Parasocial forgiveness: The roles of parasocial closeness and offense perceptions

Abstract

Recent research suggests that robust predictors of interpersonal forgiveness, such as relationship closeness and offense severity, also predict forgiveness of a parasocial target. We extend this work using an interpersonal forgiveness measure to examine forgiveness of parasocial targets as a function of parasocial closeness and offense perceptions. Across two studies, we found that pre-offense closeness was associated with greater forgiveness and current closeness, and that forgiveness significantly mediated the relationship between pre-offense and current closeness. In Study 2, we additionally found that perceptions of apology sincerity were associated with greater forgiveness and current parasocial closeness. Finally, we found that a brief measure of parasocial closeness was comparable to the Parasocial Interaction Scale in its associations with forgiveness and related outcomes. Implications regarding similarities and differences in interpersonal and parasocial forgiveness are discussed. Additionally, possible benefits of using of a brief, face-valid, and versatile set of items to assess parasocial closeness are suggested.

Keywords

Apologies apology sincerity closeness forgiveness offense severity parasocial interaction parasocial relationships sexual misconduct

In October 2017, rape allegations against film producer Harvey Weinstein catalyzed the viral spread of the #MeToo Movement, which sought to raise awareness of the prevalence and systemic tolerance of sexual misconduct, especially in the workplace (Zacharek, Dockterman, & Edwards, 2018). In the months that followed, numerous allegations against other high-profile celebrities also emerged. Many allegations drew uncontroversial condemnation but as reports of misconduct became more numerous and diverse, the public reaction became increasingly divided, and questions arose regarding whether certain cases should be considered misconduct at all. For instance, an article published in Babe.net (Way, 2018) detailed a date between “Grace” (a pseudonym) and comedian/actor Aziz Ansari, in which Grace described feeling pressured to engage in unwanted sexual acts. Some saw this incident as a useful case through which to discuss affirmative consent (Cooney, 2018), while others felt that Ansari was unfairly demonized for a “bad date” (Weiss, 2018).

While this diversity of perspectives is undoubtedly attributable to a variety of personal, political, and social variables, one lens through which these reactions might be usefully examined involves the connections that people felt with the accused individuals, which might have subtly influenced how they perceived the celebrities’ behaviors. Those who followed Ansari’s work and felt close to him, for instance, might have been more willing to construe his behavior charitably compared to someone without such a strong connection.

In the interpersonal domain, researchers have developed a rich understanding of how relationship variables and offense characteristics influence reactions to transgressive behavior. However, in the domain of parasocial relationships, or the enduring connections that people sometimes develop with media personalities (Horton & Wohl, 1956), these dynamics have not been extensively studied. We add to the small number of existing studies on this topic by examining reactions to offenses perpetrated by parasocial relationship partners as a function of well-studied predictors of interpersonal forgiveness.

The importance of understanding these dynamics in the parasocial domain is twofold. First, celebrities are influential figures. Audiences look to celebrities as role models, and as such, their behaviors and opinions have meaningful impacts upon the actual behavior of fans (Brown, 2015; Fraser & Brown, 2009; Kosenko, Binder, & Hurley, 2016), as well as on the public discourse surrounding important issues (Boykoff & Goodman, 2009; Thrall et al., 2008). Thus, understanding how people perceive and respond to the transgressive behavior of celebrities has important implications per se.

Second, comparisons between parasocial and interpersonal relationships can provide meaningful insights regarding the place that parasocial relationships occupy in people’s psychosocial experiences. Such insights can improve our understanding of the ways in which parasocial relationships are qualitatively distinct from two-sided social relationships and the ways in which they essentially function as quantitatively less intense—but nonetheless social—relationships. The extent to which parasocial relationship partners are treated similarly to our “real-life” friends (and enemies) in how we interpret and react to their bad behavior provides one useful point of comparison.

Closeness, commitment, and interpersonal forgiveness

Forgiveness centrally involves repairing the attitudinal damage caused by an offense toward an offender (McCullough, Fincham, & Tsang, 2003). To a greater extent than behavioral correlates of forgiveness (e.g., seeking revenge or reconciliation), attitude change following a transgression has just as much relevance in a one-sided relationship as in an interpersonal one. Just as one’s attitude toward a friend may be damaged when (s)he commits an interpersonal transgression, one’s attitude toward a parasocial target may likewise be damaged if (s)he engages in bad behavior, and inclinations to forgive this person are affected by many factors.

Closeness with one’s offender is among the most robust predictors of interpersonal forgiveness (see Riek & Mania, 2012, for a meta-analytic review). In part, the importance of forgiveness in an interpersonal context is its capacity to facilitate the maintenance of valuable relationships despite the hurtful moments that inevitably occur within them. Numerous investigations have shown that greater relationship closeness, satisfaction, and commitment are associated with more forgiveness of one’s partner (e.g., Brown & Phillips, 2005; Karremans & Van Lange, 2004; McCullough et al., 1998).

Research has indicated that one mechanism through which closeness and forgiveness are related involves victims’ motivations to perceive offenses and attempts at restitution charitably. For instance, perceived severity of offenses (another robust predictor of forgiveness; e.g., Fincham, Jackson, & Beach, 2005) is inversely associated with relationship satisfaction between the offender and victim (Schumann, 2012), even after accounting for the objective severity of the offense. Similarly, Finkel, Rusbult, Kumashiro, and Hannon (2002) found that greater commitment predicted more charitable attributions about offenses, which in turn led to greater forgiveness. Additionally, while apologies have been found in general to facilitate forgiveness (McCullough et al., 1998; Zechmeister, Garcia, Romero, & Vas, 2004), apologies delivered by partners in satisfying relationships are perceived as more sincere than those in less satisfying relationships (Schumann, 2012), and thus more effectively facilitate forgiveness and restitution. Such findings suggest that close relationships change perceptions of offenses in a manner that promotes forgiveness and relationship maintenance.

Forgiveness in parasocial relationships

There are theoretical reasons to believe that forgiveness, like many other social processes, may operate similarly in interpersonal and parasocial domains. A growing body of research demonstrates that parasocial and interpersonal relationships overlap substantially in how they are experienced (Derrick, Gabriel, & Hugenberg, 2009; Rubin & McHugh, 1987), their positive effects (such as increased belonging, Derrick et al., 2009), and their negative consequences (e.g., from relationship loss; Cohen, 2003; Daniel & Westerman, 2017; Eyal & Cohen, 2006; Lather & Moyer-Guse, 2011). Thus, it seems reasonable to suggest that dynamics relating to offenses and their aftermath might also work similarly in parasocial and interpersonal relationships.

However, there are important differences between parasocial and interpersonal relationships that might change the closeness–forgiveness relationship. As mentioned above, part of the reason that closeness and commitment facilitate interpersonal forgiveness is that close relationships are valuable and costly to lose (McCullough et al., 1998; Rusbult, Verette, Whitney, Slovik, & Lipkus, 1991). While parasocial relationships certainly provide benefits (e.g., feelings of belonging), the dissolution of a parasocial relationship lacks many of the costs of interpersonal loss. For instance, friendships and romantic relationships entail an exchange of tangible benefits (e.g., financial help and advice) that is absent in a one-sided relationship. People might therefore have less motivation to interpret transgressive behavior charitably in close parasocial relationships than in interpersonal relationships.

There is still relatively little quantitative research on parasocial forgiveness, but a few studies have provided findings that mirror those from interpersonal research. For instance, Cohen (2010) conducted a direct comparison between participants’ anticipated reactions to moral, trust, and social violations committed by media figures and friends and found that anticipated reactions to trust and social violations did not differ by relationship type, but that participants expected moral violations to result in larger reductions in closeness within mediated relationships than friendships. Cohen attributed the latter difference to the lower commitment in mediated relationships and speculated that people may be more motivated to attribute moral violations in friendships (compared to mediated relationships) to external factors rather than dispositions. This interpretation is consistent with findings about charitable behavioral attributions in interpersonal relationships and suggests a point of commonality between offense-related dynamics in parasocial and interpersonal relationships.

In an investigation of parasocial liking and forgiveness, Hu, Young, Liang, and Guo (2017) presented fabricated news stories about a widely liked or disliked celebrity engaging in a transgression and found that liked figures were granted more forgiveness and received less blame than disliked figures, consistent with interpersonal findings. Furthermore, Bostwick and Lookadoo (2017) showed that Cleveland residents who had stronger parasocial relationships with LeBron James experienced greater distress when he left Cleveland to play for Miami in 2008, mirroring the observation that betrayals within close relationships are more distressing than those committed by acquaintances or strangers (McCornack & Levine, 1990; McCullough et al., 1998). In short, evidence suggests that parasocial and interpersonal closeness exhibit similar relationships to offense reactions.

Existing evidence regarding how parasocial closeness relates to forgiveness can be strengthened and expanded in a number of ways. First, the methods that have been employed thus far do not consistently resemble those used in interpersonal forgiveness research. For instance, Hu and colleagues (2017) assessed forgiveness via a 3-item measure of their own design, which assessed: “forgiveness” for the target, intentions to “remain loyal,” and intentions to cease parasocial interaction (reversed) with the target. While this appears to be a reasonable measure under some conceptualizations of forgiveness, there would be value in using a measure with established construct validity in the interpersonal domain. Especially if one is interested in understanding the similarities between parasocial and interpersonal forgiveness dynamics, it is useful to maximize the comparability of measures across domains.

Second, Hu and colleagues (2017) experimentally manipulated parasocial liking and offense features rather than examining variability in real transgressions and relationships. Replicating their findings using naturally occurring transgressions to which participants had real reactions would complement the strengths of their experimental designs. Thus far, parasocial forgiveness studies using real-world transgressions have been either qualitative (Finsterwalder, Yee, & Tombs, 2017; Sanderson & Emmons, 2014) or limited in scope (Bostwick & Lookadoo, 2017).

Finally, existing research does not speak directly to why parasocial closeness might be related to forgiveness. As discussed above, interpersonal research suggests that factors like intimacy and relationship satisfaction change offense perceptions in a way that is conducive to forgiveness and reconciliation. The extent to which parasocial closeness facilitates forgiveness through similar mechanisms is an open question, and one that would be beneficial to examine centrally, as it might have broader implications regarding how parasocial relationships function in people’s psychosocial experiences.

In two studies, we examined associations between parasocial closeness and forgiveness for real-world offenses, using methods that (1) complement previous experimental methods by increasing ecological validity and (2) closely resemble those employed in interpersonal forgiveness research. In Study 1, participants answered questions about one of three celebrities who were accused of sexual misconduct, and we tested whether offense perceptions mediated the relationship between their felt closeness with the target and their level of forgiveness. In Study 2, participants recalled an offense by someone with whom they had a parasocial relationship, and we again tested whether offense perceptions mediated the relationship between parasocial closeness and forgiveness. Additionally, we examined in Study 2 whether perceptions of post-offense behavior (e.g., apology sincerity) mediated the relationship between parasocial closeness and forgiveness.

Study 1

We assessed reactions to three sexual misconduct scandals involving Aziz Ansari (Way, 2018), Louis C.K. (Ryzik, Buckley, & Kantor, 2017), and Kevin Spacey (Vary, 2017). We selected these specific targets for a few reasons. First, these cases were similar in important ways (e.g., the accusations were recent at the time of data collection and publicized around the same time, between October 29, 2017 and January 16, 2018; all scandals involved sexual misconduct allegations). Second, the public response to the allegations was mixed, leading us to anticipate wide variability in the variables of interest. Finally, we intentionally sampled offenses that on their faces differed in severity, to ensure representation across the spectrum of this important offense characteristic.

To the extent that forgiveness dynamics are similar in the parasocial and interpersonal domains, we expected to observe several well-documented relationships.
H1: Self-reported closeness before the offense (H1a) and after the offense (H1b) is positively associated with forgiveness.

H2: Self-reported closeness before the offense (H2a) and after the offense (pre-offense 2b) is inversely related to attributions of intentionality/responsibility.

H3: Greater attribution of intentionality/responsibility to the target is inversely related to forgiveness.

H4: Pre-offense parasocial closeness predicts current closeness through its relationship with offense perceptions and forgiveness.

Method

Participants

In order to recruit participants familiar with the target scandals, we advertised our study on three “Subreddits” (themed message boards on www.reddit.com) devoted to the three targets or projects for which they are well-known (i.e., /r/HouseOfCards/ for Kevin Spacey, r/PandR/ for Aziz Ansari, and /r/louisck/ for Louis C. K.). We also recruited from the Facebook pages of two current events podcasts (serious inquiries only and opening arguments) that discussed the scandals, and thus were likely to have knowledgeable audiences. All recruitment messages were posted with moderator permission. Of the 477 people who opened our study link, 245 (51%) did not answer enough questions to complete an attentional check, leaving 232 participants. After excluding participants with insufficient knowledge of the target incident or inattentive/dishonest response patterns (specific exclusion criteria described under “Validity Checks”), 47 (20%) participants were excluded, resulting in a final N = 185.

One-hundred and forty-seven (79%) participants completed demographic information. Of these, participants were predominantly young to middle-aged adults (92, or 63%, selected “18–29 years old”; 53, or 36%, selected “30–49 years old”), White (114 or 78%), male (104 or 71%), and well-educated, with 59% (86) reporting completion of a bachelor’s degree or higher and an additional 27% (39) reporting completion of “some college.”

Materials and procedure

Overview

We administered our study using Qualtrics survey software and collected responses for 2 months (April 8, 2018 to June 6, 2018). Participants were invited via posts in Subreddits or Facebook pages to participate in a study of their “reactions to celebrity sexual misconduct scandals.” Participants answered questions on one of the three targets¹ (for Aziz Ansari, n = 60; Louis C.K., n = 66; and Kevin Spacey, n = 59) regarding their pre- and post-offense closeness with the target, their perceptions of the offense, and their forgiveness of the target.² In order to avoid possible ceiling effects on parasocial closeness, the celebrity target each participant saw was randomly assigned, rather than matching the page topic. Table 1 displays descriptive statistics for all measures.

Table 1.
Descriptive statistics forgiveness, closeness, and offense perceptions.

Study 1 Study 2

M (SD) N M (SD) N

Forgiveness 4.66 (1.54) 185 3.12 (1.43) 184

Pre-offense closeness 3.09 (1.80) 184 2.88 (1.68) 184

Current closeness 2.58 (1.83) 184 1.70 (1.27) 184

Intent/responsibility 4.61 (1.61) 185 5.86 (0.96) 184

Subjective severity 5.54 (1.32) 184

Apology sincerity 2.70 (1.18) 66

PSI 2.67 (1.33) 184

Note. PSI = Parasocial Interaction Scale; all variables measured on a 1–7 scale, with the exception of apology sincerity (1–5 scale).

Parasocial closeness

Participants responded to 3 items on a 1 (not at all) to 7 (extremely) scale indicating how “close,” “committed,” and “connected” they felt to the target prior to learning of the allegations (α = .88) and an additional 3 items assessing the same feelings “now” (α = .95). From these, we computed means of pre-offense and current closeness. This measure is consistent with previous studies on interpersonal forgiveness, which have used one or a small set of face-valid items to assess felt closeness with an offender (e.g., Brown & Phillips, 2005; Tsang, McCullough, & Fincham, 2006).

Attributions of intentionality

We assessed attributions of intentionality and responsibility via a mean of 5 items (α = .84; on a scale of 1, not at all, to 7, completely): “To what extent do you think this person’s offense was intentional?” “To what extent do you think this person’s offense was pre-meditated?” “How personally responsible for this event was this person?” “The accused could have prevented his current situation,” and “The scandals were caused by the actions of the accused.”

Parasocial forgiveness

Existing parasocial forgiveness studies have not utilized validated forgiveness measures, likely because most scales focus on motivations, cognitions, and behaviors that imply or require a reciprocal relationship. However, Brown and Phillips’ (2005) State Forgiveness Scale has been widely used, well-validated, and because of its attitudinal focus, it is easily adapted for parasocial relationships.

Adapting the State Forgiveness Scale required three minor modifications. We altered the first clause of “Even though his/her actions hurt me, I do not feel ill-will toward him/her” to “Despite his/her actions” so as not to reference an interpersonal infraction. “If I saw this person again, I would try to avoid interacting with him/her” was modified to reflect a parasocial form of avoidance: “I avoid watching his/her work.” Finally, from “I hope this person gets what’s coming to them for what they did to me [reversed],” we removed “to me.” The original forms of other 4 items were applicable to parasocial relationships (e.g., “I dislike this person [reversed],” “I have forgiven this person”). Responses were on a 7-point Likert-type scale (from 1 = strongly disagree to 7 = strongly agree), from which we computed a mean composite (α = .91).

Validity checks

Given the lack of control in online data collection, we included several attentional checks. We interspersed 3 items into our questionnaires that asked participants to respond (on a scale of 1 = strongly disagree to 7 = strongly agree) to the statement “I have been answering these questions honestly and accurately” (one of the three statements was reversed, i.e., “I have not been…”). Participants who did not at least “somewhat agree” that they were being honest on all 3 items were excluded, which was the case for 9 of 232 (4%).

We also included three free-response knowledge checks, asking participants how many people had accused the target, the nature of the target’s misconduct, and what, if any, response the target made to the allegations. If a participant did not respond (as 26 did or 11%) or indicated that they were unfamiliar with the incident (e.g., “No idea”), their response was excluded. Twelve (5%) participants indicated either a lack of incident knowledge or demonstrated insincere responding. For instance, one excluded participant claimed that Louis C.K. was accused by “69” women, that the nature of the accusation was “FAJE [sic] NEWS,” and that C.K.’s response was “Boop.” We were therefore skeptical of this participant’s willingness to respond sincerely to other questions.

Results

H1–H3: Correlations between closeness, forgiveness, and offense attributions

Table 2 displays zero-order correlations between all variables. Consistent with interpersonal research, forgiveness was positively correlated with pre-offense and current closeness, supporting H1a and H1b. Current closeness, but not pre-offense closeness, was inversely related to attributions of intentionality/responsibility. Thus, H2b was supported, but H2a was not. Forgiveness was also inversely related to attributions of intentionality/responsibility, supporting H3.

Table 2.
Zero-order correlations between all continuous variables.

Study 1 1 2 3

1. Forgiveness

2. Pre-offense closeness .29*

3. Current closeness .67* .70*

4. Intent/responsibility −.58* −.08 −.33*

Study 2 1 2 3 4 5 6

1. Forgiveness

2. Pre-offense closeness .35*

3. Current closeness .65* .60*

4. Intent/responsibility −.57*** −.14* −.44*

5. Subjective severity −.53* −.07 −.39* .51*

6. Apology sincerity .67*** .26* .49* −.29* −.42

7. PSI .76* .55* .71* −.42* −.38* .67***

Note. PSI = Parasocial Interaction Scale. Study 1: N = 184–185; Study 2: N = 184, with the exception of apology sincerity (N = 66).

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

H4: Offense attributions and forgiveness as mediators of pre-offense and current closeness

As discussed previously, forgiveness can be conceptualized as a process of attitude change, and one that is central to post-offense maintenance of relationships. Thus, one might expect greater closeness to motivate offense perceptions that facilitate forgiveness, and that forgiveness would in turn facilitate greater post-offense closeness. We tested a mediation model consistent with this chain of influence using “Model 6” from PROCESS 3.0 (Hayes, 2017), which evaluates multiple mediators operating serially (see Figure 1). This model estimates three indirect paths, one through each mediator individually and one through both mediators in sequence. All variables were mean-centered prior to analysis. Note that PROCESS uses point estimates and confidence intervals rather than p-values to evaluate indirect paths (Hayes, 2017); for this reason, no p-values are reported below in association with indirect paths.

Figure 1.
Conceptual depiction of serial mediation model tested in H6 (see Table 3).

The results of this analysis indicated that neither the indirect path through attributions of intentionality/responsibility nor the path through attributions and forgiveness in sequence was significant. However, the path through forgiveness was significant, consistent with the notion that pre-offense intimacy facilitates post-offense levels of intimacy through its promotion of forgiveness (see Table 3).

Table 3.
Study 1: Mediation analysis of the relationship between pre-offense (IV) and current closeness (DV), mediated by offense attributions and forgiveness.

B SE t p-Value 95% CI

Pre-offense closeness (IV)

IV to M1 −.09 .08 −1.12 .27 −.23 .05

IV to M2 .29 .07 4.13 <.0001 .15 .43

Total effect .76 .06 13.12 <.0001 .64 .87

Direct effect .59 .05 13.19 <.0001 .51 .68

Intent/responsibility (M1)

Direct effect .01 .05 .22 .83 −.09 .11

Indirect effect −.00 .01 −.01 .01

Forgiveness (M2)

Direct effect .55 .06 10.05 <.0001 .44 .66

Indirect effect .14 .04 .06 .20

IV → M1 → M2 → DV

Indirect effect .03 .02 −.02 .07

Note. Confidence intervals (95%) were constructed using 5,000 bootstrapped samples; N = 184.

It is worth highlighting that although offense attributions exhibited a zero-order relationship with current closeness, this relationship did not account for unique variance in the mediation model. This suggests that any variance in current closeness accounted for by such attributions is explained by forgiveness. We used a simple, single-mediator model (“Model 4”) to formally examine whether the relationship between offense attributions and current closeness was fully mediated by forgiveness. Indeed, although the total path from attributions to current closeness was significant (B = −.35, p < .001, 95% CI: −.49 to −.20), the indirect effect through forgiveness (B = −.43, 95% CI: −.54 to −.33) accounted entirely for this effect (direct effect: B = .09, p = .23, 95% CI: −.05 to .23).

Discussion

These results support the contention that, as in interpersonal relationships, our reactions to transgressions within parasocial relationships are partly shaped by our preexisting feelings of closeness with the offenders. Pre-offense closeness was positively associated with current closeness through its relationship with forgiveness, and current closeness was associated with more charitable attributions of intentionality/responsibility through its relationship with forgiveness. Again, these findings are consistent with those indicating that charitable offense perceptions are a mechanism through which forgiveness may facilitate interpersonal relationship maintenance (e.g., Finkel, Rusbult, Kumashiro, & Hannon, 2002).

However, the fact that offense perceptions did not mediate the relationship between pre-offense and current closeness, either by itself or in conjunction with forgiveness, is interesting. It is possible that parasocial closeness is not as powerful as interpersonal closeness in shaping our perceptions of bad behavior. Intuitively, it would not be surprising if this were the case; even very positive parasocial relationships provide limited access to someone’s private character and behavior, and thus should be more tenuous than a close interpersonal relationship.

That said, more evidence would be necessary in order to rule out features of our study as explanations for this difference. First, we only examined sexual misconduct incidents, to which people may not react as they do to transgressions in general. Perhaps in this domain of offenses, intentionality factors less into whether a behavior is considered forgivable. Second, it is possible that our retrospective measure of pre-offense closeness resulted in a testing effect that would have obscured any potential mediation by offense perceptions. In other words, perhaps participants anchored their ratings of pre-offense closeness on their current feelings of closeness and did not accurately recall their pre-offense level of closeness. If so, we would have underestimated the mediating role of offense perceptions. Finally, our participants’ reactions to these offenses may not resemble those of the general population. People who frequent Subreddits devoted to these targets might have had distinctive characteristics that altered the relationships of interest. Thus, we considered it prudent to replicate these findings in a more general sample.

Study 2

In order to address concerns in Study 1, we conducted a replication with complementary methods of offense sampling and participant recruitment. Rather than advertising our study to participants who were likely to have close parasocial relationships with particular targets, we recruited a sample from Prolific Academic (www.prolific.co), an online research platform that is similar to Amazon’s Mechanical Turk (MTurk), but was designed by and for behavioral and cognitive researchers. People who are interested in participating in compensated research studies may sign up to be part of Prolific’s participant pool, and after completing a preliminary demographics questionnaire may browse available studies. Notably, Prolific has taken steps to address concerns that have been raised about MTurk regarding both data quality (e.g., Chandler, Paolacci, Peer, Mueller, & Ratliff, 2015) and fair compensation (Semuels, 2018).

To address generalizability concerns regarding the three offenses we selected for Study 1, we asked participants to choose a parasocial offense on which to report. Additionally, we assessed participants’ perceptions of offense severity and offenders’ post-offense behavior (i.e., apologies and apology sincerity), both of which are important sources of variance that were neglected in Study 1.

Our first four hypotheses were the same as Study 1, and predictions regarding subjective severity were also included mirroring those for attributions of intentionality/responsibility (H2 and H3). We also added the following hypotheses regarding post-offense behavior.
H5: Forgiveness is greater for offenders who apologized compared to those who did not.

H6: For offenders who apologized, parasocial closeness is positively correlated with perceived sincerity of the apology.

H7: For offenders who apologized, the relationship between pre-offense closeness and current closeness is mediated by forgiveness and apology sincerity.

Finally, we considered it important to directly address the validity of our parasocial closeness measure. Given that we did not use a standard assessment of parasocial relationship strength, we wanted to compare it to the more established Parasocial Interaction Scale (PSI) (Rubin, Perse, & Powell, 1985), which is perhaps the most popular measure of parasocial relationship strength, but has some practical and theoretical limitations (see Dibble, Hartmann, & Rosaen, 2015, for an in-depth critique).

First, in the context of brief online studies (particularly those requiring participant payment), the value of every additional question must be weighed against its monetary cost and waning participant attentiveness. Thus, a 3-item measure of closeness may be preferable to the 15-item PSI when survey length is a concern. Second, while it has been extensively demonstrated that the PSI assesses liking for a parasocial target, many items do so indirectly, by asking participants to report on aspects of their “interaction” experiences (e.g., “When I’m watching the program [target] is on, I feel as if I am part of the group,” “I like to compare my ideas with what [target] says”). Third, the PSI must be tailored for different forms of media, as its statements do not consistently apply to parasocial interactions across media domains; as media forms continue to proliferate, this will become increasingly problematic for researchers. Finally, beyond these practical benefits, many researchers interested in parasocial relationships are specifically interested in the psychosocial consequences of experiencing a close, enduring bond with a media figure, beyond interaction experiences (Dibble, Hartmann, & Rosaen, 2015). Our measure assesses this subjective closeness using direct and face-valid items.

We tested the following hypotheses in order to evaluate the relationship between our measure and the more-established PSI and to examine the extent to which the two measures overlap in the variance for which they account in forgiveness.
H8: The PSI and current closeness are positively correlated.

H9: The PSI and current closeness demonstrate comparably strong relationships with forgiveness, its antecedents, and its consequences.

H10: The PSI and current closeness partially mediate each other’s relationships with forgiveness.

Method

Participants

We recruited 206 participants through Prolific Academic for a study on “perceptions of wrongdoing by public figures.” The only exclusion variable was that participants be fluent in English; thus, of 44,173 Prolific users registered at the time of collection, 29,954 (68%) were eligible to participate in our study. Participants were compensated for a complete and conscientious response (i.e., one that passed at least 80% of validity checks) with £1.5 (approximately US$1.95), in accordance with Prolific’s compensation standards of £5 per hour (our measures required an average of 15 min to complete).

After excluding 22 (11%) participants due to validity check failures (using the same criteria as Study 1) and failures to follow instructions (e.g., one participant reported a neighbor’s transgression), our final sample had N = 184. Our participants were roughly half male (97, 53%) and female (87, 47%), ranged from 18 years to 66 years old (M = 30.39, SD = 9.93), were predominantly White (161, 88%), and well-educated, with 149 (81%) reporting at least “some college,” and 109 (59%) reporting a college degree or higher.

Materials and procedure

Participants first read a prompt designed to elicit transgressions that varied in terms of severity and were committed by targets with varying levels of pre-offense closeness.

Please take a moment to think back upon an example from the past few years in which a celebrity or public figure did something that you found wrong, hurtful, offensive, or a betrayal; the wrongdoing could be an action or an inaction – something this person said or did, or something (s)he failed to do. It could be something relatively minor, something more severe, or anything in between.

The person could be an actor or other performer, YouTuber, podcast host, politician, local news anchor, and so on but please try to think of someone of whom you had knowledge prior to the wrongdoing (in other words, don’t choose someone that you first heard about as a result of their wrongdoing). Take a moment to recall this incident, and then answer the questions below.

Participants then briefly described the offense and provided the offender’s name, gender, and occupation (e.g., actor, musician, and politician).

Following these questions, pre-offense (α = .95) and current (α = .97) closeness, attributions of intentionality/responsibility (α = .66), and parasocial forgiveness (α = .91) were all assessed as in Study 1. Subjective severity was assessed via a mean of 4 items (α = .88; on a scale of 1 = not at all to 7 = completely): “How severe were these incidents?” “How hurtful to you was this event at the time it occurred?” “How serious did you feel this offense was, at the time it occurred?” and “How offensive to you were this person’s actions or inactions at the time?” Participants also indicated whether or not the offender had apologized. If the offender had apologized, participants rated how sincere, remorseful, and repentant they perceived the apology to be on a 1 (not at all) to 5 (completely) scale. Responses to these three questions were averaged to create a composite sincerity variable (α = .97). Participants additionally completed the 15-item PSI (Rubin et al., 1985; α = .95), with response options from 1 (strongly disagree) to 7 (strongly agree). All measures, aside from the initial prompt and offense description, were presented in a randomized order.

Results

Descriptive statistics

Table 1 includes descriptive statistics for all variables. Compared to our Study 1 sample, participants in Study 2 reported lower average levels of forgiveness and parasocial closeness, as well as greater attributions of intentionality/responsibility. These differences are unsurprising given the different recruitment strategies in these studies.

H1–H3: Correlations between closeness, forgiveness, and offense attributions

All significant correlations from Study 1 were replicated (see Table 2), and additionally, the correlation between pre-offense closeness and attributions of intentionality/responsibility, which was not significant in Study 1, was significant (although weak) in Study 2. We also found that perceptions of offense severity were negatively associated with forgiveness and parasocial closeness. Thus, H1–H3 were supported.

H4: Offense perceptions as mediators of closeness and forgiveness

We repeated the mediation analysis in Study 1, with attributions of intentionality/responsibility and forgiveness specified as serial mediators (“Model 6”). Additionally, we ran a similar analysis replacing attributions with perceived offense severity. In both models, the only significant indirect effect between pre-offense and current closeness was through forgiveness; neither offense attributions nor perceived severity were significant mediators of this relationship (see Supplementary Materials). However, in contrast to Study 1, the direct paths between both attributions and perceived severity and current closeness were significant. Thus, offense perceptions accounted for unique variance in current closeness beyond the variance accounted for by pre-offense closeness.

As in Study 1, we ran two single-mediator models (“Model 4”) to assess whether forgiveness mediated the relationships of offense attributions and subjective severity with current closeness. Our results precisely replicated those of Study 1 for both variables. Offense attributions had a significant relationship with current closeness (B = −.44, p < .001, 95% CI: −.58 to −.31) that was fully accounted for by its indirect effect through forgiveness (B = −.34, 95% CI: −.46 to −.23; direct effect: B = −.10, p = .13, 95% CI: −.24 to .03). Similarly, subjective severity’s relationship with current closeness (B = −.39, p < .001, 95% CI: −.53 to −.26) was fully accounted for by the indirect effect through forgiveness (B = −.33, 95% CI: −.45 to −.22; direct effect: B = −.06, p = .35, 95% CI: −.19 to .07).

Thus, we replicated the essential findings from Study 1 that forgiveness significantly mediated the relationships between pre-offense and current closeness, as well as the relationships between offense perceptions and current closeness. Additionally, our mediation analysis suggested that offense perceptions did account for unique variance within the model, which is more consistent with interpersonal research than were the results of Study 1.

H5–H6: Perceptions of post-offense behavior

Sixty-six of our participants reported that the offender had apologized, whereas 81 reported no apology (the remaining 37 were unsure whether an apology had been offered). An independent samples t-test supported H5, indicating that participants were more forgiving of offenders who had apologized (M = 3.81, SD = 1.31) than those who had not (M = 2.39, SD = 1.22), t(145) = 6.80, p < .001. Furthermore, consistent with H6, perceived sincerity of the apology was positively correlated with both pre-offense closeness and current closeness (see Table 2).

To test H7, we computed a single-mediator model (“Model 4”) to examine whether apology sincerity significantly mediated the relationship between pre-offense closeness and forgiveness. While the relationship of pre-offense closeness was no longer significant after accounting for apology sincerity (B = .05, p = .53, 95% CI: −.12 to .23), the indirect effect through apology sincerity was not significant (B = .15, 95% CI: −.01 to .30). We ran a similar analysis replacing forgiveness with current closeness as the outcome variable and obtained conceptually similar results: The indirect effect through apology sincerity was not significant (B = .09, 95% CI: −.00 to .23), and in this case, pre-offense closeness retained a significant direct relationship with current closeness (B = .59, p < .0001, 95% CI: .40 to .55). Thus, H7 was not supported.

H8–H10: “Para-closeness” versus the parasocial interaction scale

Finally, we compared the relationships of our focal variables with current closeness and the PSI to evaluate the convergent validity of our parasocial measure with a more established one. Current closeness and the PSI were strongly correlated with each other, consistent with H8, and they demonstrated comparable relationship strength with virtually all variables of interest (supporting H9; see Table 2), reinforcing our contention that our measure predicts variables relevant to parasocial relationship strength at least as well as the PSI.

To examine H10, we conducted two single-mediator analyses (using PROCESS 3.0, “Model 4”) with forgiveness as the predictor. In the first analysis, the PSI was treated as the mediator and current closeness was treated as the outcome; in the second, current closeness was instead treated as the mediator and PSI as the outcome (see Supplementary Materials for a summary of results). Both measures had strong, significant relationships with each other and forgiveness, and both partially mediated forgiveness’s relationship with the other. However, the PSI accounted for about twice as much variance as a mediator (P_M = .60) as did closeness (P_M = .32). These results support our contention that our measure of closeness overlaps with the PSI in the construct that it assesses, but also suggest that they are distinguishable. The PSI, as discussed previously, measures constructs beyond closeness, which is reflected in these results.

Discussion

These results reinforce and extend those from Study 1 in a few critical ways. First, they suggest that the overall patterns we observed in Study 1 were not due to features of the offenses we selected, nor due to our participant recruitment strategy. Indeed, the majority of our findings in Study 2 were identical to Study 1. Thus, we can make confident inferences about a number of similarities in how forgiveness dynamics operate in parasocial and interpersonal contexts, as well as at least one point of divergence (discussed below).

Second, we added two important variables in Study 2 that were not included in Study 1: perceptions of both offense severity and apologies. Both of these variables had strong relationships with forgiveness and current closeness, mirroring interpersonal findings. However, neither explained significant variance in the relationship between pre-offense closeness and either forgiveness or current closeness. This is inconsistent with the suggestion that people with strong parasocial relationships are more likely to forgive offenses because they are motivated to see offenses as less severe or to see post-offense attempts at restitution as more sincere.

General discussion

These studies add to the literatures on forgiveness and parasocial relationships in several ways. First, they add to the growing number of demonstrations that parasocial relationship dynamics overlap consequentially with interpersonal relationships. As in numerous interpersonal studies, we found that parasocial forgiveness and relationship maintenance following an offense were associated with greater pre-offense closeness, lesser subjective severity of the offense, lesser attributions of intentionality/responsibility, and more positive perceptions of post-offense behavior.

Second, they contribute to knowledge of the mechanisms by which parasocial relationships are maintained following a transgression; to our knowledge, we are the first to apply mediation models to examine whether offense perceptions and forgiveness help to explain how people maintain parasocial relationships after a transgression. We provide evidence that forgiveness mediates the relationship between pre-offense and current closeness, supporting its role in the maintenance of close relationships, parasocial, or otherwise.

However, in both studies, we observed a point of divergence from interpersonal forgiveness research. Findings suggest that close interpersonal relationships prompt charitable perceptions of the bad behavior of significant others (Finkel et al., 2002; Paleari, Regalia, & Fincham, 2003), and that these attributions are a critical component of relationship maintenance following a transgression. Similarly, Cohen (2010) speculated that we may be more motivated to attribute friends’ than media personalities’ transgressions to external factors in service of relationship maintenance. Our findings suggest a somewhat different picture. In both studies, perceptions of the offense and post-offense behavior did have positive relationships with forgiveness and current parasocial closeness, but they did not play a mediating role between these variables and pre-offense closeness. This is an interesting difference, and one that should be probed further. If offense perceptions and attributional mechanisms do not link parasocial relationship strength with forgiveness or relationship maintenance, what does?

Limitations

One limitation of both studies was our use of a retrospective pre-offense closeness measure. While this was unavoidable, it obviously limits what can be inferred from this variable. Participants may have used current feelings of closeness to anchor their ratings of pre-offense closeness, which could have obscured the effects of other variables on changes in closeness. Alternatively, it is possible that the scandals distorted participants’ memories of how close they used to feel with the targets, and thus there was potential for “backward causality” that would not have been possible if we had obtained closeness rating before the offenses. However, retrospective reporting has been used throughout forgiveness research (e.g., Brown & Phillips, 2005; McCullough et al., 1998; Tsang, McCullough, & Fincham, 2006), so we can at least have confidence that our results can be compared with such findings.

Studies 1 and 2 also had complementary limitations regarding their samples. In Study 1, the generalizability of our sample was questionable because we recruited from Subreddits about the targets. This increased the likelihood that they related to the targets in ways that are not necessarily representative of “typical” parasocial relationships. Not everyone who feels connected to Aziz Ansari through watching his comedy specials is necessarily motivated to belong to a community of similarly interested individuals to discuss him or his work. Thus, there are likely some unique characteristics in this sample for which we did not account.

In Study 2, there was some diminished variability in parasocial closeness. Not only was closeness lower on average, but there were more participants who were not at all close and fewer who were very close, with 49 participants (27%) reporting the lowest possible closeness rating and only 9 (5%) reporting very close ratings (≥6), compared to 37 (20%) and 19 (10%), respectively, in Study 1. However, these sampling problems clearly were not substantial enough to significantly distort the overall results, as findings across them were largely consistent.

Future directions

As discussed above, further research regarding the mechanisms that link parasocial closeness with forgiveness would be useful. Our results suggest that offense perceptions and behavioral attributions might not be the primary mechanisms by which forgiveness operates to maintain parasocial relationships. Perhaps the extent to which parasocial interactions are perceived as separable from the private lives of parasocial targets could be a fruitful mechanism to explore. For example, if one primarily “interacts” with a musician through his/her music, it might be possible to compartmentalize that music from the knowledge that the musician engaged in transgressive behavior. It is also possible that there are individual differences in the extent to which particular people are able to successfully engage in this compartmentalization.

Finally, further research examining the utility of our parasocial closeness measure would be useful. Not only is our measure brief and face-valid, but its items can be easily applied to a variety of targets with no need of modification. For certain research questions, these features could provide advantages over other parasocial measures. The nature of these items makes it easy to assess participants’ important parasocial relationships in general, rather than limiting responses to a favored television character, celebrity, or other specific media personality. Thus, for studies relating to the ubiquity and importance of parasocial relationships in people’s lives, this generality could be particularly useful. If one is interested in how interpersonal belonging relates to dependence on parasocial relationships, it would be prudent to assess participants’ strongest parasocial relationships, which could be in any (or multiple) domain(s), making a domain-general measure advantageous.

Given these potential advantages, we believe that more research with these items would be valuable. The findings presented here only speak to this measure’s usefulness in a narrow context, so examining how well felt closeness predicts other important parasocial phenomena will be important in making the case for the measure’s utility. Additionally, our studies ignored differences between relationships beyond degree of closeness (i.e., whether the participant feels close to the target because of sexual attraction, admiration, identification, etc.) that might moderate how closeness relates to other phenomena. Extending the current findings by examining possible moderators like these would also be valuable. Furthermore, we performed only preliminary comparisons between closeness and one other parasocial measure. Future studies should evaluate the distinctiveness of parasocial closeness from related constructs, such as those measured by the Parasocial Interaction Scale, the Transportability Scale (Greenwood, 2008), and Retrospective Imaginative Involvement (Slater, Ewoldsen, & Woods, 2018). We recognize that something may be lost in assessing only feelings of closeness, as opposed to the multifaceted constructs assessed by some of these other measures. However, we contend that felt closeness per se is often the construct of interest in studies of how and to what extent parasocial and interpersonal relationships resemble one another and fulfill similar psychosocial functions.

Conclusion

Our study contributes meaningfully to the literatures on the effects of transgressive behavior in parasocial and interpersonal relationships. Several of our findings suggest similarities between the two domains: forgiveness appears to be important for the maintenance of parasocial relationships in the wake of a transgression, as it is for interpersonal relationships. Furthermore, the transgressor’s post-offense behavior (e.g., offering a sincere apology) appears to be a powerful facilitator of forgiveness and parasocial relationship maintenance. However, some findings suggest differences between interpersonal and parasocial forgiveness dynamics. Most notably, the degree of closeness reduction was not explained by participants’ offense perceptions. More research establishing the conditions under which this difference is observed and the mechanisms that explain it would further our understanding of the functional and conceptual overlap between parasocial and interpersonal relationships.

Supplemental material

Supplemental Material, Osterman_and_Hecmanczuk_-_Supplementary_Materials - Parasocial forgiveness: The roles of parasocial closeness and offense perceptions

Supplemental Material, Osterman_and_Hecmanczuk_-_Supplementary_Materials for Parasocial forgiveness: The roles of parasocial closeness and offense perceptions by Lindsey L. Osterman and Theresa A. Hecmanczuk in Journal of Social and Personal Relationships

	Study 1	Study 2
Forgiveness	4.66 (1.54)	185	3.12 (1.43)	184
Pre-offense closeness	3.09 (1.80)	184	2.88 (1.68)	184
Current closeness	2.58 (1.83)	184	1.70 (1.27)	184
Intent/responsibility	4.61 (1.61)	185	5.86 (0.96)	184
Subjective severity			5.54 (1.32)	184
Apology sincerity			2.70 (1.18)	66
PSI			2.67 (1.33)	184

Study 1	1	2	3
1. Forgiveness
2. Pre-offense closeness	.29***
3. Current closeness	.67***	.70***
4. Intent/responsibility	−.58***	−.08	−.33***
Study 2	1	2	3	4	5	6
1. Forgiveness
2. Pre-offense closeness	.35***
3. Current closeness	.65***	.60***
4. Intent/responsibility	−.57***	−.14*	−.44***
5. Subjective severity	−.53***	−.07	−.39***	.51***
6. Apology sincerity	.67***	.26*	.49***	−.29***	−.42**
7. PSI	.76***	.55***	.71***	−.42***	−.38***	.67***

	B	SE	t	p-Value	95% CI
Pre-offense closeness (IV)
IV to M1	−.09	.08	−1.12	.27	−.23	.05
IV to M2	.29	.07	4.13	<.0001	.15	.43
Total effect	.76	.06	13.12	<.0001	.64	.87
Direct effect	.59	.05	13.19	<.0001	.51	.68
Intent/responsibility (M1)
Direct effect	.01	.05	.22	.83	−.09	.11
Indirect effect	−.00	.01			−.01	.01
Forgiveness (M2)
Direct effect	.55	.06	10.05	<.0001	.44	.66
Indirect effect	.14	.04			.06	.20
IV → M1 → M2 → DV
Indirect effect	.03	.02			−.02	.07

Footnotes

Acknowledgements

The authors sincerely thank Ryan Brown, Chris Buchholz, Travis Carter, Guan-Soon Khoo, Danielle Findley-Van Nostrand, and anonymous reviewers for their helpful feedback and assistance on earlier drafts of this manuscript.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Lindsey L. Osterman

Open research statement

As part of IARR’s encouragement of open research practices, the authors have provided the following information: This research was not pre-registered. The data used in the research are available. The data can be obtained by emailing the first author at osterman@roanoke.edu. The materials used in the research are available. The materials can be obtained by emailing the first author at osterman@roanoke.edu.

Supplemental material

Supplemental material for this article is available online.

Notes

References

Bostwick

E. N.

Lookadoo

K. L.

(2017). The return of the king: How Cleveland reunited with LeBron James after a parasocial breakup. Communication and Sport, 5, 689–711.

Boykoff

M. T.

Goodman

M. K.

(2009). Conspicuous redemption? Reflections on the promises and perils of the ‘celebritization’ of climate change. Geoforum, 40, 395–406.

Brown

R. P.

Phillips

(2005). Letting bygones be bygones: Further evidence for the validity of the Tendency to Forgive scale. Personality and Individual Differences, 38, 627–638.

Brown

W. J.

(2015). Examining four processes of audience involvement with media personae: Transportation, parasocial interaction, identification and worship. Communication Theory, 25, 259–283.

Chandler

Paolacci

Peer

Mueller

Ratliff

K. A.

(2015). Using nonnaive participants can reduce effect sizes. Psychological Science, 26, 1131–1139.

Cohen

E. L.

(2010). Expectancy violations in relationships with friends and media figures. Communication Research Reports, 27, 97–111.

Cohen

(2003). Parasocial breakups: Measuring individual differences in responses to the dissolution of parasocial relationships. Mass Communication and Society, 6, 191–202.

Cooney

(2018, January). The Anziz Ansari allegation has people talking about ‘affirmative consent.’ What’s that? Time. Retrieved from http://www.time.com

Cuddy

A. J. C.

Rock

Norton

M. I.

(2007). Aid in the aftermath of Hurricane Katrina: Inferences of secondary emotions and intergroup helping. Group Processes and Intergroup Relations, 10, 107–118.

10.

Daniel

E. S.

Jr Westerman

D. K.

(2017). Valar Morghulis (All parasocial men must die): Having nonfictional responses to a fictional character. Communication Research Reports, 34, 143–152.

11.

Dibble

J. L.

Hartmann

Rosaen

S. F.

(2015). Parasocial interaction and parasocial relationship: Conceptual clarification and a critical assessment of measures. Human Communication Research, 42, 21–44.

12.

Derrick

J. L.

Gabriel

Hugenberg

(2009). Social surrogacy: How favored television programs provide the experience of belonging. Journal of Experimental Social Psychology, 45, 352–362.

13.

Eyal

Cohen

(2006). When good friends say goodbye: A parasocial breakup study. Journal of Broadcasting and Electronic Media, 50, 502–523.

14.

Fincham

F. D.

Jackson

Beach

S. R. H.

(2005). Transgression severity and forgiveness: Different moderators for objective and subjective severity. Journal of Social and Clinical Psychology, 24, 860–875.

15.

Finkel

E. J.

Rusbult

C. E.

Kumashiro

Hannon

P. A.

(2002). Dealing with betrayal in close relationships: Does commitment promote forgiveness? Journal of Personality and Social Psychology, 82, 956–974.

16.

Finsterwalder

Yee

Tombs

(2017). Would you forgive Kristen Stewart or Tiger Woods or maybe lance Armstrong? Exploring consumers’ forgiveness of celebrities’ transgressions. Journal of Marketing Management, 33, 1204–1229.

17.

Fraser

B. P.

Brown

W. J.

(2009). Media, celebrities, and social influence: Identification with Elvis Presley. Mass Communication and Society, 5, 183–206.

18.

Greenwood

D. N.

(2008). Television as escape from self: Psychological predictors of media involvement. Personality and Individual Differences, 44, 414–424.

19.

Hayes

A. F.

(2017). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New York, NY: Guilford Press.

20.

Horton

Wohl

R. R.

(1956). Mass communication and para-social interaction: Observations on intimacy at a distance. Psychiatry: Journal for the Study of Interpersonal Processes, 19, 215–229.

21.

Young

Liang

Guo

(2017). An investigation into audiences’ reactions to transgressions by liked and disliked media figures. Psychology of Popular Media Culture, 7, 484–498.

22.

Karremans

J. C.

Van Lange

P. A. M.

(2004). Back to caring after being hurt: The role of forgiveness. European Journal of Social Psychology, 34, 207–227.

23.

Kosenko

K. A.

Binder

A. R.

Hurley

(2016). Celebrity influence and identification: A test of The Angelina Effect. Journal of Health Communication: International Perspectives, 21, 318–326.

24.

Lather

Moyer-Guse

(2011). How do we react when our favorite characters are taken away? An examination of a temporary parasocial breakup. Mass Communication and Society, 14, 196–215.

25.

McCornack

S. A.

Levine

T. R.

(1990). When lies are uncovered: Emotional and relational outcomes of discovered deception. Communications Monographs, 57, 119–138.

26.

McCullough

M. E.

Fincham

F. D.

Tsang

J. A.

(2003). Forgiveness, forbearance, and time: The temporal unfolding of transgression-related interpersonal motivations. Journal of Personality and Social Psychology, 84, 540–557.

27.

McCullough

M. E.

Rachal

K. C.

Sandage

S. J.

Worthington

E. L.

Jr Brown

S. W.

Hight

T. L.

(1998). Interpersonal forgiving in close relationships: II. Theoretical elaboration and measurement. Journal of Personality and Social Psychology, 75, 1586–1603.

28.

Paleari

Regalia

Fincham

F. D.

(2003). Adolescents’ willingness to forgive their parents: An empirical model. Parenting, 3, 155–174.

29.

Riek

B. M.

Mania

E. W.

(2012). The antecedents and consequences of interpersonal forgiveness: A meta-analytic review. Personal Relationships, 19, 304–325.

30.

Rubin

R. B.

McHugh

M. P.

(1987). Development of parasocial interaction relationships. Journal of Broadcasting and Electronic Media, 31, 279–292.

31.

Rubin

R. B.

Perse

E. M.

Powell

R. A.

(1985). Loneliness, parasocial interaction, and local television news viewing. Human Communication Research, 12, 155–180.

32.

Rusbult

C. E.

Verette

Whitney

G. A.

Slovik

L. F.

Lipkus

(1991). Accommodation processes in close relationships: Theory and preliminary empirical evidence. Journal of Personality and Social Psychology, 60, 53–78.

33.

Ryzik

Buckley

Kantor

(2017, November). Louis C.K. is accused by 5 women of sexual misconduct. The New York Times. Retrieved from http://www.nytimes.com

34.

Sanderson

Emmons

(2014). Extending and withholding forgiveness to Josh Hamilton: Exploring forgiveness within parasocial interaction. Communication and Sport, 2, 24–47.

35.

Schumann

(2012). Does love mean never having to say you’re sorry? Associations between relationship satisfaction, perceived apology sincerity, and forgiveness. Journal of Social and Personal Relationships, 29, 997–1010.

36.

Semuels

(2018, January). The internet is enabling a new kind of poorly paid hell. The Atlantic. Retrieved from www.theatlantic.com

37.

Slater

M. D.

Ewoldsen

D. R.

Woods

K. W.

(2018). Extending conceptualization and measurement of narrative engagement after-the-fact: Parasocial relationship and retrospective imaginative involvement, Media Psychology, 21, 329–351.

38.

Tsang

McCullough

M. E.

Fincham

F. D.

(2006). The longitudinal association between forgiveness and relationship closeness and commitment. Journal of Social and Clinical Psychology, 25, 448–472.

39.

Thrall

A. T.

Lollio-Fakhreddine

Berent

Donnelly

Herrin

Paquette

… Wyatt

. (2008). Star power: Celebrity advocacy and the evolution of the public sphere. Politics, 13, 362–385.

40.

Vary

A. B.

(2017, October). Actor Anthony Rapp: Kevin Spacey made a sexual advance toward me when I was 14. BuzzFeed News. Retrieved from www.buzzfeed.com

41.

Way

(2018, January). I went on a date with Aziz Ansari. It turned into the worst night of my life. Babe. Retrieved from www.babe.net

42.

Weiss

(2018, January). Aziz Ansari is guilty. Of not being a mind reader. The New York Times. Retrieved from http://www.nytimes.com

43.

Zacharek

Dockterman

Edwards

H. S.

(2018, January). Time person of the year 2017: The silence breakers. Time. Retrieved from http://www.time.com

44.

Zechmeister

Garcia

Romero

Vas

S. N.

(2004). Don’t apologize unless you mean it: A laboratory investigation of forgiveness and retaliation. Journal of Social and Clinical Psychology, 23, 532–564.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB

	Study 1		Study 2
	M (SD)	N	M (SD)	N
Forgiveness	4.66 (1.54)	185	3.12 (1.43)	184
Pre-offense closeness	3.09 (1.80)	184	2.88 (1.68)	184
Current closeness	2.58 (1.83)	184	1.70 (1.27)	184
Intent/responsibility	4.61 (1.61)	185	5.86 (0.96)	184
Subjective severity			5.54 (1.32)	184
Apology sincerity			2.70 (1.18)	66
PSI			2.67 (1.33)	184