Communication Research

http://crx.sagepub.com/

Perceived Realism: Dimensions and Roles in Narrative Persuasion

Hyunyi Cho, Lijiang Shen and Kari Wilson
Communication Research published online 12 July 2012
DOI: 10.1177/0093650212450585

The online version of this article can be found at:

http://crx.sagepub.com/content/early/2012/09/09/0093650212450585

Published by:

http://www.sagepublications.com

Additional services and information for Communication Research can be found at:

Email Alerts: http://crx.sagepub.com/cgi/alerts

Subscriptions: http://crx.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> OnlineFirst Version of Record - Sep 10, 2012

OnlineFirst Version of Record - Jul 12, 2012

What is This?

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

450585
50585Cho et al.Communication Research
© The Author(s) 2011

Reprints and permission: http://www.

CRXXXX10.1177/00936502124

sagepub.com/journalsPermissions.nav

Communication Research

Perceived XX(X) 1­–24

© The Author(s) 2012
Reprints and permission:
Realism: Dimensions sagepub.com/journalsPermissions.nav
DOI: 10.1177/0093650212450585
and Roles in Narrative http://crx.sagepub.com

Persuasion

Hyunyi Cho1, Lijiang Shen2,

and Kari Wilson3

Abstract
Perceived realism may be a crucial message characteristic facilitating narrative-based
persuasion. This study examined dimensions of perceived realism and their roles in
narrative persuasion. Data based on responses to messages on three topics showed that
perceived realism was multidimensional. Its dimensions included plausibility, typicality,
factuality, narrative consistency, and perceptual quality. Plausibility predicted emotional
involvement, but not identification. Typicality predicted identification, but not emotional
involvement. Narrative consistency and perceptual quality predicted message evaluation.
Emotional involvement, identification, and message evaluation, in turn, predicted attitudes.
Implications for theory, research, and message design pertinent to narrative persuasion
are discussed.

Keywords
perceived realism, narratives, persuasion, identification, emotional involvement, message
evaluation, multilevel SEM

A growing volume of communication research investigates the persuasive impact of narra-

tives. Studies have found that narratives influence individuals’ beliefs and behaviors rele-
vant to health of self and others (e.g., Morgan, Movius, & Cody, 2009; Moyer-Gusé &
Nabi, 2010). Studies have also found that narratives influence individuals’ attitudes toward
public policies for addressing social issues (e.g., Slater, Rouner, & Long, 2006). A focus of

1
Purdue University, West Lafayette, IN, USA
2
University of Georgia, Athens, GA, USA
3
Raclin School of the Arts, Indiana University South Bend, IN, USA

Corresponding Author:
Hyunyi Cho, Brian Lamb School of Communication, Purdue University, 100 N. University Street, West
Lafayette, IN 47907, USA
Email: hcho@purdue.edu

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

2 Communication Research XX(X)

the extant research on narrative persuasion has been its process. This stream of research has
investigated how the audience’s involvement in narratives facilitates persuasion. Specifi-
cally, studies examined the nature and role of narrative involvement (e.g., Green & Brock,
2000; Morgan et al., 2009), audience and channel factors influencing narrative involve-
ment (Appel & Richter, 2010; Durkin & Wakefield, 2008), and paths between narrative
involvement and outcomes (e.g., Dunlop, Wakefield, & Kashima, 2010; Moyer-Gusé &
Nabi, 2010).
Building on the existing research focusing on the processing of narratives, this study
focuses on the characteristics of narratives that may promote processing. Knowledge about
the characteristics of narrative messages, in addition to the processing of messages, may be
important. Communication theory of narrative persuasion should be able to explain not
only the process, but also the characteristics of narratives that predict the process. Such
knowledge, in turn, can inform the development of narrative messages for motivating indi-
viduals’ behaviors and societal actions.
Recent research suggests that a message characteristic important to narrative persuasion
may be perceived realism (Busselle & Bilandzic, 2008; Green, 2004; Larkey & Hecht,
2010). Although the importance of perceived realism has long been recognized in com-
munication research, the construct has yet to achieve conceptual consensus and corre-
sponding precision in measurement. Scholars have argued that perceived realism may be a
multidimensional construct (e.g., Busselle & Greenberg, 2000; Dorr, 1983; Hall, 2003;
Hawkins, 1977; Potter, 1986; Pouliot & Cowen, 2007). Little research, however, has
empirically investigated the multidimensional nature of perceived realism and the role that
the dimensions may play in the persuasion process. Lack of consensus in conceptualization
and operationalization may have produced inconsistent findings regarding the role of the
construct in persuasion (Andsager, Austin, & Pinkleton, 2001; Pouliot & Cowen, 2007). A
clear understanding of perceived realism may be essential to advancing theory and research
on narrative persuasion. Thus, the goal of this study is to examine the dimensions of per-
ceived realism and their roles in narrative persuasion.

Narrative Persuasion and Perceived Realism

Narratives are stories that consist of a “sequence of thematically and temporally related
events” involving characters (Adaval & Wyer, 1998, p. 208; Fisher, 1987). Narrative per-
suasion is a message strategy to influence beliefs, attitudes, and behaviors through stories.
Research has shown the effectiveness of narrative messages in various formats. These
formats range from short story–based advertisements in which persuasion intent is explicit
(e.g., Chang, 2008; Escalas, 2004) to long entertainment education dramas where persua-
sion intent is implicit (e.g., Singhal, Cody, Rogers, & Sabido, 2003).
Earlier communication research on narrative effects focused on entertainment-education.
Perhaps, as a result, the absence of explicit persuasive intent or persuasive messages has
been considered essential to narrative effectiveness. However, more recent communication
research recognizes that narratives can be used with explicit persuasive intent. For example,
in suggesting how narrative persuasion can be used for cancer control and prevention, a

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 3

group of scholars define narratives as: “a representation of connected events and characters
. . . and contains implicit or explicit messages about the topic being addressed’’ (Kreuter et al.,
2007, p. 222). Similarly, Green (2008) writes, “Stories are often associated with entertain-
ment, but they can also be used to convey serious information ranging from company poli-
cies to heath advice” (p. 47). One can think of many instances of narrative persuasion in
which persuasive intent is explicit. For example, many cancer survivors are willing to share
their stories to help other people prevent and manage cancer. These stories are delivered in
formats ranging from interpersonal communication, small group discussion, videos in tar-
geted and tailored health interventions, to testimonials and public service announcements
(PSAs) on YouTube.
Research suggests that implicit persuasion intent is important for entertainment-
education, but may not be for other forms of educational messages. For example,
Moyer-Gusé and Nabi (2010) found that perceived persuasive intent predicted reac-
tance (Brehm, 1966; Dillard & Shen, 2005) in their dramatic narrative / entertainment-
education condition; however, the relationship between perceived persuasive intent and
reactance was nonsignificant in their nonnews / informational education condition.
Together, these findings may suggest that being in an informative education structure,
in which persuasive intent is explicit, may not always make the message vulnerable to
reactance. Perhaps more importantly, the findings suggest that being in a narrative
structure or entertainment education format may not necessarily make the message
impervious to reactance.
An important line of inquiry in narrative persuasion, then, may be to investigate the
within-narrative variations in perceived and actual message characteristics that determine
positive and negative message effects. Perceived realism may be one such characteristic.
Research on persuasion and media effects suggests that perceived realism may be a narra-
tive characteristic stimulating emotional involvement (Busselle & Bilandzic, 2008; Green,
2004; Larkey & Hecht, 2010), identification (Larkey & Hecht, 2010; Livingstone, 1990;
Potter, 1986; Zillmann, 1980), and message evaluation (Cho & Boster, 2008).
Generally, perceived realism is the audience’s judgment of the degree to which the nar-
rative world is reflective of the real world (Gerbner & Gross, 1976). However, just as the
real world is not unidimensional, the judgment of realism of a narrative may not be unidi-
mensional. Scholars have conceptualized that the audience may evaluate narrative repre-
sentations with respect to divergent facets of the real world (e.g., Busselle & Greenberg,
2000; Dorr, 1983; Hall, 2003; Hawkins, 1977; Potter, 1986; Pouliot & Cowen, 2007).
Although a number of studies have offered conceptual discussions about the poten-
tial multidimensionality of perceived realism, few empirically investigated the dimen-
sions. A more comprehensive perspective integrating divergent conceptualizations of
perceived realism may be Hall’s (2003). Hall synthesized previous conceptualizations
of perceived realism and collected and analyzed qualitative data of audience members’
own descriptions of how they judge realism in narrative media to propose an integrative
perspective on perceived realism. Specifically, Hall proposes that the construct may
include the dimensions of plausibility, typicality, factuality, narrative consistency, and
perceptual quality.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

4 Communication Research XX(X)

Dimensions of Perceived Realism

Perceived Plausibility
Plausibility is considered to be a fundamental requirement of perceived realism, and refers
to the degree to which narrative presentations of behaviors and events could possibly occur
in the real world (Hall, 2003). For example, the soap opera Days of Our Lives could be
perceived as unreal because it involves devils and demons, but the soap opera The Young
and the Restless could be perceived as real because it is about the lives of two families,
although all soap operas are fictional (Hall, 2003). Plausibility may be similar to Dorr’s
(1983) notion of possibility, which concerns whether or not the events portrayed in a nar-
rative could actually happen in reality.

Perceived Typicality
Typicality refers to the degree to which narrative portrayals appear to fall within the
parameter of the audience’s past and present experiences (Hall, 2003). The movie Erin
Brockovich may be perceived as unreal because the experience of Erin Brockovich pre-
sented in the movie may be outside the boundary of experiences that the audience them-
selves have had (Hall, 2003). Typicality may be similar to the concept of social expectation
of Hawkins (1977). Social expectation is about the degree to which people and events in
a narrative are perceived to be similar to those in one’s real life (Hawkins, 1977). Dorr’s
(1983) concept of probability, the degree to which a narrative depiction is considered
representative of reality, may be similar to typicality as well.
Typicality is distinct from plausibility (Dorr, 1983; Hall, 2003). Actions and events in a
narrative may be perceived as plausible but not typical. The actions and events could be
construed to possibly happen to someone somewhere at some point, but they may be
deemed beyond the range of the everyday experience of the audience. Typicality, therefore,
is a more stringent evaluation criterion of realism than plausibility (Hall, 2003).

Perceived Factuality
Factuality refers to the degree to which a narrative is perceived to portray a specific indi-
vidual or event in the real world: The narratives of Erin Brockovich and Schindler’s List
are factual; they are based on real persons and events (Hall, 2003). Hawkins’ (1977) notion
of “magic window” may be related to factuality. Magic window refers to the phenomenon
in which children believe that individuals in the media actually exist and the events in the
media actually took place (Hawkins, 1977). Magic window is pertinent only to children,
however (Dorr, 1983). Factuality is distinct from other dimensions of realism. The narra-
tives of Erin Brockovich and Schindler’s List may be factual, but unlikely to be typical
(Hall, 2003).
Factuality has sometimes been viewed as representative of realism. For example, to
examine the effects of realism, earlier studies presented the same violent media content to
participants as either fact or fiction (e.g., Feshbach, 1976; Reeves, 1978). Pouliot and

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 5

Cowen (2007) distinguished factual realism from psychological realism. Factual realism is
the audience’s judgment regarding whether individuals and events in the media are made
up or not; psychological realism is the degree to which the individuals and events in a nar-
rative are similar to those in reality.
Plausibility, typicality, and factuality judgments may focus on narratives’ proximity to
one’s reality (Hall, 2003). However, the judgments of narrative consistency and perceptual
quality may focus more on the properties of the narratives than their proximity to one’s
reality, as discussed below.

Perceived Narrative Consistency

Narrative consistency is the degree to which a story and its elements are judged to be
congruent and coherent, and without contradictions (Hall, 2003). Narrative consistency
may be similar to Fisher’s (1987) concept of narrative probability in his narrative para-
digm theory. Narrative probability refers to a narrative’s internal and structural consis-
tency; it concerns how well a story “hangs together” (Fisher, 1987). Judgment of narrative
consistency can be independent of the judgments of a narrative’s plausibility, typicality, or
factuality (Hall, 2003). Participants in Hall’s focus groups said that Buffy the Vampire
Slayer was not relatable to their real-life experience, but the story itself was internally
coherent. Brinson and Brown (1997) evaluated the television public service announce-
ments of the Centers for Disease Control and Prevention’s America Responds to AIDS
campaign, and argued that the messages would be persuasive due to their high narrative
consistency.

Perceived Perceptual Quality

Perceptual quality refers to the degree to which the audio, visual, and other manufactured
elements of a media narrative comprise a convincing and compelling portrayal of the real-
ity, independent of the consideration whether the content of the narrative is related or
relevant to the audience’s real-world experience (Hall, 2003). For example, the partici-
pants in Hall’s focus groups perceived the dinosaurs in Jurassic Park as real, although
dinosaurs do not exist anymore and are unlikely to be encountered in the real world. On
the basis of the theory and research discussed above, we posit that perceived realism is a
multidimensional construct, comprised of plausibility, typicality, factuality, narrative con-
sistency, and perceptual quality (Hypothesis 1).

Role of Perceived Realism in Narrative Persuasion

An important implication of the multidimensionality of perceived realism would be that
the dimensions of perceived realism may be differentially related to the mediators of nar-
rative persuasion. As discussed in detail below, extant theory suggests mediators of narra-
tive persuasion may include identification, emotional involvement, and message evaluation
(Green & Brock, 2000; Larkey & Hecht, 2010; Moyer-Gusé, 2008; Slater & Rouner, 1996;

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

6 Communication Research XX(X)

Identification

Emotional Attitudes
Perceived Realism
Involvement

Message
Evaluation

Figure 1. Overall model of the role of perceived realism in narrative persuasion

2002). Data consistent with the postulations have been observed (e.g., Cho & Boster,
2008; Morgan et al., 2009; Moyer-Gusé & Nabi, 2010; Slater et al., 2006).
Integrating these theoretical perspectives and research findings, we propose a model
explicating the role of perceived realism in narrative persuasion. Overall, this model posits
that perceived realism of narratives will produce persuasion through three indirect routes,
that is through identification, emotional involvement, and message evaluation (see Figure 1).
Furthermore, it posits that dimensions of perceived realism will be differentially related to
the mediators of persuasion.

Identification
Identification refers to the bond, connection, or relatedness that the audience perceives
with the character of a narrative (e.g., Basil, 1996; Campbell & Babrow, 2004; Hoffner &
Buchanan, 2005; Slater & Rouner, 2002). This sense of connection can be predicted by
perceived similarity between the two (Slater et al., 2006 ;Slater & Rouner, 2002) and
accompanied by the audience’s experiential adoption of the character’s perspectives
(Cohen, 2001). Although identification has been viewed as an important factor determin-
ing the effectiveness of narrative persuasion (Larkey & Hecht, 2010; Moyer-Gusé, 2008;
Slater & Rouner, 2002), the kinds of message characteristics that facilitate identification
have yet to be investigated.
Scholars suggest that perceived realism may be associated with perceived identification
(e.g., Larkey & Hecht, 2010; Livingstone, 1990; Potter, 1986; Zillmann, 1980). The more
the audience perceives the character and the events involving the character as real, the
greater the likelihood of identifying with the character. Seen in this light, perceived realism
may be one of the preconditions of identification. Furthermore, if perceived realism is
multidimensional, then the dimensions may differentially predict identification. Of the five
dimensions of perceived realism, typicality may be more likely to predict identification
than any other dimension.
Typicality is more likely to predict identification because this perceived realism dimen-
sion considers the degree to which a narrative represents a commonly occurring event in

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 7

the real world. Events that are typical are widely distributed in society and are more likely
to fall within the range of the audience’s past and present experiences, as compared to
event that are not typical. For these reasons, typicality may be construed to reflect the audi-
ence’s own experience more than any other dimensions.
Some events in narratives may be real, but not common (e.g., murder, avalanche;
Shapiro & Chock, 2003). These rare but real events may meet other requirements of real-
ism, plausibility and factuality. Plausibility concerns possible occurrence, while factuality
concerns actual occurrence. However, neither of these realism judgments require a high
prevalence of the event in real life, nor do they concern the common experience of the audi-
ence. The remaining dimensions of perceived realism may concern the story itself more
than the relatedness between the story and reality. That is, narrative consistency is about
whether the story is coherent in its internal structure, and perceptual quality is about
whether the story possesses authentic audiovisual properties. Thus, we hypothesize that
typicality is more likely to predict identification than other dimensions (Hypothesis 2).

Emotional Involvement
The audience’s involvement in the narrative has been described with various terms includ-
ing absorption (Slater & Rouner, 2002), message engagement (Larkey & Hecht, 2010),
and transportation (Green & Brock, 2000). These are different labels for the same con-
struct, narrative involvement, the process in which the audience focuses on the narrative
and is influenced by the narrative (Larkey & Hecht, 2010; Moyer-Gusé, 2008).
Importantly, central to this process is the audience’s emotional involvement in the nar-
rative (Slater & Rouner, 2002). Growing evidence indicates the importance of emotional
involvement in narrative persuasion (e.g., Busselle & Bilandzic, 2009; Morgan et al.,
2009; Stephenson, 2001). Little is known about what characteristics of narrative messages
foster emotional involvement, however.
Heightening perceived realism of a narrative may be one way of facilitating emotional
involvement in the narrative (Larkey & Hecht, 2010). But because perceived realism may
be multidimensional, research has to investigate which of the dimensions of perceived real-
ism may foster emotional involvement.
The judgment of plausibility, typicality, or factuality may not be central to emotional
involvement. For example, Green and Brock (2000) found that whether a narrative was
labeled as fiction or fact did not influence the level of involvement in the narrative, sug-
gesting that factuality is not critical to emotional involvement. Pouliot and Cowen (2007)
found that exposure to fictional narratives (i.e., films) elicited a stronger emotional arousal
than factual narratives (i.e., documentaries). Hall (2003) suggests that the lack of plausibil-
ity, typicality, or factuality in a narrative may not prohibit emotional involvement: For
example, the movie Forrest Gump is not plausible, typical, or factual, but Hall’s partici-
pants reported having emotional responses to it. Collectively, these findings suggest that
plausibility, typicality, and factuality may not be central to emotional involvement.
Existing theory and research suggest that narrative consistency and perceptual quality
are more relevant to emotional involvement than plausibility, typicality, and factuality. For

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

8 Communication Research XX(X)

example, high narrative consistency of Buffy the Vampire Slayer motivated some partici-
pants of Hall’s (2003) focus groups to follow the series. Somewhat similarly, Slater and
Rouner (2002) posited that “story line appeal” would predict emotional involvement.
Perceptual quality may also facilitate emotional involvement (Slater & Rouner, 2002). For
example, irrespective of its plausibility, typicality, or factuality, viewers of the film Jurassic
Park are more likely to be absorbed into the actions and events of the movie narrative and
empathize with the predicaments of the protagonist family when the dinosaurs look real
than not.
Overall, plausibility, typicality, and factuality address the likelihood of occurrence of
the story in reality. Relative to plausibility, typicality, and factuality, the dimensions of nar-
rative consistency and perceptual quality address the quality of the story itself more than
the likelihood of the story’s occurrence in reality. High likelihood of occurrence may not
be sufficient to make a story emotionally involving. For example, a story can be high in
typicality and factuality by conveying an event that is widely distributed in society and has
actually happened. Such depiction of routines of everyday life may not necessarily be
interesting or stimulate emotional interests from the audience.
It is possible that stories high in narrative consistency and perceptual quality can be low
in plausibility, typicality, and factuality. But such stories, that describe the reality well with
its internal structure and perceptual properties, may be better able to gain emotional inter-
ests of the audience than those that are merely high in likelihood of occurrence (i.e., high
in plausibility, typicality, and factuality). Thus, we hypothesize that narrative consistency
and perceptual quality are more likely to predict emotional involvement than other dimen-
sions (Hypothesis 3a–b).

Message Evaluation
In addition to emotional involvement and identification, message evaluation may play a
role in the process and effects of narrative persuasion (Slater et al., 2006; Slater & Rouner,
1996, 2002). Specifically, Slater and Rouner (2002) posit that narrative involvement and
identification directly predict “net polarity of responses regarding embedded persuasive
content,” which then predicts attitudes/behaviors (p. 178). Slater et al.’s postulation on the
net polarity of responses regarding embedded persuasive content may suggest that the
audience thinks about the persuasive content and evaluates the narrative in relation to its
persuasive content at some point during the process leading to persuasion. Furthermore,
the postulation may suggest that message evaluation is an important variable to examine
in narrative persuasion research.
Message evaluation is conceptually defined as the assessment of “persuasive potential”
of the message (Dillard, Weber, & Vail, 2007, p. 617). Colloquially, message evaluation
may concern how good a story is. Slater and Rouner (1996, 2002) have operationalized the
net polarity of responses as the valence of thoughts on persuasive content qualitatively
indicated in the thought-listing task (Cacioppo & Petty, 1981). Importantly, Slater and
Rouner (1997, as cited in 2002, p. 184) found that the qualitatively indicated favorable
thoughts on the persuasive content were positively associated with quantitatively assessed
message evaluation.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 9

Message evaluation may differ from pro- or counterarguments, although the distinction
has not always been clear in previous research. The focus of the message evaluation is the
message and its “persuasive potential,” where the focus of pro- or counterarguments is the
issue advocated in the message (Dillard, Weber, & Vail, 2007; Slater & Rouner, 1996). It
may be useful to investigate the role of message evaluation over cognitive elaboration (i.e.,
counter- or pro arguments) in the narrative persuasion process. Cognitive elaboration during
message processing is relevant to argument-based persuasion, but such cognitive elabora-
tion during message processing is less likely during narrative-based persuasion (Green &
Brock, 2000; Slater & Rouner, 2002). Research found that counterarguing, negatively
valenced cognitive elaboration, is suppressed during narrative message processing (e.g.,
Moyer-Gusé & Nabi, 2010; Slater & Rouner, 1997, 2002; also Green & Brock, 2000). But
Slater and Rouner (1996) argue that even the counterattitudinal audience may still notice the
quality of the message even while counterarguing against the issue advocated. Existing find-
ings suggests that positive evaluations of the quality of the narrative may mediate persua-
sion, even among counterattitudinal audiences (Slater & Rouner, 1996; Stephenson, 2002).
Message evaluation has been an important construct in the general persuasion literature.
Dillard, Shen, and Vail (2007) demonstrated that message evaluation is a causal antecedent
of attitudes, suggesting that influencing message evaluation may be one way to increase
persuasion. Previous research found that perceived realism influenced message evaluation
(e.g., Cho & Boster, 2008; Fishbein, Hall-Jamieson, Zimmer, von Haeften, & Nabi , 2002).
Research, however, has not examined which specific dimensions of perceived realism pre-
dict message evaluation.
Typicality may matter to the evaluation of narrative messages because it concerns the
degree to which the story is representative of the commonly understood reality. In compari-
son, the perception that the story could happen (i.e., plausibility), or has actually happened
(i.e., factuality), may not provide a sufficient reason to deem it of high quality. Narrative
consistency may also foretell the quality of the narrative. According to Fisher (1987), struc-
tural coherence of a narrative is one of the criteria that the audience employs to evaluate
whether the narrative has “good reasons,” which then provides assurance for adhering to
the advice offered in the narrative. In addition, perceptual quality may predict a positive
evaluation of the narrative. In Andsager et al.’s (2001) study, many college students noted
the low production quality of antialcohol messages, while distinguishing the production
quality from the persuasive intent of the messages. Slater and Rouner (2002) postulated
that production quality of a narrative influences the audience’s absorption to the narrative,
which in turn influences persuasion by the narrative.
Compelling and convincing audio, visual, and other sensory elements of a mediated
narrative message may bolster the audience’s judgment of the quality of the narrative. We
hypothesize that typicality, narrative consistency, and perceptual quality will predict mes-
sage evaluation (Hypothesis 4a–c)

Attitudes
Identification, emotional involvement, and message evaluation fostered by perceived realism
may predict attitudes. Research suggests that identification encouraged by exposure to

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

10 Communication Research XX(X)

narratives produce persuasion. For example, identification with a narrative character pre-
dicted intentions to perform the behavior exhibited by the character (i.e., smoking; Dal Cin,
Gibson, Zanna, Shumate, & Fong, 2007). Research has indicated that emotional involve-
ment, induced by exposure to narratives, predicts persuasion. For example, sympathetic
emotional response to individuals and events in narrative antidrug messages has predicted
antidrug attitudes (Stephenson, 2001). Similarly, emotional involvement in narratives about
organ donation predicted intentions to become an organ donor (Morgan et al., 2009). Finally,
Dillard, Weber, and Vail’s (2007) meta-analysis found that positive evaluation of persuasive
messages predicts attitudes toward the behavior advocated in the messages. We hypothesize
that identification, emotional involvement, and message evaluation predicted by the posited
dimensions of perceived realism, in turn, will predict attitudes (Hypothesis 5a–c).

Method
Design and Messages

Overall, a 3 (topic) × 5 (message) mixed design was employed. The first factor, topic, was
between-subjects and the second factor, message, was within-subjects. Each participant
was randomly assigned to one of the three topic conditions: antidrinking PSAs, antimeth
PSAs, or beer ads. The antimeth PSAs were from the Montana Meth Project. Each par-
ticipant viewed five different messages of the same topic. The order of presentation of the
messages was randomized.
The three topics were employed to have variance in perceptions of realism and responses
related to the dependent measures. From a pool of narrative messages addressing the three
topics, five messages were chosen for each topic with a goal of eliciting a range of per-
ceived realism and its five posited subdimensions. For example, one of the antidrinking
PSAs begins with a party scene. An excessively drunk man trips and falls on a rug, collid-
ing with a pregnant woman, causing her to hit her belly on the edge of a counter. The
woman and her husband are then seen in the doctor’s office, where the doctor is telling
them that they lost the baby. The scene ends with the woman crying and embracing her
husband. In an antimeth PSA, the scene begins by panning a young girl on a bed in a motel
with nothing but her underwear on. An older man is seen zipping his pants and leaving the
room handing something to a younger man outside the door. The voiceover says, “I love
my boyfriend. We’ve been together since like the 8th grade. He’s my best friend. He takes
care of me” as the boyfriend goes and sits next to the girl who is crying with smeared
makeup. The scene ends with him opening up the bag of meth given to him by the older
man and starting to use it. In a beer advertisement (ad), three men are drinking at a bar.
After all of them have taken a sip, they start walking backward through time. The scenes
keep changing as they go further and further back in time, including being cavemen frozen
during the ice age, being apes and dinosaurs, and finally emerging as little amoeba.
Although PSAs and ads may represent minimal-length narratives, the fact that skilled
advertisers can produce such narratives that are meaningful in situation and character illus-
trates the audience’s inclination to process narratives (Slater & Rouner, 2002).

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 11

Participants and Procedures

Participants (N=315) were recruited from communication classes at a large Midwestern
university. Participants’ ages ranged from 18 to 45 (M=20.39, SD=2.43). About 50.8% of
the participants were male. The majority of participants were White, comprising 81.7%.
The remainder consisted of African Americans (2.9%), Asian (11.6%), Hispanic (2.9%),
and other (1%). Participants received extra credit.
The study was conducted in a laboratory using Medialab software. Each experimental
session was attended by approximately 8 to 10 students. Participants were informed about
the purpose, procedure, and anonymous nature of the study prior to the study. Participants
viewed the messages individually in a booth equipped with a headset. After viewing each
ad, participants completed measures of perceived realism of the messages and dependent
variables. Each of the orders of presentation of perceived realism measures and dependent
measures was randomized. Each session lasted for about 45 min.

Measures
All measures were given on a Likert-type scale ranging from 1 = strongly disagree to
7 = strongly agree unless noted otherwise. Items were developed to measure the posited
five dimensions of perceived realism: plausibility, typicality, factuality, narrative consis-
tency, and perceptual quality. Hall’s (2003) synthesis of previous conceptualizations of
perceived realism and reports of verbatim quotes from her focus group participants
served as the basis of developing items. See the appendix for a full description of the
perceived realism measures. Table 1 presents means, standard deviations, and reliability
alphas for all measures.
Emotional involvement was assessed with four items based on Hall (2003) and Morgan
et al. (2009). The items were as follows: “The ad had an effect on my emotions,” “The ad
made me emotional,” “I could feel the emotions the event portrayed,” and “I was emotional
while watching the ad.”
Identification was measured with four items based on Campbell and Babrow (2004).
Items were as follows: “I could relate to the situation described in the ad,” “I could identify
with what was shown in the ad,” “I could relate to the event portrayed in the ad,” and “I
could relate to what was shown in the ad.”
Message evaluation was assessed with Cacioppo, Petty, and Morris’s (1983) five-item
scale. Items were as follows: “I feel that the ad was effective,” “This ad was persuasive,”
“I feel that the ad made its point effectively,” “I feel that the ad was convincing,” and “I feel
that the ad was compelling.”
Attitude toward alcohol use was measured with the following three sets of bipolar
adjectives: “good/bad,” “pleasurable/not pleasurable,” “fun/not fun.” Attitude toward meth
use was measured with the same three sets of adjectives. The response scale ranged from 1
to 7 in which higher scores indicated more negative attitudes toward using the substance.
In data analysis, the values obtained from the beer ad condition were recoded so that all
associations could be in the same direction.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

12 Communication Research XX(X)

Table 1. Zero-Order Correlations between Perceived Realism Dimensions, Identification,

Emotional Involvement, Identification, Message Evaluation, and Attitudes

Variables M SD 1 2 3 4 5 6 7 8 9
1 Plausibility 4.61 1.61 .90
2 Typicality 3.43 1.45 .64 .84
3 Factuality 3.84 1.54 .65 .58 .89
4 Narrative consistency 5.27 1.10 .32 .19 .27 .82
5 Perceptual quality 5.31 1.17 .48 .35 .40 .55 .90
6 Emotional involvement 4.20 1.60 .38 .29 .33 .29 .38 .92
7 Identification 2.83 1.74 .29 .39 .29 .14 .22 .12 .95
8 Message evaluation 5.23 1.33 .40 .34 .30 .52 .56 .49 .24 .94
9 Attitudes 2.81 1.71 .05 .06 .05 –.04 –.01 –.22 –.23 –.07 .81
Note: Alpha reliabilities on the diagonal. Top five rows are factor correlations.

Results
Data Analysis Strategy
Recall that each participant was randomly assigned to a topic and then saw the five mes-
sages presented in a random order. Thus in this mixed design, message-specific variables
such as perceived realism, emotional involvement, identification, and message evaluation
were repeated across different measurement occasions defined by messages nested within
each message topic. Attitudes were assessed at the message topic level. To explore the
dimensions of perceived realism and to test our hypotheses, the multilevel analysis
approach was adopted, in general, with individuals at the higher level and measurement
occasions as repeated measures clustered within individuals. Multilevel confirmatory fac-
tor analyses (CFA) were performed to assess the factor structure of perceived realism in
LISREL 8.80 (Mels, 2004). Multilevel structural equation modeling (SEM) was per-
formed to assess the contributions of perceived realism dimensions to attitudes through
identification, emotional involvement, and message evaluation (both direct and indirect
effects; see Figure 1). In both multilevel CFA and SEM, demographic variables and mes-
sage topics were controlled for.

Factor Structure of Perceived Realism

Hypothesis 1 predicted that the items for perceived realism would have a five-factor
structure. Individuals’ responses to each message (five in total) regarding the items for
the components of perceived realism, namely, plausibility, typicality, factuality, narra-
tive consistency, and perceptual quality, were submitted to LISREL 8.80 for multilevel
CFA, with the individual specified as the group variable. Basically, a two-group analysis
was performed. The difference between a typical two-group SEM model and a multilevel

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 13

Table 2. Fit Indices of the Multilevel Confirmatory Factor Analyses Models

Model df χ2 RMSEA
First-order single factor 378 10,577.54 .19
First-order five factor 359 1,838.88 .07
Second-order single factor 368 2,198.43 .09
Second-order two factora 355 1,984.84 .09
Note: aOne factor comprises plausibility, typicality, and factuality; the other comprises narrative consis-
tency and perceptual quality.

SEM model lies in that the two groups in the former case are independent of each other;
whereas the two groups in the latter are not: Group 1 estimates the model that is within-
subjects and Group 2 estimates the model that is between-subjects (with the grouping
variable designated). Four multilevel CFA models were estimated: (a) a single-factor
model, where all the items were specified to load on a single latent factor (i.e., perceived
realism); (b) a multidimensional model, where the items were specified to load on the
respective dimensions only, and the five dimensions were allowed to be associated with
each other; (c) a second-order single-factor model, where the five dimensions in model
(b) were specified to load on a second-order factor (perceived realism), rather than asso-
ciated with each other; and (d) a two-factor second-order model, where plausibility,
typicality and factuality were specified to load on one second-order factor and narrative
consistency and perceptual quality on the other. The two second-order factors were
allowed to be associated with each other. Table 2 presents the model-fit indices of these
multilevel CFA models. These results showed that the first-order five-factor model
yielded the best fit to the data and was obtained as the factor model. Figure 2 presents
the factor structure of that obtained model. The associations among these five factors,
along with their means and standard deviations are presented in Table 1 (top five rows).
These results demonstrated evidence in favor of Hypothesis 1.

The Role of Perceived Realism in Persuasion

Table 1 presents the zero-order correlations between these variables. These correlations
might be biased because the interdependency in the data structure due to repeated mea-
sures was not controlled for in zero-order correlations. The standard errors for parameter
estimates would have been inaccurate and significance tests biased in unknown ways if
such interdependence had not been controlled for. Such biases can be addressed in multi-
level analyses, which we applied in the current study. The zero-order correlation matrix is
provided for informational purposes only.
To test Hypothesis 2 to Hypothesis 5, multilevel structural equation modeling analyses
were performed.1 The five perceived realism dimensions were specified as exogenous vari-
ables and allowed to be associated with each other. Emotional involvement, identification,

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

14 Communication Research XX(X)

Item 1

.93 Item 2
.96
Plausibility .88 Item 3
.57
.77 Item 4

Item 5

Item 1
.50
Item 2
Typicality .59
.74
Item 3

Item 1
.70
Factuality .77 Item 2
.74
Item 3

Item 1

.67 Item 2
Narrative .69
Consistency .56 Item 3
.53
.58
Item 4

Item 5

Item 1
.75
Production .87 Item 2
Quality .87
.62 Item 3
.69
Item 4

Item 5

Figure 2. Factor structure of perceived realism dimensions

message evaluation, and attitudes were specified as endogenous variables. Figure 3 pres-
ents the obtained model: χ2 =71.22, df = 24, RMSEA = .035. All paths in Figure 3 were
statistically significant (p < .05).

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 15

Emotional
Plausibility .19 -.10 Attitude
involvement

Typicality
.11

-.15
.09
.32
Factuality

.30

-.08
.19
.07

Narrative Identification
Consistency
.11
.27
.07

Perceptual Message
.27
Quality Evaluation

Figure 3. Model test results

Note: Associations between the exogenous variables are not shown in the figure for the sake of clarity.

Hypothesis 2 predicted that typicality is more likely to predict identification than the
other four dimensions. Results from the multilevel SEM model showed that the magnitude
of the path from typicality to identification was .32. Plausibility and narrative consistency
did not have a direct path to identification. The direct paths from factuality and perceptual
quality were both .07. Pairwise comparisons of these coefficients were both significant at
p < .001, supporting Hypothesis 2.
Hypothesis 3 predicted that narrative consistency and perceptual quality are more likely
to predict emotional involvement than the other three dimensions. The direct path from nar-
rative consistency to emotional involvement was .09 and that from perceptual quality to
emotional involvement was .19. Typicality did not have a direct path to emotional involve-
ment. However, the direct path from plausibility to emotional involvement was .19 and that
from factuality to emotional involvement was .11. The path from plausibility to emotional
involvement (.19) was significantly greater than the path from narrative consistency to emo-
tional involvement (.09, p < .001). Therefore, Hypothesis 3 was not supported.
Hypothesis 4 predicted that typicality, narrative consistency, and perceptual quality
would predict message evaluation. The direct paths from narrative consistency to message
evaluation and from perceptual quality to message evaluation were both .27. No significant
association between typicality and message evaluation was found. In addition, emotional
involvement (.30) and identification (.11) emerged as significant predictors of message
evaluation. These results provided partial support for Hypothesis 4.
Hypothesis 5 predicted that identification, emotional involvement, and message evalu-
ation would predict attitudes. The direct path from identification to attitudes was –.15, the
path from emotional involvement to attitudes was –.10, and the path from message evalu-
ation to attitudes was –.08. Hypothesis 5 was supported.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

16 Communication Research XX(X)

Discussion
The results show that perceived realism is a multidimensional construct. The dimensions
of perceived realism include plausibility, typicality, factuality, narrative consistency, and
perceptual quality. The results also show that the different dimensions of perceived realism
are differentially related to the mediators (identification, emotional involvement, message
evaluation) and outcome (attitudes) of narrative persuasion.
Although most of the results were consistent with our expectations, some findings
were not. Whereas we predicted that narrative consistency and perceptual quality would
be stronger predictors of emotional involvement than other perceived realism dimen-
sions, plausibility emerged as a stronger predictor of emotional involvement than narra-
tive consistency. This finding may suggest that plausibility, the most fundamental form
of perceived realism, is necessary to stimulate the audience’s emotional involvement in
narrative messages in addition to perceptual quality and narrative consistency. Whereas
we predicted that typicality, narrative consistency, and perceptual quality would produce
positive message evaluation, no significant association between typicality and message
evaluation was found. Only narrative consistency and perceptual quality predicted posi-
tive message evaluation. The prevalence of a narrative event in the real world does not
seem to be a criterion that the audience employs to evaluate the persuasive potential of
the narrative message.

Dimensions and Roles of Perceived Realism

Overall, these findings offer an intriguing first look at the roles of perceived realism in the
narrative persuasion process. The results demonstrate that plausibility and typicality play
distinct roles in narrative persuasion. Plausibility was associated with emotional involve-
ment, but not with identification or message evaluation. Thus, plausibility predicted nar-
rative persuasion only indirectly through emotional involvement.
In contrast, typicality showed no association with emotional involvement but predicted
identification. Therefore, typicality differs from plausibility. These results may suggest
that plausibility, a basic criterion with which the audience judges realism (Hall, 2003), is
necessary to promote emotional involvement, but not sufficient to foster identification.
However, typicality, a more rigid criterion of realism than plausibility (Hall, 2003), may
be necessary to promote identification, but not emotional involvement.
Together, these results suggest that the emotional involvement- and identification-
mediated routes to narrative persuasion are predicted by different dimensions of perceived
realism. Factuality differed from plausibility and typicality in that it showed associations
with both emotional involvement and identification. However, the associations were smaller
in size than the associations that plausibility and typicality exhibited with emotional involve-
ment and identification. These results may suggest that the role of factuality in narrative
persuasion may be less important than the role of plausibility and identification.
Broadly, narrative consistency and perceptual quality behaved differently from plausi-
bility, typicality, and factuality. Narrative consistency and perceptual quality directly pre-
dicted message evaluation, whereas plausibility, typicality, and factuality showed no direct

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 17

association with message evaluation. Plausibility, typicality, and factuality predicted mes-
sage evaluation only indirectly by means of emotional involvement. In contrast, narrative
consistency and perceptual quality indirectly predicted message evaluation through emo-
tional involvement, in addition to the direct associations.
The role of perceptual quality found in this study may be noteworthy. For example, the
association between perceptual quality and emotional involvement was larger in size than
the association between narrative consistency and emotional involvement. The results
may suggest that the perceptual or sensory quality of narrative messages cannot be
neglected in persuasion efforts. It is possible that the results are because of this study’s use
of very brief advertising-formatted narratives, which may have directed the audience’s
attention to more perceptual quality than narrative consistency. But it is possible that per-
ceptual quality enhances longer narratives, such as movies, as well. Dramas with gritty
perceptual quality may be better able to foster absorption than those with compromised
perceptual quality. Whereas narrative consistency is not visible or audible, perceptual
quality is. These considerations may be relevant to explaining the stronger association
with emotional involvement found for perceptual quality than for narrative consistency.
Furthermore theorizing about perceptual quality, as well as empirical replication and
extension about its role in other formats of narrative messages, is in order.
Because this study employed cross-sectional data, the assumption of the causal order of
the variables should be checked. To address the possibility of a reverse causation, we tested
an alternative model in which emotional involvement, identification, and message evalua-
tion predicted perceived realism dimensions, which in turn predicted attitudes. A test of the
model indicated a poor fit: RMSEA=.31 NFI=.72, CFI=.72, RMR=.29, GFI=.78. Thus,
evidence for a reverse causation was not found.

Implications
Communication theory and research should be able to explain not only the process of nar-
rative persuasion, but also the characteristics of narratives that are responsible for the
process. Scientific and pragmatic impact of the knowledge regarding the process of narra-
tive persuasion will be enhanced when it is combined with knowledge about the factors
predicting the process. By examining the dimensions and roles of perceived realism, this
study sought to advance the understanding of both the characteristics of narrative mes-
sages and the process of narrative persuasion that they facilitate.
This study may add to the current theory of narrative persuasion in the following ways.
The results may suggest that there are three routes to narrative persuasion and that differ-
ent perceived realism dimensions differentially activate the routes. Although a growing
body of research has documented the importance of emotional involvement in the process
of narrative persuasion, little research has examined the characteristics of narratives that
activate emotional involvement. The results of this study show that plausibility is impor-
tant to facilitate emotional involvement. Similarly, although existing theory considers
identification central to narrative persuasion, few studies empirically examined the spe-
cific characteristic of narratives that facilitate identification. The results show that typical-
ity is important to fostering identification.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

18 Communication Research XX(X)

Another route to narrative persuasion found in this study is via message evaluation,
lending support for Dillard, Shen, & Vail (2007) postulation that influencing message
evaluation can be a useful persuasion strategy. Evaluation of narratives may also be rele-
vant to Fisher’s (1987) concept of “good reasons,” which provide grounds for changing
beliefs or actions. The perceived realism dimensions of narrative consistency and percep-
tual quality directly predicted message evaluation. Narrative consistency and perceptual
quality may be concerned more with the characteristics of the narrative itself, whereas the
dimensions of plausibility, typicality, and factuality may be concerned more with the nar-
rative’s connection to reality. The results of this study suggest that message evaluation
may be influenced more by narratives’ internal structure and flow and sensory quality
than the narratives’ external allegiance to reality. Future research should further investi-
gate the role of message evaluation in narrative persuasion. Particularly interesting may
be to explore its role in facilitating what Larkey and Hecht (2010) called “social prolifera-
tion” in which people share or relay good stories, which in turn fosters the diffusion of
persuasive content. Online “viral” messages are an example of social proliferation of
narratives (Larkey & Hecht, 2010).
These findings offer implications for narrative message design. Specifically, the
implications are that depending on the routes to narrative persuasion, different perceived
realism dimensions should be emphasized. The results may also improve perceived real-
ism research in related domains including media uses and effects. For example, in media
uses research, perceived realism is conceptualized to predict the media audience’s para-
social relationship with media figure (Horton & Wohl, 1956; Giles, 2002). In media
effects research, cultivation theory (Gerbner & Gross, 1976) is predicated on the assump-
tion that television content is perceived as real by viewers. The multidimensionality of
perceived realism found in this study may suggest avenues to refine the current concep-
tualizations about the role of perceived realism in these frameworks of media uses and
effects. In addition, the perceived realism dimensions may be usefully employed in
media literacy education (e.g., Austin, Pinkleton, & Funabiki, 2007) to assist the audi-
ence with discerning the difference between the media world and the real world.
The instruments developed in this study represent an advance over the existing global
measures of perceived realism. Future research involving perceived realism would benefit
by identifying the specific facets of perceived realism and by employing the instruments
validated in this study. The instruments can assist investigators by further examining the
dimensions’ relationships with other variables pertinent to the process of narrative persua-
sion. Formative evaluation of narrative messages can be assisted with the instruments vali-
dated in this study. Message designers can use the instruments as a diagnostic tool to gauge
whether the message activates the intended dimensions of perceived realism corresponding
to the intended routes to narrative persuasion.

Limitations and Suggestions for Future Research

Because this study was conducted in a lab with college students, its external validity may
be limited. Additional validation studies of the scale developed in this study would benefit

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 19

from involving more representative samples in more natural settings. This study used
PSAs and ads as experimental stimuli; the topics focused on antidrug and beer attitudes.
Future research can benefit by investigating different formats and topics of narrative mes-
sages. Factors such as channels and genres of media can also be considered in future
research. Due to the short length of the PSAs and ads, the intensity and duration of emo-
tional involvement and identification might have been limited as compared to entertainment-
education. However, as Slater and Rouner (2002) note, “if predicted effects can be found
in such impoverished stimuli, we can certainly anticipate finding similar but stronger
effects in well-developed entertainment-education stimuli” (p. 183).
The results show that the associations between the theoretical predictors and attitudes
were small in size. The effect sizes should be interpreted in context. This study employed
short narratives including PSAs and ads, which are typically 30 s in length. This length
pales to that of longer narratives such as television dramas and movies. The topics of the
messages were drug and alcohol use. Research in various fields has suggested that attitudes
toward these behaviors may be a function of not only exposure to persuasive messages but
also various sociocultural factors. Nonetheless, antidrug PSAs and alcohol ads are a highly
visible form of communication strategy in everyday media environments. The examination
of the perceived realism of the PSAs and ads is especially important because these mes-
sages are frequently perceived to present unrealistic outcomes (Andsager et al., 2001).
Although this study shows that perceived realism is one of the characteristics of PSAs and
ads determining their effectiveness, given the small effects, other characteristics influenc-
ing the effectiveness should also be investigated.
With the scale developed in this study, future research should experimentally manipu-
late the different dimensions of perceived realism. Using the measures, investigators can
select stimulus materials that vary in specific dimensions of perceived realism to further
examine their roles in the process of persuasion.
Finally, whereas identification is about the perceived connection between the audience
and the character of a narrative, the measurement of identification in this study assessed the
perceived connection with the situation depicted in the narrative. Although the situation of
the character is a primary determinant of the identification with the character (Trope,
1986), the measure of identification in this study is less than desirable and future research
should address this limitation.

Conclusion
This study may improve the current understanding of perceived realism and narrative
persuasion. The results show that plausibility, typicality, factuality, narrative consistency,
and perceptual quality are distinct dimensions of perceived realism. Plausibility predicted
the indirect route mediated by emotional involvement, while typicality predicted the indi-
rect route mediated by identification. Narrative consistency and perceptual quality pre-
dicted the indirect route mediated by message evaluation. These findings provide
implications for narrative persuasion theory, narrative message design, perceived realism
measurement, and media uses and effects research.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

20 Communication Research XX(X)

Appendix
Items for the Perceived Realism Scale
Plausibility

1. The ad showed something that could possibly happen in real life.

2. The event in the ad portrayed possible real-life situations.
3. The story in the ad could actually happen in real life.
4. Never in real life would what was shown in the ad happen.
5. Real people would not do the things shown in the ad.

Typicality
1. Not many people are likely to experience the event portrayed in the ad.
2. The ad portrayed an event that happens to a lot of people.
3. What happened to the people in the ad is what happens to people in real world.

Factuality
1. The ad was based on facts.
2. The ad showed something that had really happened.
3. What was shown in the ad had actually happened.

Narrative Consistency
1. The ad showed a coherent story.
2. The story portrayed in the ad was consistent.
3. Parts of the ad were contradicting each other.
4. The story portrayed in the ad made sense.
5. The event in the ad had a logical flow.

Perceptual Quality
1. The visual elements of the ad were realistic.
2. The audio elements of the ad were realistic.
3. The acting in the ad was realistic.
4. The scenes in the ad were realistic.
5. I felt that the overall production elements of the ad were realistic.

Acknowledgments
The authors wish to thank the two anonymous reviewers of the manuscript for their insightful
comments that improved this manuscript substantially.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship,
and/or publication of this article.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 21

Funding

The authors received no financial support for the research, authorship, and/or publication of this
article.

Note
1. Although demographic variables and message topics were controlled in both multilevel CFA and SEM,
some readers might be interested to know if there are differences among the topics and between genders
in the hypothesized associations (Hypothesis 2 to Hypothesis 5). Additional subanalyses per topic and
gender were conducted, for a total of six separate multilevel SEM tests. There were minor variations in
the hypothesized associations across the three topics. There was no gender difference within each topic.
Detailed results are available upon request.

References
Adaval, R., & Wyer, R.S. (1998). The role of narrative in consumer information processing. Journal of
Consumer Psychology, 7, 207-245
Andsager, J. L., Austin, E. W., & Pinkleton, B. E. (2001). Questioning the value of realism: Young adults’
processing of messages in alcohol-related public service announcements and advertising. Journal of
Communication, 51, 121-142.
Appel, M., & Richter, T. (2010). Transportation and need for affect in narrative persuasion: A mediated
moderation model. Media Psychology, 13, 101-135.
Austin, E. W., Pinkleton, B. E., & Funabiki, R. P. (2007). The desirability paradox in the effects of media
literacy training. Communication Research, 34, 483-506.
Basil, M. D. (1996). Identification as a mediator of celebrity effects. Journal of Broadcasting & Electronic
Media, 40, 478-495.
Brehm, J. W. (1966). A theory of psychological reactance. New York, NY: Academic Press.
Brinson, S. L, & Brown, M. H. (1997). The AIDS risk narrative in the 1994 CDC campaign. Journal of
Health Communication, 2, 101-112.
Busselle, R., & Bilandzic, H. (2008). Fictionality and perceived realism in experiencing stories: A model of
narrative comprehension and engagement. Communication Theory, 18, 255-280.
Busselle, R., & Bilandzic, H. (2009). Measuring narrative engagement. Media Psychology, 12, 321-347.
Busselle, R. W., & Greenberg, B. S. (2000). The nature of television realism judgments: A reevaluation of
their conceptualization and measurement. Mass Communication & Society, 3, 249-268.
Cacioppo, J. T., & Petty, R. E. (1981). Social psychological procedures for cognitive response assessment:
The thought-listing technique. In T. V. Merluzzi, C. R. Glass, & M. Genest (Eds.), Cognitive assessment
(pp. 309-342). New York, NY: Guilford.
Cacioppo, J. T., Petty, R. E., & Morris, K. J. (1983). Effects of need for cognition on message evaluation,
recall, and persuasion. Journal of Personality & Social Psychology, 45, 805-818.
Campbell, R. G., & Babrow, A. S. (2004). The role of empathy in responses to persuasive risk communication:
Overcoming resistance to persuasive HIV prevention messages. Health Communication, 16, 159-182.
Chang, C. (2008). Increasing mental health literacy via narrative advertising. Journal of Health Communi-
cation, 13, 37-55.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

22 Communication Research XX(X)

Cho, H., & Boster, F. J. (2008). First and third person perceptions on anti-drug ads among adolescents. Com-
munication Research, 35, 169-189.
Cohen, J. (2001). Defining identification: A theoretical look at the identification of audiences with media
characters. Mass Communication & Society, 4, 245-264.
Dal Cin, S., Gibson, B., Zanna, M. P., Shumate, R., & Fong, G. T. (2007). Smoking in movies, implicit
associations of smoking with the self, and intentions to smoke. Psychological Science, 18, 559-563.
Dillard, J. P., & Shen, L. (2005). On the nature of reactance and its role in persuasive health communication.
Communication Monographs, 72, 144-168.
Dillard, J. P., Shen, L., & Vail, R. G. (2007). Does perceived message effectiveness cause persuasion or vice
versa? 17 consistent answers. Human Communication Research, 33, 467-488.
Dillard, J. P., Weber, K. M., & Vail, R. G. (2007). The relationship between the perceived and actual effec-
tiveness of persuasive messages: A meta-analysis with implications for formative campaign research.
Journal of Communication, 57, 613-631.
Dorr, A. (1983). No shortcuts to judging reality. In J. Bryant & D. R. Anderson (Eds.), Children’s under-
standing of television: Research on attention and comprehension (pp. 199-220). New York, NY: Aca-
demic Press.
Dunlop, S. M., Wakefield, M., & Kashima, Y. (2010). Pathways to persuasion: Cognitive and experiential
responses to health promoting mass media messages. Communication Research, 37, 133-164.
Durkin, S., & Wakefield, M. (2008). Interrupting a narrative transportation experience: Program placement
effects on responses to antismoking advertising. Journal of Health Communication, 13, 667-680.
Escalas, J. E. (2004). Imagine yourself in the product: Mental stimulation, narrative transportation, and
persuasion. Journal of Advertising, 33, 37-48.
Feshbach, S. (1976). The role of fantasy in response to television. Journal of Social Issues, 32, 71-85.
Fishbein, M., Hall-Jamieson, K., Zimmer, E., von Haeften, I., & Nabi, R. (2002). Avoiding the boomerang:
The need for experimental tests of the relative effectiveness of anti-drug public service announcements
prior to their use in a national campaign. American Journal of Public Health, 92, 238-245.
Fisher, W. R. (1987). Human communication as narration: Toward a philosophy of reason, value, and
action. Columbia, SC: University of South Carolina Press.
Gerbner, G., & Gross, M. (1976). Living with television: The violence profile. Journal of Communication,
26, 172-194.
Giles, D. C. (2002). Parasocial interaction: A review of the literature and a model for future research. Media
Psychology, 4, 279-305.
Green, M. C. (2004). Transportation into narrative worlds: The role of prior knowledge and perceived real-
ism. Discourse Processes, 38, 247-266.
Green, M. C. (2008). Research challenges in narrative persuasion. Information Design Journal, 16, 47-52.
Green, M. C., & Brock, T. C. (2000). The role of transportation in the persuasiveness of public narratives.
Journal of Personality and Social Psychology, 79, 701-721.
Hall, A. (2003). Reading realism: Audiences’ evaluations of the reality of media texts. Journal of Com-
munication, 53, 624-641.
Hawkins, R. P. (1977). The dimensional structure of children’s perceptions of television reality. Communi-
cation Research, 4, 299-320.
Hoffner, C. & Buchanan, M. (2005). Young adults’ wishful identification with television characters: The
role of perceived similarity and character attributes. Media Psychology, 7, 325-351.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

Cho et al. 23

Horton, D., & Wohl, R. (1956). Mass communication and para-social interaction: Observation of intimacy
at a distance. Psychiatry, 19, 188-211.
Kreuter, M. W., Green, M. C., Cappella, J. N., Slater, M. D., Wise, M. E., Storey, D., . . . Woolley, S. (2007).
Narrative communication in cancer prevention and control: A framework to guide research and applica-
tion. Annals of Behavioral Medicine, 33, 221-235.
Larkey, L. K., & Hecht, M. L. (2010). A model of effects of narrative as culture-centric health promotion.
Journal of Health Communication, 15, 114-135.
Livingstone, S. M. (1990). Interpreting a television narrative: How different viewers see a story. Journal of
Communication, 40, 72-85.
Mels, G. (2004). LISREL 8.7 for Windows: Getting started guide. Lincolnwood, IL: Scientific Software
International, Inc.
Morgan, S. E., Movius, L., & Cody, M. (2009). The power of narratives: The effect of organ donation enter-
tainment television storylines on the attitudes, knowledge, and behaviors of donors and non-donors.
Journal of Communication, 59, 135-151.
Moyer-Gusé, E. (2008). Toward a theory of entertainment persuasion: Explaining the persuasive effects of
entertainment-education messages. Communication Theory, 18, 407–425.
Moyer-Gusé, E., & Nabi, R. L. (2010). Explaining the persuasive effects of narrative in an entertainment
television program: Overcoming resistance to persuasion. Human Communication Research, 36, 25-51.
Potter, W. J. (1986). Perceived reality and the cultivation hypothesis. Journal of Broadcasting & Electronic
Media, 30, 159-174.
Pouliot, L., & Cowen. P. S. (2007). Does perceived realism really matter in media effects? Media Psychol-
ogy, 9, 241-259.
Reeves, B. (1978). Perceived TV reality as a predictor of children’s social behavior. Journalism Quarterly,
55, 682-695.
Shapiro, M. A., & Chock, T. M. (2003). Psychological processes in perceiving reality. Media Psychology,
5, 163-198.
Singhal, A., Cody, M., Rogers, E. M., & Sabido, M. (2003). (Eds.). Entertainment education and social
change: History, research, and practice. Mahwah, NJ: Erlbaum.
Slater, M. D., & Rouner, D. L. (1996). Value-affirmative and value-protective processing of alcohol educa-
tion messages that include statistical or anecdotal evidence. Communication Research, 23, 210-235.
Slater, M. D., & Rouner, D. (1997, May). The processing of narrative fiction containing persuasive content
about alcohol use: Effects of gender and outcome. Paper presented at the annual meeting of the Inter-
national Communication Association, Montreal, Canada.
Slater, M. D. & Rouner, D. (2002). Entertainment-education and elaboration-likelihood: Understanding the
processing of narrative persuasion. Communication Theory, 12, 173-191.
Slater, M. D., Rouner, D., & Long, M. A. (2006). Television dramas and support for controversial public
policies: Effects and mechanisms. Journal of Communication, 56, 235-252.
Stephenson, M. T. (2001). Examining adolescents’ response to antimarijuana PSAs. Human Communication
Research, 29, 343-369.
Stephenson, M. T. (2002). Sensation seeking as a moderator of the processing of anti-heroin PSAs. Com-
munication Studies, 53, 358-380.
Trope, Y. (1986). Identification and inferential processes in dispositional attribution. Psychological Review,
93, 239-257.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013

24 Communication Research XX(X)

Zillmann, D. (1980). Anatomy of suspense. In P. H. Tannenbaum (Ed.), The entertainment functions of

television (pp. 133-163). Hillsdale, NJ: Erlbaum.

Bios
Hyunyi Cho, PhD, is an associate professor in the Brian Lamb School of Communication at
Purdue University.

Lijiang Shen, PhD, is an associate professor in the Department of Speech Communication at

the University of Georgia.

Kari Wilson, PhD, will be an assistant professor of Mass Communication in the Raclin School
of the Arts at Indiana University South Bend beginning the fall of 2012.

Downloaded from crx.sagepub.com at INDIANA UNIV on November 23, 2013