Introduction

Information and communication technologies have expanded opportunities for online consumers to access information and stay connected, but online communications also increase risks of confronting cyberbullying by individuals or groups using electronic or digital media to send harmful, discomforting, aggressive messages (Tokunga, 2010). Compared with traditional bullying, cyberbullying causes instant, widespread, and permanent damage to victims’ well-being, mental health, and tendencies toward risky behavior (Kowalski et al., 2014; Tokunaga, 2010). Indeed, almost half of social media users experience cyberbullying and face pervasive risks of suicidal ideations and deaths (Hinduja & Patchin, 2019; Patchin, 2019, Safety Net, 2018).

The increasing societal damages have prompted governments, nonprofit organizations, and brands to create and conduct anti-cyberbullying campaigns. Brands such as Facebook call for state and local lawmakers to undertake legal actions (Kay, 2021). The U.S. Department of Education developed a framework of common components found in anti-cyberbullying state laws, policies, and regulations (Stop-Bullying, 2022). Automated techniques are being used to monitor and analyze cyberbullying. For instance, Apple developed and installed a digital keyboard that uses artificial intelligence to identify offensive words and then deploy pop-up alerts to suggest that users reconsider objectionable words in emails, texts, or social media posts (Selyukh, 2015). Cybersmile Foundation (2021) offers a full curriculum of interactive modules designed to increase awareness and skills for counteracting cyberbullying and suitable for all age groups and abilities. We lack evidence, however, that those good intentions work as intended. Indeed, meta-analyses commonly indicate that anti-cyberbullying programs have unclear or minor effectiveness (Blaya, 2019; Gaffney et al., 2019; Zych et al., 2015).

Interventions fail partly because traditional persuasion techniques generally fail to consider that cyberbullying has unique characteristics of anonymity, remoteness, and diffused responsibility so that users are disinhibited and deindividuated (Lowry et al., 2016). Moreover, cyberbullies cause no direct physical harm and may lack malicious intent (Kwak et al., 2021). Perpetrators and spectators generally escape accountability. Consequently, traditional top-down, direct remedies in which perpetrators passively receive external interventions may unintentionally backfire by prompting psychological reactance and denial (Ashktorab & Vitak, 2016; Jansen & Van Schaik, 2018), indicating imperative needs for rigorous research that develops and tests strategies for motivating consumers to support anti-cyberbullying campaigns (Blaya, 2019; Gaffney et al., 2019; Zych et al., 2015).

We suggest that self-persuasion will be more effective than external interventions and that hypocrisy induction is a novel way to motivate online users to become voluntarily engaged in anti-cyberbullying CSR campaigns. Hypocrisy induction is a method for making people aware that their past actions have violated their moral standards, causing psychological discomfort and eliciting intrinsic and autonomous motivations to correct past wrongdoings through future behavior (Aronson et al., 1991). Hypocrisy induction is deeper, longer lasting, and less likely to backfire than traditional persuasion techniques (Aronson, 1999; Fointiat et al., 2008; Stone et al., 1994). Our first objective is to test whether hypocrisy induction produces voluntary support for a brand-sponsored anti-cyberbullying CSR campaign.

The second objective of this paper is to further elaborate hypocrisy induction effects by identifying and testing potential moderators. Hypocrisy induction is known to encourage some wrongdoers to rectify bad behavior, but it also risks evoking avoidance strategies (Effron, 2014; Fointiat et al., 2008). Most hypocrisy induction research, however, has focused on approach and rectification and has identified moderators that determine curative actions only (Gamma et al., 2020). Consequently, in 30 years of hypocrisy induction studies, only a few have explored why it sometimes fails (Dickerson et al., 1992; Rubens et al., 2015; Stone & Fernandez, 2008) and even fewer have investigated rectification versus disengagement. By further elaborating hypocrisy induction effects in the cyberbullying context, we address that gap and provide practical and strategic benefits for creators of anti-cyberbullying campaigns.

Informed by the moral regulation theory (Janoff-Bulman et al., 2009; Sheikh & Janoff-Bulman, 2010), we suggest that regulatory focus explains varying responses to hypocrisy induction. Predominantly promotion-focused individuals are motivated by hopes, dreams, goals, and self-improvement; predominantly prevention-focused individuals are motivated by duties, obligations, safety, and avoidance of negative consequences (Aaker & Lee, 2001; Avnet & Higgins, 2006; Higgins et al., 1994). This paper argues that when hypocrisy induction causes discomfort, promotion-focused individuals are likely to try to relieve it with approach behavior, while prevention-focused individuals will choose avoidance and disengagement.

Another objective is to elucidate the mechanisms underlying moderating effects of regulatory focus. Researchers have generally overlooked types of emotional experiences (Priolo et al., 2019) although emotion, rather than deliberate reasoning, is the basis for moral judgment (Greene & Haidt, 2002; Teper et al., 2015). We suggest that guilt and shame are distinct emotions that explain the moderating effect of regulatory focus. Guilt has been consistently shown to motivate remedial, corrective actions such as apologies, confessions, and prosocial behaviors, while shame has been shown to motivate hiding actions such as denial, withdrawal, and escape (Baek & Yoon, 2017; Pounders et al., 2018; Tangney et al., 2007). Moral regulation studies have associated promotion (prevention) focus with guilt (shame), desires to achieve positive outcomes (avoid negative outcomes), a focus on what should (not) be done, and behavioral activation (inhibition) (Janoff-Bulman et al., 2009; Sheikh & Janoff-Bulman, 2010). Thus, we propose that guilt (shame) explains why promotion (prevention)-focused individuals follow approach (avoidance) strategies when hypocrisy induction makes them realize that they have acted inconsistently with behavior they advocate.

In this article, we report the results of two studies showing that regulatory focus and emotions determine when and why hypocrisy induction will be effective in anti-cyberbullying CSR campaigns.

Theoretical Background

Cyberbullying Intervention and Prevention Strategies

Researchers have taken three broad approaches to examine cyberbullying. One shows outcomes associated with cyberbullying; that is, pervasive negative effects on victims’ well-being, risky behavior, and mental health (reviews from Kowalski et al., 2014; Tokunaga, 2010). Another shows determinants of cyberbullying such as moral beliefs and values (Lazuras et al., 2013), gender & age (Tokunaga, 2010), sociodemographics (Vandebosch & Van Cleemput, 2009), sensation seeking (Graf et al., 2019), peer group contexts (Shim & Shin, 2016), parental practices (Gómez-Ortiz et al., 2018), bystander responses (Wang, 2021), and cross-cultural factors (Barlett et al., 2014). A third branch of cyberbullying research develops and tests intervention and prevention strategies comprised of technical, legal, and educational categories. In the following sections, we systematically review three cyberbullying interventions used by governments, organizations, and brands.

Interventions often use technical approaches. Information and communication technology (ICT) companies such as YouTube and Facebook are acutely aware that they need to burnish their public images by showing that they oppose cyberbullying. They issue ethical declarations of commitment; use automated detection to scrutinize, filter, and block cyberbullying content; and allow users to report abuse or flag problematic content. Unfortunately, contents spread more rapidly than reports; filters can be tricked; and users may fail to recognize cyberbullying content (Blaya, 2019). User participation is essential, so technical approaches are obviously limited.

Another approach is through legal interventions. For example, Canada has mandated imprisonment and monetary penalties (Public Safety Canada, 2021). The U.S. government lacks direct federal laws but has established the StopBullying agency to enforce rules and consequences for cyberbullying in schools (Stop-Bullying, 2021). Texas passed “David’s Law” in 2017 (Texas State, 2020) enabling companies to unmask and charge cyberbullies with misdemeanors. Papa Johns and 3 M brands have anti-cyberbullying campaigns supporting legal remedies (David Legacy Foundation, 2021). Nevertheless, legislators need consistent and clear definitions, but cyberbullying definitions are inevitably untimely because they constantly evolve as technology advances. Broad definitions provide inadequate legal guidance for enforcers and risk possibilities that simple misbehavior will be overly prosecuted (Yang & Grinshteyn, 2016). Moreover, people must trust in cyberbullying laws, but civil society, especially youths, tend to doubt that the legal system can properly address online defamation and hate crimes (Wigerfelt et al., 2015). Cyberbullies know that their comments are anonymous and that they are likely to escape direct feedback and responsibility (Slonje & Smith, 2008). Consequently, they tend to lack empathy or fear of legal consequences (El Asam & Samara, 2016). Heavy-handed deterrents generate immediate but short-lived effects and can cause backlashes such as psychological reactance and denial (Fointiat et al., 2008; Stone et al., 1994). A strong political impediment is that legislation limits free speech (Chung et al., 2020).

Governments, organizations, and brands can use a longer-term, softer approach by providing empathy training that teaches people to eschew haters and harmful online behaviors (Blaya, 2019) or providing educational collaborative programs for adolescent participants (Ashktorab & Vitak, 2016). For instance, Intel, Twitter, and Brita brands partnered with the Cybersmile Foundation (2021) to support “Stop Cyberbullying Day” in which celebrities increase awareness and teach skills for counteracting cyberbullying. Education-based interventions are well intentioned but we lack empirical evidence of their effectiveness (Blaya, 2019). For instance, BullyDown, a 7.5-week intervention program using text messages to discourage cyberbullying, failed to evaluate behavioral impacts (Ybarra et al., 2016). Similarly, Samsung’s (2020) “Blue Elephant” youth cyberbullying prevention education project failed to provide data evaluating success in its broad goals to reach three million students and decrease cyberbullying from the current 30 percent rate to 3 percent (Bharti, 2021). That is, most education programs focus mainly on building and providing educational resources without developing evaluation tools. Their long-term perspectives also suggest the lack of immediate effects. Without rigorous evaluation and with potential time lags, we cannot know whether brands can use education programs to change attitudes and behaviors.

In summary, our review of current interventions revealed that cyberbullying is a new line of research. Ortega-Ruiz et al. (2012) were first to provide empirical evidence for intervention programs, but few have examined how brands can prevent cyberbullying. Discussions of hard and soft approaches have failed to fully consider anonymity or disinhibition in cyberbullying (Lowry et al., 2016) and lack reliable evaluation tools for explaining causal mechanisms (Blaya, 2019; Zych et al., 2015). By telling consumers how to behave, legal and education approaches are short-lived “top-down” methods susceptible to backfire (Barlett et al., 2017). Consequently, meta-analyses indicate a dearth of anti-cyberbullying programs, a lack of clarity regarding their effectiveness, and an imperative need for rigorous research that develops and tests strategies for motivating support (Blaya, 2019; Gaffney et al., 2019; Zych et al., 2015).

When consumers encounter traditional, direct persuasion techniques, they are aware of being manipulated and resist them. Instead, we argue that anti-cyberbullying campaigns need approach in which consumers actively persuade themselves to change their attitudes or behaviors. Self-persuasion techniques are indirect, stronger, deeper, and longer lasting because they convince people that they are internally motivated to change their attitudes or behaviors (Aronson, 1999). Effects of self-persuasion have been documented in marketing (Bernritter et al., 2017), education (Canning & Harackiewicz, 2015), health communication (Baldwin et al., 2021; Loman et al., 2018), green behavior (Odou et al., 2019), interpersonal relationships (Wong et al., 2022), and computer science domains (Krsek et al., 2022). As a type of self-persuasion, we suggest that hypocrisy induction evokes internal motivations to change behavior specifically in anti-cyberbullying CSR campaigns.

Hypocrisy Induction

Based on arguments that cognitive dissonance strongly motivates attitudinal and behavioral change, Aronson et al. (1991) first introduced the hypocrisy induction paradigm as a method for encouraging college students to use condoms. Hypocrisy induction is a sequential procedure for producing cognitive dissonance. Study participants are first asked to advocate socially desirable behaviors. Then they are reminded that they have acted in ways that transgressed against those normative standards. In recognizing their hypocrisy, participants feel discomforting cognitive dissonance. They desire to resolve it by aligning future behavior with their advocacy (Stone & Fernandez, 2008). Researchers use subtle ways to induce hypocrisy and promote prosocial resolutions, such as using video cameras or voice recorders to document the signing of pledges or petitions or asking participants to write and then read normatively muddled messages (Dickerson et al., 1992; Fointiat et al., 2008; Priolo et al., 2016). A meta-analysis of 38 hypocrisy induction studies revealed an overall 95% confidence interval for Cohen’s d, from 0.33 to 0.46 (Priolo et al., 2019). Furthermore, when Morongiello and Mark (2008) replicated Aronson et al.’s (1991) first hypocrisy induction study, 222 participants in many labs demonstrated reliable effects, with a 95% confidence interval for d from 0.59 to 0.74. Effects were also demonstrated in a study of 1377 consumers in naturally occurring settings (Gamma et al., 2020) regarding health (Peterson et al., 2008), civic behavior (Aronson et al., 1991; Stone et al., 1994), environmental protection (Dickerson et al., 1992; Gamma et al., 2020; Kim et al., 2022a, 2022b; Priolo et al., 2016), shopping behavior (Rubens et al., 2015), and compliance with COVID-19 measures (Kim & Ryoo, 2022).

Despite promising findings, two gaps in the literature prevent conclusions that hypocrisy induction will always benefit prosocial campaigns. First, individuals have different strategies for dealing with dissonance. If they attribute moral failure to internal weakness,Footnote 1 they may try approach strategies for rectifying actions or avoidance strategies for detaching from the situation (Back & Yoon, 2017; Janoff-Bulman et al., 2009; Sheikh & Janoff-Bulman, 2010). Moral regulation research has documented both reactions, but the hypocrisy induction literature has failed to consider alternative avoidance/behavioral inhibition for resolving dissonance. Some studies found similar reactions in both hypocrisy and control conditions (Dickerson et al., 1992; Rubens et al., 2015; Stone & Fernandez, 2008) but assumed that situations determine effects, without further theoretical explanations. Thus, hypocrisy research may draw inconsistent results because it lacks a cohesive theoretical approach for unifying self-regulation strategies.

Second, the literature lacks strong evidence for multiple claims that hypocrisy induction causes cognitive dissonance, manifested as psychological discomfort, that strongly drives prosocial behaviors (Kenworthy et al., 2011). The lack of studies, heterogeneous emotion measures, and mixed findings indicate a need for clarified assumptions and substantial examinations of how people reduce dissonance. For instance, a meta-analysis examined 27 studies indicating that participants in both hypocrisy and control conditions reported similar levels of psychological discomfort, but only 7 measured psychological discomfort (Priolo et al., 2019). Varying operationalizations of psychological discomfort could have added error. For instance, some studies used scales combining questions related to psychological discomfort, irritability, and distress (Yousaf & Gobet, 2013), while others used Elliot and Devine’s (1994) discomfort subscale (Priolo et al., 2016). Also, conflicting findings based on approach and avoidance may reduce the overall power of the positive effect of hypocrisy induction, so that mediation becomes insignificant. In the following sections, we suggest that regulatory focus is a moderator that determines moral regulation strategies and that discrete emotions of guilt and shame further explain campaign support.

Regulatory Focus as Moderator

Regulatory focus is the basic motivational orientation that affects how individuals pursue their goals (Chernev, 2004; Higgins, 2002; Higgins & Pinelli, 2020). Promotion focus is associated with the pursuit of an ideal self through success, accomplishment, and progress; prevention focus is associated with the avoidance of risks, losses, and regression through protection, prevention, safety, and maintenance of the status quo (Higgins, 1997; Sung & Choi, 2011).

How does regulatory focus affect moral self-regulation strategies after individuals undergo hypocrisy induction in which they become aware of their moral discrepancy, experience dissonance, and try to reduce or inhibit inner conflict (Cornwell & Higgins, 2016)? We argue that regulatory focus determines moral regulation tactics, based on theorizations that promotion (prevention) focused reactions are associated with prescriptive (proscriptive) moral regulations (Janoff-Bulman et al., 2009; Sheikh & Janoff-Bulman, 2010). When promotion-focused people are presented with positive moral referents, their sensitivity to advancement (Förster et al., 1998; Higgins et al., 1994; Scholer & Higgins, 2008) will make them view a prosocial campaign as an opportunity to align behaviors with accepted norms. In contrast, prevention-focused individuals will adhere to their efforts to minimize negative consequences, maintain the status quo, avoid threatening situations, and retreat from negative feelings (Förster et al., 1998; Higgins et al., 1994; Scholer & Higgins, 2008).

To understand how regulatory focus determines moral regulation tactics, imagine a scenario in which bystanders observe cyberbullying on SNSs. Both fail to intervene and suffer later regret. When presented with a subsequent anti-cyberbullying campaign, promotion-focused bystanders are likely to want to make reparations and restore moral worth, while prevention-focused bystanders will want to avoid the negative emotional end-state and will simply detach from the offer. The likelihood of supporting anti-cyberbullying campaigns is simply a consequence of regulatory focus. Our theorization assumes that most consumers desire to maintain positive moral self-views (Mazar et al., 2008; Ryoo, 2022; Ryoo et al., 2020; Schwabe et al., 2018), although consumer behavior often fails to uphold that belief.

Evidence supports our predictions. For instance, in conjunction with regulatory focus, Schwabe et al. (2018) tested whether initial immoral behaviors lead to subsequent moral or immoral behaviors. They found that strongly promotion-focused consumers acted more morally after they made immoral decisions, because they were more likely to deviate from the status quo, were more open to change, and were more likely to seek risk (Chernev, 2004; Liberman et al., 1999; Zhou & Pham, 2004). However, strongly prevention-focused consumers adhered to initial immoral behavior because they placed such value on avoiding negative consequences they may confront by deviating from the status quo (Chernev, 2004). Being inclined toward stability and conservative strategies (Zhou & Pham, 2004), they even continue to act immorally. Other studies have documented tendencies to avoid correction and maintain the status quo (Zhang et al., 2014). In studying relationship conflicts, Rodrigues et al. (2019) found that promotion-focused individuals seek constructive resolution strategies such as accommodation, negotiation, and blame avoidance, while prevention-focused individuals maintain the status quo by distancing during conflict and refraining from discussions.

Combining our arguments, we hypothesize:

H1

Hypocrisy induction (vs. control) will generate favorable responses toward a brand-sponsored anti-cyberbullying campaign.

H2

Regulatory focus will moderate hypocrisy induction effects. Under a dominant promotion mindset, participants in the hypocrisy induction (vs. control) condition will express more favorable responses toward the campaign. However, under a dominant prevention mindset, hypocrisy induction will fail to change responses toward the campaign.

Guilt and Shame as Mediators

When people realize that they have transgressed against or deviated from internal moral values (Hosser et al., 2008; Kenworthy et al., 2011), they form negative self-evaluations of guilt or shame (Baek & Yoon, 2017; Kim et al., 2022a, 2022b; Pounders et al., 2018; Sheikh & Janoff-Bulman, 2010; Tangney et al., 2007). Both emotions are highly correlated but conceptually distinct (Tangney et al., 2007). Guilt forces transgressors to focus on what they did, evoking tenseness, remorse, regret, and responsibility for violating moral values; in contrast, shame forces transgressors to focus on who they are, evoking a flawed, worthless, powerless, inadequate, incompetent, and unworthy self-image (Lewis, 1971; Tangney et al., 2007). Both involve internal attributions, but guilt draws self-blame for a specific act, while shame casts aspersions on overall character (Baek & Yoon, 2017; Pounders et al., 2018; Tangney et al., 2007).

Considerable evidence indicates that emotions have a self-regulatory function and that guilt and shame are particularly central in moral regulation (Greene & Haidt, 2002; Onweze et al., 2014; Teper et al., 2015; Tian & Li, 2021). How might guilt and shame relate to hypocrisy induction reactions among individuals guided by approach or avoidance moral regulation strategies? Based on arguments that guilt (shame) is distinctly associated with prescriptive (proscriptive) moral regulation (Schmader & Lickel, 2006; Sheikh & Janoff-Bulman, 2010), we propose that predominantly promotion (prevention)-focused individuals will react to hypocrisy induction with guilt (shame) and will rectify bad behavior by supporting an anti-cyberbullying campaign (by detaching from the situation).

Guilt occurs when people primarily focus on past moral failure and interpret transgressions according to prescriptive, positive end-states, and moral shoulds (Sheikh & Janoff-Bulman, 2010). They are then motivated to take controllable, malleable, reparable actions (Baek & Yoon, 2017), such as making amends, offering reparations, confessing and apologizing, and improving social relationships through active social engagement (Lim & Hong, 2022; Pounders et al., 2018; Tangney et al., 2007). In contrast, shame occurs when people interpret transgressions according to proscriptive, negative end-sates and moral should nots (Sheikh & Janoff-Bulman, 2010). They are then motivated to withdraw, isolate, hide, and escape (Lewis et al., 1992; Pounders et al., 2018; Sheikh & Janoff-Bulman, 2010; Tangney et al., 2007) because they perceive past immoral actions as evidence that they are in uncontrollable, persistent, unredeemable, personal states (Baek & Yoon, 2017; Tracy & Robins, 2006).

In sum, prescriptive (proscriptive) morality evokes guilt (shame) and corrective actions (efforts to escape negative implications about personal character). Considering that regulatory focus determines moral regulation, we can assume that guilt (shame) explains why hypocrisy induction drives approach-activation (avoidance-inhibition) tendencies. Studies of regulatory focus and guilt/shame support our hypothesis. For example, studies showed that promotion-focused consumers felt guilt and adopted prosocial behaviors in contexts including safe driving (Pounders et al., 2018), sharing beneficial knowledge (Fang, 2017), helping strangers in need (Sheikh & Janoff-Bulman, 2010), and conserving water (Baek & Yoon, 2017). Others showed that prevention-focused consumers felt shame and were reluctant to help strangers (Sheikh & Janoff-Bulman, 2010), to share helpful knowledge (Fang, 2017), or to take curative actions (Lickel et al., 2005). Shame was also related to low self-efficacy (Skinner & Brewer, 2002). Therefore, we hypothesize:

H3

When promotion (prevention) is dominant, participants in the hypocrisy induction (vs. control) condition will indicate higher guilt (shame).

In H2, hypocrisy induction is predicted to positively affects participation in an anti-cyberbullying campaign only in the promotion condition, not the prevention condition. This suggests that only guilt, not shame, has a significant indirect effect on willingness to engage in the campaign. Thus, we hypothesize

H4

Guilt will mediate the positive effect of hypocrisy induction among participants in the promotion condition.

Figure 1 shows our conceptual model. Before collecting data, we obtained ethical approval from the University’s Institutional Review Board. Before participating, all participants provided informed consent and were free to withdraw at any time.

Fig. 1
figure 1

Conceptual design and hypotheses

Full size image

Study 1

Study 1 was intended to provide initial evidence that hypocrisy induction prompts favorable responses toward a brand-sponsored cyberbullying prevention campaign (H1), and that regulatory focus moderates the hypocrisy effect (H2). Study participants viewed a fictitious ad in which a brand shows corporate social responsibility (CSR) by encouraging online users to participate in an anti-cyberbullying campaign.

Sample and Research Design

A 2 (hypocrisy induction: presence vs. absence) × 2 (regulatory focus: promotion vs. prevention) between-subject design was used. We recruited 176 participants (Mage = 38.23 years, 82 women) from Amazon’s MTurk, an appropriate source because cyberbullying occurs in online contexts.

Stimuli Development and Manipulation

Hypocrisy Induction

To induce hypocrisy, researchers have traditionally used paper–pencil tasks that (1) reinforce participants’ current beliefs about issues, (2) make them form commitments regarding the issues, and (3) cause them to realize their hypocrisy in past behavior that was antithetical to their beliefs and commitments (Aronson et al., 1991; Dickerson et al., 1992). For the experiment, we developed a 3-page anti-cyberbullying CSR advertising campaign ostensibly sponsored by a fictitious brand. Each page was an interactive advertising format corresponding to a phase of hypocrisy induction. The first page featured a statement intended to intensify disdain toward cyberbullying: “Cyberbullying leaves deep emotional scars on victims.” The second page corresponded with the second commitment phase with the message: “Please leave some kind words for victims of cyberbullying.” To enhance realism, participants were informed that their supportive messages would be used in a video clip designed to console victims. The third page, corresponding with the third phase, featured questions designed to remind participants that they had engaged in past cyberbullying, contrary to the values stated in the first two phases: (1) “Have you ever tried to help someone who was being bullied online?” (2) “Have you ever paid attention to victims of cyberbullying?” (3) “Have you ever donated to cyberbullying victims?” and (4) “Have you ever stood by a victim of online bullying?” To answer the questions, participants clicked on checkboxes: 1 = yes, 2 = no, or 3 = do not want to report. We expected that hypocrisy would be triggered when participants recollected past transgressions that were inconsistent with moral stands reported in the first two steps. Hypocrisy would not be induced among participants who lacked a cyberbullying history, so those who answered yes to all four questions were directed to the end page and eliminated from the data set. Participants in the non-hypocrisy condition did not view the third page.

Regulatory Focus

After participants finished the three phases of hypocrisy induction, they viewed ads for an anti-cyberbullying CSR campaign framed to manipulate a promotion or prevention regulatory focus. Consistent with prior research (Kim and Ryoo, 2022; Sung & Choi, 2011), the promotion-focused ad message used the slogan, “Free victims from cyberbullying,” followed by promotive messages “Let’s move toward cyber-utopia. Pay attention to their suffering.” The prevention-focused ad message used the slogan, “Prevent victims from being cyberbullied,” followed by preventive messages, “Let’s move away from cyber-dystopia. Do not ignore their suffering.” Other than regulatory focus messages, the two advertisements were the same.

Procedure and Measures

Participants were randomly assigned to one of four conditions. At their own pace, they viewed an anti-cyberbullying CSR campaign. Then they answered three items as a manipulation check for hypocrisy induction: (1) “My behavior was inconsistent with what I should do,” (2) “My behavior contradicted my ethical beliefs,” and (3) “My behavior might seem hypocritical” (α = 0.78; Stone & Fernandez, 2008). The hypocrisy induction was measured on a five-point Likert scale (1 = strongly disagree, 5 = strongly agree). The manipulation check for regulatory focus included: (1) “I think this ad would focus on achieving good outcomes,” (2) “I think this ad would generate positive perceptions,” (3) “I think this ad would encourage efforts to achieve positive outcomes,” (4) “I think this ad would focus on preventing bad outcomes,” (5) “I think this ad would generate negative perceptions,” and (6) “I think this ad would encourage efforts to prevent negative outcomes” (Pham & Avnet, 2004), answered on a seven-point Likert-type scale (1 = strongly disagree, 7 = strongly agree). The six items were categorized into two dimensions (variance = 75%); the first three reflect promotion foci (α = 0.85); the last three reflect prevention foci (α = 0.81).

Willingness to support the campaign (WTS) and willingness to donate (WTD) to cyberbullying victims were then measured as dependent variables. Specifically, to measure WTS, participants responded to (1) “How likely are you to support this campaign?” (2) “How inclined are you to support this campaign?” and (3) “How willing are you to support this campaign?” (α = 0.91; White & Peloza, 2009). To measure WTD, they responded to “The likelihood that I would donate to this victim is…” (1) very unlikely/very likely (2) impossible/possible and (3) improbable/probable (α = 0.89; Baek & Yoon, 2017). WTS was measured on a seven-point Likert-type scale (1 = strongly disagree, 7 = strongly agree); WTD was measured on a seven semantic differential scale. The correlation between the two dependent variables was 0.65 (p < 0.001). After completing demographic measures, participants were debriefed and thanked.

In this and the following studies, we carefully ensured data quality by blocking fraudulent respondents in Qualtrics surveys (Kennedy et al., 2020). Before the main experiment, we administrated attention checks as recommended for ensuring good data. Specifically, participants answered two questions related to a short story we provided: (1) “What is the dog’s name?” (2) “Why did the family love the dog?” Participants who failed to answer one or more questions correctly were directed to the end page and excluded from the data.

Results

Manipulation Check

Regarding the manipulation checks of hypocrisy induction, participants in the hypocrisy condition reported higher levels of hypocrisy perception (M = 3.12, SD = 1.44) than those in the non-hypocrisy condition (M = 2.28, SD = 1.43, t(174) = 3.85, p < 0.001). A repeated-measures analysis of variance (ANOVA) revealed a significant main effect for regulatory focus conditions (F(1, 174) = 10.99, p = 0.001), but not for regulatory focus scales (F(1, 174) = 0.86, p > 0.1). We found a significant interaction (F(1, 174) = 121.48, p < 0.001). Specifically, participants thought that the promotion-focused ad highlighted the pursuit of intended goals (M = 5.72, SD = 1.4) rather than the avoidance of failure to achieve intended goals (M = 4.08, SD = 1.68, t(87) = 7.88, p < 0.001). However, they perceived the prevention-focused ad to be related more to the avoidance of failure (M = 5.23, SD = 1.62) rather than the pursuit of goals (M = 3.3, SD = 1.93, t(87) = − 7.78, p < 0.001). Thus, the manipulations of hypocrisy and regulatory focus were valid and effective.

Hypothesis Testing

To test our hypotheses, we ran a multivariate analysis of variance (MANOVA) for WTS and WTD. The main effect of hypocrisy induction (WTS: F(1, 172) = 8.54, p < 0.001; WTD: F(1, 172) = 6.9, p < 0.01) showed that participants who received the hypocrisy induction treatment showed higher WTS (MHypo = 4.93 vs. MNon-hypo = 4.3) and WTD (MHypo = 5.24 vs. MNon-hypo = 4.66) than participants excluded from the treatment, supporting H1. The main effect of regulatory focus was also significant (WTS: F(1, 172) = 8.98, p < 0.001; WTD: F(1, 172) = 7.37, p < 0.01); that is, the promotion-focused ad produced higher WTS (Mpromotion = 4.91 vs. Mprevention = 4.3) and WTD (Mpromotion = 5.22 vs. Mprevention = 4.65) than the prevention-focused ad. As expected, significant hypocrisy induction × regulatory focus interactions emerged for WTS (F(1, 172) = 11.97, p = 0.001) and WTD (F(1, 172) = 18.77, p < 0.001). Planned contrasts (Fig. 2) further showed that hypocrisy induction increased WTS (MHypo = 5.67 vs. MNon-hypo = 4.26; t(86) = 5.34, p < 0.001) and WTD (MHypo = 6.08 vs. MNon-hypo = 4.48; t(86) = 5.57, p < 0.001) when the campaign was framed with a promotion focus. However, hypocrisy induction failed to generate differential effects on WTS (MHypo = 4.24 vs. MNon-hypo = 4.36; p > 0.1) and WTD (MHypo = 4.45 vs. MNon-hypo = 4.85; p > 0.1) when the campaign was framed with a prevention focus. Figure 3 illustrates these interaction effects predicting the relationship between hypocrisy induction and WTS and WTD for two different regulatory focus types by utilizing a data visualization application called InterActive (McCabe et al., 2018) for better data transparency and nuance understanding. As expected, hypocrisy induction (vs. absence) is positively associated with WTS and WTD for those in the promotion focus (WTS: b = 0.71, 95% CI = [0.4 to 1.02]; WTD: b = 0.8, 95% CI = [0.48 to 1.13]), but not for those in the prevention focus (WTS: b = − 0.06, 95% CI = [− 0.37 to 0.25]; WTD: b = − 0.2, 95% CI = [− 0.52 to 0.12]). Thus, H2 was supported.

Fig. 2
figure 2

Interaction between hypocrisy induction and regulatory focus on WTS and WTD (Study 1). Note: *** means the difference is significant at the .001 level

Full size image
Fig. 3
figure 3

Depictions of the interaction effect between hypocrisy induction and regulatory focus on WTS and WTD (Study 1). Note: Hypocrisy induction is coded as 1 (presence) or -1 (absence). Each graph shows the computed 95% confidence region (shared area), the full range of observed data (gray circles), and the threshold at which the association between hypocrisy induction and WTS and WTD changes as a function of regulatory focus (diamond)

Full size image

Discussion in Brief

Study 1 examines whether hypocrisy induction encourages ethical behavior among online users. We develop an interactive ad to deliver a virtual anti-cyberbullying campaign and find that hypocrisy induction is persuasive only when campaigns are framed with a promotion rather than a prevention focus. The results suggest that when consumers are made aware that they have misbehaved and are then exposed to a promotion-focused ad, they respond favorably to an offer for a socially responsible activity.

Study 2

Our goal in Study 2 was to examine guilt and shame as underlying mechanisms (H3 and H4) and to replicate Study 1 using a different regulatory focus manipulation. Message framing or separate tasks can be used to induce regulatory focus (Higgins et al., 1994; Kees et al., 2010). In Study 1, we provided practical implications for advertisers by using message framing in an advertising campaign, but the focus on advertising can blur whether regulatory focus really evokes discrete emotions and discomfort that then prompts behavior consistent with the regulatory focus. Also, despite the rigorous message manipulation, we used two advertisements. To examine whether results would remain consistent if participants viewed the same advertisement, in Study 2, we manipulated regulatory focus through a separate writing task irrelevant to the campaign (Pham & Avnet, 2004; Semin et al., 2005). If regulatory focus drives moral regulation strategies following hypocrisy induction, the same advertisement should yield varying reactions. If regulatory focus continues to hold its moderating effect across manipulations, it should strongly indicate robust conceptualization. To ensure rigor, we measured and controlled for confounding variables such as prior experience of being bullied, involvement in ethical issues, and affect intensity.

Sample and Research Design

A 2 (hypocrisy induction: presence vs. absence) × 2 (regulatory focus: promotion vs. prevention) between-subject design was used. We recruited 293 participants (Mage = 44.25 years, 138 women) from Amazon’s MTurk.

Procedure

The procedure and measures were the same as in Study 1, with some exceptions. We manipulated regulatory focus by assigning writing tasks, which are considered reliable methods for manipulation in studies of consumer choice and cause-related marketing (Pham & Avnet, 2004; Semin et al., 2005). First, rather than view an advertisement to manipulate regulatory focus as in Study 1, participants performed a priming task as an ostensibly separate study. Participants in the promotion focus condition responded to the prompt: “Imagine that you are the kind of person who tries to be a good friend in your close relationships. What strategy would you use to meet your goal?” Participants in the prevention focus condition responded to the prompt, “Imagine that you are the kind of person who tries to avoid being a bad friend in your close relationships. What strategy would you use to meet your goal?” (Semin et al., 2005). Recall that advancement, success, and accomplishment are most salient for a promotion focus, while protection, prevention, and safety are most salient for a prevention focus (Sung & Choi, 2011).

After participants completed the priming task, they answered items designed to check the regulatory focus manipulation: (1) “I care about positive achievements,” (2) “I tend to focus on the bright side,” (3) “I care about positive outcomes,” (4) “I care about preventing bad occurrences,” (5) “I am primarily concerned about avoiding the dark side,” and (6) “I believe it is important to prevent negative outcomes” (Pham & Avnet, 2004) on a seven-point Likert-type scale (1 = strongly disagree, 7 = strongly agree). The six items were categorized into two dimensions (variance = 73%); the first three reflect a promotion focus (α = 0.84); the last three reflect a prevention focus (α = 0.79).

After participants viewed the anti-cyberbullying campaign at a self-paced rate, they reported guilt with four items: (1) “I feel guilty about cyberbullying” (2) “I feel responsible for cyberbullying” (3) “I feel accountable for cyberbullying” and (4) “I feel guilty for victims of cyberbullying” (α = 0.89; Coulter & Pinto, 1995). Shame was measured with three items: (1) “I feel ashamed for victims of cyberbullying” (2) “I feel embarrassed about cyberbullying issues” and (3) “I feel humiliated about failing to help victims of cyberbullying” (α = 0.82; Coulter & Pinto, 1995). Participants then answered questions concerning willingness to donate to victims (WTD), using the items from Study 1.

Theories and empirical evidence suggest that individual traits may affect outcomes in ethical contexts (Chen & Moosmayer, 2020). We measured and controlled for ethical sensitivity (three items; e.g., “I notice ethical issues around me”; Chen & Moosmayer, 2020), dispositional affect intensity (six items; e.g., “When I do wrong, I feel strong shame and guilt”; Geuens & Pelsmacker, 2002), and prior victimization experiences (two items; e.g., “I’ve experienced some form of cyberbullying”; Kuusisto et al., 2012) as potential covariates measured on a seven-point Likert-type scale (1 = strongly disagree, 7 = strongly agree). The scales are well-established, widely used in ethical contexts, and appropriate for our research.

Results

Manipulation Check

Manipulation check results showed that participants in the hypocrisy induction condition (M = 3.08, SD = 1.39) reported higher hypocrisy scores than those in the non-hypocrisy induction condition (M = 2.01, SD = 1.29, t(291) = 6.86, p < 0.001). Also, a repeated-measures analysis of variance (ANOVA) revealed a significant main effect for regulatory focus conditions (F(1, 291) = 7.96, p < 0.01), but not for regulatory focus scales (F(1, 291) = 0.43, p > 0.1). A significant interaction between regulatory focus conditions and scales (F(1, 291) = 61.63, p < 0.001) indicated that participants in the promotion focus condition reported higher promotion focus scores (M = 5.08, SD = 1.57) and lower prevention focus scores (M = 4.07, SD = 1.45, t(152) = 5.93). Conversely, participants in the prevention focus condition reported higher prevention focus scores (M = 5.36, SD = 1.14) and lower promotion focus scores (M = 4.5, SD = 1.72, t(139) = -5.2, p < 0.001). Thus, manipulations of hypocrisy induction and regulatory focus were effective and successful.

Willingness to Donate to Victims of Cyberbullying (WTD)

An ANOVA was performed to test hypotheses. After controlling for three covariates (ps > 0.1), both hypocrisy induction (MHypo = 4.9 vs. MNon-hypo = 4.29; F(1, 286) = 8.64, p < 0.01) and regulatory focus (Mpromotion = 4.92 vs. Mprevention = 4.2; F(1, 286) = 13.95, p < 0.001) had a significant main effect on WTD, again supporting H1. As predicted, hypocrisy induction and regulatory focus had significant two-way interactions for WTD (F(1, 286) = 12.3, p < 0.001). Planned contrasts further showed that participants in the promotion focus condition who received the hypocrisy induction treatment exhibited higher WTD (MHypo = 5.54 vs. MNon-hypo = 4.32; F(1, 148) = 22.36, p < 0.001). However, among participants in the prevention focus condition, hypocrisy induction did not differentiate WTD (MHypo = 4.14 vs. MNon-hypo = 4.26; F(1, 135) = 0.2, p > 0.1), supporting H2.

Guilt and Shame

We performed a MANOVA to better understand how hypocrisy induction interacts with regulatory focus to arouse guilt and shame. After controlling for covariates,Footnote 2 both hypocrisy induction (MHypo = 4.65 vs. MNon-hypo = 3.11; F(1, 286) = 108.12, p < 0.001) and regulatory focus (Mpromotion = 3.6 vs. Mprevention = 4.12; F(1, 286) = 15.97, p < 0.001) had significant main effects on guilt. Hypocrisy induction and regulatory focus had a significant two-way interaction in arousing guilt (F(1, 286) = 8.4, p < 0.01). Planned contrasts (Fig. 4) further indicated that those in the hypocrisy induction condition experienced more guilt than those in the non-hypocrisy condition in both types of regulatory focus (promotion: MHypo = 4.18 vs. MNon-hypo = 3.03; F(1, 148) = 25.24, p < 0.001; prevention: MHypo = 5.21 vs. MNon-hypo = 3.2; F(1, 135) = 98.27, p < 0.001). That is, the interactive pattern was found for guilt when regulatory focus was compared within the hypocrisy induction or non-hypocrisy induction condition. Specifically, participants in the hypocrisy induction condition and the prevention condition (M = 5.21) experienced higher guilt in comparison with those in the promotion condition (M = 4.18, F(1, 135) = 19.22, p < 0.001). However, within the non-hypocrisy condition, regulatory focus failed to affect guilt levels (Mpromotion = 3.03 vs. Mprevention = 3.2; F(1, 148) = 0.69, p > 0.1). Replicating these results, the visual depiction of the interaction (Fig. 5) shows that when prevention focus is dominant, the positive slope is stronger (b = 1, 95% CI = [0.79 to 1.21]) than that when promotion focus is dominant (b = 0.56, 95% CI = [0.35 to 0.77]).

Fig. 4
figure 4

Interaction between hypocrisy induction and regulatory focus on guilt and shame (Study 2). Note: *** means the difference is significant at the .001 level

Full size image
Fig. 5
figure 5

Depictions of the interaction effect between hypocrisy induction and regulatory focus on guilt and shame (Study 2). Note: Hypocrisy induction is coded as 1 (presence) or -1 (absence). Each graph shows the computed 95% confidence region (shared area), the full range of observed data (gray circles), and the threshold at which the association between hypocrisy induction and guilt and shame changes as a function of regulatory focus (diamond)

Full size image

After controlling for covariates (ps > 0.1), only hypocrisy induction (MHypo = 3.55 vs. MNon-hypo = 3.1; F(1, 286) = 5.92, p < 0.05) had a significant main effect on shame. Regulatory focus had a nonsignificant main effect on shame (Mpromotion = 3.21 vs. Mprevention = 3.43; F(1, 286) = 2.33, p > 0.1). Hypocrisy induction and regulatory focus interactively influenced shame (F(1, 286) = 11.62, p = 0.001). As Fig. 4 shows, participants in the prevention condition who received hypocrisy induction (M = 4.07) experienced more shame than those who did not receive hypocrisy induction (M = 2.89; F(1, 135) = 16.09, p < 0.001). However, hypocrisy induction did not change shame among those in the promotion condition (MHypo = 3.12 vs. MNon-hypo = 3.3; F(1, 148) = 0.59, p > 0.1). Figure 5 also shows that hypocrisy induction (vs. absence) is positively associated with shame for those in the prevention focus (b = 0.58, 95% CI = [0.3 to 0.87]), but not for those in the promotion focus (b = − 0.12, 95% CI = [− 0.4 to 0.17]).

In sum, the findings support H3: promotion (prevention) focus led to greater guilt (shame) among participants in the hypocrisy induction condition, compared with those in the non-hypocrisy induction condition.

The Moderated Mediation

To test whether guilt and shame mediate the interaction between hypocrisy induction and regulatory focus on WTD, we performed a moderated mediation analysis using a PROCESS (Hayes, 2017; Model 8, 5000 bootstraps; Fig. 6). Hypocrisy induction, regulatory focus, guilt, shame, and WTD were entered into the model as an independent variable, a moderator, first mediator, second mediator, and a dependent variable, respectively. The model also included ethical sensitivity, distortional affect intensity, and prior victimization experience as covariates, but they did not significantly influence the results. Thus, we discuss the covariates no further.

Fig. 6
figure 6

Moderated mediation analysis (Study 2). Note: * p < .05

Full size image

Regarding guilt and shame as mediators, the moderated mediation test revealed that only guilt had significant indirect effects (IEs) on the interaction of hypocrisy and regulatory focus on WTD; moderated mediation had a significant index level (Index = − 0.09, 95% CI = − 0.2 to − 0.01). Specifically, guilt was a significant mediator for those in the promotion (b = 0.11, 95% CI = 0.02 to 0.22) and those in the prevention (b = 0.19, 95% CI = 0.03 to 0.38). That is, higher levels of guilt mediated the positive effect of hypocrisy induction on WTD among those in the promotion condition (b = 0.5, 95% CI = 0.23 to 0.77), while extremely higher levels of guilt among those in the prevention condition resulted in the nonsignificant effect of hypocrisy induction on WTD (b = − 0.28, 95% CI = − 0.58 to 0.02). As expected, shame did not generate significant IEs on WTD due to the insignificant impact of shame on WTD; the index of moderated mediation failed to reach significance (Index = − 0.04, 95% CI = − 0.13 to 0.05), suggesting that higher levels of shame among those in the prevention focus condition prevented hypocrisy induction from having a positive effect. The moderated mediation supported H4.

Discussion in Brief

Study 2 again supports our prediction that hypocrisy induction encourages online users to support anti-cyberbullying campaigns (H1), but only for those who have a dominant promotion rather than prevention focus (H2). Guilt and shame are the underlying forces: hypocrisy induction evokes higher guilt (shame) for participants who are predominantly promotion (prevention) focused (H3). Guilt, but not shame, is associated with campaign support (H4): when participants with promotion mindsets confront moral failure, their guilt drives them toward joining a brand-sponsored anti-cyberbullying campaign. Conversely, when participants with prevention mindsets confront moral failure, they avoid further shame by ignoring the issue.

An interesting finding is that hypocrisy induction caused prevention-focused participants to feel the most guilt, and the extreme levels of guilt did not translate to higher campaign support. The finding aligns with research showing that moderate guilt motivates relatively simple and direct approach reparations (Pounders et al., 2018), but extreme guilt activates lower self-efficacy, neutralization, denial, and complex avoidance actions (Lickel et al., 2005; Skinner & Brewer, 2002). Guilt and shame are so interrelated that they may occur simultaneously. Extremely strong guilt can arouse anger, annoyance, irritation, or even shame (Coulter & Pinto, 1995; Pounders et al., 2018), suggesting that intolerable guilt evokes shame and avoidance. Supporting this, a supplementary analysis indicated that those with higher levels of guilt experienced higher shame (b = 0.47, SE = 0.06, t = 8.97, p = 0.000). Thus, Study 2 provides empirical evidence that regulatory focus causes varying levels of discrete guilt (shame) determining desires to make amends (hide).

Study 2 further strengthens the rigor of our research by applying a different method of regulatory focus manipulation and controlling for confounding variables. The study measures prior victimization experiences, ethical issue sensitivity, and affect intensity as potential confounding facts, but finds no impact on WTD. We speculate that, in a controlled experimental setting, the manipulated main independent variables had immediate strong impacts on the dependent variable, decreasing the power of possible confounding effects.

General Discussion

Theoretical Implications

Our studies extend the depth and breadth of current scholarship on cyberbullying, hypocrisy induction, regulatory focus, and guilt and shame effects. First, by showing that hypocrisy induction is an appropriate, effective, self-persuasive technique that generates voluntary and spontaneous support of anti-cyberbullying campaigns, we resolve difficulties associated with theories and studies of traditional offline bullying (Hinduja & Patchin, 2019; Slonje & Smith, 2008). Importantly, we go beyond simply applying theory to a new context and beyond current dissonance accounts. Hypocrisy induction evokes autonomous behavioral changes and attraction to alternatives for alleviating discomfort (Priolo et al., 2019), but over-zealousness may have inflated bias regarding its effects (Gamma et al., 2020; Priolo et al., 2019). We address a research void and provide a more comprehensive theoretical framework by taking the moral regulation perspective and testing regulatory focus as a moderator. We confirm that promotion-focused people are internally motivated to make amends and seek positive outcomes such as by supporting prosocial campaigns, making donations, or purchasing cause-related products. In contrast, prevention-focused people are motivated to avoid negative feelings and detach from making amends.

Second, by examining guilt and shame, our work deepens understanding of cyberbullying, moral regulation, and regulatory focus. Researchers concur that induced hypocrisy evokes dissonance and passingly acknowledge that the discomfort drives desires to change behavior (Priolo et al., 2019), without identifying the discrete emotions involved. The only exception is the work by Kim et al. (2021). However, their examination of guilt and shame with hypocrisy induction is not intended to test moral self-regulation theories; instead, emotional responses were simply the consequences of different advertising appeals (e.g., self-benefit vs. other benefit). Aligned with the moral emotion literature, we show that guilt (shame) is associated with approach motivations, prescriptive moral regulation, and desires to amend behavior (avoidance motivations, proscriptive moral regulation, and desires to escape) (Lewis et al., 1992; Pounders et al., 2018; Sheikh & Janoff-Bulman, 2010; Tangney et al., 2007). Hypocrisy research did not attempt to distinguish shame and guilt in accordance with general approach-avoidance regulatory systems, but we consolidate the research and present possibilities for future investigations.

Third, we contribute to regulatory focus theory studies that have extensively examined effects on goal compatibility (Aaker & Lee, 2001; Baek & Reid, 2013; Baek & Yoon, 2017; Kees et al., 2010; Kim, 2006; Sung & Choi, 2011). Regulatory fit theory explains that prosocial campaign ads should match self-regulatory goals (Aaker & Lee, 2006; Avnet & Higgins, 2006; Cesario et al., 2004). Rather than focusing on campaign support as the end-state for hypocrisy induction, we focus on how regulatory focus influences moral regulation in response to ethical dissonance. Our examination counterargues a “one-size fits all” single-level theory of motivation assuming that approach/avoidance comprises prescriptive/proscriptive morality, regulatory focus, and activating/inhibiting behavior. Instead, we show that regulatory focus is unique and functionally independent from prescriptive/proscriptive morality (Cornwell & Higgins, 2016).

Ethical Implications

Our study on the effectiveness of hypocrisy induction in combating cyberbullying has significant implications for the realm of business ethics. Today, a key factor for business success is answering the question “Why do we do what we do?” When companies recognize that their fundamental role is to provide a societal benefit for humanity rather than cater to shareholders, profits follow (Bhattacharya, 2019). As cyberbullying has become a prevalent and destructive phenomenon, this suggests that organizations need to have a moral obligation to utilize their resources in an ethical and responsible manner to address this issue.

This paper demonstrates that organizations can leverage hypocrisy induction to garner support for anti-cyberbullying campaigns, but only if they consider the consumer’s regulatory focus and the emotions it elicits. For instance, if the target audience has a dominant promotion focus, guilt should be evoked through hypocrisy induction to motivate support for anti-cyberbullying campaigns. Conversely, shame may encourages avoiding participation in anti-cyberbullying campaigns, so caution is warranted if the company targets consumers with a dominant prevention focus.

Moreover, this research raises crucial ethical considerations regarding the role of businesses in promoting social causes. While it is commendable for companies to address cyberbullying through marketing and advertising efforts, they must also be mindful of the ethical implications of their actions. For example, inducing negative emotions through hypocrisy may increase consumer support for anti-cyberbullying campaigns, but it could also be viewed as unethical by some consumers, leading to negative publicity, loss of consumer trust, and harm to the company’s reputation (Fointiat et al., 2008; Tangney et al., 1998).

Overall, this study adds to the ongoing academic discourse on the evolution of business ethics and the need for ethical practices in various areas of social activities (Greenwood & Freeman, 2017; Islam & Greenwood, 2021). By providing a deeper understanding of the contextual and multifaced consumer reactions to hypocrisy induction, this paper offers organizations and managers a unique vantage point on the world, allowing them to apply ethical principles to their business practices aligned with their value and priorities. This study underscores the importance of comprehending ethics as a tangible and applicable practice in real-world scenarios (Islam & Greenwood, 2021), such as countering cyberbullying.

Managerial Implications

Understanding how industry professionals perceive, process, and approach ethics in marketing is important (Richardson-Greenfield & La Ferle, 2021). In this regard, our research has practical marketing implications. Brands are advised to conduct CSR campaigns that discourage increasingly pervasive cyberbullying, especially because platforms are criticized for using advanced technologies that encourage unethical online behavior (Cohen-Almagor, 2018). Indeed, about 83% of online users blame the easy accessibility of the internet and of digital devices (BBC, 2018) and are demanding that companies conduct voluntary anti-cyberbullying campaigns to support victims and raise awareness. We also demonstrate that social media and online platforms can enhance consumer–company connections and brand reputations by using hypocrisy induction strategies in CSR campaigns. After marketers induce hypocrisy, they can provide separate tasks or verbal interactions such as drag-and-drop or check boxes (Aronson et al., 1991) that will make consumers perceive that they are spontaneously rather than externally motivated to support prosocial campaigns. Practitioners may question the ethics of using hypocrisy induction to generate guilt. We argue that promotion-focused consumers will welcome information that highlights positive outcomes, goals, and achievements. Thus, hypocrisy induction strategies should use appeals that activate promotion foci.

Limitations and Future Research

Our work has limitations that may guide future research. First, researchers can study other ways to apply hypocrisy induction strategies beyond cyberbullying campaigns, such as social campaigns to raise awareness about rudeness, cyber trolling, or illegal downloads. Our findings imply that hypocrisy induction will succeed in any social causes that require intrinsic and autonomous motivation of the consumer in contrast with direct, heavy-handed external deterrents.

Second, future research may explore other factors affecting hypocrisy induction, such as interdependent self-construal, which sensitizes individuals to violations of moral values (Bedford & Hwang, 2003; Tangney et al., 2007) and increases awareness of social norms (Brockner et al., 2005). Believing that interpersonal relationships affirm basic moral values, interdependent individuals are more likely to feel ethical dissonance and intense regret when they realize they have disrupted social harmony and unity (Tangney et al., 2007), making promotion focus less influential. Conversely, predominantly independent self-construal might increase receptivity to prevention focus. Construal levels (Trope & Liberman, 2010) are another consideration. For example, when consumers read concrete details about negative events, they may perceive greater risk and severity (Blum, 2008). Their guilt might heighten if they feel psychologically close to victims. As extreme levels of guilt are associated with shame, promotion focus may lose its positive effect when hypocrisy induction is paired with low-level construal.

Third, considering that prevention-focused consumers are motivated to avoid shame, how can hypocrisy induction encourage them to support anti-cyberbullying campaigns? We need further investigation of factors that cause shame to have positive effects. Prevention-focused individuals might feel less shame if they received benevolent affirmations (Ahmed & Braithwaite, 2006) or were encouraged to perceive that character is malleable (Breines & Chen, 2012; Goetz et al., 2010; Steele, 1988). Thus, future research could examine ways to alleviate shame so that hypocrisy induction would be more effective for prevention mindsets.

Fourth, our experiment was executed via a virtual interactive advertisement and failed to clearly capture actual prosocial behaviors. Attitudes aligned consistently with behavior, but controlled environments may cause participants to exaggerate their pro-sociality (Levitt & List, 2007). Future research should pursue investigations in real-world settings. Besides, although we tried to generalize the findings, our sample has WEIRD issues in coming largely from “western, educated, industrialized, wealthy, and democratic” societies (Henrich et al., 2010). Cultural differences are known to influence ethical motivations and tendencies (Kim & Johnson, 2013). Thus, future research should extend the sampling group and incorporate cultural diversity.

Lastly, some may question the appropriateness, ethicality, or responsibility of using advertisements that leverage unpleasant emotions to modify and prevent unethical behavior. Philosophers have long debated whether ends justify means. We leave judgment to readers or to future research that further examines moral implications.