• Aucun résultat trouvé

Chapitre 4 The interplay of trait-related antecedents and state empathy: Empathic responding to anti-

4.1 Résumé

Les stratégies d’intervention qui se basent sur l’empathie pour lutter contre la discrimination se sont avérées efficaces pour réduire les stéréotypes négatifs et pour améliorer les relations intergroupes. Cependant, très peu est connu sur les caractéristiques individuelles ou les traits de personnalité qui peuvent influencer positivement une réponse empathique dans le contexte de la réception publicitaire. Dans cette étude, les antécédents inhérents à l’individu (c.-à-d. empathie dispositionnelle et le sexe) de la réponse empathique (cognitive et affective) aux publicités sociales anti-discrimination ont été examinés. Les résultats suggèrent que le sexe joue un rôle important dans les réponses empathiques affectives, mais qu’il n’a pas d’influence sur les réponses empathiques cognitives. Aucune association significative n’a été observée entre l’empathie dispositionnelle et les réponses empathiques des participants, bien qu’une association marginale ait été observée entre la dimension cognitive de l’empathie dispositionnelle et la réponse empathique cognitive. Cet article fournit des preuves supplémentaires sur l’importance des caractéristiques personnelles pour favoriser les réactions empathiques envers l’exogroupe dans le contexte spécifique de la réception publicitaire.

Mots clés : réponse empathique, empathie dispositionnelle, discrimination, publicités sociales,

relations intergroupes, évaluation continue

4.2 Abstract

Intervention strategies that rely upon empathy to tackle discrimination have shown promise in reducing negative stereotyping and improving intergroup relations. However, very little is known about individual characteristics or traits that can positively influence an empathic response in the media reception context. In this study, trait-related antecedents (gender and trait empathy) of empathic responses (cognitive and affective) to anti-discrimination public service announcements (PSAs) were examined. Results suggest that gender seems to play a significant role in the affective empathic responses but does not seem to have an impact on the cognitive empathic responses. No significant associations were observed between trait empathy and the participants’ empathic responses, although a marginal

association between cognitive trait empathy and the cognitive response was observed. This article provides additional evidence on the importance of personal characteristics in fostering empathic reactions toward the outgroup, particularly in the media reception context.

Keywords: state empathy, trait empathy, discrimination, public service announcements, intergroup

relations, continuous response measurement

Despite decades of efforts to eradicate discrimination and prejudice and interventions to raise awareness of these issues, they are still part of everyday life of most societies (Feagin, 2010; Whitley & Kite, 2013). In light of recent events (anti-immigration protests around the world, racial tensions in the USA, racial profiling and race-related deaths in Canada and the US, and so on) there is an urgent need for more effective interventions that promote social cohesion and harmonious intergroup relations. One approach that has yielded promising results is to foster empathy for discriminated groups (Batson & Ahmad, 2009; Dovidio, Johnson, et al., 2010; Eisenberg et al., 2010; Finlay & Stephan, 2000; McKeever, 2015; Pedersen, Beven, Walker, & Griffiths, 2004; Stephan & Finlay, 1999). Several explanations have been proposed to account for the effects of empathy on intergroup relations. For instance, some authors suggest that the effects of perspective taking- which corresponds to the cognitive dimension of empathy and involves adopting the perspective of another person or putting oneself in that person’s shoes - have shown to have significant and positive effects on prejudice reduction (Galinsky & Ku, 2004; Galinsky & Moskowitz, 2000; Vescio et al., 2003; Weyant, 2007). Research has also shown that perspective taking can help overcome a multitude of cognitive barriers that impede attempts to improve intergroup attitudes and relations (Davis & Maitner, 2010; Wang et al., 2014; Yzerbyt, Dumont, Gordijn, & Wigboldus, 2002). According to Stephan and Finlay (1999), one of the main benefits of perspective taking is that it “leads people to see that they are less different from members of the other group than they thought they were. It may also lead them to perceive that they themselves and members of the other group share a common humanity and a common destiny” (p.735). This, in turn, can lead to a greater self-other merging (Pagotto, 2010). Among intervention tools widely used to promote and to raise awareness of certain behaviors, public service announcements (PSAs) are a promising avenue for changing attitudes and behaviors and are often an integral part of larger interventions aimed at improving individual and collective well-being (Abroms & Maibach, 2008; Atkin & Rice, 2013b; Vrij & Smith, 1999). In the context of anti-discrimination strategies, this type of advertising has the potential to induce empathic responses towards discriminated groups (Donovan & Leivers, 1993; Pedersen et al., 2004). However, in order to enhance the probability that

an empathic reaction is induced, the development of the PSA messages needs to go through various stages, the first of which is the formative research stage where all the necessary information about the targeted audiences is collected (Atkin & Freimuth, 2013). In addition, adequate message tailoring and audience segmentation require a deeper knowledge of individual characteristics that are likely to have an effect on fostering an empathic response. Our research aims to uncover which personal characteristics are likely to have an impact on an empathic response in the context of anti- discrimination PSA reception.

4.3 Theoretical Framework

Empathy as a concept continues to stir debate among those who venture into this field (Bagozzi & Moore, 1994; Davis, 1996; Duan & Hill, 1996; Reniers, Corcoran, Drake, Shryane, & Völlm, 2011; Zillman, 2013). The main issue lies in the dimensionality of the concept. Consequently, some researchers have defined empathy in purely cognitive terms, placing emphasis on deliberate cognitive processes such as perspective taking (Dymond, 1950; Ickes, 1993; Katz, 1963; Weyant, 2007). Others have conceived this concept as a solely affective phenomenon, ascribing purely emotional reactions to empathic responding (e.g., having the same or similar affective reactions as the person with whom one is empathizing, feeling sympathy, concern, tenderness, etc. for the needy other) (Batson, 2011; Mehrabian, Young, & Sato, 1988; Stotland, 1969; Zillman, 2013). Despite this lack of consensus, there is general agreement among the contemporary empathy researchers that these two dimensions characterize empathy (Chory-Assad & Cicchirillo, 2005; Davis, 2006; Duan & Hill, 1996; Mooradian, Davis, & Matzler, 2011; Shen, 2010a). There is the cognitive dimension or cognitive empathy, which is commonly characterized by the cognitive process of perspective taking and an affective dimension, which entails an emotional reaction similar to or concordant with the target’s experience (Lietz, Gerdes, Sun, Geiger, & Alex, 2011; Strayer, 1987). Additionally, it is important to note that there are usually two lines of inquiry about empathy – the study of trait or state empathy. On the one hand, researchers have studied empathy as a relatively stable personality trait (Davis, 2006; Mooradian et al., 2011). In the media reception context, this can be translated as a personality trait that “shapes responsiveness to a message” (Campbell and Babrow, 2004, p.162). For instance, it is assumed that individual differences in trait empathy may influence the arousal of responses and processes related to empathy (Davis, 1983b, 1996; Eisenberg et al., 1989; Hall & Bracken, 2011). For example, in the media viewing context, Chory-Assad and Cicchirillo (2005) found cognitive trait empathy to be associated with viewers’ identification with media characters. In research on responses to graphic horror movies, scholars have

argued that receivers with higher dispositional affective empathy were more likely to experience emotions of fear and distress in response to the suffering of others (Hoffner, 2009; Tamborini, Stiff, & Heidel, 1990). Similarly, Hoffner and Levine (2005) found affective trait empathy to be negatively correlated with the enjoyment of violence and graphic horror. In addition, it has also been hypothesized that the ability and tendency to spontaneously take the perspective of others is an important antecedent or precursor of empathic responding (Eisenberg et al., 1991). In the studies that bear on the issue of discrimination and negative stereotyping, it is also assumed that individuals who have higher levels of trait empathy are less likely to make negative judgments about the members of an outgroup (Batson & Ahmad, 2009; Eisenberg et al., 2010; Hansson et al., 1978). From this perspective, empathy toward a media (PSA) protagonist is assumed to be aroused if a receiver has a natural tendency to adopt another’s perspective and have an affective reaction similar to the one that the other person is experiencing.

Also with regard to trait empathy, some studies have shown that females tend to score higher on the affective trait empathy scale and are more likely to experience empathic concern over others’ misfortunes (Davis, 1996; Trobst, Collins, & Embree, 1994). The empathy-as-a-state approach, on the other hand, allows researchers to further investigate the effects of situational factors or stimuli leading to an empathic response (Campbell, 1998; Escalas & Stern, 2003). However, to the best of our knowledge, no previous research has focused on the association between personal characteristics such as gender and dispositional (trait) empathy and empathic (state) responses to anti-discrimination PSAs. Furthermore, most of the previous research on empathic responses has used situational manipulations (e.g., instructions to put oneself in the needy person’s shoes and/or to imagine how they feel) to induce an empathic response (Batson, Early, et al., 1997; Pagotto, 2010; Toi & Batson, 1982). This highlights the need to uncover trait-related antecedents other than instructions to take the perspective of another that may contribute to triggering an empathic reaction towards anti- discrimination media content. In fact, previous studies have highlighted the importance of trait-related antecedents (i.e. trait empathy, gender) in audience segmentation and message-tailoring (Shen, 2010b). Therefore, the results of the current study may help shed some light on audience features that are most likely to play a role in triggering empathic reactions to anti-discrimination PSAs.

In the field of psychology of emotions, researchers have generally taken into account two types of antecedents that may have an impact on triggering an emotional response: 1) antecedents that are linked to individual characteristics (personality traits or dispositions); and 2) antecedents that are

related to situational factors (stimulus related) (Ellsworth, 2009; Izard, 1991; Lazarus, 1991). In the current study, however, we focus solely on individual characteristics likely to play a role in fostering an empathic reaction towards anti-discrimination PSAs.

Although some researchers have studied empathy in the media context (Campbell & Babrow, 2004; Chory-Assad & Cicchirillo, 2005; McKeever, 2015; Shen, 2010a, 2011), in most cases it was studied in the context of direct interactions. Admittedly, an interaction between individuals is not of the same nature as an interaction between a receiver and media content. In order to accurately capture the particularity of being empathetic toward media protagonists the term “virtual empathy” was coined (Daignault, 2007; Daignault & Paquette, 2009; Reny Delisle, 2013). Virtual empathy takes into consideration the particularities of an interaction between a receiver and mediated content. It is a type of empathy that, while entailing no substantive difference in meaning from the psychological construct regarding its underlying processes, is distinguished by the (media) context in which it is embedded. Consequently, virtual empathy reflects the ability of a receiver to take the perspective of a protagonist (or an individual presented in the media), to project oneself into the other’s situation, to feel and understand the protagonist’s emotions all the while maintaining an awareness of the distinction between the self and the other (Daignault, 2007). This innovative concept of empathy – specific to the field of media communications – is therefore better adapted to the study of its effects in persuasion. A great volume of research has been dedicated to gender differences in trait empathy and the results tend to show a higher propensity among females for this disposition (Davis, 1996; Eisenberg & Lennon, 1983; Hoffman, 1977; Strauss, 2004; Toussaint & Webb, 2005). In a study on gender differences in empathic responses to violent media content, Kobach and Weaver (2012) found that females reported having stronger reactions than males and seemed to be equally able to empathize both with real and fictional content. In the same vein, a study on empathic responses to video stimuli found that women were more likely to be concerned for people in negative situations (Kuypers, 2017). In the research field of altruism and charitable giving, it has also been concluded that higher trait empathy scores in women are one of the reasons why women are more prone to prosocial behavior (Batson, 2011; Eisenberg & Miller, 1987; Smith, 2003).

Very little research has examined gender differences and intergroup attitudes and relations. One study that examined the association between gender differences and attitudes towards ethnic minorities, found no difference between genders (Hughes & Tuch, 2003). To our knowledge, we are the first to

examine the association between gender and empathic responses generated by anti-discrimination PSAs. This is an important characteristic to keep in mind for segmenting and targeting different populations when designing anti-discrimination interventions (French, 2017; Lotenberg et al., 2011). S The objective of the current paper is to explore trait-related antecedents of empathic responding in the context of anti-discrimination PSA reception. More specifically, this study is aimed at exploring the association between the (virtual) trait empathy and gender and their effects on the empathic response (state empathy).

4.4 Method

4.4.1 Participants

Sixty-one participants took part in this study. All of the participants were of European Canadian descent, Canadian citizens and born in Canada. In total, twelve viewing groups were conducted. The participants were mainly recruited through the local university’s mailing list. We also resorted to Kijiji – an online classified advertising service in the province of Quebec, Canada. Participants were compensated 15$ (CAN) for their participation in the study. The study was approved by the local university’s Ethics Committee and all the participants signed the consent form.

4.4.2 Stimulus material

The stimulus material consisted of 16 PSAs aimed at tackling discrimination. To obtain this set of advertisements, we initially identified 32 PSAs (some of which pertained to the same advertising campaign) on the Internet using keywords such as: anti-discrimination campaigns, anti-racism

campaigns, anti-discrimination PSAs, pro-immigration campaigns, pro-diversity PSAs, etc. We subsequently eliminated 16 PSAs that were judged to be not enough affectively laden. In the second step, three independent experts in communication and emotion rated the remaining PSAs in order to determine the valence of the conveyed emotions, the empathy arousal potential of the PSAs and their intelligibility. The first indicator - the valence of the emotions and the advertising tone – were evaluated according to the facial expressions that the PSA protagonists conveyed, since facial expressions are considered to be one of the main communicative signals of emotional states (Ekman, 1993; Ekman & Friesen, 1978; Ekman et al., 1972; Ekman & Keltner, 1997; Harrison, Wilson, & Critchley, 2007). In addition to the facial expressions, the raters also examined the verbal cues (e.g., Words can kill; I am

exacerbating a positive or negative emotional advertising tone (Morris & Boone, 1998). The second indicator was labelled empathy potential which referred to the advertising parameters that could facilitate adopting the protagonist’s perspective and understating and sharing their emotions (e.g., first person point of view). The third and final indicator was the PSA intelligibility, which corresponds to the PSA’s clarity and elements that make it easy to understand. The three coders compared their evaluations for each PSA and discussed their analyses to determine a final classification by reaching a consensus. A fourth coder was involved when an agreement could not be reached. Ten of the selected PSAs presented positive scenarios and portrayals of discriminated groups (e.g., the positive contributions of ethnic minorities to the host society, similarities between the ingroup and the outgroups, etc.), and 6 presented negative scenarios and the negative consequences of discrimination (e.g., the negative impacts of discrimination on personal and social welfare, etc.). Within this set of PSAs, 10 were printed ads and 6 were videos, we added French subtitles to all of the PSAs that were in a language other than French. The stimulus material is presented in the Appendix.

4.4.3 Measures

4.4.3.1 Virtual empathy (trait)

The virtual empathy questionnaire assessed the affective and cognitive dimensions of empathy. This previously validated instrument (Reny Delisle, 2013), aimed at measuring trait empathy, is a combination of two multidimensional empathy scales: the Interpersonal Reactivity Index (IRI) (Davis, 1983a) and the Empathy Assessment Index (EAI) (Lietz, Gerdes, Sun, Geiger, & Alex, 2011). The study questionnaire comprised 60 items – 30 to assess the affective dimension of empathy (e.g., When

I watch a TV PSA depicting drunk driving, domestic violence, discrimination, etc. I can feel the protagonists’ emotions) and 30 to assess the cognitive dimension (e.g., When I watch a TV PSA

depicting drunk driving, domestic violence, discrimination, etc. I have a difficult time imagining how I would feel if the same things were happening to me). These constructs were assessed using a 5-point Likert-type scale. The study questionnaire also gathered information on demographic items (age and gender). Participants self-reported gender and we coded it as 0 (male) and 1 (female).

4.4.3.2 State empathy (empathic responses)

The intensity of state empathy as a response to the exposure to the stimulus material was assessed with a continuous response measurement (CRM) technique - a type of system designed for measuring moment-by-moment responses to dynamic media content (Biocca, David, & West, 1994; West &

Biocca, 1996). In this way, the intensity of the emotional responses is reported by participants on a 0 to 100 scale using a handset. The instrument is equipped with a screen that gives visual feedback to the participants by indicating the scores in real time (e.g., from 0 to 100). This measurement technique collects data one per second per participant as the participants are viewing the PSAs (West & Biocca, 1996). Since it has been suggested that retrospective message evaluations (e.g., self-reported questionnaires) may not capture the complexity of persuasive communication processes and the spontaneity of participants’ responses, this technique is better adapted for capturing changes in psychological states during media reception (Gunter, 2000; Yzer, Vohs, Luciana, Cuthbert, & MacDonald, 2011).

4.4.4 Procedure

Upon arrival, participants were asked to complete the study questionnaire. After practicing with the CRM handset while watching a 30-second anti-discrimination test PSA, participants were exposed to the study’s anti-discrimination PSAs. The participants were exposed to the same set of PSAs twice – once in order to assess the cognitive dimension of their empathic responses and once to assess their affective responses. To minimize the habituation effects with the stimulus material, the sequence of the PSAs was randomly changed between the two viewings. To evaluate the cognitive dimension, the participants were given the following instructions: “As the PSA progresses, use the rotary button from

<0> (minimum intensity) to <100> (maximum intensity) to indicate to what extent you are able to put yourself in the discriminated person’s shoes”. The assessment instructions for the affective dimension of empathy read as follows: “As the PSA progresses, use the rotary button from <0> (minimal intensity)

to <100> (maximum intensity) to indicate to what extent you are able to feel the discriminated person’s emotions”. Participants were instructed to keep the rotary button at zero <0> if they didn’t feel anything or if they simply could not put themselves in the discriminated person’s shoes. The poster campaigns were shown for 15 seconds and to simplify the procedure, since poster campaigns are not dynamic in nature, the participants were instructed to indicate only the highest intensity felt during exposure. To avoid an answer pattern, the order in which the dimensions were assessed was switched up between groups. Although the CRM collects data second per second, we only considered the highest assessment values that the participants indicated since we are mainly interested in the intensity of the empathic responses generated by the stimuli and the participants’ capacity to experience an intense empathic reaction.

4.4.5 Statistical procedure

Regarding the virtual empathy concept, three distinct scores per participant were calculated: 1) a score pertaining to the affective dimension of empathy, which corresponds to the average of the answers given to the 30 questions relating to this dimension (Cronbach’s α = .88), 2) a score pertaining to the cognitive dimension of empathy, the average of the answers to the 30 questions on perspective taking (α = .71); and 3) a global score of virtual empathy calculated from the average of all the answers of the