This article was downloaded by: [University of Bristol]

On: 21 February 2015, At: 08:54

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Communication Quarterly
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/rcqu20

An Examination of the Narrative

Persuasion with Epilogue through the
Lens of the Elaboration Likelihood Model
a b c
Rebekah Lane , Ann Neville Miller , Christopher Brown & Natalie
d
Vilar
a
Performing Arts Department , Full Sail University
b
Nicholson School of Communication , University of Central Florida
c
Communication Department , Valencia State College
d
Orlando Fashion Week
Published online: 22 Aug 2013.

To cite this article: Rebekah Lane , Ann Neville Miller , Christopher Brown & Natalie Vilar (2013) An
Examination of the Narrative Persuasion with Epilogue through the Lens of the Elaboration Likelihood
Model, Communication Quarterly, 61:4, 431-445, DOI: 10.1080/01463373.2013.799510

To link to this article: http://dx.doi.org/10.1080/01463373.2013.799510

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
 Communication Quarterly
Vol. 61, No. 4, September–October 2013, pp. 431–445

An Examination of the Narrative

Persuasion with Epilogue through
the Lens of the Elaboration
Likelihood Model
Rebekah Lane, Ann Neville Miller,
Downloaded by [University of Bristol] at 08:54 21 February 2015

Christopher Brown, & Natalie Vilar

This study synthesized components of the elaboration likelihood model with recent
theorizing on persuasion through narrative. Specifically, it examines the relationship
of perceived salience, transportation, and character identification with respect to narra-
tive, argument, and combination (narrative þargument) messages. Two hundred
fifty-five university students were randomly assigned to 1 of the 3 conditions, read a blog
about the institution of an exit exam policy at an unnamed university, and filled out an
online questionnaire. As anticipated, salience was found to mediate the relationship
between transportation and acceptance of the message when narrative messages were
followed by epilogues. Participants in the narrative condition were more accepting of
the message than participants in the argument condition.

Keywords: Character Identification; Elaboration Likelihood Model; Epilogues;

Narrative; Transportation

Various scholars have asserted that narrative persuasion does not operate in the same
way as argument-based persuasion (e.g., Cohen, 2001; De Wit, Das, & Vet, 2008;
Kopfman, Smith, Ah Yun, & Hodges, 1998; Slater & Rouner, 2002). Recipients of

Rebekah Lane (M.A., University of Central Florida, 2010) is an instructor in the Performing Arts Department at
Full Sail University. Ann Neville Miller (Ph.D., University of Georgia, 2005), is an assistant professor in the
Nicholson School of Communication at the University of Central Florida. Christopher Brown (M.A., Carnegie
Mellon, 2013) is an adjunct instructor in the Communication Department at Valencia State College. Natalie Vilar
(B.A., University of Central Florida, 2011) is a public relations assistant at Orlando Fashion Week. Correspon-
dence: Ann Neville Miller, Nicholson School of Communication, University of Central Florida, P.O. Box
161344, Orlando, FL 32816; E-mail: ann.miller@ucf.edu

ISSN 0146-3373 print/1746-4102 online # 2013 Eastern Communication Association

DOI: 10.1080/01463373.2013.799510
 432 R. Lane et al.
narratives, they suggested, are motivated by the desire to enjoy the characteristics of a
good narrative, rather than the self-interest that is typical of attention to argumenta-
tive persuasion. Thus, rather than issue involvement being key to persuasion, as is
posited by classic persuasion theories like the elaboration likelihood model (ELM;
Petty & Cacioppo, 1979), several theorists identify transportation (also referred to
as narrative involvement or absorption; Green & Brock, 2000; Slater & Rouner,
2002) and identification with characters (Cohen, 2001; Dunlop, Wakefield, &
Kashima, 2010; Moyer-Guse & Nabi, 2010) as crucial factors in the level of per-
suasion a narrative will generate. With persuasion processes in narrative and
non-narrative persuasion being construed as operating so differently, it has been dif-
ficult to integrate the two theoretically or to predict the impact of epilogues and other
combinations of the two types of persuasion.
We suggest that these seemingly disparate processes can be theoretically integrated
Downloaded by [University of Bristol] at 08:54 21 February 2015

via the predictions of ELM. According to that theory, if one wanted to maximize the
power of a single message, it would make sense to attempt to persuade an audience
through the central route. However, for messages that audience members are not likely
to view as relevant to themselves (i.e., those in which their issue involvement is low),
central route processing is unlikely. Means must be found to increase their sense of
issue involvement. Slater and Rouner’s (2002) extended elaboration likelihood model
(E-ELM) proposes that narrative and non-narrative persuasion are effected by differ-
ent processes. However, ELM itself may allow for a way of considering the two within a
single persuasive framework. It may be that exposure to a transportive, persuasive nar-
rative will increase the respondent’s perception of salience, or issue involvement. Once
issue involvement has been established, participants may process more centrally when
presented with a similarly themed non-narrative epilogue.
The central purpose of this study, therefore, was to investigate the relationship
between transportation, character identification, perceived salience, and acceptance
of a message presented via narrative followed by epilogue. In the process we set
out to supply a means of conceptualizing the persuasive effects of narratives, argu-
ments, and combination messages under a unified theoretical framework: ELM.

Literature Review
The ELM
The ELM purports to explain recipient reaction to persuasive messaging (Hinyard &
Kreuter, 2007). As a dual-process persuasion model, it predicts the likelihood that a
person will think about a persuasive message, as well as the outcome of the attempted
persuasion (Petty & Cacioppo, 1984). According to the ELM, a message may be pro-
cessed one of two ways: either centrally or peripherally (Petty, Cacioppo, &
Schumann, 1983). When a person processes a message through the central route,
he or she generally thinks carefully about the message and the arguments. The person
decides whether the arguments presented are convincing enough to lead him or her
to shift attitudes on the issue in question (Petty & Cacioppo, 1984). When processing
 Communication Quarterly 433

a message peripherally, a person relies on cues from the message that may or may not
have anything to do with the actual arguments. For example, a person may report a
more positive attitude toward a specific toothpaste brand after viewing an ad for that
toothpaste featuring an attractive spokesperson. Another example might be a person
choosing a hotel based on the number of amenities listed in an advertisement,
although the list includes standard items like air conditioning and television (Petty
& Cacioppo, 1981). The term elaboration in the theory derives from the amount of
issue-relevant thinking, or elaboration, that a person dedicates to a persuasive mess-
age. Much elaboration is the hallmark of central route processing; little elaboration is
peripheral processing. ELM does not predict that central processing will lead to any
greater initial persuasion than will peripheral processing. Persuasion through central
route processing is, however, thought to be less susceptible to decay and counterar-
guing following message exposure (Petty, Haugtvedt, & Smith, 1995).
Downloaded by [University of Bristol] at 08:54 21 February 2015

ELM suggests that individuals will only process a message centrally if they are both
sufficiently able and motivated to do so (Petty & Cacioppo, 1984). The factor that is
most frequently mentioned as inducing motivation is personal involvement in an
issue, or issue involvement (Petty & Cacioppo, 1979). Suggesting that a message con-
tains information that directly involves the goals and aspirations of an individual can
induce high issue involvement (Petty & Cacioppo, 1979). We are interested here in
issue involvement as a dependent variable—that is, how one constructs a persuasive
message in such a way as to increase issue involvement and thereby promote central
processing. To distinguish between these two types of involvement, we refer to our
dependent variable as perceived salience.
Although ELM has shown predictive value for effects of non-narrative persuasive
messaging, researchers have reported difficulty measuring narrative impact with the
scales and methods generally used in non-narrative ELM studies (Green & Brock,
2000; Slater & Rouner, 2002). For example, ELM suggests that issue-relevant thought
following a message is indicative of the persuasion that has taken place in the indi-
vidual (Petty & Cacioppo, 1979). This assertion is typically tested by asking parti-
cipants to list their thoughts immediately following message exposure, then coding
those thoughts according to message relevance and favorability toward the message
(Brock, 1967). Thoughts reported after exposure to narratives, however, tend to be
emotion- and story-relevant as opposed to issue-relevant and, therefore, have not
been useful for assessing central processing (Green & Brock, 2000).
Slater and Rouner (2002) have attempted to align narrative and non-narrative
persuasion theory by conceptualizing an alternate version of ELM for narrative per-
suasion efforts. Their E-ELM incorporates the concepts and logic found in the ELM
and uses them to guide investigations into narrative persuasion. E-ELM suggests that
whereas issue involvement is an important factor in non-narrative messaging,
involvement with the narrative is more important for effectiveness of a narrative.
In ELM terms, the likelihood that a person will process a narrative is a type of elab-
oration likelihood. According to Slater and Rouner, this is determined by two factors:
(a) the level of transportation, or absorption in the story world, and (b) the degree of
character identification, or response to characters within the narrative text.
 434 R. Lane et al.
Transportation
Theorizing on the effects of reader or viewer immersion into a story world began
nearly two decades ago when Gerrig (1993) coined the term transportation to dis-
tinguish between the individual’s experience in response to a narrative, as opposed
to the response to dense information-based messages. He suggested that narrative
and non-narrative messages inherently induced different experiences. Slater (1997)
later used the term engagement to describe the level of involvement an individual felt
with the storyline of the narrative. It was Green and Brock’s (2000) development of a
validated transportation scale, however, that gave impetus to the study of transpor-
tation as a key to narrative effects (Slater & Rouner, 2002). Their goal was to measure
emotional and cognitive responses, as well as the mental imagery and lack of aware-
ness of surroundings experienced by participants who read narrative messages (Green
Downloaded by [University of Bristol] at 08:54 21 February 2015

& Brock, 2000). Results of a series of four experiments indicated that transportation
into a narrative was associated with positive evaluations of the characters in the story,
as well as with belief change consistent with story assertions (Green & Brock, 2000).
Transportation has been successfully used to predict persuasion and attitude shifts
on a range of topics, particularly belief change that contradicts a participant’s general
ideology (Slater & Rouner, 2002). It has also been shown to influence belief change
on controversial public issues like the death penalty (Slater, Rouner, & Long, 2006)
and participants’ support and confidence in public knowledge (Appel & Richter,
2007). Transportation has also been proposed as an explanation for the finding that
audiences sometimes accept as truth even simple falsities that contradict common
knowledge when these statements are embedded in a narrative (Appel & Richter,
2007; Green & Brock, 2000). Green (2004) suggested that this may be because indi-
viduals seek out narratives for their entertainment value and personal enjoyment,
part of which occurs through transportation. Monitoring narratives for accuracy
and other details reduces the transportive experience, an action that may in turn
reduce the enjoyment of the narrative itself (Green & Brock, 2000).

Character Identification
In addition to transportation, researchers on narrative persuasion have cited charac-
ter identification as a mediating variable in the persuasive process. Character identi-
fication has long been considered an important factor in social cognitive theory,
based on the idea that people are more receptive to emulating behavior modeled
by people who are similar to them (Bandura, 1986). The current understanding of
character identification was articulated by Cohen (2001) in an attempt to compile
disparate findings and concepts into a single comprehensive definition. Cohen drew
on writings as early as Maccoby, Wilson, and Burton’s (1958) discussions of viewer
identification with film characters. He also included studies that focused on individ-
ual aspects of character identification, like isolating specific character traits (in this
case, aggressiveness), and viewer violence (Huesmann, Eron, Lefkowitz, & Walder,
1984). The resulting definition of identification with a character is ‘‘an imaginative
 Communication Quarterly 435

process invoked as a response to characters presented within mediated texts’’ (Cohen,

2001, p. 250).
Cohen (2001) described identification in terms of experiential processes, and dis-
tinguished it from perceived similarity, liking, and modeling, as these responses to
characters are more spectator-like in nature. Character identification, in contrast,
includes feeling with—not about—the character and internalizing the character’s
point of view. Similar to descriptions of transportation, this type of identification
involves loss of self-awareness in favor of awareness of the character, such that audi-
ence members feel as though they are actually experiencing the events in the story
with the character and internalizing those experiences, as opposed to simply watching
the character experience the events (Cohen, 2001). Cohen theorized that the experi-
ence of identifying with the character leads audience members to feel understanding
and empathy with the character in both emotional and cognitive ways. Thus, parti-
Downloaded by [University of Bristol] at 08:54 21 February 2015

cipants who expressed more identification with the characters have been found to
evidence less counter-arguing and greater perception of vulnerability (Moyer-Guse
& Nabi, 2010).
Our first hypothesis attempted to confirm the relationship proposed by previous
theorizing regarding the relationship of transportation and character identification
with perceived salience of the issue to participants:

H1: In the narrative and combination conditions, perceived salience will be positively
related to (a) transportation and (b) character identification.

Our second hypothesis extended that expectation to a comparison with

participants who received a non-narrative message:

H2: Persons in the narrative and combination conditions will perceive the message as
more salient than persons in the argument condition.

Following Narrative with Non-Narrative Persuasion

Sometimes narrative and non-narrative formats are used together in an effort to
maximize the effectiveness of persuasive messages. Many entertainment–education
campaigns have employed epilogues (Singhal & Rogers, 1999, 2004), non-narrative
monologues typically delivered by an emotional authority figure from the narrative.
Epilogues summarize the issues covered in the narrative, ask the audience rhetorical
questions, and offer suggestions for follow through within the local area (Sabido,
2004; Vaughn et al., 2000). Despite the popularity of the epilogues, solid research
findings on combining narrative and non-narrative messages are only beginning to
emerge. Allen et al. (2000) examined the effectiveness of combining narrative and
statistical message formats by exposing college students to one of four persuasive
message conditions: neither statistical nor narrative evidence, statistical evidence,
narrative evidence, and both narrative and statistical evidence. Although the
message type did not affect the perceived credibility of the message, it did affect
 436 R. Lane et al.
attitude toward the issue. Results showed that the most persuasive condition was
the one that combined narrative and statistical evidence, followed by the message
including only statistical evidence, then the message including only narrative
evidence.
More recently, Feeley, Marshall, and Reinhart (2006) pretested undergraduates
regarding their opinions toward organ donation then exposed them to two messages
in sequence: either narrative followed by statistics or statistics followed by narrative.
Participants completed a thought-listing task and survey measures after reading the
second message. Results showed that students listed more thoughts following their
first exposure to a message, regardless of the message format, and that students found
the narrative message more causally relevant, more positive, and more credible than a
newspaper editorial style message. Both of these studies make clear that at least in
certain circumstances the addition of narrative to argumentative evidence may
Downloaded by [University of Bristol] at 08:54 21 February 2015

increase persuasion, although they do not provide any explanation for the mech-
anism by which that might happen.
We suggest that one explanation for the effectiveness of a combination narrative
þ argument message may be that narratives operate to increase perceived salience
of an issue (Braverman, 2008). Then, when non-narrative persuasive messages are
introduced to these audiences for whom salience of the issue has already been heigh-
tened, individuals may be more inclined to process them centrally. This possibility is
supported by Green’s (2006) and others’ (Busselle & Bilandzic, 2009) assertion that
transportation into a narrative may have the ability to help abstract concepts translate
into real life, both by providing concrete imagery and assisting with mental simula-
tion of the described events. Sood (2002) asserted that when audience members inter-
act with and reflect on a narrative they are using the narrative at hand to make sense
of their own life narratives. Her emphasis on cognitive, as well as affective, aspects of
transportation suggests a different connection between transportation and ELM than
that posited by E-ELM: Narrative involvement (i.e., transportation and character
identification) could serve to increase issue involvement (perceived salience), which
could then facilitate central processing of a subsequent non-narrative message.
Figure 1 graphically represents this process.
In line with this reasoning, we advanced the following hypotheses related to the
narratives with epilogues:

H3: In the combination condition, acceptance of the message will be positively related
to (a) transportation and (b) character identification.
H4: In the combination condition, perceived salience will mediate the relationship
between acceptance of the message and transportation.

Figure 1 Proposed model of processing narrative with epilogue.

 Communication Quarterly 437

Because existing literature did not provide sufficient guidance to advance a

hypothesis regarding character identification, we also posed a research question:

RQ1: In the combination condition, will perceived salience mediate the relationship
between acceptance of the message and character identification?

Finally, given the inconsistent findings of previous comparisons of statistical

versus narrative evidence, we asked a final research question:
RQ2: What differences will there be between message conditions with respect to accept-
ance of the message?
Downloaded by [University of Bristol] at 08:54 21 February 2015

Method
Participants
A convenience sample of 261 university undergraduates was recruited from large,
undergraduate classes at a large, Southeastern university. Six cases in which two or
more scales were not completed were removed from analysis. Of the remaining
255 cases, 37.6% of the participants were men and 62.0% were women, and one
did not indicate his or her sex. The racial=ethnic breakdown of the sample was
8.2% Asian, 7.8% African American, 63.9% White, 16.9% Hispanic, 2.7% mixed,
and .4% ‘‘other.’’ Participants ranged in age from 18 to 24 years (M ¼ 18.51,
SD ¼ 0.98). Most participants were freshmen (77.6%), with 14.1% sophomores,
4.3% juniors, and 3.9% seniors. Participants came from over 50 different majors.
Most participants were offered extra credit for completion of the experiment, with
alternate extra credit assignments available for those <18 years of age. Ethical per-
mission for the study was obtained from the university institutional review board.

Procedure
Students were notified about the study through an emailed announcement forwarded
by their instructors. They were provided with a link to the survey, which was hosted
on SurveyMonkey.com1. This approach provided high externally validity because
students were reading messages in the same way they would typically read blogs: at
any time in any location where they had computer access and an Internet connection.
After reading an informed consent screen, participants read an introduction
explaining that they were taking part in a study evaluating audience responses to stu-
dent blogs. After responding to a few demographic questions they were randomly
assigned to one of the three experimental conditions. Once they finished reading
the blog, participants were directed to measures of acceptance of the message, per-
ceived salience, transportation, character identification, and writing quality. After
students completed the survey they were given the option of receiving extra credit
by clicking on a link to a different survey. The survey recording student names for
 438 R. Lane et al.
extra credit was not connected to the survey responses for the experiment, but could
be accessed only by individuals who finished the experiment. Therefore, participant
anonymity was maintained.

Measures
Message condition. Participants were randomly assigned to one of three message
conditions: narrative, argument, or narrative þ argument (combination). The points
made in all conditions were adapted from arguments that had been judged to be
strong in early ELM studies by Petty and Cacioppo (e.g., 1979). Participants in the
argument condition read a blog describing the facts about a proposal to institute exit
exams at the university and providing straightforward arguments for the benefit of
such exams. Participants in the narrative condition read a blog in which a student
Downloaded by [University of Bristol] at 08:54 21 February 2015

described his own experiences with exit exams in the form of a narrative, ultimately
concluding that the requirement for exit exams at his university (unnamed) had ben-
efitted him. The same benefits were woven into the plot of the blog. The positive role
model, negative role model, transitional role model structure suggested for entertain-
ment–education media by Singhal and Rogers (1999) was achieved by having the
blogger include information about friends’ responses to the exit exam policy as he
reflected on his own experiences. The positive role model happily complied with
the requirement; the negative role model found a loophole and chose not to comply
with the requirement; the transitional role model initially planned not to comply
with the requirement, but changed his mind over the course of the blog. Characters
who complied with the requirement were rewarded, and the character who did not
comply with the requirement experienced negative results. Participants in the com-
bination condition read the narrative blog with the addition of an ‘‘epilogue’’ at
the end summarizing the benefits of university exit exams and referring to credible
sources.
Perceived salience. Perceived salience was measured via Katt’s (2003) adaptation of
Zaichkowsky’s (1985) six-item semantic differential scale. A statement asked parti-
cipants to indicate what they thought about the idea of a university requiring all stu-
dents to take exit exams before graduation. Bipolar pairs included ‘‘of concern to
me=not of concern to me,’’ ‘‘matters to me=does not matter to me,’’ and
‘‘relevant=irrelevant.’’ Previous reliabilities of the scale were .88 and .84 (Katt,
2003). Cronbach’s alpha for this study was .83 (M ¼ 3.47, SD ¼ 1.44; range ¼ 1.00–
7.00).
Acceptance of the message. Participant response to the message was measured with
the opinion items scale used by Park, Levine, Westerman, Orfgen, and Foregger
(2007). Participants responded to items on a semantic differential scale inquiring into
their general opinion about the issue of exit exams. Items included ‘‘oppose–
support,’’ ‘‘disfavor–favor,’’ and ‘‘disagree–agree.’’ In previous use this scale
produced a coefficient alpha of .98 (Park et al., 2007). Cronbach’s alpha for this study
was .94 (M ¼ 3.14, SD ¼ 1.54; range ¼ 1.00–7.00). Examination of the histogram of
 Communication Quarterly 439

this variable indicated responses were not normally distributed. Therefore, we ran a
Kruskall–Wallis test as a failsafe for analyses of variance (ANOVAs) using this vari-
able. No alternate analysis was deemed necessary for regression analyses because his-
tograms of standardized residuals for the narrative and combination conditions
appeared normally distributed.

Character identification. Identification with the character was assessed via a scale
developed by Cohen (2001) and tested by Busselle and Bilandzic (2009). Items
included, ‘‘At important moments in the story, I could feel the emotions the
characters=people in the story felt,’’ and ‘‘I understood the reasons why the
characters=people in the story did what they did.’’ Participants were asked to rate
all items on a 7-point scale, where 1 indicated that the statement did not represent
their opinion about the blog they had just read (strongly disagree), and 7 indicated that
Downloaded by [University of Bristol] at 08:54 21 February 2015

the statement strongly represented their opinion about the blog they had just read
(strongly agree). Reliability for the scale in Busselle and Bilandzic’s study was .72.
Cronbach’s alpha for this study was .84 (M ¼ 4.83, SD ¼ 1.02; range ¼ 2.38–7.00).

Transportation. Transportation was measured by a modification of the transpor-

tation scale developed by Green and Brock (2000). The 12 items included, ‘‘I could
picture myself in the scene of the events described in the blog,’’ and ‘‘I wanted to
learn how the narrative=blog ended.’’ Participants were asked to rate all items on a
7-point scale, with 1 indicating that the statement did not represent their opinion
about the narrative they just saw, and 7 indicating that the statement strongly
represented their opinion about the narrative they had just seen. In the original stu-
dies, a Cronbach’s alpha of .76 was reported at the original presentation of the scale
(Green & Brock, 2000). Reliability for the scale in this study was .82 (M ¼ 3.73,
SD ¼ 0.97; range ¼ 1.00–6.27).

Passage quality. To control for the possibility that the quality of writing in the dif-
ferent messages might influence our results we adapted a measure of passage quality
developed by Pinkleton, Austin, and Fujioka (2001). Participants responded on a
7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). Items included,
‘‘This passage can be trusted,’’ and ‘‘This passage was of high quality.’’ In the original
study, a Cronbach’s alpha of .92 was reported for the production value scale. Cron-
bach’s alpha for in this study was .80 (M ¼ 3.58, SD ¼ 1.08; range ¼ 1.00–6.50).
The experiment was pretested in a laboratory environment with two groups of 10
students. Based on the responses to the pretest, the survey was updated, and several
items were adjusted.

Results
Preliminary Analyses
To verify that perceived writing quality was constant across experimental conditions,
we performed a one-way ANOVA with message condition as the independent
 440 R. Lane et al.
variable and writing quality as the dependent variable. No significant differences
emerged (F ¼ 1.81, p ¼ .166). As a test of the manipulation, we compared transpor-
tation scores between the groups. We assumed that transportation should be higher
in both narrative conditions than in the non-narrative condition. A one-way ANOVA
confirmed that this was the case (F ¼ 4.96, p ¼ .008). Pairwise comparisons indicated
that transportation scores for both the narrative (M ¼ 3.78, SE ¼ 0.10) and combi-
nation (M ¼ 3.95, SE ¼ 0.11) conditions significantly higher than transportation
scores in the argument condition (M ¼ 3.48, SE ¼ 0.10).

Hypotheses and Research Questions

H1 predicted that in the narrative and combination conditions, perceived salience
would be positively related to both transportation and character identification. We
Downloaded by [University of Bristol] at 08:54 21 February 2015

tested H1 by splitting the file according to message condition and running Pearson’s
correlations. Table 1 displays the resulting correlation matrix. As indicated, in the
narrative condition, both transportation and character identification were positively
related to salience. However, in the combination condition, a statistically significant
positive relationship emerged only between transportation and salience. H1 was par-
tially supported.
H2 predicted that persons in the narrative and combination conditions would per-
ceive the message as more salient than persons in the argument condition. We tested
H2 using a one-way ANOVA with a planned contrast between the two narrative con-
ditions and the non-narrative condition. Although the sample mean for salience was
highest within the combination condition (M ¼ 3.70, SE ¼ 0.16) as compared to the
narrative (M ¼ 3.38, SE ¼ 0.14) and argument (M ¼ 3.50, SE ¼ 0.14) conditions, the
difference was not statistically significant (F ¼ 1.12, p ¼ .329). Therefore, H2 was not
supported.

Table 1 Correlations Among Transportation, Character Identifi-

cation, Salience, and Acceptance of Message
Variable 1 2 3 4

Narrative
Salience 1.00 .22
.24
.28


Acceptance of message 1.00 .05 .05
Transportation 1.00 .66


Character identification 1.00
Combination
Salience 1.00 .44

.36

.15
Acceptance of message 1.00 .35

.05
Transportation 1.00 .44


Character identification 1.00




p < .05; p < .01.
 Communication Quarterly 441

H3 predicted that in the combination condition acceptance of the message would

be positively related to transportation and character identification. As indicated in
Table 1, this hypothesis was supported.
H4 predicted that in the combination condition, perceived salience would mediate
the relationship between acceptance of the message and transportation, and RQ1
asked whether salience would also mediate the relationship between acceptance of
the message and character identification. No relationship emerged between accept-
ance of the message and character identification in the combination condition; there-
fore, the answer to RQ1 was clearly ‘‘no.’’ To test H4, we used Judd and Kenney’s
(1981) three-step procedure for demonstrating mediation using regression analysis
to estimate the parameters of the possible causal chain. Transportation was a signifi-
cant predictor of acceptance of the message (adjusted R2 ¼ .04; b ¼ 0.35; t ¼ 3.07,
p ¼ .003), transportation also predicted salience (adjusted R2 ¼ .12; b ¼ 0.36;
Downloaded by [University of Bristol] at 08:54 21 February 2015

t ¼ 3.21, p ¼ .002), and the relationship between transportation and acceptance of

the message became statistically insignificant when salience was held constant
(adjusted R2 ¼ .08; b ¼ 0.22; t ¼ 1.94, p ¼ .056). Therefore, salience did operate as a
mediator between transportation and acceptance of the message. H4 was supported.
Finally, RQ2 asked what difference there would be between message conditions
with respect to acceptance of the message. To answer this question we ran a one-way
ANOVA with message condition as the independent variable and persuasion as the
dependent variable. The omnibus test was marginally significant (F ¼ 2.99,
p ¼ .052; Kruskal–Wallis, p ¼ .069), and post hoc comparisons indicated that parti-
cipants in the narrative condition (M ¼ 3.38, SD ¼ 1.52) were significantly more
positive toward the message than participants in the non-narrative condition
(M ¼ 2.83, SD ¼ 1.54; p ¼ .017; Mann–Whitney U, p ¼ .025). Neither differed signifi-
cantly from the combination condition (M ¼ 3.21, SD ¼ 1.54; Mann–Whitney U, null
hypotheses not rejected).

Discussion
Synthesizing components of the ELM (Petty & Cacioppo, 1984) with recent theoriz-
ing on persuasion through narrative (Green & Brock, 2000; Moyer-Guse, 2008;
Moyer-Guse & Nabi, 2010; Slater & Rouner, 2002), we tested components of a model
representing processing of narratives with epilogues. We expected that when indivi-
duals were highly transported, they would perceive messages as more salient. As
anticipated and in line with previous research and theory (Braverman, 2008; Cohen,
2001; Green, 2006), this expectation was confirmed. Among participants in both
narrative and combination conditions, high transportation scores were positively
associated with perceived salience of the message.
We further conjectured that when narratives were followed by non-narrative
epilogues, perceived salience would lead participants to process epilogues more cen-
trally, resulting in acceptance of the message. Establishing these associations was a
first step toward future testing of whether such throughput factors could be linked
to increased central processing, a key tenet of ELM (Petty & Cacioppo, 1981). Such
 442 R. Lane et al.
a series of studies would address calls for research on how narrative messages interact
with non-narrative messages to persuade (Green, 2006; Hinyard & Kreuter, 2007;
Slater & Rouner, 2002), as well as supply a means of conceptualizing the persuasive
effects of both narrative and argument under a unified theoretical framework: ELM.
Our expectations regarding the mechanisms associated with persuasion through
narratives with epilogues were supported. In the combination condition, perceived
salience did act as a mediator between transportation and acceptance of the mess-
age—that is, when participants were more highly transported, they perceived the
issue as more salient to them, and they responded more positively to the message.
Thus, relationships between three of the four elements in our proposed mediational
chain (see Figure 1) were confirmed. Although we did not measure central processing
in this study, unpublished data (Lane, Miller, Brown, & Vilar, 2011) indicates that
epilogues do lead to greater central processing. The full model has yet to be tested,
Downloaded by [University of Bristol] at 08:54 21 February 2015

but when present results are combined with the implications of Lane et al.’s raw data,
it appears likely that the full causal chain will be supported. Measurement of central
processing of narrative messages and epilogues would be to be the next step toward
confirming the persuasive paths we have outlined in this article.
In addition to transportation, character identification has also been theorized as a
key to making issues more salient to narrative audience. Although character identi-
fication was indeed related to perceived salience among our participants in the nar-
rative condition, it is not clear why that was not the case in the combination
condition. Perhaps the shift of the blogger into the argument mode caused parti-
cipants to drop any identification they might have developed with the character. Epi-
logues in entertainment–education programming are often provided by one of the
actors from the narrative (Sabido, 2004; Vaughn et al., 2000). Future research should
compare this type of epilogue with epilogues provided by other sources, in terms of
acceptance of the message, as well as the mediating variables of transportation and
character identification.
Several limitations of this study should be mentioned. As with many persuasive
studies, ours was limited in that we used a single issue to test our hypotheses. There
are obvious pragmatic reasons for this, and we employed a topic that has been his-
torically useful in studies of ELM, but this choice means that we cannot be certain
which effects we observed may be confined to our specific, or similar, topics
(O’Keefe, 2002). In addition, the choice to allow students to complete the survey
on their own time in whatever location they chose undoubtedly increased external
validity, but it may also have decreased internal validity as students may have dis-
cussed the survey with friends while completing it. Finally, ELM posits that a major
advantage of the central processing of persuasion is that any persuasion effected is
less likely to decay over time and more resistant to counter-persuasion than per-
suasion that takes place via the peripheral route (Boninger, Brock, Cook, Gruder,
& Romer, 1990; Haugtvedt & Strathman, 1990). Our study only tapped participants’
immediate response to the blogs they read. The relative stability of persuasion over
time in the three message conditions and its relationship to transportation and
salience should be tested in future research.
 Communication Quarterly 443

Narratives with epilogues are common tools in health and development com-
munication. In addition to the talk-back technique common in entertainment–edu-
cation, audiences may also view embedded prosocial storylines in entertainment
media and be directed to Web sites, Facebook1 pages, or Twitter accounts. Face-to-
face interventions, like puppet shows and theater groups, often follow up dramas
with handouts or other non-narrative materials. Understanding the impact of narra-
tive in these situations, and recognizing the different that a combination approach
can make is essential if planners are to design non-narrative materials that capitalize
on those effects. Results of this study provide initial data on the mechanisms by
which such techniques are processed by audience members.

References
Downloaded by [University of Bristol] at 08:54 21 February 2015

Allen, M., Bruflat, R., Fucilla, R., Kramer, M., McKellips, R. D., & Spiegelhoff, M. (2000). Testing
the persuasiveness of evidence: Combining narrative and statistical forms. Communication
Research Reports, 77, 331–336.
Appel, M., & Richter, T. (2007). Persuasive effects of fictional narratives increase over time. Media
Psychology, 10, 113–134.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood
Cliffs, NJ: Prentice Hall.
Boninger, D., Brock, T. C., Cook, T. D., Gruder, C. L., & Romer, D. (1990). Discovery of reliable
attitude change persistence resulting from a transmitter tuning set. Psychological Science, 1,
268–271.
Braverman, J. (2008). Testimonial versus informational persuasive messages: The moderating effect
of delivery mode and personal involvement. Communication Research, 35, 666–694.
Brock, T. C. (1967). Communication discrepancy and intent to persuade as determinants of
counterargument production. Journal of Experimental Social Psychology, 3, 296–309.
Busselle, R., & Bilandzic, H. (2009). Measuring narrative engagement. Media Psychology, 12,
321–347.
Cohen, J. (2001). Defining identification: A theoretical look at the identification of audiences with
media characters. Mass Communication and Society, 4, 245–264.
De Wit, J. B. F., Das, E., & Vet, R. (2008). What works best: Objective statistics or a personal
testimonial? An assessment of the persuasive effects of different types of message evidence
on risk perception. Health Psychology, 27, 110–115.
Dunlop, S., Wakefield, M., & Kashima, Y. (2010). Pathways to persuasion: Cognitive and experiential
responses to health-promoting mass media messages. Communication Research, 37, 133–164.
Feeley, T. H., Marshall, H. M., & Reinhart, A. M. (2006). Reactions to narrative and statistical writ-
ten messages promoting organ donation. Communication Reports, 19, 89–100.
Gerrig, R. J. (1993). Experiencing narrative worlds. New Haven, CT: Yale University Press.
Green, M. C. (2004). Transportation into narrative worlds: The role of prior knowledge and per-
ceived realism. Discourse Processes, 38, 247–266.
Green, M. C. (2006). Narratives and cancer communication. Journal of Communication, 56, S163–S183.
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.
Haugtvedt, C. P., & Strathman, A. J. (1990). Situational product relevance and attitude persistence.
Advances in Consumer Research, 17, 766–769.
Hinyard, L. J., & Kreuter, M. W. (2007). Using narrative communication as a tool for health beha-
vior change: A conceptual, theoretical, and empirical overview. Health Education Behavior,
34, 777–792.
 444 R. Lane et al.
Huesmann, L. R., Eron, L. D., Lefkowitz, M. M., & Walder, L. O. (1984). The stability of aggression
over time and generations. Developmental Psychology, 20, 1120–1134.
Judd, C. M., & Kenny, D. A. (1981). Process Analysis: Estimating mediation in treatment
evaluations. Evaluation Review, 5(5), 602–619.
Katt, J. A. (2003). Student perceptions of relevance (Unpublished doctoral dissertation). University of
Central Florida, Orlando, FL.
Kopfman, J. E., Smith, S. K., Ah Yun, J., & Hodges, A. (1998). Affective and cognitive reactions to
narrative versus statistical organ donation messages. Journal of Applied Communication
Research, 26, 279–300.
Lane, R., Miller, A., Brown, C., & Vilar, N. (2011). An investigation into the role of perceived salience
in narrative, argument, and combination messages (Unpublished pilot study). University of
Central Florida, Orlando.
Maccoby, E. E., Wilson, W. C., & Burton, R. V. (1958). Differential movie-viewing behavior of male
and female viewers. Journal of Personality, 26, 259–267.
Moyer-Guse, E. (2008). Toward a theory of entertainment persuasion: Explaining the persuasive
Downloaded by [University of Bristol] at 08:54 21 February 2015

effects of entertainment–education messages. Communication Theory, 18, 407–425.

Moyer-Guse, E., & Nabi, R. L. (2010). Explaining the effects of narrative in an entertainment tele-
vision program: Overcoming resistance to persuasion. Human Communication Research, 36,
26–52.
O’Keefe, D. J. (2002). Persuasion: Theory and research (2nd ed.). Newbury Park, CA: Sage.
Park, H. S., Levine, T. R., Westerman, C. Y. K., Orfgen, T., & Foregger, S. (2007). The effects of
argument quality and involvement type on attitude formation and attitude change: A
test of dual-process and social judgment predictions. Human Communication Research, 33,
81–102.
Petty, R. E., & Cacioppo, J. T. (1979). Issue involvement can increase or decrease persuasion by
enhancing message-relevant cognitive responses. Journal of Personality and Social
Psychology, 37, 1915–1926.
Petty, R. E., & Cacioppo, J. T. (1981). Issue involvement as a moderator of the effects on attitude of
advertising content and context. Advances in Consumer Research, 8, 20–24.
Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on responses to argument quantity
and quality: Central and peripheral routes to persuasion. Journal of Personality and Social
Psychology, 46, 69–81.
Petty, R. E., Cacioppo, J. T., & Schumann, D. (1983). Central and peripheral routes to
advertising effectiveness: The moderating role of involvement. Journal of Consumer Research,
1, 135–146.
Petty, R. E., Haugtvedt, C. P., & Smith, S. M. (1995). Elaboration as a determinant of attitude
strength: Creating attitudes that are persistent, resistant, and predictive of behavior. In
R. E. Petty & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences
(pp. 93–130). Mahwah, NJ: Erlbaum.
Pinkleton, B. E., Austin, E. W., & Fujioka, Y. (2001). Effects of perceived beer and PSA quality on
high school students’ alcohol-related beliefs and behaviors. Journal of Broadcasting &
Electronic Media, 45, 575–597.
Sabido, M. (2004). The origins of entertainment education. In A. Singhal & E. Rogers (Eds.),
Entertainment–education and social change: History, research, and practice (pp. 61–74).
Mahwah, NJ: Erlbaum.
Singhal, A., & Rogers, E. (1999). Entertainment–education: A communication strategy for social
change. Mahwah, NJ: Erlbaum.
Singhal, A., & Rogers, E. (2004). Entertainment–education and social change: History, research, and
practice. Mahwah, NJ: Erlbaum.
Slater, M. (1997). Persuasion processes across receiver goals and message genres. Communication
Theory, 7, 125–148.
 Communication Quarterly 445

Slater, M., & Rouner, D. (2002). Entertainment–education and elaboration likelihood:

Understanding the processing of narrative persuasion. Communication Theory, 12, 173–191.
Slater, M., Rouner, D., & Long, M. (2006). Television dramas and support for controversial public
policies: Effects and mechanisms. Journal of Communication, 56, 235–252.
Sood, S. (2002). Audience involvement and entertainment–education. Communication Theory, 12,
153–172.
Vaughn, P. W., Regis, A., & St. Catherine, E. (2000). Effects of an entertainment–education radio
soap opera on family planning and HIV prevention in St. Lucia. International Family Plan-
ning Perspectives, 26, 148–157.
Zaichkowsky, J. L. (1985). Measuring the involvement construct. Journal of Consumer Research, 12,
341–352.
Downloaded by [University of Bristol] at 08:54 21 February 2015