especially incidents that may traditionally have been viewed as “private,” such as domestic violence (Levine et al., 2020; McMahon & Banyard, 2012; Palmer et al., 2018). Bystander Intervention Models in Online Environments Social interaction has changed significantly since Latané and Darley's (1970) seminal work. In particular, the rise of technological advancements in communication, such as social media, has changed societal norms, behaviors, and social environments. There are numerous benefits to social media, such as widening one's social networks and providing an outlet for communication and knowledge generation. However, the ubiquitous nature of social media today also has the potential to facilitate antisocial online behavior (e.g., Hayes, 2019; Southworth, Finn, Dawson, Fraser, & Tucker, 2007; Woodlock, 2017), including the targeting of minority populations such as racial minorities and members of the LGBTQI communities. Antisocial and cybercrime behaviors like online harassment can have detrimental impacts on the social, mental, and physical wellness of victims (Bates, 2017; Henry et al., 2020; Powell, Scott, Henry, & Flynn, 2020; Powell, Henry, & Flynn, 2018; Tennant, Demaray, Coyle, & Malecki, 2015; Widgerson & Lynch, 2013). These consequences not only lead to stigmatization, they may contribute to psychological distress (Wright, 2016) and suicide (Bauman, Toomey, & Walker, 2013; Sullivan, Annest, Simon, Luo, & Dahlberg, 2015). Although the emergence of social media has undoubtedly changed the ways and environment in which we communicate with each other, the classic bystander effect may continue to inhibit bystanders from intervening online (Dempsey, Sulkowski, Nichols, & Storch, 2009; Olenik-Shemesh, Heiman, & Eden, 2017; Sticca & Perren, 2013). One possible source of bystander inaction in online environments may be ambiguity in cybervictimization. Cybervictimization, along with the online social environment they occur within, is relatively new. As a result, the norms and values for observing and responding to status updates in social media continue to evolve, which may create uncertainty for witnesses regarding what is and is not acceptable behavior online (Banyard, Plante, & Moynihan, 2005). In much the same way that perpetrators can benefit from anonymity in online environments, bystanders of online victimization may be able to avoid the cognitive conflict that often accompanies witnessing a potential emergency (Dempsey et al., 2009). Intervention may be particularly unlikely when the bystander does not know the victim well (Palmer et al., 2018). Finally, the lack of physical presence can make it complicated for bystanders to intervene in the emergency. This is especially the case when considering a lack of parental socialization toward online defending behavior reported by many adolescents (DeSmet et al., 2014). Witnesses inclined to physically intervene in offline situations (e.g., putting their body in front of a victim) may be uncertain about the best way to intervene in online environments. Despite these differences, however, it appears BIMs apply in both settings (Hayes, 2019). Much of the research on bystander intervention in online environments has centered around cyberbullying, sexual violence, and intimate partner violence (IPV; Hayes, 2019). This research has generally found that bystander helping behavior, often taking the form of defending the targeted person, can be predicted by a bystander's social group, attitudes, and situational influences. More specifically, research has found that online bystander intervention is most likely to occur when the bystander is female (Allison & Bussey, 2017; Lambe, Della Cioppa, Hong, & Craig, 2019), older (Hayes, 2019; Olenik-Shemesh et al., 2017), and where they enjoy social support or popularity (Lambe et al., 2019; Olenik-Shemesh et al., 2017). Bystanders are less likely to intervene when they hold negative views of the victim (Hayes, 2019) or decline to speak up against prejudice (Moisuc, Brauer, Fonseca, Chaurand, & Greitemeyer, 2018). Finally, the norms and values of a bystander's environment, including those learned from family (DeSmet et al., 2014; Lambe et al., 2019) or school (Allison & Bussey, 2016, 2017), influence bystander intention to defend, as well as actual defending behaviors in cyberbullying events. Interestingly, though, the ubiquity of social media might also provide a protective effect for victims. For example, victims of cybercrime and online harassment may have more social support than they have offline. To the extent that others might engage in the BIM in online environments, social media may contribute to the prevention of online victimization (e.g., positive bystander effect), as well as curbing the adverse effects of such victimization. Indeed, some scholars have found evidence of a positive bystander effect in online settings (Armstrong & Mahone, 2017). For example, online intervention is more likely, regardless of group size, when witnesses perceive the emergency as being more clearly dangerous or there is a relationship between the bystander and witnesses or victim. This can occur when online bystanders are aware of the dangers associated with online victimization (Armstrong & Mahone, 2017) or share a common identity or group identity (DeSmet et al., 2016; Lambe et al., 2019). Considerations for Studying Bystander Intervention in Social Media A variety of complexities arise when applying offline BIMs to online incidents. As BIMs rely heavily on the environment to explain bystander action, it follows that the theoretical pathways used to explain responses to emergencies may differ based on whether they occur offline or online. In this section, we review some of the ways that traditional BIMs may need further examination in research before application to online social environments. The first step of BIMs is that one must observe the incident (Latané & Darley, 1970). In online settings, especially through social media, it may be difficult to determine when bystanders become aware of potential emergencies (Dillon & Bushman, 2015; McMahon & Banyard, 2012). For example, many social networking sites (SNSs) have a “wall” of posts where comments from followers are brought to the user's attention. SNS walls can be a public “place” used to bully, embarrass, and shame or defend, support, and praise other SNS users. Because of the breadth of information one sees on social media, online bystanders might become aware of incidents days after it occurred (McMahon, Palmer, Banyard, Murphy, & Gidycz, 2017), thereby removing the immediacy of observing the incident from the classic BIM. In light of the asynchronous nature of the online environment, this may be a key consideration for future research. The second step of the BIM is that the bystander must define the incident as an emergency (Latané & Darley, 1970). As outlined earlier, this process can be influenced by the reactions of others present at the incident (Darley & Latané, 1968). BIMs that apply in the online context may need additional research to clarify the cognitive processes by which bystanders evaluate danger online, including content and other signals for help (Schachter, Greenberg, & Juvonen, 2016). The keystone of traditional BIMs has been the influence of bystander group size on intervention (Latané & Darley, 1970). Presently, much of the research into online bystander intervention has focused on how diffusion of responsibility affects bystander responses to online emergencies, with a general conclusion being that the presence of online bystanders discourages bystander action (Dempsey et al., 2009; Hayes, 2019; Sticca & Perren, 2013). Although these studies provide an excellent starting point for online bystander intervention research, there remain several important questions for understanding bystander effects online. For example, how do online bystanders become aware of the presence of other bystanders? The anonymity of many online social environments may affect a person's ability to diffuse responsibility if they are not certain others are available to assist the victim. Conversely, anonymity in social media may allow the bystander to rationalize inaction by assuming others are present and observing the same emergency, therefore someone else may step in. Larger bystander group size has been shown to discourage intervention by simply having bystanders think about the presence of other witnesses (Garcia et al., 2002). Therefore, it may be easy for online bystanders to assume the presence of other witnesses regardless of ambiguous evidence. In addition to applying the classic bystander effect to harassment and violence, research into online bystander intervention should also consider the possibility of positive bystander effects (Fischer et al., 2006). When the victim is in clear danger, such as violence, a larger number of witnesses actually facilitates bystander intervention, regardless of the assumed personal cost to bystanders. This finding has been attributed to both the less ambiguous nature of the threat to a victim in violent encounters (Piliavin, Rodin, & Piliavin, 1969) and the improved chance of successful intervention given the support of other bystanders (Fischer et al., 2011). This makes recognizing danger in online incidents particularly relevant when designing strategies to prevent cybervictimization. Applications to Technology-Facilitated Violence and Abuse BIMs can inform the public about TFVA, educate users about how to identify online problems, and provide recommendations for intervention (Moynihan et al., 2015). Evaluations of school-based BIM programs have proved successful in preventing violence in educational settings, especially among high school (Coker, Bush, Brancato, Clear, & Recktenwald, 2019; Polanin et al., 2012) and college students (Banyard, Moynihan, & Plate, 2007). Although most school-based bystander intervention programs do not use all five stages of the BIM, Nickerson, Aloe, Livingston, & Feeley (2014) suggested relying more on BIM to develop school-based, online bystander programs that explicitly teach the five steps of BIMs to decrease online bystander apathy. BIM-driven programs could facilitate the identification of TFVA, interpretation of TFVA as situations requiring intervention, assumption of responsibility for defending the victim, and learning the skills necessary to intervene safely. Currently, few online bystander intervention programs target TFVA. The BIM can inform strategies to reduce bystander apathy and redirect passive attitudes into acceptable online behavior. Bystander intervention can reduce TFVA by interfering in violent situations before they escalate. Recent research on online bystanders have identified factors that increase the likelihood for intervention such as gender, age, perceived danger, familiarity with the victim, and attitudes (Allison & Bussey, 2016, 2017; Bastiaensens et al., 2014; Hayes, 2019; Olenik-Shemesh et al., 2017). Unfortunately, many of these factors are difficult to implement in programming (i.e., knowing victim, social popularity, and classroom). Furthermore, existing bystander intervention programming, especially with relation to sexual assault prevention, may be difficult to generalize to cases beyond cisgendered, heterosexual male offenders, and female victims (Dunn, Bailey, & Msosa, 2020). Based on existing research, then, intervention training should focus on the factors that can be changed through programming, like helping attitudes and the perceived danger of emergency. To apply the first step of BIM (Latané & Darley, 1970), observation of the incident can be promoted by raising awareness to TFVA behaviors and warning signs. BIMs could also teach users to identify alarming content and TFVA as an emergency (Step 2). A key decision point to address in online bystander intervention programming will likely be the identification of the situation as an emergency. Recent research has suggested that many forms of TFVA reviewed in this chapter (and beyond), may be perceived by both the offender and witnesses as attempts at humor or “having fun” gone awry (Retter & Shariff, 2012; Shariff & DeMartini, 2016; Shariff, Wiseman, & Crestohl, 2012). Although this research has focused primarily on the motivation of the offender, this finding may also apply to bystander assessments of situations as an emergency. If bystanders believe that some event is a poorly executed joke, or they can quickly rationalize it as such, they can define the situation as a non-emergency and avoid responsibility for action. Multiple strategies may be needed to encourage bystanders to accept responsibility for intervention (Step 3). These strategies should promote social norms that encourage empathy for victims of TFVA (Nickerson et al., 2014), as well as increase self-efficacy for intervention among bystanders (Pöyhönen, Juvonen, & Salmivalli, 2012). Alternatively, online bystander intervention programs should consider how internet users perceive the presence of others to combat diffusion of responsibility and pluralistic ignorance, and identify ways to address these challenges (Darley & Latané, 1968; Latané & Darley, 1970). Along with recognizing and accepting responsibility for TFVA incidents, BIM holds that bystanders must have a plan for action (Step 4) and be able to intervene (Step 5; Latané & Darley, 1970). Indeed, bystanders who are better trained in how to respond may be better prepared and therefore, more likely to assist in online emergencies effectively and safely (Coker et al., 2011). These strategies should focus on building knowledge and skills regarding how best to act. Although social media may promote opportunities for intervention (such as flagging or reporting TFVA), some aspects of social media like anonymity appear to hinder bystander intervention (Dempsey et al., 2009; Sticca & Perren, 2013). Therefore, programming should present tangible ways to intervene in specific types of TFVA. Application #1: Technology-Facilitated Suicidality Suicidality has long garnered attention, but only recently have online presentations of suicidality been considered a path to prevention. Some research suggests that adolescents who have suicidal ideations or engage in self-harm may share their feelings and practices on SNSs (Luxton, June, & Fairall, 2012; Marchant et al., 2017). Moreover, suicidality has been linked to social media use (Mok, Jorm, & Pirkis, 2015) and cybervictimization (Hinduja & Patchin, 2010). Witnesses to suicidal or self-harming posts can escalate self-harm through demeaning comments. As such, bystanders to online suicidal ideation and self-harm represent an opportunity to reduce and prevent suicidality. To observe and intervene in suicidal threats or ideations online, internet users should be trained to identify direct and indirect warning signs (Marchant et al., 2017). For example, programs might include ways to teach users to identify risky hashtags (e.g., #cutting), take them seriously, and respond appropriately. To accomplish raising the perceived seriousness of online suicidal ideation, programs should incorporate peers into awareness campaigns. For example, peer-facilitators could confront the acceptance of jokes about suicide and rally others to treat all threats seriously. Peer-led bystander intervention exposure has previously been implemented with success in student populations, especially when many students were trained to be peer leaders (Pickering et al., 2018). Even if online suicidal threats are perceived as worthy of intervention, though, BIMs predict that bystanders will be less likely to respond if one can diffuse responsibility with other witnesses (Latané & Darley, 1970). To encourage internet users to accept responsibility, programs can focus on activities that promote empathy for suicidal persons and self-efficacy in intervention, both of which have been shown to discourage diffusion of responsibility online (Nickerson et al., 2014; Pöyhönen et al., 2012). Even if the witness identifies the situation as an emergency, however, the stress and anxiety produced in individuals intervening in suicidality may create a specific obstacle to bystander action for this form of TFVA (Ting, Jacobson, & Sanders, 2011). In offline settings, the classic bystander apathy effect reverses so that witnesses become less likely to intervene when alone if the emergency poses a danger to the witness (Fischer & Greitemeyer, 2013). Therefore, BIMs intended to address technology-facilitated suicidality might also need to consider ways to help potential interveners process and cope with such anxiety. For example, social media campaigns can educate their users on the importance of intervention despite one's discomfort. In addition, programs to encourage bystander intervention in online environments can include services for interveners should their experience prove to be traumatic. Peer-facilitators can also role play scenarios to help internet users develop strategies for effective intervention, which can empower bystanders in the face of an emergency. Effective plans for intervention can develop through skills-training, access to effective resources, and education on what works and what worsens the situation. Training programs should guide internet users through scenarios to explore direct and indirect options for intervening. Direct interventions for online suicidality could include approaches to talking to persons threatening self-harm online. Alternatively, bystanders may offer support and resources intended to help suicidal persons (e.g., suicide prevention phone lines or licensed counselors). Indirect measures would include strategies for bringing attention to the suicidal content, such as contacting emergency services in an immediate crisis or reporting alarming content on social media platforms. The benefits for bystander prevention in online suicidality are cumulative in the sense that intervention from one bystander may encourage others to act as well (Darley & Latané, 1968; Latané & Nida, 1981). Application #2: Image-Based Sexual Abuse Another source of TFVA that has grown from the ubiquity of social media has been image-based sexual abuse (IBSA). IBSA refers to dissemination of sexualized materials without the subject's consent (Bond & Tyrrell, 2018; Henry & Flynn, 2020). Although still early in its research development, scholars have noted significant health consequences for IBSA that range from symptoms of psychological distress to suicidality (Bates, 2017; Henry, McGlynn, Flynn, Johnson, Powell, & Scott, 2020; Powell et al., 2020; Powell et al., 2018). Despite recognition of the consequences of IBSA for victims, policymakers and police have reported difficulties designing effective responses to IBSA (Bond & Tyrrell, 2018; Henry et al., 2020; Henry, Flynn, & Powell, 2018). This difficulty stems from ambiguity in the definition of IBSA, especially in determining what material should be considered illicit, nonconsensual, and sexual (Bond & Tyrrell, 2018; Flynn & Henry, 2019; Walker & Sleath, 2017). In light of these limitations, bystander intervention may be useful in discouraging IBSA. IBSA grows in online environments, wherein bystanders not only ignore the victimization but also reward its production (Henry & Flynn, 2020; Walker & Sleath, 2017). In addition, bystanders might assume the sexual material was published consensually and may not accurately recognize it as a form of victimization. Furthermore, if bystanders perceive the IBSA as a “joke” or venting frustration (Shariff & DeMartini, 2016), the situation may be defined as a non-emergency. Consequently, intervention would be perceived to be unnecessary. To the degree that bystander apathy contributes to the proliferation of IBSA, then, we may be able to quell its use, support police efforts to take action, and raise public awareness to the methods and conse