Mechanisms in narrative persuasion

Available online at www.sciencedirect.com

ScienceDirect Journal of Consumer Psychology 27, 1 (2017) 11 – 22

Research Article

Reflecting on the journey: Mechanisms in narrative persuasion Anne Hamby a,⁎, David Brinberg b , Kim Daniloski b a b

Zarb School of Business, Hofstra University, 124 Weller Hall, Hempstead, NY 11550, USA Pamplin College of Business, Virginia Tech, 2016 Pamplin Hall, Blacksburg, VA 24061, USA Accepted by Amna Kirmani, Editor; Associate Editor, Hans Baumgartner Received 12 January 2015; received in revised form 22 June 2016; accepted 23 June 2016 Available online 1 July 2016

Abstract Recent work demonstrates that narratives persuade via mechanisms distinct from other persuasive message formats. The present work draws from the discourse processing and communication literature to introduce a construct of retrospective reflection as an additional mediator in narrative persuasion. Retrospective reflection represents self or other-relevant memories evoked by transportation into a story, which corroborates and extends story-implied beliefs into the reader's world. The reported studies indicate that retrospective reflection is distinct from transportation, mediates the relationship between transportation and various persuasion-related outcomes, and predicts these outcomes beyond transportation. The current work also examines the influence of personal relevance (Study 2) and cognitive load (Study 3) to better understand the role of retrospective reflection in narrative persuasion. © 2016 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved. Keywords: Narrative; Transportation; Reflection; Persuasion

Stories play a central role in human life. We actively seek out and pay for story-based entertainment, but we dread didactic, expository lectures of the same length. The internet's growth has facilitated the exchange of opinions between large numbers of geographically widespread consumers who often describe experiences in a story format with rich descriptions of actors and scenes. Businesses frequently present their “Brand Story” as a way to deepen connections with customers, and advertisements often feature stories displaying a product. Narratives' presence in daily life has long been recognized, but only recently have contemporary scholars in psychology, marketing, and communication focused on understanding narratives' persuasive influence. Transportation theory (Green & Brock, 2000) has been the dominant perspective to account for narrative persuasion, and describes a process entailing emotional engagement, generation of mental imagery, and attention. Research has demonstrated that the transported “traveler” can return changed ⁎ Corresponding author. E-mail addresses: [email protected] (A. Hamby), [email protected] (D. Brinberg), [email protected] (K. Daniloski).

by the reading experience and is more likely to believe statements or positions illustrated in the story, in some cases despite knowledge that the story is fictional (Van Laer, De Ruyter, Visconti, & Wetzels, 2014). Questions remain regarding how transportation achieves its persuasive effects. Transportation encapsulates the trip into the story world; what causes the audience member's reemergence into the real world as a changed individual? Readers bring their real-world knowledge and beliefs to help them comprehend and become transported into the story (Zwaan & Radvansky, 1998). If stories do not persuade through elaboration on the strength of arguments, as studies have demonstrated (Escalas, 2007), what mechanisms underlie the conversion of the story-world experience into one's beliefs about the real world? We posit and explore a cognitively based pathway linking transportation to persuasion-related outcomes. We describe this mechanism as a retrospective reflection process that integrates elements encountered in the story into one's understanding of the real world. The present work contributes to the field of narrative persuasion by introducing retrospective reflection as a mechanism

http://dx.doi.org/10.1016/j.jcps.2016.06.005 1057-7408/© 2016 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.

12

A. Hamby et al. / Journal of Consumer Psychology 27, 1 (2017) 11–22

to understand how transportation influences persuasion-related outcomes. This research represents an effort to define and measure the retrospective reflection process, building a more comprehensive model of narrative persuasion. The paper proceeds as follows: we first review past narrative persuasion research in consumer psychology to provide background for the construct of retrospective reflection. We then define retrospective reflection and distinguish it from related constructs. Finally, we develop a model relating transportation, retrospective reflection, and persuasion-related outcomes, and evaluate this model by testing the influence of personal relevance and cognitive load on the transportation and retrospective reflection processes in narrative persuasion. Brief overview of narrative persuasion research in consumer psychology Social information is often shared and acquired by consumers using a narrative (Adaval & Wyer, 1998). A narrative format describes events in a chronological sequence, relating information about characters engaged in actions (Bruner, 1986). This format contrasts with messages that persuade by an appeal to arguments and are decontextualized (i.e., have no reference to actors or time; Deighton, Romer, & McQueen, 1989). Several recent meta-analytic reviews of narrative persuasion point to a growing body of empirical evidence in the fields of consumer psychology and marketing that highlights narratives as a persuasive communication platform (Van Laer et al., 2014; Shen, Sheer, & Li, 2015). Many of these efforts have examined narrative transportation—“an integrative melding of attention, imagery, and feelings, focused on story events” (Green, 2004, p. 247)—as the mechanism that underlies narrative persuasion, as well as the context factors (e.g., Wang & Calder, 2006) and individual differences (Argo, Zhu, & Dahl, 2008; McFerran, Dahl, Gorn, & Honea, 2010) that influence the evoked level of transportation. Work on narrative persuasion demonstrates that consumers respond favorably to stories because of the pleasurable nature of narrative processing (Van Laer et al., 2014) and willingly expend cognitive effort to engage with them (Nielsen & Escalas, 2010). Past work has found that responses to narrative ads were unaffected by argument strength and that narratives reduced counter-arguments compared to an argument-based message (Escalas, 2007). Narratives are processed more holistically than argument-based messages, which affects information integration and response to interrupting stimuli. Adaval and Wyer (1998), for example, found negative information exerted less effect when embedded in a narrative than identical information in an argument-based format, which evokes a piece-meal form of processing. When narrative processing is interrupted, this disturbance can have a negative influence on persuasion-related outcomes. For example, Wang and Calder (2006, 2009) found ads that interrupt transportation when placed in the middle of an entertainment narrative were evaluated negatively, but ads at the end of the transportation experience benefit from the transference of positive affect associated with the transportation experience. In contrast, Nelson, Meyvis, and Galak (2009) found interrupting

ads reduced hedonic adaptation to the narrative (enjoyment declining over time), but enhanced the overall narrative experience. Collectively, this work demonstrates that persuasion via narratives is a nuanced process, distinct from persuasion via arguments or advocacy messages, and important because of narratives' prevalence and influence on consumers' lives. Conceptual framework Research in psychology, consumer behavior, and communication provides support for the influence of transportation on persuasion-related outcomes and has explored factors that strengthen or attenuate its influence. “What is still lacking… are answers to the question of how transportation affects persuasion” (Appel & Malečkar, 2012 p. 26). The studies reported in this paper evaluate a model that incorporates a new construct—retrospective reflection—to connect engagement with a narrative and the outcomes typically associated with narrative persuasion. Retrospective reflection: A process to create meaning We propose the reader recalls past episodes related to the story and connects the story-world to the reader's “real” world to better understand and corroborate the meaning of the narrative. We define this process as retrospective reflection, the recall of self or other-relevant memories evoked by transportation into a story, which validate and extend story-implied beliefs into the reader's world. For example, Mick and Buhl (1992) examined how personal background (memories described as “life themes”) guided the meaning derived from visual narrative ads. They found the meanings derived from these ads were affected by “the attempt to make the (story) world congruent with their own lives” (Mick & Buhl, 1992, p. 333). That is, retrospective reflections evoked by the message affected the interpretation of the message's meaning. These reflections included both experiences that happened to the person directly and stories heard from others, such as experiences shared by friends (Woodside, Sood, & Miller, 2008). Past work has shown meaning is derived distinctly from arguments and narratives. Deighton et al. (1989) and Escalas (2007) each compared argument and narrative-based messages and found different facets are emphasized for persuasion (e.g., strong arguments for rhetorical messages versus story verisimilitude for narratives). Collectively, these studies provide evidence a reflective process about the self and others is used to derive meaning and differs for narrative and argument-based messages. This process of sense-making is also consistent with Fisher's narrative paradigm (1989), in which people understand others and justify their own behavior by telling a plausible story rather than producing evidence or constructing rational arguments. According to this paradigm, people make decisions and accept stories based on “good reasons” that fit with their narrative explanation rather than rational tests of objective facts. For example, a student may be more likely to sign up for a study-abroad trip after reading about another student's experience than after reading statistics


about how study abroad enhances job prospects. Busselle and Bilandzic (2008) note the distinction between narrative rationality and external realism: “the power of narrative is not diminished by readers' knowledge that the story is invented…those that engage us most often are both fictional and unrealistic” (2008, p. 256). Stories need not be “real” to influence, but they must correspond with the reader's beliefs and experiences to be accepted. Retrospective reflection is a process through which this correspondence is created by bringing to mind similarities between the story world and observed experience (personal or mediated). Past research provides indirect evidence to support the process of retrospective reflection. In discourse psychology research, for example, individuals recall story-congruent memories in response to a story and create linkages between the characters in the story and these memories (Larsen & Seilman, 1988). Larsen and László (1990) describe this process as “personal resonance,” in which textual elements remind readers of real-world encounters and enhance story enjoyment. Similarly, Strange and Leung (1999) describe how “evoked reminding of related experiences in a reader's personal or mediated past” (p. 437) frames later thinking about the story and lends the story subjective credibility. In persuasion research, Dunlop, Wakefield, and Kashima (2009) found transportation into an anti-smoking narrative influenced intention to quit smoking through a reflection process (e.g., “To what extent were you reminded of your own experiences while viewing the ad?”). The process of “reminding” and personal resonance each highlight the importance of memories of the self and others in narrative persuasion. In the next section, we describe and distinguish self-referencing processes from retrospective reflection.

13

events while mixing in hypothetical elements” (Escalas, 2007, p. 422). Narrative self-referencing and retrospective reflection differ on several features. The former occurs as an antecedent to transportation, whereas the latter occurs subsequent to and as a consequence of entering a story world. Narrative self-referencing involves the creation of a story about the self and influences via affective transference; retrospective reflection includes the recall of memories related to the self and others, such as relating a story one has read to an experience recounted by a neighbor. The function of each process in narrative persuasion also differs. Narrative self-referencing allows an individual to create a story, which evokes positive affect. Retrospective reflection involves the recall of self- and other-related memories, and provides a foundation to contrast the individual's mental model of the narrative (Zwaan & Radvansky, 1998) with the individual's mental model of the “real world.” Beliefs implied in the story are accepted as valid by corroborating the model of the story with the model of the real world (Larsen & Seilman, 1988) through retrospective reflection (Hamby & Brinberg, in press). Narrative self-referencing serves to pull oneself into or help generate the story world. Development of hypotheses Our model views a narrative as prompting transportation into a story, which evokes retrospective reflection on that story, which then enhances persuasion. The focus of the current work is to measure and assess this process, and evaluate retrospective reflection as an additional mechanism in narrative persuasion (see Fig. 1).

Distinguishing retrospective reflection from self-referencing Self-referencing focuses on relating information to the self (Burnkrant & Unnava, 1995) and provides potential linkages between words in a narrative and related structures in memory. The increased availability of one's own experiences increases self-related cognitive responses (i.e., generation of thoughts about oneself). Self-referent thoughts are evoked by a range of stimuli (such as argument-based messages and images of individuals similar to the message receiver; Debevec & Romeo, 1992), facilitate the linking of thoughts about the stimulus to the self, and persuade through dual-route cognitive response processes (Burnkrant & Unnava, 1995). Escalas (2007) distinguished between analytical self-referencing and narrative self-referencing, describing the latter as an individual's recall or generation of autobiographical memories in a story format. Narrative self-referencing evokes a mechanism of affective transference; that is, the positive experience of thinking about the self enhances persuasion-related outcomes. Narrative self-referencing is an antecedent to entering the story world (e.g., one may be transported into one's self-focused story) and entails the generation of a story about oneself (West, Huber, & Min, 2004; Hung & Wyer, 2011) in “rehearsals of likely future events, fantasizing about less likely future events, realistically re-experiencing past events, or reconstructing past

Transportation and retrospective reflection Fisher (1989) argued that, as born storytellers, we naturally match a story to our own beliefs and experiences to establish its fidelity and to learn how the story can guide our future actions. This matching occurs reflexively and illustrates a retrospective reflection process; we consider how the values in a narrative are consistent with our own values to determine how the story can serve as a guide for our actions. Fisher contrasts narrative fidelity with the traditional paradigm of the rational world, in which people are logical and make decisions on the basis of evidence and lines of argument. Unlike the critical elaboration that occurs in response to argument-based messages and entails active evaluations, narrative processes (both transportation and retrospective reflection) are not experienced as effortful (Nielsen

Fig. 1. Conceptual framework.

14


H3. Retrospective reflection will mediate the relationship between transportation and persuasion-related outcomes.

Personal relevance Personal relevance has been examined as an important determinant of motivation to process information contained in a rhetorical message (Celsi & Olson, 1988): “Only when the message addresses an issue of personal relevance is the receiver likely to process the information carefully” (Ajzen, Brown, & Rosenthal, 1996, p. 44). Retrospective reflection differs from personal relevance in that the former is a process of message interpretation while the latter represents a state of correspondence between the target message and self. Numerous studies demonstrate the limited effect of personal relevance on transportation or absorption into a narrative because readers suspend disbelief to be transported into a narrative (see Green, 2004 for an exception to this finding). Correspondence between a protagonist and reader on various facets (e.g., attending the same school as the reader: de Graaf, 2014; Quintero Johnson, Harrison, & Quick, 2013; Wheeler, Green, & Brock, 1999) does not affect transportation. Personal relevance, however, does influence persuasion-related outcomes (Prentice, Gerrig, & Bailis, 1997; Green, Garst, Brock, & Chung, 2006), suggesting it may influence other aspects of the narrative persuasion process. In the argument-based literature, personal relevance moderates the relationship between argument strength and elaboration. One is motivated to process a message when personally relevant, and the relationship between argument strength and elaboration is stronger than when the message is less relevant (Petty & Cacioppo, 1986). This finding may extend to the context of narratives, though the processes differ. Humans prefer consistency and are motivated to reconcile past experiences with new information (Abelson, 1968). The more personally relevant the story, the more motivated one will be to create consistency between the story and surrounding world. This motivation strengthens the relationship between transportation and retrospective reflection, which involves matching the mental model of the narrative with the mental model of the reader's world. In other words, we propose personal relevance will enhance the transportation—retrospective reflection relationship because one will be more motivated to create consistency by reconciling the mental model evoked through transportation with the real world. One could argue personal relevance would reduce the strength of the relationship between transportation and retrospective reflection. There is less need to reflect on the familiar (i.e., highly personally relevant) to comprehend its meaning, but retrospective reflection may facilitate message understanding when personal relevance is low. However, we propose that retrospective reflection is not evoked simply to understand the meaning of the narrative. It is a process of validation or corroboration by which one substantiates the stance of the narrative by linking the mental model of the narrative and the mental model of the reader's real world. We propose:

Beyond the test of our basic model, we will incorporate personal relevance to evaluate a more comprehensive model of narrative persuasion (Study 2) and will apply a cognitive load to demonstrate retrospective reflection and transportation are distinct constructs (Study 3).

H4. Personal relevance moderates the relationship between transportation and retrospective reflection. When personal relevance is high, the relationship between transportation and retrospective reflection is stronger than when personal relevance is low.

& Escalas, 2010) because people naturally use narratives to create meaning (Fisher, 1989). The more transported one becomes into a story world, the more a comprehensive mental model is needed to corroborate or interpret the story, resulting in more motivation to construct meaning from these engaging experiences (Arnould & Price, 1993). When stories depart from the norm (a hallmark of transporting stories), there is a compelling need to make sense of them. This is done, according to Bruner (1986), through comparison to other stories. We propose: H1. Transportation is significantly and positively related to retrospective reflection. Retrospective reflection and beliefs Retrospective reflection is a form of narrative comparison between the story one has just experienced and one's memories of self or of others evoked by the narrative. A reader accepts the beliefs implied in the story through generating correspondence between the story world and recalling consistent memories from one's real world “story bank.” For example, a story about a consumer's experience with a pharmaceutical (e.g., side effects of nausea) can be compared with stories one has heard from friends about their experiences with the same pharmaceutical. The convergence across the stories (e.g., the reader's friends also experienced nausea) enhances acceptance of other beliefs implied in the story (e.g., the pharmaceutical is effective at curing a particular ailment). In narrative processing, the reader's orientation is confirmatory and entails credulous reception in which one recruits story-consistent memories of the self and others. Readers are more likely to recall story consistent thoughts because of the additional effort needed to falsify a message (Gilbert, 1991). Narrative persuasion research has established that readers are less likely to generate thoughts counter to a story's stance. For example, Niederdeppe, Kim, Lundell, Fazili, and Frazier (2012) found that participants who read a narrative about a controversial topic were more likely to generate sourceintended processing (elaborations consistent with the message content) than counter-elaborations, suggesting retrospective reflections are more likely to produce story-consistent content that corroborates story consistent beliefs. We expect retrospective reflection will be positively related to persuasion-related outcomes and will mediate the relationship between transportation and persuasion-related outcomes. Formally, we propose: H2. Retrospective reflection will be significantly and positively related to persuasion-related outcomes.


Retrospective reflection, transportation, and cognitive resources Transportation is a cognitively demanding process. Green and Brock (2000) and Wentzel, Tomczak, and Herrmann (2010) demonstrated that a cognitive load induced via reading instructions reduced transportation and corresponding belief outcomes. We have proposed that retrospective reflection also requires cognitive resources and occurs in response to transportation. Because these processes are proposed to operate in succession, we can examine whether transportation and retrospective reflection are distinct by applying a cognitive load at different points during narrative processing. Specifically, a cognitive load applied throughout narrative processing should diminish transportation, retrospective reflection, and persuasion-related outcomes because each process would be suppressed. A cognitive load applied after story exposure (once the reader has reached the story end) should diminish retrospective reflection and persuasion-related outcomes, but should not diminish transportation (which occurs while reading). H5. A cognitive load applied during and after reading a narrative will reduce transportation, retrospective reflection, and beliefs. A cognitive load applied only after reading the narrative will reduce retrospective reflection and beliefs, but not transportation. We also examine the effect of cognitive load on the relationships between transportation, retrospective reflection, and beliefs. As suggested in Hypothesis 3, we propose retrospective reflection is a mechanism of narrative persuasion; that is, transportation influences belief outcomes through retrospective reflection when processing is unconstrained. When a load is applied throughout exposure to the narrative, we predict retrospective reflection will not mediate the transportation—belief relationship; that is, suppressing both transportation and retrospective reflection inhibits both engagement in the story and consequent connecting of the story to the self and related others. Similarly, imposing a load immediately after reading should diminish retrospective reflection's mediating role, since the process is diminished by the load. We propose: H6. Retrospective reflection will not mediate the relationship between transportation and beliefs when a cognitive load is applied throughout the narrative persuasion process or after reading the narrative.

15

Study 1a: Developing the measure of retrospective reflection Method Pilot work: Identifying measures for retrospective reflection In two pilot studies, undergraduate participants (N = 93, Mage = 20.4, 56% female; N = 169, Mage = 20.9, 38% female, respectively) read health-related narratives and reported their thoughts and feelings after reading. We developed categories to organize participants' responses about story content. We used descriptive coding to capture the themes expressed that represented our conceptualization of retrospective reflection and identified four main ideas: “The story reminded me of some of my personal experiences,” “Some parts of the story reminded me of people I know,” “Some parts of the story reminded me of situations other people I know have been in,”, and “I thought about other people I know while reading the story.”

Stimuli and procedure A total of 172 undergraduates (Mage = 20.9, 62% female) were asked to read a story which focused on an individual's health-related reactions to home pesticide use. Examination of the amount of time spent reading the narrative indicated no respondents spent less than 10 s or were more than 3 standard deviations beyond the mean, resulting in the inclusion of all respondents in the analyses.

Measures We adapted Appel, Gnambs, Richter, and Green's (2015) short form of the transportation scale (Green & Brock, 2000): “While I was reading the narrative, I could easily picture the events in it taking place,” “I could picture myself in the scene of the events described in the narrative,” “I was mentally involved in the narrative while reading it,” “I wanted to learn how the narrative ended,” “The narrative affected me emotionally.” Participants' responses to these items and the four retrospective reflection items were measured using 7-point Likert scales.

Results Overview of studies In Study 1a and 1b, we examine our measures of transportation and retrospective reflection using a CFA to assess whether the indicators represent each construct. In Study 1c, we explore the predictive validity of the model. In Study 2, we manipulate personal relevance to examine its role in narrative persuasion and to further examine the predictive validity of the narrative persuasion model. In Study 3, we use a cognitive load manipulation to examine the role of retrospective reflection in narrative persuasion.

We used AMOS 20 to conduct a CFA with 5 indicators of transportation and 4 indicators of retrospective reflection. We set the factor variances to 1 and freely estimated all loadings. Table 1 contains the factor loadings for each item on each construct. The CFA provided a good fit, suggesting the measures represent each construct distinctly (χ2(26) = 40.22, p = .04; CFI = .97; RMSEA = .06; Pclose = .35). In addition, the AVE values for transportation and retrospective reflection were .47 and .54, which were greater than the shared variance (.14) between the two, providing evidence of discriminant validity (Fornell & Larcker, 1981).

16


Table 1 Study 1a–c factor loadings. Study 1a

Study1b

Study 1c

Item

Factor loading

SE

CR value

Factor loading

SE

CR value

Factor loading

SE

CR value

Trans1 Trans2 Trans3 Trans4 Trans5 Reflec1 Reflec2 Reflec3 Reflec4

.86(.68) 1.04 (.76) .89 (.55) 1.09(.81) .81(.58) .84(.47) 1.36(.79) 1.42(.82) 1.35(.79)

.09 .10 .12 .09 .11 .14 .12 .12 .12

9.37** 9.34** 7.23** 11.70** 7.58** 6.03** 11.43** 12.07** 11.37**

.49(.50) .66(.51) .92(.59) .96(.79) .54(.38) 1.17(.68) 1.61(.85) 1.81(.96) 1.61(.86)

.12 .17 .20 .15 .19 .18 .18 .17 .18

3.95** 3.96** 4.67** 6.38** 2.92** 6.40** 8.89** 10.72** 8.96**

.77(.75) 1.14(.89) .33(.18) .77(.78) 1.01(.55) 1.16(.63) 1.58(.85) 1.74(.94) 1.50(.82)

.07 .08 .14 .06 .13 .12 .11 .10 .11

11.42** 14.42** 2.39* 12.05** 7.74** 9.45** 14.22** 16.91** 13.57**

Standardized coefficients in parentheses. * p b .05 ** p b .01.

Study 1b: Developing the measure of retrospective reflection: Replication

Study 1c: Predictive validity of transportation and retrospective reflection in narrative persuasion

Study 1b is a conceptual replication of Study 1a using a different narrative and different participants to assess our measures of transportation and retrospective reflection.

Method

Method Undergraduates (N = 73, Mage = 20.7, 62% female) read a narrative about an individual's development of allergies in response to GMO food. Participants completed the same transportation and retrospective reflection measures used in Study 1a. Examination of the amount of time reading the narrative indicated no respondents spent less than 10 s reading or were more than 3 SD beyond the mean, resulting in the inclusion of all respondents in the analyses.

We used a consumer, rather than health-oriented, narrative to evaluate the robustness of our model across narrative content. Amazon mTurk workers (N = 200) participated through the Qualtrics survey platform. We removed participants who spent less than 10 s or 3 SD over the mean time viewing the narrative, resulting in the removal of six participants, and yielding a final sample of 194 participants (Mage = 41.3, 60.8% female). Participants read a narrative about an individual's experience with a product called the Knee Defender, designed to create space for tall individuals flying on airplanes. Participants completed the same transportation and retrospective reflection measures used in Studies 1a and b, in addition to an assessment of the participant's attitude toward the knee defender product, evaluated on 7-point bad–good, uninteresting–interesting, useless–useful scales.

Results Results We conducted a CFA with multiple indicators of transportation and retrospective reflection and set the factor variances to 1 to freely estimated all loadings. Table 1 contains the factor loadings for each item on each construct. The CFA provided a good fit, (χ2(26) = 36.93, p N .05; CFI = .96; RMSEA = .08; Pclose = .22), and the AVE values for transportation (.33) and retrospective reflection (.71) were greater than the shared variance between the two (.12), again providing support for the discriminant validity of our measures.

Discussion Studies 1a and 1b offer evidence the transportation and retrospective reflection measures adequately discriminate these two constructs across narrative content. In Study 1c, we examine whether the relationships among transportation, retrospective reflection, and story-consistent attitude are consistent with the narrative persuasion model (see Fig. 1).

Model of narrative persuasion: Structural and measurement model We used AMOS 20 to evaluate the full model presented in Fig. 1. The indices suggested adequate model fit (χ2(52) = 109.28; p b .01; CFI = .96; RMSEA = .07; SRMR = .08). The path coefficients between the latent constructs were consistent with expectations. Transportation predicted retrospective reflection (β = .53; CR = 4.09; p b .01), and retrospective reflection predicted attitude toward the product (β = .48; CR = 4.49; p b .01). Each indicator of transportation and retrospective reflection loaded significantly on the relevant factor. The addition of a direct path from the transportation latent factor to attitude did not significantly improve model fit (χ2(51) = 108.84), nor was the path coefficient significant. A bootstrap test of mediation found the 95% confidence interval of transportation's indirect effect on attitude did not contain 0 [β = .27, 95% CI: .12, .45], indicating that retrospective reflection did mediate the transportation and attitude relationship.


Discussion

17

the protagonist's school. This story was edited from a blogger's post about debt accrued during college.

Studies 1a and 1b represented an initial assessment of retrospective reflection in narrative persuasion, providing evidence that transportation and retrospective reflection are distinct constructs robust across narrative content and participant characteristics. Study 1c provided support for the hypothesized model of narrative persuasion: transportation predicted retrospective reflection, which in turn predicted attitude toward the focal product, and retrospective reflection mediated these paths (supporting H1–H3). In Studies 2 and 3, we use an experimental approach to further test the role of retrospective reflection in narrative persuasion.

Measures Participants completed the transportation (α = .72) and retrospective reflection items (α = .81) and responded to four storyconsistent belief measures about credit cards and debt (see Methodological Appendix), which were combined into a composite (α = .81). Finally, participants completed a manipulation check measure: “The character in the story shared some of the experiences that I did, in terms of college life” on a 7-point Likert scale. Results

Study 2: The influence of personal relevance on the transportation–retrospective reflection relationship Study 2 manipulates the personal relevance of a story to examine how it influences the relationship between transportation and retrospective reflection. Method Amazon mTurk workers (N = 106) who qualified for the study based on a pre-screening question (i.e., a valid response to the question, “where did you go to university or college?”) completed the study online and were placed in a low (high) personal relevance condition. We removed participants that spent less than 10 s or more than 3 SD from the mean time viewing the story, resulting in the removal of 5 participants (N = 101, Mage = 36.2, 37% female). We inserted the participants' alma mater in the story, a common approach to manipulate personal relevance in narrative persuasion (Wheeler et al., 1999; Green, 2004). Participants in the low personal relevance condition read about an individual who accrued debt during her college years at an unnamed university. Participants in the high personal relevance condition read the same story with the name of their university inserted as

Table 2 Study 2 Regressions.

Personal relevance as a moderator of the transportation– retrospective reflection relationship An independent samples t-test indicated participants in the high personal relevance condition were more likely to view the story as related to their college experience than participants in the low personal relevance condition (Mrelevant = 5.28 and Mnot relevant = 4.41; t(99) = 2.46, p b .05). Neither transportation nor retrospective reflection differed between conditions. We used PROCESS (model 7; see Hayes, 2013) to test our hypotheses. We mean-centered transportation and retrospective reflection, and then regressed retrospective reflection onto transportation, personal relevance (treated as a dummy variable, 0 = low, 1 = high; Hardy, 1983), and their product term. See Table 2. The resulting model accounted for a substantial proportion of the variation in retrospective reflection (R2 = .35, F(3,97) = 17.46, p b .01). The coefficient of the product term was statistically significant (β = .45, t = 1.99, p b .05), indicating the slopes for the low and high personal relevance conditions differed. Transportation had a significant effect on retrospective reflection when personal relevance was low (β = .55, t = 3.92, p b .01), but a stronger effect when personal relevance was high (β = 1.00, t = 5.70, p b .01; see Fig. 2). We replicated this analysis using the continuous manipulation check measure rather than a dichotomous dummy variable to represent personal relevance and obtained equivalent results (i.e., a significant interaction between transportation and the manipulation check measure of personal relevance, β = .16, t = 2.53, p b .05).

Study 2, Model A: Regression on retrospective reflection

(Constant) Transportation Relevance T×R

β

Se

t

2.08 0.55 − 2.26 0.45

.78 .14 1.28 .23

2.65** 3.89** − 1.76 1.99**

β

Se

t

4.53 0.06 0.26

.46 0.10 .07

9.89** .60 3.65**

Study 2, Model B: Regression on belief measure

(Constant) Transportation Reflection ** p b .01 * p b .05.

Fig. 2. Study 2: personal relevance and transportation interaction.

18


Retrospective reflection as predictor of the belief outcome and as mediator of the transportation–belief relationship Next, we regressed transportation and retrospective reflection on the belief composite. The model was significant (R2 = .19, F(2,98) = 11.76, p b .01). Retrospective reflection had a significant effect (β = .26, t = 3.65, p b .01) on the belief composite; transportation's effect on the belief composite was not significant. We also tested a model in which transportation alone predicted the belief composite. The adjusted R2 was lower than the model with retrospective reflection (R2 = .08), indicating retrospective reflection does contribute beyond transportation in narrative persuasion. Finally, personal relevance did not moderate the transportation–beliefs or retrospective reflection–beliefs relationships. We then examined retrospective reflection as a mediator of the transportation–belief relationship. When personal relevance was high, the 95% CI of the indirect effect did not contain 0 [β = .27, 95% CI: .05, .64]. When personal relevance was low, the 95% CI of the indirect effect also did not contain 0 [β = .15, 95% CI: .03, .37]. This analysis indicates that retrospective reflection mediated the influence of transportation on the belief composite in both personal relevance conditions. The confidence interval of the Index of Moderated Mediation (Hayes, 2013), a test of the equality of the indirect effects of the two groups, did not contain 0 [β = .12, 95% CI: .01, .42], indicating the indirect effect was stronger in the high personal relevance condition. Discussion These results replicated the significance of the transportation– retrospective reflection and retrospective reflection–beliefs paths (H1-H3). The results also support H4, which predicted that high personal relevance would strengthen the relationship between transportation and retrospective reflection relative to a low-relevance condition. These results are consistent with prior work suggesting transportation can cue memories related to elements of the story, and readers reconcile their past experiences with new information when a story is more personally relevant. The results also suggest retrospective reflection allows the reader to corroborate the story by connecting it to the real world. When personally relevant, a reader can relate elements of the story to events in her life or the lives of others. She may then accept other beliefs implied by the story that she has not experienced. This perspective is related to work indicating that once a source's credibility is established through confirmation of a shared experience, an individual is more likely to accept other statements made by the source (Brumbaugh, 2002). The current work also provides insight into prior research on the influence of personal relevance in narrative persuasion. While work has found that personal relevance does not influence transportation directly (Wheeler et al., 1999; Quintero Johnson et al., 2013; de Graaf, 2014), other work (Prentice et al., 1997; Green et al., 2006) suggests personal relevance does influences persuasion-related outcomes. The present findings indicate that when the narrative is more relevant to the reader, the relationship between transportation and retrospective reflection is stronger,

Fig. 3. Study 3: cognitive load conditions.

and the indirect effect of transportation on belief change is larger in the high (versus low) personal relevance condition. Finally, the influence of personal relevance also provides additional evidence for the construct validity of retrospective reflection; specifically, its nomological relation to other constructs in the narrative persuasion process (Cronbach & Meehl, 1955). Study 3: The effect of cognitive load on the transportation– retrospective reflection relationship Study 3 used a cognitive load manipulation to differentiate the constructs of transportation and retrospective reflection, and to evaluate retrospective reflection's role in narrative persuasion. Previous work has indicated that a cognitive load introduced when reading a narrative reduced transportation (through reading instructions: Green & Brock, 2000; a number memorization task: Wentzel et al., 2010), which resulted in a corresponding reduction in story consistent beliefs. A cognitive load administered after a participant completes reading a story cannot influence transportation, because transportation occurs only when engaged in reading the narrative. If retrospective reflection is a consequence of transportation, a cognitive load imposed after reading the story will diminish retrospective reflection and story-consistent beliefs. If retrospective reflection is not evoked or is not distinct from transportation, however, a cognitive load imposed after reading the story will not diminish story-consistent beliefs. Method Participants read a story which scrolled vertically up the screen and described a protagonist who develops melanoma. This approach was used to control when the cognitive load was introduced (see Fig. 3). A pretest (N = 94) indicated the use of a static (traditional) or scrolling format did not lead to differences in transportation (F(1,92) = .55, n.s., Mscrolling = 4.73, Mstatic = 4.90) or retrospective reflection (F(1,92) = 1.18, n.s., Mscrolling = 4.11, Mstatic = 3.82). Table 3 Study 3 means. Transportation No load Load during Load after

1

5.44 (.79) 4.23 (1.44) 4.98 (1.22)1

Reflection

Susceptibility

4.94 (1.34) 3.72 (1.43)1 4.23 (1.62)1

4.35 (1.21) 3.66 (1.38)1 3.74 (1.05)1

Cells that share a superscript do not differ at p b .05.


The current study had three conditions: a load-throughout condition (which should diminish the mean level of transportation, retrospective reflection, and beliefs); a load-after-reading condition (which should diminish the mean level of retrospective reflection and beliefs); and a no-load condition (in which transportation, retrospective reflection, and beliefs should be significantly higher than either cognitive load condition). In all conditions, a string of numbers scrolled across the bottom of the screen (adapted from Gilbert, 1991). In the load-throughout condition, participants counted how often a certain number appeared on the screen throughout the experimental session. In the load-after-reading condition, participants were instructed to ignore the numbers until the story ended, at which point they were prompted to start counting. In the no-load condition, participants were instructed to ignore the numbers on the screen.

Procedure Undergraduate participants (N = 123) were assigned to one of the three conditions. Participants read instructions describing their condition and were told that they would read a story that scrolled up the screen. Participants who received the load manipulations were told that they would be asked to report their total count of the numbers that also scrolled across the screen. Immediately following the instructions, participants were presented with an open-ended question that asked them to describe their assigned task. We used response to this question to assess whether participants understood and were attentive to the task requirements. Some participants (N = 11) did not provide an adequate response to this question and were removed (e.g., “don't know,” “no idea”), yielding 112 participants (Mage = 21.1, 39.3% female). When ready, participants clicked to access the story. As the last sentence of the story scrolled up the screen, participants were reminded to wait/continue counting/start counting (in the no-load/load-throughout/load-after-reading conditions) until the page advanced. The webpage auto-advanced 15 s after the story ending in all conditions; all participants spent the same amount of time viewing the story/scrolling number page (180 s total).

Measures Participants completed the transportation (α = .83) and retrospective reflection items (α = .91). Perceived susceptibility was assessed with the question: “The likelihood that I will contract melanoma is:” extremely low (1)–extremely high (7). We also measured negative emotion and skepticism to assess whether the cognitive load inadvertently evoked these states, which would confound the hypothesized effects.

19

Effects of cognitive load on constructs Cognitive load and transportation A one-way ANOVA on transportation indicated significant differences (F(2, 109) = 9.41, p b .01). LSD post hoc tests indicated that transportation in the load–throughout condition was significantly lower (M = 4.23) than the load-after-reading (M = 4.98; p b .01) and no-load (M = 5.44; p b .01) conditions. Consistent with our predictions, transportation did not differ in the load-after-reading and no-load conditions: A cognitive load imposed after reading a passage did not influence transportation. Cognitive load and retrospective reflection A one-way ANOVA on retrospective reflection indicated significant differences (F(2, 109) = 5.99, p b .01). LSD post hoc tests indicated retrospective reflection in the load-throughout (M = 3.72) and the load-after-reading (M = 4.23) conditions were significantly lower than in the no-load condition (M = 4.94, p b .01, for both). The load-throughout and load-after-reading conditions were not different from each other, supporting our prediction: a cognitive load after reading did reduce retrospective reflection compared to a no-load condition. Cognitive load and belief outcome A one-way ANOVA on perceived susceptibility indicated significant differences (F(2109) = 3.61, p b .05). LSD post hoc tests indicated that the means were significantly lower in the load-throughout (M = 3.66) and load-after-reading conditions (M = 3.74) than the no-load condition (M = 4.35, p b .05 for both; see Table 3 for a summary of the means), suggesting a reduction in retrospective reflection leads to a corresponding reduction in story-consistent beliefs. Effect of cognitive load on relationships between constructs We used a bootstrap analysis (PROCESS model 75; see Hayes, 2013) to examine retrospective reflection as a mediator of the transportation–belief relationship, and whether cognitive load moderates this mediated relationship. Model 75 entails a path from transportation to retrospective reflection and retrospective reflection to belief, and it includes the load conditions as moderators of each path (see Fig. 4). We dummy coded the experimental conditions, first treating the no-load condition as the

Results A one-way ANOVA indicated the manipulation did not lead to significant differences in negative emotion (F (2109) = .02, n.s.) or skepticism (F (2109) = .75, n.s.).

Fig. 4. Study 3: PROCESS Model 75: overall model test with no-load as reference group.

20


reference group (Jaccard & Turrisi, 2003). The results replicated the predictions of H1 and H2: Transportation was significantly related to retrospective reflection (β = .73, t = 2.76, p b .01), and retrospective reflection was significantly related to the belief measure (i.e., susceptibility; β = .37, t = 2.46, p b .05). The retrospective reflection–belief relationship was qualified by a significant interaction (β = −.42; t = 2.22, p b .05) between retrospective reflection and the dummy coded loadafter condition. In the no-load condition, the relationship between retrospective reflection and beliefs was positive and significant (β = .37, t = 2.76, p b .01). In the load-after condition, the relationship between retrospective reflection and the belief measure was not significant (β = −.04, t = −.40, p N .05). The transportation–retrospective reflection (β = −.01, t = −.04, p N .05) and retrospective reflection–belief (β = −.05, t = −.17, p N .05) relationships did not differ between the load-throughout and no-load conditions, nor did the transportation–retrospective reflection relationship (β = −.27, t = −.17, p N .05) differ between the load-after and no-load conditions, which we consider in the discussion. Fig. 4 displays the underlying model, treating the no-load condition as the reference group. Fig. 5 contains a simplified version of Fig. 4, showing path coefficients derived by treating each condition as a reference group in PROCESS Model 75 for the transportation–retrospective reflection, retrospective reflection–beliefs, and transportation– beliefs relationships. We treated retrospective reflection as a mediator of the transportation–belief path and examined the conditional indirect effect in each experimental condition. In a no-load condition, the 95% confidence interval did not contain zero [β = .27, 95% CI: .01, .80]. In the load-throughout and load-after conditions, the confidence intervals did contain zero ([β = − .02, 95% CI: −.21, .11]; [β = .07, 95% CI: −.24, .40], respectively), indicating that retrospective reflection does not mediate the transportation– belief path when a load is applied. Discussion We used a cognitive load manipulation to explore the contributions of retrospective reflection to narrative persuasion and examine whether transportation and retrospective reflection are distinct constructs. When a cognitive load was applied throughout the session, the mean levels of transportation, retrospective reflection and the belief were lower. In contrast, participants who read the story with no cognitive load reported higher levels of transportation, retrospective reflection, and belief. When a cognitive load was applied after exposure to the

Fig. 5. Study 3: path coefficients with each condition as the reference group.

story, retrospective reflection and the belief, but not transportation, were diminished, supporting H5. This pattern of results provides evidence that retrospective reflection is distinct from transportation and occurs subsequent to transportation. In addition, respondents in the load-after-reading condition had significantly lower beliefs compared to those in the no-load condition (when retrospective reflection was not diminished), illustrating the importance of retrospective reflection in narrative persuasion. When retrospective reflection (but not transportation) was diminished in the load-after condition, persuasion was also reduced to a similar level as the load throughout condition in which both processes were diminished. Retrospective reflection mediated the transportation–belief path in the no-load condition, replicating support for H3. When retrospective reflection was diminished by a cognitive load (in either load condition), it did not mediate transportation–belief path, supporting H6. As expected, a load applied after reading the story disrupted the process of retrospective reflection, diminishing the relationship between retrospective reflection and belief compared to a no-load condition. Unexpectedly, the relationship between transportation and retrospective reflection was not affected in the load-after condition. This finding suggests the transition between transportation and retrospective reflection may be gradual rather than a sharp demarcation; some retrospective reflection may have occurred toward the end of engagement with the narrative, during a time when no load had been applied. In addition, the load-throughout condition did not diminish the relationship between the transportation– retrospective reflection or the retrospective reflection–belief path compared to the no-load condition. Participants in the load throughout condition were asked to allocate their cognitive resources to two tasks—reading the narrative and counting the number of instances of a number scrolling on the screen, prior to starting each task. These participants appeared to adapt a multitasking strategy and partitioned their resources to each task, resulting in fewer resources allocated to each task (Logan & Gordon, 2001; Kiesel et al., 2010). The load-throughout condition resulted in a reduced mean level of transportation and retrospective reflection, but in no meaningful change in the relations between transportation, retrospective reflection, and beliefs from a no-load condition. General discussion Green and Brock (2000) present the mechanism of transportation to explain the effect of narratives on persuasion-related outcomes. Most research subsequent to Green and Brock (2000; see Van Laer, de Ruyter, Visconti, & Wetzels, 2014 for a review) has focused on the scope and limits of the transportation process. One unaddressed question is the mechanism(s) that account for the relationship between transportation and persuasion-related outcomes. The current work develops a measure to assess retrospective reflection and demonstrates retrospective reflection mediates the relationship between transportation and persuasionrelated outcomes. Two experimental studies provide evidence that transportation and retrospective reflection are distinct by showing (1) personal relevance moderates their relationship and


(2) a cognitive load applied following the reading of the story diminishes retrospective reflection but not transportation. Theoretical and applied implications First, the current work defined and developed a measure of retrospective reflection as a process to aid a reader's understanding of a story. This process represents a reader's effort to evaluate the message by comparing the mental model of the story with memories related to the self and others. A story that resonates with the audience through evoked “remindings” (Larsen & László, 1990) enhances appreciation of the narrative, substantiates and legitimizes the narrative (Bruner, 1986; Fisher, 1989), and increases persuasion. The current work suggests this process follows transportation. When this process was actively diminished through a cognitive load, persuasion was diminished and retrospective reflection did not mediate the transportation–belief relationship. Second, the present work deepens our understanding of how narratives persuade. Previous work on narrative persuasion focused primarily on transportation as the sole persuasive mechanism, linking it directly to persuasion-related outcomes and characterizing it as a “magical,” transformative process. The current work adds to this model, proposing retrospective reflection as a mediator of the transportation–persuasion relationship. Past work has focused on narratives' affective pathways to persuasion. Retrospective reflection adds a more cognitively oriented process and creates a refined understanding of narrative persuasion. Third, retrospective reflection also helps to explain the effect of a narrative's personal relevance on belief change. Prior work (e.g., Wheeler et al., 1999; Quintero Johnson et al., 2013; de Graaf, 2014) found no direct effect of personal relevance on transportation, but personal relevance does affect the transportation–retrospective relationship. Higher personal relevance increases motivation to align the story message to the reader's real world and achieve consistency. While the current work suggests that retrospective reflection mediates the relationship between transportation and beliefs regardless of personal relevance, the indirect effect of transportation on beliefs was stronger in the high personal relevance condition than in the low personal relevance condition. Fourth, the current work provides insight into how a cognitive load can influence aspects of the narrative persuasion process. The load task applied in Study 3 diminished but did not eliminate the narrative processes, illustrating their robustness (Green & Brock, 2000). Further, “the narrative persuasion process” should be construed beyond processes that occur while reading the story (during transportation) to include post-narrative processes (retrospective reflection). The current research also has implications for persuasionoriented practitioners. For example, persuasion-related outcomes can be inhibited (or enhanced) by manipulating factors that influence retrospective reflection. Distraction, as instantiated using a cognitive load, reduced persuasion. This effect highlights the importance of allowing time following message exposure to enable the process of retrospective reflection to occur. For example, many advertising narratives are followed

21

immediately by other stimuli. Shortening the narrative and allowing time for retrospective reflection may enhance the persuasiveness of the message. We also anticipate efforts could encourage retrospective reflection, such as instructing readers to consider people in their lives who shared experiences described in the narrative. For example, readers can be instructed to connect memories about their experiences to the characters and events in the narrative. We would expect the result of these efforts to enhance retrospective reflection, resulting in more persuasion. Limitations and future directions The current approach to measure transportation is intended to capture a holistic, retrospective evaluation of the experience without disrupting the ongoing process. We applied the same measurement perspective to assess retrospective reflection. This approach has some limitations, including possible recall biases that may diminish participants' accuracy when reporting their experience. We anticipate future research can address potential biases introduced by using self-report measures and apply real-time assessments, such as the use of physiological measures, f-MRIs, and real-time dial-meter technology used in public opinion research, to gain a more nuanced understanding of how the experience of transportation (e.g., its range and duration) relates to the process of retrospective reflection and persuasion. The current work explored the construct of retrospective reflection in the context of consumer or health-oriented narratives. Work on narrative persuasion has been conducted in a range of fields, and we did not examine whether retrospective reflection leads to persuasion in an entertainment-oriented context. The relations among the narrative persuasion processes may differ in such contexts, and future work should explore the relevance of this distinction. The complexity in the structure, construction, and delivery of narrative messages introduce many potential factors that can affect the relationships among transportation, retrospective reflection, and persuasion-related outcomes. For example, retrospective reflection may differ as a function of modality. Visual narratives may reduce the amount of retrospective reflection compared to their written counterparts because of the self-paced nature of reading and the readers' enhanced ability to create mental representations of the narrative world that are connected to their own world. Other contextual or individual-level variables factors, such as a participant's goal when reading the narrative, may also influence narrative processes. A specific, concrete goal related to decision making may influence the types of details one attends to in a narrative and the memories recalled. We anticipate future research will create more complex conceptual models to better capture the many factors that influence narrative persuasion. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.jcps.2016.06.005.

22


References Abelson, R. P. (1968). Theories of cognitive consistency: A sourcebook. Chicago: Rand McNally. Adaval, R., & Wyer, R. S. (1998). The role of narratives in consumer information processing. Journal of Consumer Psychology, 7(3), 207–245. Ajzen, I., Brown, T. C., & Rosenthal, L. H. (1996). Information bias in contingent valuation: Effects of personal relevance, quality of information, and motivational orientation. Journal of Environmental Economics and Management, 30(1), 43–57. Appel, M., & Malečkar, B. (2012). The influence of paratext on narrative persuasion: Fact, fiction, or fake? Human Communication Research, 38(4), 459–484. Appel, M., Gnambs, T., Richter, T., & Green, M. C. (2015). The transportation scale—short form (TS-SF). Media Psychology, 18(2), 243–266. Argo, J. J., Zhu, R., & Dahl, D. W. (2008). Fact or fiction: An investigation of empathy differences in response to emotional melodramatic entertainment. Journal of Consumer Research, 34(5), 614–623. Arnould, E. J., & Price, L. L. (1993). River Magic: Extraordinary experience and the extended service encounter. Journal of Consumer Research, 20(1), 24–45. Brumbaugh, A. M. (2002). Source and nonsource cues in advertising and their effects on the activation of cultural and subcultural knowledge on the route to persuasion. Journal of Consumer Research, 29(2), 258–269. Bruner, J. (1986). Actual minds, possible worlds. Cambridge, MA: Harvard University Press. Burnkrant, R. E., & Unnava, H. R. (1995). Effects of self-referencing on persuasion. Journal of Consumer Research, 22(1), 17–26. Busselle, R., & Bilandzic, H. (2008). Fictionality and perceived realism in experiencing stories: A model of narrative comprehension and engagement. Communication Theory, 18(2), 255–280. Celsi, R. L., & Olson, J. C. (1988). The role of involvement in attention and comprehension processes. Journal of Consumer Research, 2(15), 210–224. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52(4), 281–302. de Graaf, A. (2014). The effectiveness of adaptation of the protagonist in narrative impact: Similarity influences health beliefs through self-referencing. Human Communication Research, 40(1), 73–90. Debevec, K., & Romeo, J. B. (1992). Self-referent processing in perceptions of verbal and visual commercial information. Journal of Consumer Psychology, 1(1), 83–102. Deighton, J., Romer, D., & McQueen, J. (1989). Using drama to persuade. Journal of Consumer Research, 3(16), 335–343. Dunlop, S. M., Wakefield, M., & Kashima, Y. (2009). Pathways to persuasion: Cognitive and experiential responses to health-promoting mass media messages. Communication Research, 37(1), 133–164. Escalas, J. E. (2007). Self-referencing and persuasion: Narrative transportation versus analytical elaboration. Journal of Consumer Research, 33(4), 421–429. Fisher, W. R. (1989). Human communication as narration: Toward a philosophy of reason, value, and action. Columbia: University of South Carolina Press. Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 1(18), 39–50. Gilbert, D. T. (1991). How mental systems believe. American Psychologist, 46(2), 107–119. Green, M. C. (2004). Transportation into narrative worlds: The role of prior knowledge and perceived realism. Discourse Processes, 38(2), 247–266. Green, M. C., & Brock, T. C. (2000). The role of transportation in the persuasiveness of public narratives. Journal of Personality and Social Psychology, 79(5), 701–721. Green, M. C., Garst, J., Brock, T. C., & Chung, S. (2006). Fact versus fiction labeling: Persuasion parity despite heightened scrutiny of fact. Media Psychology, 8(3), 267–285. Hamby, A., & Brinberg, D. (2016). A conceptual framework of narrative persuasion. Journal of Media Psychology (in press). Hardy, M. A. (1983). Regression with dummy variables. Thousand Oaks, CA: Sage. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. Guilford Press.

Hung, I. W., & Wyer, R. S., Jr. (2011). Shaping consumer imaginations: The role of self-focused attention in product evaluations. Journal of Marketing Research, 48(2), 381–392. Jaccard, J., & Turrisi, R. (2003). Interaction effects in multiple regression. Newbury Park: Sage. Kiesel, A., Steinhauser, M., Wendt, M., Falkenstein, M., Jost, K., Philipp, A. M., & Koch, I. (2010). Control and interference in task switching—A review. Psychological Bulletin, 136(5), 849–874. Larsen, S. F., & László, J. (1990). Cultural–historical knowledge and personal experience in appreciation of literature. European Journal of Social Psychology, 20(5), 425–440. Larsen, S. F., & Seilman, U. (1988). Personal remindings while reading literature. Text-Interdisciplinary. Journal for the Study of Discourse, 8(4), 411–430. Logan, G. D., & Gordon, R. D. (2001). Executive control of visual attention in dual-task situations. Psychological Review, 108(2), 393–434. McFerran, B., Dahl, D. W., Gorn, G. J., & Honea, H. (2010). Motivational determinants of transportation into marketing narratives. Journal of Consumer Psychology, 20(3), 306–316. Mick, D. G., & Buhl, C. (1992). A meaning-based model of advertising experiences. Journal of Consumer Research, 3(19), 317–338. Nelson, L. D., Meyvis, T., & Galak, J. (2009). Enhancing the televisionviewing experience through commercial interruptions. Journal of Consumer Research, 36(2), 160–172. Niederdeppe, J., Kim, H. K., Lundell, H., Fazili, F., & Frazier, B. (2012). Beyond counterarguing: Simple elaboration, complex integration, and counterelaboration in response to variations in narrative focus and sidedness. Journal of Communication, 62(5), 758–777. Nielsen, J. H., & Escalas, J. E. (2010). Easier is not always better: The moderating role of processing type on preference fluency. Journal of Consumer Psychology, 20(3), 295–305. Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlag. Prentice, D. A., Gerrig, R. J., & Bailis, D. S. (1997). What readers bring to the processing of fictional texts. Psychonomic Bulletin & Review, 4(3), 416–420. Quintero Johnson, J. M., Harrison, K., & Quick, B. L. (2013). Understanding the effectiveness of the entertainment-education strategy: An investigation of how audience involvement, message processing, and message design influence health information recall. Journal of Health Communication, 18(2), 160–178. Shen, F., Sheer, V. C., & Li, R. (2015). Impact of narratives on persuasion in health communication: A meta-analysis. Journal of Advertising, 44(2), 105–113. Strange, J. J., & Leung, C. C. (1999). How anecdotal accounts in news and in fiction can influence judgments of a social problem's urgency, causes, and cures. Personality and Social Psychology Bulletin, 25(4), 436–449. Van Laer, T., De Ruyter, K., Visconti, L. M., & Wetzels, M. (2014). The extended transportation-imagery model: A meta-analysis of the antecedents and consequences of consumers' narrative transportation. Journal of Consumer Research, 40(5), 797–817. Wang, J., & Calder, B. J. (2006). Media transportation and advertising. Journal of Consumer Research, 33(2), 151–162. Wang, J., & Calder, B. J. (2009). Media engagement and advertising: Transportation, matching, transference and intrusion. Journal of Consumer Psychology, 19(3), 546–555. Wentzel, D., Tomczak, T., & Herrmann, A. (2010). The moderating effect of manipulative intent and cognitive resources on the evaluation of narrative ads. Psychology and Marketing, 27(5), 510–530. West, P. M., Huber, J., & Min, K. S. (2004). Altering experienced utility: The impact of story writing and self-referencing on preferences. Journal of Consumer Research, 31(3), 623–630. Wheeler, C., Green, M. C., & Brock, T. C. (1999). Fictional narratives change beliefs: Replications of Prentice, Gerrig, and Bailis (1997) with mixed corroboration. Psychonomic Bulletin & Review, 6(1), 136–141. Woodside, A. G., Sood, S., & Miller, K. E. (2008). When consumers and brands talk: Storytelling theory and research in psychology and marketing. Psychology and Marketing, 25(2), 97–145. Zwaan, R. A., & Radvansky, G. A. (1998). Situation models in language comprehension and memory. Psychological Bulletin, 123(2), 162–185.