‘ 'm‘J.
0W

5

’ ”- ml.
5%;545573355

 

3:15

,.
.-
my

.. ,
”'4- .,. ..L. w
“mam“.‘i —’(:_
'mflny r w‘ m
"" ,2: A -
. .7,
,.

..9~

.. r-—

I“!

4
I-

~_...

w..- .w
. .5.

~-

Ir
G

 

 

This is to certify that the
dissertation entitled

THE EFFECTIVENESS OF ENTERTAINMENT-EDUCATION AS
MEDIA HEALTH CAMPAIGNS: THE EFFECTS OF
ENTERTAINMENT NARRATIVE AND IDENTIFICATION ON
HIV/AIDS PREVENTIVE BEHAVIOR

presented by

Byoungkwan Lee

has been accepted towards fulfillment
of the requirements for the

Ph. D. degree in Mass Media

 

 

Qfie ‘

 

 

Major Professor’s Signature

«JLIH/ 1; $0.1

 

Date

MSU is an Affinnative Action/Equal Opportunity Institution

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

MARIE 2373
31,—1.8 r,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 C'JCIRC/DaIeDue.06$~p.15

THE EFFECTIVENESS OF ENTERTAINMENT-EDUCATION AS MEDIA
HEALTH CAMPAIGNS: THE EFFECTS OF ENTERTAINMENT NARRATIVE
AND IDENTIFICATION ON HIV/AIDS PREVENTIVE BEHAVIOR
By

Byoungkwan Lee

A DISSERTATION
Submitted to
Michigan State University
in partial fiJlfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Mass Media

2004

ABSTRACT

THE EFFECTIVENESS OF ENTERTAINMENT-EDUCATION AS MEDIA
HEALTH CAMPAIGNS: THE EFFECTS OF ENTERTAINMENT NARRATIVE
AND IDENTIFICATION ON HIV/AIDS PREVENTIVE BEHAVIOR
By

Byoungkwan Lee

Entertainment-education (E-E), defined as “the intentional placement of
educational content in entertaining messages” to attempt individual and social change,
has received attention as a new effective tool of social change by not only communication
scholars but also media campaign practitioners. Since many studies have confirmed the
effects of E-E on behavioral changes, E-E researchers have turned their attentions into its
process; how audiences change their beliefs, attitudes, or behaviors as a result of
exposure to E-E programs based on various theoretical explanations such as speigl 0

learning/social cognitive theory, thebextended Elaboration Likelihood Model, wience

 

 

igyolvement, and gama theory

Given that understanding the processing of persuasive content in narrative
messages can provide important insights to explain why E—E has effects, examining the
roles of narrative and identification in the process of BB intervention might have a
significant implication in the development of EB research. With this primary interest in

the roles of narrative en ement and identification with characters of E-E programs, the

current study will provide an explanatory model for specific preventive health behavior,

 

 

 

Specifically, condom use, including other relevant theoretical variables from social

learning/cognitive theory, audience involvement, and the Health Belief Model.

A structural equation model proposed in this study represents how the
engagement of narrative affects the degree of audience members’ identification with
character, which is a predictor of perceived susceptibility to and perceived severity of
HIV/AIDS, perceived benefits minus perceived barriers of HIV/AIDS preventive
behaviors, and self-efficacy toward safer sex behavior. In turn, the model predicts that
these four factors positively affect individuals’ future safer sex behavior. Finally, five
socio-demographic variables (gender, age, education, marital status, and number of sexual
partner) will be controlled. This study will use survey data from listeners of a radio serial
drama for HIV/AIDS prevention in Ethiopia. A hybrid structural equation model will be
estimated using LISREL program, after confirming good-fit measurement model.
Findings obtained from this study will substantially contribute not only to the theoretical
accumulation of E-E literature but also to the creation of practical design for more

effective E-E intervention.

Dedicated to my father, Chun-Sik Lee

iv

ACKNOWLEDGEMENTS

This study would not have been possible without the guidance and assistance of
numerous individuals.

First and foremost I must thank my advisor, Dr. Charles T. Salmon, for his
unflagging support. He has been a source of strength and support throughout my graduate
studies at Michigan State University. He gave me academic guidance and kept me
employed throughout my graduate studies. I was able to enjoy and survive my graduate
training through his support, encouragement, and patience. I will never forget the many
thought-provoking conversations we had concerning different projects, and his support and
advice in other areas of my academic life.

I would also like to express my appreciation to the members of my committee. A
very special thanks goes to Dr. Ron Tamborini, who was also my adviser during my
masters program. His expertise, questions, and kind critiques gave me direction for this
work. Above all things, I deeply appreciate that he gave me an opportunity to stay at
Michigan State University. I also express my sincere gratitude to Dr. Sandi Smith and Dr.
Hairong Li. Their encouragement and insightful comments were crucial to the final
dissertation.

I also want to thank Dr. Kim Witte, who allowed me to use the precious data for
this work, and Dr. Rachel Smith, who gave me invaluable information for the statistical
analysis of this work.

I must thank my friends, Karen M. Lancendorfer, Kiyoung Lee, and Euidong Yoo,
who provided me with constant encouragement and assistance, without which this work

could not have been completed.

Thanks also to my family for their emotional support throughout the long, long
process; especially to my mother for continually helping me in every way possible.

I will never find words enough to express the gratitude that I owe to H]. Each and
every time I wanted to give up, she found the magic words that helped me to finish.

Finally, my lovely son, Minyoung, who had to exchange what be mostly loves in
the world for my degree, must possess all of my achievement. He has been my greatest

source of inspiration and motivation.

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................... ix
LIST OF FIGURES .............................................................................................................. x
CHAPTER 1 — Introduction ................................................................................................... 1
CHAPTER 2 —— Literature Review and Hypotheses .............................................................. 5
The Persuasive Effects of Narrative Content ......................................................................... 5
Narrative Engagement and Identification .............................................................................. 8
The Effects of Identification on Health Beliefs ................................................................... 15
Perceived Susceptibility and Perceived Severity ................................................................ 22
Perceived Benefits, Perceived Barriers, and Perceived Self-Efficacy ................................ 23
The Effects of Socio-Demographic Variables ...................................................................... 25
Proposed Model for the Effects of EB Intervention on Safer Sex Behavior ...................... 27
CHAPTER 3 — Method ........................................................................................................ 30
An Overview of Journey of Life, the Entertainment-Education Program ........................... 30
Survey Data .......................................................................................................................... 32
Measures .............................................................................................................................. 33
Narrative Engagement ............................................................................................. 33

,t Identification with Characters .................................................................................. 37
Perceived Susceptibility to AIDS/HIV .................................................................... 37
Perceived Severity to AIDS/HIV ............................................................................. 37
Perceived Benefit ..................................................................................................... 38
Perceived Barriers .................................................................................................... 38
Perceived Self-Efficacy ........................................................................................... 38

Future Safer Sex Behavior ....................................................................................... 39
Socio-demographics ................................................................................................. 39
CHAPTER 4 — Results ......................................................................................................... 41
Correlational Analyses ......................................................................................................... 41
Path Analyses: Testing the Hypothesized Model .................................................................... 45
Model Specification and Parameter Estimates .................................................................... 45
Model Respecification ......................................................................................................... 49
Modified Model 1 .................................................................................................... 51
Modified Model 2 .................................................................................................... 57

vii

CHAPTER 5 — Discussion ................................................................................................... 63

The Roles of Narrative Engagement and Identification in E-E Intervention ....................... 63
Influences of Engagement with Narrative on Individuals’ Beliefs ...................................... 68
Limitations of the Present Study .......................................................................................... 70
Suggestions for Future Directions of EB:

Issue Placement as an Alternative Strategy of EB .................................................................. 73
APPENDIX A - Synopsis of “Journey of Life” ................................................................... 77
APPENDIX B - Description of the Main Characters of Journey of Life ............................ 78
APPENDIX C - Respondents’ Demographic Characteristics .............................................. 79
APPENDIX D - Modification Indices for BETA ................................................................ 80
APPENDIX E - Mean Differences

among Identifications with Three Main Characters ............................................................. 81
APPENDIX F - Family Planning & HIV/AIDS Communication Survey, 2002 ................. 82
BIBLIOGRAPHY .............................................................................................................. 107

viii

List of Tables

Table 1
Narrative Engagement Items ................................................................................................ 34

Table 2
Two-factor Model of Narrative Engagement Scales: Standardized
Factor Loadings, Standard Errors, and Goodness-of-F it Summary ..................................... 36

Table 3
Health Beliefs and Behavioral Intention Items .................................................................... 40

Table 4
Zero-Order Correlations among Exogenous and Endogenous Variables

in the Proposed Model ......................................................................................................... 44

Table 5
Goodness-of-Fits Summary among Models ........................................................................ 62

ix

List of Figures

Figure 1
Hypothetical Model for the Effects of Entertainment-Education
Intervention on Safer Sex Behaviors ................................................................................... 29

Figure 2
Path Diagram for the Recursive Model of the Effects of Entertainment-Education
Intervention on Safer Sex Behaviors ................................................................................... 47

Figure 3
Modified Recursive Path Model of the Effects of Entertainment-Education
Intervention on Safer Sex Behavior ..................................................................................... 56

Figure 4
Alternative Modified Recursive Path Model of the Effects of Entertainment-Education
Intervention on Safer Sex Behavior ..................................................................................... 61

CHAPTER 1
INTRODUCTION

As a new effective tool for social change, entertainment-education (E-E), which
refers to “the intentional placement of educational content in entertaining messages” to
effect individual and social change (Singhal & Rogers, 2002, p. 117), has attracted the
attention of not only communication scholars but also media campaign practitioners. In
particular, academic and practical interest in the effectiveness of EB intervention has
increased in the area of health communication. Since the mid-19805, over 200 E-E
interventions for health-related educational issues have been implemented (Singhal &
Rogers, 2004).

In the case of the US, government agencies have been working with the
commercial media to insert health messages into existing entertainment programming
(Keller & Brown, 2002). A report in an Internet magazine (Forbes, 2000) depicts the
govemment’s increasing interest in E-E as an outlet for health media campaigning that
substitutes for public service announcements (PSAS). According to the Internet magazine
Salon, an agency of the Clinton White House, the Office of National Drug Control Policy
(ONDCP), secretly worked with all of the commercial television networks to insert the
government's anti-drug message into TV scripts as part of the storylines of prime time
programs, such as NBC ’5 “ER,” Fox's “Beverly Hills 90210,” CBS’S “Chicago Hope,”
ABC’s “The Drew Carey Show,” and WB’S “Smart Guy” and “7th Heaven,” and paid the
networks $25 million for their cooperation (Forbes, 2000).

In developing countries, the E-E strategy for social change has been conducted

more actively than that in the US. Since Miguel Sabido developed the EB strategy in
Mexico based on his analysis of the audience effects of Simplemente Maria, a 1969
Peruvian television soap opera, E-E has been considered a popular and effective
communication strategy for social change in many developing countries: Brazil (Pastina,
Patel, & Schiavo, 2004), Egypt (Abdulla, 2004), India (Sood, 2002) (Singha1& Rogers,
2001), South Africa (Usdin & Singhal, 2004), and Turkey (Yaser, 2004). For example,
about 125 E-E programs for public health interventions were conducted by the support of
the Johns Hopkins University’s Center for Communication Programs (JHU/CCP)
between 1986 and 2002 in over 40 countries (Piotrow & Fossard, 2004).

The growing institutional interest in BE through popular media might be based
on a conviction that the inserting of campaign messages in entertainment programming is
as effective as (or more effective than) paid advertising or the PSA. Several advantages of
EB intervention over traditional media campaigns can be indicated. First, the use of
entertainment for educational messages can be more effective than traditional media
campaigns insofar as the lack of entertainment in many traditional campaigns may have
prompted audiences to avoid or to ignore the messages (Greenberg et a1., 2004). Fictional
dramatic presentation in broadcast media, as Bandura (1997) indicated, is an

extraordinarily effective tool to achieve changes, which has widespread social impact

”WI-4H?“
fl ..... w — —v "w o. . ~ ~ .7

 

 

.,
I ,,. "M «m,‘v-~F"-

 

 

because broadcast media can reach huge numbers of people over a prolonged period and

W T «MM-Am...

{3- encourage audience members to immerse themselves in the lives of the televised models.9

' Wyn—v — u- -

 

 

 

- M ‘Wr .—

Second, there is no reason that media should not teach prosocial behavior if it
teaches antisocial behavior (Baran, Chase, & Courtright, 1979). AS E-E offers audiences

a more balanced perspective regarding antisocial and prosocial content, their attitudinal

and behavioral changes that result from antisocial messages can be decreased (Greenberg
et a1., 2004). Thus, as Greenberg and his colleagues (2004) argued, a prosocial message
disseminated by E-E programs through popular media can play the role of a counter-
message that offsets the negative influence of popular media on audiences.

More importantly, the narrative content of an E—E program can be considered an
important advantage of E-E over traditional media campaigns. Narrative is an important
element of television entertainment programs such as the serial drama, the sitcom, or the
mini-series, and therefore, E-E necessarily includes the use of narratives. Greenberg and
his colleagues (2004) assert that through its “seamless” integration of information and
diversion, E-E can “circumvent” audiences’ perceptual defenses against persuasion
attempts such as PSA, Web banner ads, or advocacy pamphlets (p. 195). The narrative
content of EB programs is an element for the circumventing (or the suppressing) of
audiences’ perceptual defenses against persuasive messages. Thus, counterarguments to a
persuasive message can be suppressed to the extent that the recipient of the message is
absorbed or transported by a narrative (Slater & Rouner, 2002).

The primary goal of the present study is to investigate the persuasive effects of
narrative content in the entertainment program of popular media in order to explain why
E-E messages have the effects. Many studies have confirmed that E-E affects behavioral
changes (Kincaid, Yun, Piotrow, & Yaser, 1993; Piotrow, Kincaid, Rimon II, & Rinehart,
1997; Singhal & Rogers, 1999; Sypher, McKinley, Ventsam, & Valdeavellano, 2002;
Valente, Kim, Lettenmaier, Glass, & Dibba, 1994). Whether E-E has an effect, as Singhal
and Rogers (1999) noted, is no longer an issue. In fact, based on various theoretical

explanations such as social learning/social cognitive theory (Bandura, 2004; Papa,

Singhal, Law, Pant, Sood, Rogers, & Shefner-Rogers, 2000), the Elaboration Likelihood

Model (Slater & Rouner, 2002), audiencg‘inxolgernmenLQae & Lee, 2004; Sood, 2002),

uNWJC‘A r’: uusipam gnu A~~prqu4~oyu ”“2". r: um.

and mmeemneiehteeseseee 5e.M999eem.l9§$.!9¥193142 299.2), E-E researchers
have turned their attention to the process underlying EB and, thus, to how audiences
change their beliefs, attitudes, or behaviors as a result of exposure to E-E programs.
Given that an understanding of the processing of persuasive content in narrative
messages can provide important insights that will help researchers to explain why E-E
has certain effects (Slater & Rouner, 2002), an examination of the roles of narrative and
identification in the process of EB intervention might have a significant implication in
the development of E-E research. With this primary interest in the role of narrative in E-E
intervention, the current study will provide an explanatory model for specific preventive
health behavior, thus, condom use, including other relevant theoretical variables from
social learning/cognitive theory, audience involvement, and the Health Belief Model
(HBM). Therefore, this study proposes and tests a recursive structural equation model
with specific hypotheses in order to examine the roles of narrative engagement and
identification with the characters of an E-E program and these roles’ relations with other
variables that affect condom use for the prevention of the Human Immunodeficiency
Virus/Acquired Immune Deficiency Syndrome (HIV/AIDS). Survey data drawn from
listeners of a radio serial drama for HIV /A[DS prevention in Ethiopia are analyzed for the
study. Through this attempt, the current study might substantially contribute not only to
the theoretical accumulation of E-E literature but also to the creation of practical design

for more effective E-E intervention.

CHAPTER 2

LITERATURE REVIEW AND HYPOTHESES

The Persuasive Eflects of Narrative Content

Narratives refer to stories, accounts, tales, or descriptions. Humans are socially
and culturally conditioned into understanding the narrative form (Shankar & Goulding,
2001). Narrative theory views two modes of knowing in the construction of reality, the
paradigmatic mode and the narrative mode (Bruner, 1990). While the paradigmatic mode,
which is found in the natural sciences, attempts to establish universal truth conditions
through propositions and laws, the narrative mode is a product of the literary and
interpretive social sciences and emphasizes the recounting and the explaining of events
through storytelling (Langley, 1995). Narrative thinking as a process is “the translation of
one’s own and others’ experiences into stories that integrate facts, perceptions, emotions,
intentions, actions, and consequences into coherent meaning”(Cole, 1997, p. 331). Cole
(1997) noted that “narrative thought makes it possible to interpret events by constructing
coherent causal patterns of meaning that blend what is known about a situation (facts)
with relevant conjecture (imagination)” (p. 333). Through narratives, we learn about who
we are, our history, and our culture (Shankar & Goulding, 2001 ).

Although narrative impact has long been a focus in some disciplines such as
sociology, communications, humanities, and political science, it has been overlooked or '
even discounted within psychology (Brock, Strange, & Green, 2002). As a result, as
Brook and others (2002) argued, the scientific studies of persuasion within psychology

have directed considerably more attention to rhetoric rather than to narrative even though

narrative is much more prevalent than rhetoric in our everyday life. Rhetorical persuasion
refers to a variety of messages such as editorials, advertisements, public campaigns, and
political speeches in which arguments are adduced for an advocated opinion or position
(Green & Brock, 2002).

Narrative persuasion can be compared with the Elaboration Likelihood Model
(ELM) (Petty & Cacioppo, 1986), a well-known model for rhetorical persuasion, in terms
of several theoretical differences. First, these two persuasion approaches have a different
role regarding personal relevance in message processing. In the ELM, the personal
relevance of a message or issue is an important variable that affects the motivation to
process a persuasive message. Thus, personal relevance increases a person’s motivation
for engaging in a “diligent consideration of the issue-relevant information” (Petty et a1.,
1983, p. 143). Green and Brock (2002) argued, however, that there is no steady effect of
personal relevance for fictional narratives (Green & Brock, 2002). For example, a finding
(Prentice, Gerri g, & Bailis, 1997) showed that weakly supported arguments have less
impact in more personally relevant conditions (familiar settings), whereas other findings
(Weeler, Green, & Brock, 1999) showed that weakly supported arguments strongly affect
beliefs regardless of the familiarity (personal relevance) of the narrative setting.
Accordingly, in light of these findings, the ELM limits intensive processing to personally
relevant communications and allows for narrative persuasion to be differentiated from the
ELM (Green & Brock, 2000).

Although a need for cognition plays an important moderating role in rhetoric-
based persuasion, need for cognition does not seem to moderate attitude or behavior

changes in narrative-based persuasion because it is not helpful to conceive of evoked

images as independently strong or weak (Green & Brock, 2002). Several experiments by
Green and Brock (2000) found that the need for cognition was associated with neither
beliefs nor character evaluations in response to narratives. Thus, their findings suggested
that individual difference in the need for cognition is not a reliable moderator of the
impact of a story on beliefs or on character evaluation ratings.

Green and Brock (2002) indicated that the overall character of the persuasion
message makes another difference between rhetoric-based and narrative-based persuasion.
In rhetoric-based persuasion, messages consist of arguments, whereas stories in narrative-
based persuasion feature images. Green and Brock (2002) argued that “arguments can
stand alone and can be assessed for their inherent strength or weakness, whereas images
derive their strength and power from their evocation of story events” (p. 332).

Slater and Rouner (2002) argued that a clear distinction between central and
peripheral processes, which is proposed by a dual-process model such as the ELM
proposes, is no longer discemable in narrative-based persuasion because the impact of the
persuasive subtext imbedded in the narrative will be a function of the extent to which the
audience members engage with the narrative. In this respect, both engagement
(transportation or absorption in other literature)1 with the story line and identification
with characters are—instead of the issue involvement suggested by the ELM—important
variables underlying the persuasive effects of the narrative.

Therefore, in the next section, the roles of narrative engagement and identification

in the narrative contents are discussed based on the psychological mechanism related to

 

1 Engagement, absorption, and transportation are used interchangeably in this study. These three temrs
have been used to describe the same phenomenon by different researchers (Slater & Rouner, 2002).

the concepts of narrative engagement and identification. It is believed that a
psychological understanding of narrative impact can provide not only another approach
to persuasion strategy but also feasible explanations of why narratives are effective on

people’s beliefs, attitudes, or behavior.

Narrative Engagement and Identification

Although most E-E programs deal with the affective and emotional aspects of
human communication, theoretical investigations of E-E are currently emphasizing the
role of emotions (Singhal & Rogers, 2004). Oatley (2002) argued that the impact of
fictional narrative primarily stems from the recipient’s emotions that are elicited by the
vicissitudes that occur to the protagonist-agent. He views emotion as “the engine of
narrative impact” (Brock, Green, & Strange, 2002, p. 344). Through the structures of
narrative, readers construct their own narratives and then experience their own emotions
rather than those associated with the protagonists of the narrative. In other words, as
readers (or audiences) assimilate a story, their emotions are their own, rather than those
of the characters.

Oatley (2002) notes that through emotion, fictional narrative transports the readers
or audiences to the story world. He emphasizes the concepts of transportation in the
psychological processes by which narrative effects occur in audiences (or readers).
Transportation can be defined as the extent to which “individuals are absorbed into a
story or transported into a narrative world” (Green & Brock, 2000, p. 701) or “the degree
to which a message recipient is cognitively and affectively invested in a narrative” (Slater

& Rouner, 2002, p. 179). Compared to a dual-process model, their emphasis is on the role

of transportation in narrative-based persuasion. First, given that transported readers (or
audiences) are less likely to disbelieve or counterargue story claims, transportation might
reduce negative cognitive responses to a message embedded in narrative content. Second,
transportation might encourage readers to experience the fictional narrative world as a
more real world. Given that direct experience significantly affects the formation of
people’s attitudes and behaviors (Fazio & Zanna, 1981), this claim can be another
important aspect of the persuasive effects of E-E. Finally, they argued that as
transportation creates strong feelings toward story characters, the experiences or the
beliefs of the characters in the narrative content might have an enhanced influence on
readers’ beliefs.

In fact, previous studies have empirically demonstrated the persuasive effects of
narratives (Green & Brock, 2000; Kopfrnan, Smith, Yun, & Hodges, 1998; Prentice et a1.,
1997; Slater & Rouner, 1996; Stephenson & Palmgreen, 2001; Strange & Leung, 1999;
Weeler et a1., 1999). Their findings showed that narrative anecdotal arguments could be
more effective than fact-based arguments. For example, examining the cognitive and
affective reactions to organ donation messages, Kopfrnan and her colleagues (1998)
found that narratives about organ donation produced a greater influence on the
respondents’ affective reactions. Additionally, their findings showed that narratives forced
individuals to generate more total, positive, and negative emotions than statistical
evidence messages, although this effect was not statistically significant. Assessing the
influences of both the message sensation value and personal involvement with marijuana
on three types of processing (cognitive, narrative, and sensory), Stephenson and

Pahngreen (2001) found that when anti-marijuana PSA viewers were engaged in

narrative processing, they tended to be engaged in argument-related cognitive processing.
Strange and Leung (1999) examined how anecdotal accounts in news and in fiction can
influence judgments about a social problem’s urgency, causes, and cures. Their results
showed that both news and fictional stories influenced participants’ judgments about the
causes of and the solutions to the dropout problem of high school students and about the
urgency with which policy makers should attend to educational and health care reform.
Slater and Rouner (1996) examined the effects of anecdotal evidence on youths’ alcohol-
risk beliefs. According to their findings, when an alcohol education message included
anecdotes, respondents responded to the message presentation more positively.
Respondents also tended to rate the messages as being of higher persuasive and writing
quality as well as more believable in the presence of anecdotes.

Despite this empirical support, most studies have overlooked the mediating role of
identification in narrative content. Owing to this lack of accountability for a form of
identification that might mediate the effects of narrative engagement on audiences’
beliefs, attitudes, or behaviors, no important reason why narrative modes have a greater
impact than nonnarrative modes could be fully explained. Identification with characters
plays a crucial role in narrative-based persuasion because the character is the driving
force in fiction (Green & Brock, 2000, p. 702). Identification is a fundamental
psychological process in which a fictional narrative has persuasive effects on people’s
attitudes and behaviors. Freudian psychoanalysis theory views identification as “the
earliest expression of emotional tie with another person,” which results from
psychological pressures caused by the Oedipus complex (Freud, 1949, p. 60). In the

earliest stage of a male child’s life, for example, a little boy exhibits a special interest in

10

his father; he would like to grow like him and be like him, and take his father everywhere
(Freud, 1949). However, it is not necessarily that a child’s object of identification is
limited to his or her parents. Identification involves imagining being someone else and
imagining behaving like someone else (Wollheim, 1974). Thus, the object of
identification can be any other person or character in fiction. F or example, Oatley (2002)
explained the identification as a process of simulation, which provides a principal basis
on which to think about what connects the audience (or the reader) to the text. He argued
that when encountering narratives, the audiences (or readers) identify with a character
and in turn, transfer the character’s actions to their own planning processes in order both
to make sense of these sequences and to experience the emotions that the actions bring
about. In addition, Oatley (2002) claimed that identification is one of the principal
psychological processes in which emotional effects occur in readers and audiences of

narrative fiction:

The impact of fiction is largely emotional. Emotions occur as readers or watchers
construct their own versions of the story. As the simulation runs, emotions occur
to readers or watchers that depend on psychological processes such as
identification with a protagonist, sympathy for story characters, and activation of
emotional autobiographical memories that resonate with story themes. (Oatley,
2002,p.41)

In the area of media studies (e.g., Basil, 1996; Cohen, 2001; Cohen & Perse,
2003; Fraser & Brown, 2002; Hoffner & Cantor, 1991; Liebes & Katz, 1990; Livingstone,
1991; Zillmann, 1994), the role of identification has been emphasized as a central
mechanism underlying media effects. Although identification with media characters has

generally been understood to stand for either an audience member’s feelings of affinity

ll

for, fi'iendship with, similarity toward, liking of media characters, or imitation of a
character, identification has been explained and conceptualized in a variety of ways by
different theorists (Cohen, 2001). Liebes and Katz (1990) explained identification as
audiences’ responses to media characters based on three types of reactions: liking, being
liking, and wanting to be like. HOffner and Cantor (1991) also indicated that audiences’
responses to media characters have three components: liking, perceived similarity to
characters, and desire to be like characters. Oatley (2002) argued that identification
involves two factors such as “likeness to a character” and “wanting to be like that
character” (p. 62). Most studies on audience involvement (Basil, 1996; Brown, Basil, &
Bocamea, 2003; Horton & Wohl, 1956; Levy, 1979; Rubin & McHugh, 1987; Sood,
2002) have used parasocial interaction as a concept similar to identification. Parasocial
interaction originally meant quasi-relationships that audience members develop with
media performers or the characters that they portray, relationships that are Similar to those
that surface within the audience’s peer group (Horton & Wohl, 1956). Levy (1979) found
that television news viewers parasocially interact with the media personae both
affectively and cognitively. Rubin and Perse (1987) classified parasocial interaction into
three sub-dimensions: affective, cognitive, and behavioral parasocial interaction.
Affective parasocial interaction is the “degree to which audience members identify with a
particular media character,” whereas cognitive parasocial interaction is the “degree to
which audience members pay careful attention to the characters in a media message and
think about its educational content” (Singhal & Rogers, 1999, pp. 172-173).

In contrast, Cohen (2001) conceptualized identification in a specific manner. He

defines identification as “an imaginative process invoked as a response to characters

12

presented within mediated texts” (Cohen, 2001, p. 250). Citing Wollheim’s (1974)
interpretation of Freud’s thought, Cohen (2001) distinguished identification from
imitation. He argued that although identification has an internal component, imitation has
external and behavioral components. He views imitation as a consequence of
identification. He distinguishes identification from concepts such as liking, similarity, and
affinity because identification does not foster any attitudes or judgments that audiences—
based on their perceptions of characters and of themselves—make about these characters.
Furthermore, he posits that identification should be distinguished from parasocial
interaction because identification with a media characters is based on a psychological
attachment between the viewer and a character rather than on the viewer’s interaction
with a character. Despite Cohen’s (2002) arguments, it is true that identification and its
related concepts have been used interchangeably. Although further research that would
clarify the distinctions between them is required, whether identification can be
distinguished from other similar concepts might depend on how identification is broadly
conceptualized.

In spite of the arguable conceptualization of identification, a rigorous
understanding of the relationship between audiences’ engagement with a narrative and
audiences’ identification with characters might provide researchers with an important
insight into the effectiveness of EB intervention. Slater and Rouner (1997) have
identified two factors that drive people’s engagement with a narrative: narrative interest
and identification. Slater (2002) argued that an audience’s engagement with an E-E
message should be governed by the extent of the audience’s identification with characters

and by the level of narrative interest that the plot holds for the audience. Examining the

13

role of transportation in narratives, Green and Brock (2000) found that transportation had
an impact on story-consistent beliefs and on audiences’ evaluations of characters in a
narrative. Thus, their findings showed that highly transported participants not only
reported beliefs that were more consistent with those implied in the story but also felt
more positive sentiments toward the characters in a narrative.

The relationship between narrative engagement and identification (parasocial
interaction in audience involvement research) has also been explained in the audience
involvement literature.2 In the interpersonal and mass communication field, audience
involvement is defined as “direct personal experience” during message reception
(Krugrnan, 1966, p. 583), or the degree to which audience members perceive a
connection between themselves and mass media content and the degree to which the
members interact psychologically with a medium or its messages (Levy & Windahl,
1984). Identifying involvement as a mediator of responses towards prosocial behaviors
promoted by an E-E media program, Sood (2002) conceptualized parasocial interaction
and reflection as two elements that consist of audience involvement. Although the author
did not provide a clear distinction between parasocial interaction and identification, it
seems that both processes overlap (Brown & Fraser, 2004). Sood (2002) viewed
identification as a part of audience involvement that is one way in which to promote the
development of identification. Examining possible differences between identification,
parasocial interaction, and imitation, Cohen and Perse (2003) found a strong positive

correlation (r = .73) between identification and fictional involvement.

 

2 Narrative engagement (or transportation) is considered as similar to the concept of audience involvement
(Sood, 2002).

14

In summary, an audience’s consuming of narratives leads to audience’s
engagement with the narrative content, as the fictional narrative transports the audience
to the story world. In turn, as the engagement creates audience members’ strong affective
and cognitive responses to the characters in the narrative content, the experiences or the
beliefs of the characters in the narrative content might have an enhanced influence on
audiences’ beliefs. In other words, narrative engagement leads to the audience’s
identification with the characters in the narrative content. As such, this study proposes the

first hypothesis, which follows:

Hypothesis 1: Cognitive narrative engagement will be positively related to
identification with a character. Thus, the more listeners are cognitively engaged
with the narrative of the E—E program, the greater they will show identification
with the character of the program.

Hypothesis 2: Aflective narrative engagement will be positively related to
identification with a character. Thus, the more listeners are aflectively engaged

with the narrative of the E—E program, the greater they will show identification
with the character of the program.

The Effects of Identification on Health Beliefs

Social cognitive theory has emphasized identification (or perceived similarity)
with characters as a learning process through modeling. Thus, the social learning
perspective defines identification as “the process through which an individual takes on a
model’s behavior and/or personality patterns in some form” (Singhal & Rogers, 1999, p.
65). Bandura (1977) noted that identification affects the effectiveness of a mediated

message. He stated that “models presented in televised form are so effective in capturing

15

attention that viewers learn much of what they see without requiring any special
incentives to do so” (p. 25).

In the area of media effects research, as mentioned above, identification has been
emphasized to explain a process by which the media has an effect on its audience (Basil,
1996; Brown et al., 2003; Bryant & Miron, 2002; Kelman, 1961; Rubin & Step, 2000;
Rubin & McHugh, 1987). In their examination of the effects of parasocial interaction on
audiences’ listening to public affairs talk radio shows, Rubin and Step (2000) measured
three elements: perceived similarity, fiiendship, and empathy with the host of a given talk
radio Show. They found that parasocial interaction positively predicted not only that a
sense of congruence would surface between listeners’ attitudes toward societal issues and
those expressed by a talk radio host but also that listener’s behaviors would be consistent
with the suggestions of this host. Brown and others (2003) also supported the mediating
role of identification. They found that identification with Princess Diana mediated the
effects of the press’ coverage of her death on audiences’ attitudes toward the news media.
Thus, people who identified more with Princess Diana had a more negative attitude
toward the news media than those who identified less with her. Basil (1996) viewed
identification as a psychological mechanism with which to explain the effectiveness of
celebrity endorsers. Examining the effects of identification with Magic Johnson on
audiences’ sexual behavior, he found that those who identified more strongly with Magic
Johnson were more likely to change their AIDS-related attitudes and behaviors, even
though the effect of identification on perceived risk was not statistically significant. Basil
(1996) also examined the mediating role of identification in determining a viewer’s

behavior. The results of a path analysis showed that identification mediated the effects of

16

demographics, sexual habits, and knowledge of celebrities on sexual behavior regarding
AIDS.

In the context of EB, several studies have been interested in the role of
identification—also known as parasocial interaction (Bae & Lee, 2004; Papa et a1., 2000;
Sood, 2002; Sood & Rogers, 2000). Given that celebrity identification leads to changes in
issue-related awareness, values, and behavior consistent with those of the celebrity,
Brown and Fraser (2004) emphasized the advantages of using celebrities for E-E.
Examining the role of audience involvement in the effectiveness of Yinka Iinka Sukh
(TT S), a radio E-E program in India, Sood (2002) found that when audiences
demonstrated identification with TT S characters, they also related the radio serial to their
personal lives. In another study on IT S, Papa and others (2000) found that TT S as an E-E
program leads to high levels of parasocial interaction for certain audience members.
However, they did not find that an audience’s identification with characters who display
prosocial behaviors prompts the audience members to directly change their behavior.
They asserted that this behavior change, in part, depends on the extent to which these
parasocial relationships prompted conversations among listeners. Several studies showed
that the perceptual level of parasocial interaction with characters or content in E-E
programs can lead to actual behavioral participation (Bae & Lee, 2004; Sood & Rogers,
2000). Analyzing audience letters written to an E-E television soap opera in India, Sood
and Rogers (2000) identified three types of parasocial interaction: affective interaction,
cognitive interaction, and behavioral interaction. From a content analysis of a Korean
television drama’s episodes and online bulletin board messages about the episodes, Bae

and Lee (2004) also found that these three parasocial interactions with the characters or

17

the contents of the programs. According to their findings, however, the magnitude of each
parasocial interaction can vary depending on the characteristics of the issue. They found
that issue controversy within an episode was related with parasocial interaction. In other
words, when an issue was controversial, the audiences tended to respond rationally
through cognitive parasocial interaction. Conversely, when an issue was non-
controversial, the audiences tended to respond emotionally through affective parasocial
interaction.

Based on the discussions above, this study assumes that audience members’
identification with characters on E-E programs has an indirect effect—not a direct
effect—on these audience members’ changes in behavior. The social psychological
theories (e. g., learning theory, the health belief model, and the theory of reasoned action)
maintain that individual perceptions, beliefs, or values are the driving force of behavior
change (Sood, Menard, & Witte, 2004). Identification with characters in E-E programs
might prompt a process of behavior change as it influences the perceptions, beliefs, or
values of audience members. Slater (2002) argued that, as a necessary condition for
effective drama, identification with characters in a story causes audience members to
perform a cognitive rehearsal of the beliefs and values expressed or embodied by the
character.

Therefore, this study predicts that, by affecting their perceptions of or beliefs
about safer sex, audience members’ identification with the characters of EB programs
influences the degree to which the audience members change their preventive behavior.
Identification, as Schoenmakers (1988) asserted, allows the listeners to enter a fictional

world and experience emotions as if the listeners were the agents of the transformations

18

occurring in that fictional world (Cited in Gholamain, 1998, p. 5). Schoenmakers (1988)
suggested that the mechanisms of identification typically involve four processes:

introj ection (taking on the desirable attributes of an object), projection (disowning the
unwanted aspects of self by projecting them onto the fictional characters), assimilation,
and accommodation (changes in the structure of cognitive schema as a result of a new
experience) (Cited in Gholamain, 1998, p. 5). Thus, as audience members intemalize—
based on emotional attachment—the important attributes of an object (Meissner, 1972),
they might change their perceptions, beliefs, or values about the health behaviors
advocated in E-E programs. For instance, as individuals internalize the fate of a tragic
character in an E-E drama into their own lives, they might experience the seriousness of a
disease portrayed in the drama as though they were the character in the drama, who is
tragically afflicted with the disease.

From this reasoning, this study operates according to the assumption that the
degree of a listener’s identification with the characters of an E-E program might be
positively related to perceived susceptibility and perceived severity, which are
constructed in the Health Belief Model. The HBM posits that perceptions motivate action,
which is a social-psychological model based on value-expectancy theory. According to
the HBM, two main beliefs (or perceptions) predict preventive health behavior: perceived
threat and expectations (J anz & Becker, 1984; Rosenstock, 1974; Rosenstock, Strecher, &
Becker, 1994). The perceived threat of a particular illness or disease is a combination of
two factors: perceived susceptibility and perceived severity. Perceived susceptibility
refers to an individual’s subjective beliefs about how he or she is susceptible to an illness

or disease. Perceived severity is an individual ’3 subjective beliefs about how severe the

19

illness is. Two hypotheses are suggested:

Hypothesis 3-a: Identification with a character will be positively related to
perceived susceptibility to a health threat. Thus, the more listeners show
identification with the character of the E-E program for preventing HIV/AIDS, the
greater they will show perceived susceptibility to HIV/AIDS.

Hypothesis 3-b: Identification with a character will be positively related to
perceived severity of a health threat. Thus, the more listeners show identification
with the character of the E-E program for preventing HIV/AIDS, the greater they
will show perceived severity of H1 V/AIDS.

Social learning theory provides strong evidence that learning by identification is
more likely to increase both outcome expectancies and self-efficacy. Once individuals
identify themselves with media characters, they adopt the beliefs, values, and motives of
those media characters, often viewing him or her as a counselor, comforter, and role
model (Papa et a1., 2000). Identification can prompt role modeling as audiences carefully
consider the behavioral choices made by media characters (Horton & Wohl, 1956). In an
E-E program, soap opera characters “teach” audience members, via modeling, that they
can make a recommended response, thereby enhancing self-efficacy perceptions, and that
the recommended response works in averting a wanted outcome, thereby enhancing
outcome expectancies (Sood et a1., 2004, p. 125). For example, in analyzing an E-E radio
soap opera in India, Sood (2002) found a strong correlation (.75) between self-efficacy
and parasocial interaction with character. Her findings suggest that the more listeners
Show parasocial interaction with the character, the greater they show self-efficacy toward
abstaining from taking and giving a dowry. The HBM indicates that any particular health

action that will be taken depends upon beliefs regarding two expectancies: the expected

20

benefits of taking the action and the potential barriers to the particular action. Perceived
barriers refers to the potential negative aspects of a particular health action, which act as
impediments to a person’s undertaking of recommended behavior (Rosenstock, 1974). In
contrast, perceived benefits refers to perceptions that the perceived threat would be
substantially reduced by taking a specific action (Rosenstock, Strecher, & Becker, 1988).
The combined levels of susceptibility and severity provide the energy and force to act,
whereas the perception of benefits (less barriers) provides the preferred path of action
(Rosenstock, 1974). Based on the discussions above, this study predicts a positive
association between audience members’ identification with characters in the E-E program
and perceived benefit, a negative association between the audience members’
identification with a character and perceived barriers, and a positive association between

the audience members’ identification with a character and perceived self-efficacy:

Hypothesis 3-c: Identification with a character will be positively related to
perceived benefits of preventive health behaviors. Thus, the more listeners show
identification with the character of the EB program for preventing HIV/AIDS, the
greater they will show perceived benefits of safer sex behaviors.

Hypothesis 3-d: Identification with a character will be negatively related to
perceived barriers to preventive health behaviors. Thus, the more listeners show
identification with the character of the BE program for preventing HIV/AIDS, the
less they will show perceived barriers to safer sex behaviors.

Hypothesis 3-e: Identification with a character will be positively related to
perceived self-eflicacy of preventive health behaviors. Thus, the more listeners
show identification with the character of the EB program for preventing
HIV/AIDS, the greater they will show perceived self-efficacy of safer sex
behaviors.

21

Perceived Susceptibility and Perceived Severity

In the context of HIV /AIDS prevention behaviors, a few studies have examined
the associations of perceived susceptibility and perceived severity with safer sex
behaviors. The studies have hypothesized that perceived susceptibility and perceived
severity are positively related to safer sex behaviors. For instance, as individuals perceive
a threat of HIV/AIDS to be more serious and their contracting of the disease to be more
likely, they have a greater motivation (a desired reduction in anxiety levels) to do
something about this health threat (Smith, Witte, & Downs, 2003). As such, the actions
that individuals take to reduce a health threat are determined by those individuals’
perceptions of the recommended action.

However, most of the findings on the associations of perceived susceptibility and
perceived severity with safer sex behaviors have been inconsistent. While some studies
have found that perceived susceptibility is a significant positive predictor of HIV/AIDS
preventive behaviors (Basen-Engquist, 1992; Rotheram-Borus, Reid, Rosario, & Kasen,
1995; Steers, Elliott, Nemiro, Ditman, & Oskamp, 1996), others have not found any
significant association (Volk & Koopman, 2001; Wiggers, de Wit, Gras, Coutinho, & van
den Hoek, 2003; Winfield & Whaley, 2002) or have provided only partial support for the
association (Mattson, 1999; Polacsek, Celentano, O'Campo, & Santelli, 1999). The
association of perceived severity with safer sex behaviors has also shown mixed results.
Several studies found that perceived severity has a Si gnificant positive association with
safer sex behaviors (e.g., Allard, 1989; Montgomery, Joseph, Becker, Ostrow, Kessler, &
Kirscht, 1989; Yep, 1993), whereas, in other studies, it has no association with the

behaviors (e.g., Volk & Koopman, 2001; Winfield & Whaley, 2002). The inconsistency in

22

the literature might be due to unsophisticated measures (Rosenstock et a1., 1994) or to the
lack of a comprehensive model (Winfield & Whaley, 2002). As such, the next hypotheses
are intended to replicate previous findings on the associations of perceived susceptibility
and perceived severity with safer sex behaviors. This study makes the following

predictions:

Hypothesis 4—a: Perceived susceptibility to a health threat will be positively
related to intention to engage in preventive health behaviors. Thus, the more
listeners show perceived susceptibility to HI V/AIDS, the greater they will show an
intention to use condoms in the future.

Hypothesis 4-b: Perceived severity of a health threat will be positively related to
intention to engage in preventive health behaviors. Thus, the more listeners show

perceived severity of HI V/AIDS, the greater they will show an intention to use
condoms in the future.

Perceived Benefits, Perceived Barriers, and Perceived Self-Efficacy

While there have been inconsistent findings on the relationship between perceived
benefits (or response eflicacy in other literature) and HIV/AIDS prevention behavior
(Allard, 1989; DiClemente, Durbin, Siegel, Krasnovsky, Lazarus, & Comacho, 1992;
Volk & Koopman, 2001; Winfield & Whaley, 2002), perceived barriers have received
relatively consistent support as a significant negative predictor of preventive behaviors
(Volk & Koopman, 2001; Winfield & Whaley, 2002; Zimmerman & Olson, 1994). The
HBM hypothesizes that at any level of perceived threat of AIDS beyond a threshold,
HIV /AIDS prevention behavior decisions become largely a function of perceptions of
both the benefits of and the barriers to behavior change (Rosenstock et a1., 1994, p.17).

For instance, Rosenstock and his associates (1994) noted that even though the perceived

23

threat is not high, strong perceived benefits of HIV/AIDS preventive behavior might still
influence behavior change. However, if the perceived threat and benefits of HIV /AlDS
prevention behavior are not high, it is not likely that low perceived barriers would
necessarily affect the preventive behavior. As such, this study predicts the influence of

both perceived benefits and perceived barriers on HIV /AIDS preventive behaviors:

Hypothesis 4-e: Perceived benefits of preventive health behaviors will be
positively related to intention to engage in preventive health behaviors. Thus, the
more listeners show perceived benefits of AIDS/HI V preventive behaviors, the
greater they will show intention use condoms in the future

Hypothesis 4-d: Perceived barriers to preventive health behaviors will be
negatively related to intention to engage in preventive health behaviors. Thus, the
more listeners show perceived barriers to AIDS/HI V preventive behaviors, the less
they will show intention to use condoms in the future.

Individuals’ beliefs that they can motivate themselves and regulate their own
behavior might play a crucial role in the process of behavior change in whether these
individuals will consider changing habits that are detrimental to their health (Bandura,
1994). The conceptualization of perceived self-efficacy in the social learning theory
(Bandura, 1977, 1997) is another important construct predicting preventive health
behavior.3 In the context of health communication, perceived self-efficacy can be defined
as “people’s beliefs that they can exert control over their own motivation, thought
processes, emotional states, and patterns of behavior” (Bandura, 1994, p. 26). Because

behavior change depends on one’s perceived capability to cope with stress and boredom

 

3 Rosenstock and others (1988) extended the HBM to include the concept of self-efficacy in order to
strengthen its utility and explanatory power for health preventive behaviors.

24

and to mobilize one’s resources and courses of action required to meet situational
demands, efficacy beliefs affect the intention to change risk behavior, the amount of
effort expended to attain this goal, and persistence in continuing to strive in spite of
barriers and setbacks that may undermine motivation (Schwarzer & Fuchs, 1995, p. 262).

Many studies have examined and supported a hypothesis that individuals’
perceived self-efficacy is positively related to HIV/AIDS preventive behaviors (e.g.,
Mattson, 1999; Polacsek et a1., 1999; Rotheram-Borus et a1., 1995; Steers et a1., 1996;
Wong & Tang, 2001; Wulfert & Wan, 1993). The findings of the studies showed that
when individuals’ self-efficacy toward safer sex increases, they are more likely to adopt
safer sex behaviors. As such, the next hypothesis represents an attempt to replicate

previous findings on the associations of perceived self-efficacy with safer sex behaviors:

Hypothesis 4-e: Perceived self-efficacy will be positively related to intention to
engage in preventive health behaviors. Thus, the more listeners show perceived
self-efficacy of AIDS/HI V preventive behaviors, the greater they will show an
intention to use condoms in the future.

The E fleets of Socio-Demographic Variables

The effects of several socio—demographic variables on each construct in this study
have been well documented in the previous studies. Gender might be a significant
predictor of narrative engagement. One of Green and Brook’s (2000) experiments showed
that women subjects showed significantly greater transportation than men subjects. In the
research on identification, the effects of gender differences were examined (Basil, 1996;

Brown et a1., 2003). However, the studies’ findings regarding the effects of gender on

25

identification were inconsistent. For example, Basil (1996) found that men, more than
women, identified strongly with Magic Johnson. Brown and his colleagues (2003) found
that women showed a level of identification with Princess Diana that was higher than the
level showed by men. This inconsistency might be due to the fact that people are more
likely to model the behavior of others of the same gender (Bandura, 1986). The HBM
studies have also examined gender differences in health beliefs and safer sex behaviors.
Some studies found that gender was a Si gnificant predictor of safer sex behaviors,
specifically, condom use (Dinoff & Kowalski, 1999; Wilson, Manual, & Lavelle, 1991;
Winfield & Whaley, 2002). Thus, the findings showed that women respondents were less
likely than men respondents to report condom use for preventing HIV/AIDS. In contrast,
other studies have indicated no gender differences in safer sex behaviors (DiClemente et
a1., 1992; Hingson, Strunin, Berlin, & Heeren, 1990; Volk & Koopman, 2001; Wong &
Tang,2001)

Previous studies also found that age has a significant positive association with
identification (Brown et a1., 2003; Levy, 1979). In contrast, several studies have shown
that education level is negatively associated with identification (Brown et a1., 2003; Levy,
1979). Thus, respondents with lower education levels showed stronger identification with
media characters than did respondents with higher education levels. In the HBM research,
although most studies have found no effect of age on safer sex behaviors (DiClemente et
a1., 1992; Winfield & Whaley, 2002), Hingson and others (1990) found that age
significantly predicted condom use, with younger persons more likely than older persons
to use condoms, but they did not find the effect of education on respondents’ condom use.

Finally, some studies examined whether or not a respondent’s marital status and

26

number of sexual partners affect safer sex behaviors. Individuals’ marital status (Volk &
Koopman, 2001) and number of sexual partners (DiClemente et a1., 1992) were
significantly related to condom use. Thus, whereas single respondents were more likely
to use condoms during sexual intercourse, respondents who reported more sexual partners
were less likely to use condoms. On the contrary, Hingson and others (1990) did not find

a Si gnificant effect of the number of sexual partners on safer sex behavior.

Proposed Model for the Effects of E-E Intervention on Safer Sex Behavior

The main purpose of the present study is to provide an integrated model with
which to explain the process by which E-E intervention has effects on behavior change.
More specifically, in the context of health campaigns for the prevention of HIV /AIDS in
Ethiopia, this study proposes a conceptual model explaining the roles of narrative
engagement and identification in the process of E-E and intervention and their
relationships with other relevant theoretical variables from social learning/cognitive
theory, audience involvement, and the HBM. Figure 1 shows a hypothetical model based
on the relationships hypothesized above. The proposed model represents how the
engagement of narrative affects the degree of respondents’ identification with a character,
which is a predictor of perceived susceptibility to and perceived severity of HIV/AIDS,
perceived benefits and perceived barriers of HIV/AIDS preventive behaviors, and self-
efficacy toward safer sex behavior. In turn, the model predicts that these four factors
positively affect individuals’ future safer sex behavior. Finally, to properly assess the

proposed model and achieve model parsimony, five socio-demographic variables discussed

27

earlier (gender, age, education, marital status, and number of sexual partners) are controlled

in the model.

28

 

 

 

539.8 xow
cohmm 23:".

 

 

_+

 

Smog—meow oozoocon. .
22:3 8285.”. .
fleecom umzoeod .
355m cozoocod .

b___nuaoom:w cozoeon. .

 

 

4

 

 

cosmoczcmc_

 

 

 

m596m £_mmI

 

 

 

 

EoEmmmmcm

 

 

 

 

 

 

Baton. .maxom

 

 

 

o>=mtmz 3.35:2 .
938:4. mesa .952 .
coamoacw .
Om< 0
509.00 0
EoEmmmmcm
. o>=mtmz moEQSmoEoo
o>=Emoo -o_oom

 

 

 

 

 

Eggnom Snow Snow .8 535?.35 =e3a~scfl4=~8£a§8=fl he 88am— 2: .5“ .252 3932—893.— .— Baum...—

29

CHAPTER 3

METHOD

An Overview of “Journey of Life, ” the E ntertainment-Education Program

The E-E stimulus used in this study is a serial radio drama in Ethiopia, called
Journey of Life. This radio program was designed as one component of the Ethiopian
Reproductive Health Communication Proj ect's four-year Information, Education and
Communication (IEC) initiative in Family Planning and HIV/AIDS Services by
Ethiopia’s National Office of Population and the Ministry of Health, along with the
Consortium of Family Planning NGOs and Johns Hopkins University Population
Communication Services (Belete, Girgre, & Witte, 2003). The overall purpose of this
radio drama was to provide education and modeling for modern family planning methods
and for the dangers of HIV/AIDS and, thus, how to avoid these dangers. According to
formative research (Witte, Girrna, & Girgre, 2001), the HIV epidemic in Ethiopia is very
serious. Adult prevalence is estimated at 7.3% (urban 13.4%, rural 5%); 250,000 children
under age 5 are estimated to be infected with HIV.

Because literacy rates in Ethiopia are low (almost 20% of Ethiopians reported
difficulty reading, and 25% reported not reading newspapers or magazines at all), only
6% reported not listening to the radio at all, with 64% reporting daily listening (Witte et
a1., 2001). This combination of low literacy rates and the popularity of radio listening
were important reasons underlying the campaign designers’ decision to select radio as the
optimal medium through which to transmit Journey of Life. In addition, radio is

inexpensive, reaches illiterate audience members, is portable, and is available at work

30

(Belete et a1., 2003). In reality, Ethiopians cited radio as one of their most frequent
sources of information about HIV/AIDS (Belete et a1., 2003).

Journey of Life, a 26-episode weekly radio drama, targeted young Ethiopians,
aged 15-30, within the five major regions of Ethiopia: Tigray (Mekele and Adigrat);
Amhara (Gondar and Bahir Dar); Oromia (J imma and Nazareth); Southern Nations,
Nationalities, and Peoples Region (Awassa and Arba-Minch); and the Addis Ababa
Administration (Akaki and Gedam-Sefer). Each radio episode lasted 20 minutes and was
broadcast on Sunday mornings at 10:00 am. for six months. The main message in
Journey of Life was “YICHALAL” (“It is possible”). This message means that it is
possible for the people in Ethiopia to improve the quality of their own life by planning
their families and that it is possible to control the epidemic of HIV/AIDS by taking
appropriate measures to control the spread of the disease (Belete et a1., 2003). This series
represented an attempt to generate feelings of “hope” and “confidence” in the belief that
to determine one's own family Size and to protect oneself from HIV /AIDS are beneficial
practices that can be achieved by all (Belete et a1., 2003). In this serial drama, a female
police investigator, Askale, and her husband, Bahiru, were created as protagonists or
positive role models who try to educate others about health topics including HIV /AIDS
(See Appendices 1 and 2). Throughout each episode, listeners took note of this couple’s
attempts to change (with more or less success) other characters’ attitudes toward
HIV /AIDS and to take precautionary steps to protect themselves from the virus, such as

condom use and testing for HIV (Smith et a1., 2003).

31

I Survey Data

The data used in this study were collected for a final summative evaluation of the
Ethiopian Reproductive Health Communication Project's four-year [EC initiative in
Family Planning and HIV /AIDS Services. This final summative evaluation was
conducted nationwide in 2002 after the end of the transmission of the Journey of Life.
This summative survey was to evaluate the impact of Journey of Life on respondents’
sexual behavior, beliefs and practices toward family planning and HIV prevention and on
their perceptions about various services and baniers to reproductive health (Belete et a1.,
2003). There were two sets of samples in the summative evaluation, panel survey and
cross-sectional survey, and the current study used the cross-sectional survey data. This
survey data included the measures that were grounded in various theories or models, for
example, perceived susceptibility, perceived severity, perceived self-efficacy, perceived
barriers, collective efficacy, sensation seeking, gender norms, and social capital (see
Appendix F).

A cross-sectional sample of 200 respondents was selected from Woredas
(districts) 21 and 14 of zones two and five of Addis Ababa. The total sample size of 200
was equally divided between the two Woredas (100 each) and each Kebele (local village)
within each Woreda (17 each for 8 Kebeles and 16 for 4 Kebeles). All of the households
in the selected Kebeles were listed, and the households having resident members within
the age group of 15-30 years old were identified. By using these households as a sample
frame and the predetermined sample size of 16 and 17 (for the specific Kebeles)
households, sampling interval for each Kebele was determined, and the sample

households were selected and identified by using systematic random sampling. Only one

32

person, within the target age, was interviewed within each household. For the purpose of
this study, of the 200 participants, the data from 124 respondents who reported listening
to the radio serial drama and completed the questionnaires were used. For more

methodological details concerning sampling procedures, see Belete and others (2003).

Measures

Narrative Engagement. Respondents’ answers to 12 questions served as the basis
for measurements of respondents’ engagement with narrative. Each item was measured on
a 5-point Likert-type scale from 1 (strongly disagree) to 5 (strongly agree). Based on
previous studies (Green & Brock, 2000; Sood, 2002), this study expects two dimensions
of narrative engagement: the affective dimension and the cognitive dimension. Aflective
engagement is conceptually defined as the degree to which audience members
emotionally respond to a media program. For example, questions such as “Journey of Life
had a strong emotional impact on me” or “Journey of Life kept me in suspense from
episode to episode” are instances of affective engagement. In contrast, cognitive
engagement is conceptually defined as the degree to which audience members respond to
the educational influences that a media program has on their beliefs or behaviors.
“Journey of Life influenced me to protect myself against HIV/AIDS” or “Journey of Life
made me believe there were effective ways to avoid HIV/AIDS infection” can be
considered instances of cognitive engagement. Table 1 illustrates narrative engagement

items.

33

Table l. Narrative Engagement Items

 

Items M SD
Joumey of Life influenced me to protect myself 4.39 .69
against HIV/AIDS.
Journey of Life made me feel hopeless about fighting 1.52 .74
against AIDS.
Journey of Life made me feel scared about AIDS. 2.89 1.53
Journey of Life made me feel emotional. 2.97 1.44
Journey of Life had a strong emotional impact on me. 3.26 1.45
Journey of Life was an emotional program. 3.55 1.36
Journey of Life made me feel hopeful about avoiding 4.34 .78
HIV/AIDS infection.
Journey of Life made me believe there were effective 4.39 .71
ways to avoid HIV/AIDS infection.
Journey of Life kept me in suspense from episode to 4.44 .76
episode.
I learned a lot about HIV/AIDS from Journey of Life. 4.59 .68
Journey of Life provided a lot of good information 4.53 .77
about HIV/AIDS.
After listening to Journey of Life, I wanted to change 4.24 .79

my life for the better.

Scale range: l-strongly disagree to 5-strong1y agree.

To ascertain whether these two measures have a two-factor structure, both an
Exploratory Factor Analysis (EFA) and a Confirmatory Factor Analysis (CFA) were
performed. First, a principal component analysis with varimax rotation on the 12 items
was run. A minimum factor loading of .40 was employed as a criterion for including an
item to be part of a factor. Two items with low factor loading (less than .40) or multi
loaded-factor (“Journey of Life kept me in suspense from episode to episode”) were
dropped. After dropping these two items, the results of the EF A suggested a two-factor

solution with eigenvalues greater than 1, which is consistent with the hypothesized factor

34

structure. The two-factor solution accounted for 53.85% of the variance. Factor loadings
ranged from .42 to .91.

To verify this factor structure obtained from the EPA, a CPA was conducted
through use of the LISREL 8.30 SEM program (J dreskog & S6rbom, 1999) using a
maximum likelihood estimation. Listwise deletion was used for missing data. After
dropping three items with correlated errors, the total of seven items yielded a two-factor
structure with good fit indices (GFI = .95, SRMR = .06, CFI = .97, and RMSEA = .07)4,
which is consistent with the result of the EF A. Cronbach’s as of .68 and .85 were
obtained for factor 1 (cognitive dimension) and factor 2 (affective dimension),
respectively. The CFA factor loadings ranged from .39 to .78 for cognitive dimension and
fiom .70 to .92 for affective dimension. Table 2 presents the indices of model fit, factor
loadings, and standard errors. The items of each factor were summed to create an overall
cognitive engagement score (M = 17.56, SD = 2.11, max = 20.00, min = 10.00) and
affective engagement score (M = 9.79, SD = 3.71, max = 15.00, min = 3.00). The Pearson

product-moment correlation between these two factors was .14 (n.s., two-tailed, n = 119).

 

4 In general, a value higher than .90 on the CPI was considered representative of a well-fitting model; a
cutoff point value close to .95 has recently been recommended as a good model fit (Hu & Bentler, 1999). A
favorable value of the SRMR is less than .10 (Kline, 1998) or close to .09 (Hu & Bentler, 1999). In contrast,
values of .8 or above on the CPI, closer to 1.0 on the CPI, and less than or equal to .08 on the RMSEA are
considered a good fit (Kelloway, 1998; Kline, 1998).

35

Table 2. Two-factor Model of Narrative Engagement Scales: Standardized

Factor Loadings, Standard Errors, and Goodness-of-Fit Summary ‘

 

 

 

 

Cognitive Affective
Engagement Engagement
Factor Factor
Loading SE Loading SE
Journey of Life made me feel
hopeful about avoiding HIV/AIDS .39
infection. b
Journey of Life made me believe
there were effective ways to .78 .55
avoid HIV/AIDS infection.
I learned a lot about HIV/AIDS 65 44
from Joumey of Life. ’ '
After listening to Journey of Life,
I wanted to change my life for the .55 .45
beflen
Journey of Life made me feel 70
. b .
emotional.
Journey of Life had a strong 92 18
emotional impact on me. ' '
Journey of Life was an emotional 76 14
program.
Goodness-of-F it Indices c
x2 (df) GFI RMSEA SRMR CFI
19.04 (13) .95 .07 .06 .97

 

a All factor loadings are significant under p < 0.001.

b Indicator item.

° GFl, goodness-of-fit index; RMSEA, root mean square error of approximation;
SRMR, standardized root mean squared residual; CFI, comparative fit index.

36

Identification with Characters. To measure the construct of identification with a
main character, two items were adopted from previous studies (e. g., Basil, 1996; Basil &
Brown, 1997; Brown et a1., 2003): (1) “I identify with Askale”; (2) “I want to be like
Askale.” Respondents were asked to indicate on a 5-point scale (1 = strongly disagree and
5 = strongly agree) how much they agree with each of these two statements as they regard
the character (inter-item correlation = .47, p < .001). Askale, the female police
investigator mentioned above, was selected as a protagonist or positive role model. AS
described in Appendices 1 and 2, she tries to educate others about health topics including
HIV/AIDS. The sum of the scores for the two items was taken so that a single

identification score could be created.

Perceived Susceptibility to AIDS/HIV. Perceived susceptibility to HIV/AIDS was
measured by a single item. Respondents were asked to indicate on a 5-point scale (1 =
strongly disagree and 5 = strongly agree) how much they agree with the following
statement: “I am at risk for getting HIV /AIDS.” The mean of perceived susceptibility was

1.58 (SD = .83).

Perceived Severity to AIDS/HI V. Perceived severity to HIV/AIDS was measured
by respondents’ ratings on their agreement with two statements: “Getting AIDS is the
worst thing that could happen to me” and “Getting AIDS is a sure death sentence.” These
two items were measured on a 5-point Likert-type scale from 1 (strongly disagree) to 5

(strongly agree). The inter-item correlation of the measure was .50 (p < .001 , two-tailed, n

37

= 124). The two items were summed to create an overall perceived severity score (M =

7.69, SD = 2.33, max = 10.00, min = 2.00).

Perceived Benefit. A single item was used to measure perceived benefits of
condom use. Respondents were asked to rate their agreement with the statement,
“Condoms are effective in preventing HIV /AIDS infection.” This item was measured on a
5-point Likert-type scale, to one end of which was affixed “strongly disagree” and to the

other end “strongly agree.” The mean of perceived benefit was 2.66 (SD = 1.34).

Perceived Barriers. Perceived barriers were measured using three items. The
respondents rated on their agreement with three statements: “HIV/AIDS prevention
methods cost too much” and “HIV/AIDS prevention methods are inconvenient” and “It is
better not to talk about HIV or AIDS in my community.” These three items were
measured on a 5-point Likert-type scale from 1 (strongly disagree) to 5 (strongly agree).
These three items showed low Cronbach’ s as of .49 and were summed to create an

overall perceived barriers score (M = 5.14, SD = 2.01 , max = 11.00, min = 3.00).

Perceived Self-Efficacy. Perceived self-efficacy was measured by asking
respondents to characterize their beliefs about their ability to take specific actions
(condom use) to avoid HIV transmission. Respondents indicated how strongly they
agreed or disagreed with three statements, using a five-point Likert scale (1 = strongly
disagree, 5 = strongly agree): (1) “I am (would be) able to use condoms to prevent

HIV/AIDS infection,” (2) “I am able to talk to my partner about HIV /AIDS prevention

38

methods to prevent infection,” and (3) “Using condoms to prevent HIV /AIDS is (would
be) easy for me” (Cronbach’s or = .66). The three items were summed into a single index

for perceived self-efficacy (M = 11.12, SD = 2.77, max = 15.00, min = 5.00).

Future Safer Sex Behavior: Future safer sex behavior was measured by a single
item. Respondents were asked to rate their agreement with the following statement: “I
intend to use condoms to prevent HIV /AIDS infection during the next 12 months.” This
item was measured on a 5-point Likert—type scale from 1 (strongly disagree) to 5 (strongly
agree). The mean of safer sex behavior was 2.86 (SD = 1.40). Table 3 shows the items used

for measuring respondents’ health beliefs and behavioral intention related to safer sex.

Socio-demographics. Five socio-demographic variables serve as control
variables: sex, age, education, marital status, and number of sexual partners. Respondents
had a mean age of 22.04 (SD = 4.32) with a higher proportion of females (77.4%) to
males (22.6%). Surveyed on a five-point ordinal scale ranging from “Illiterate, No
schooling” to “Over 12 years or Diploma,” most respondents (about 84%) Showed an
education level of 5-12 years. 25% of respondents were “married” or “living with
partner,” whereas 68.5% of them were single. The number of sexual partners was
measured by respondsin terms of their response to the following question: “About how
many sexual partners have you had during the last 12 months (the last year).” Most
respondents had “no partners” (53.2%) or “just one partner” (45.2%). To achieve both
control and model parsimony simultaneously, these five socio-demographic variables

were residualized in the model by using multiple regressions. Appendix C shows the

39

socio-demographic characteristics of this study’s respondents.

Table 3. Health Beliefs and Behavioral Intention Items

 

 

Items M SD
Susceptibility I am at risk of getting HIV/AIDS. 1.58 .84
Severity Getting AIDS is the worst thing that 4.15 1.16
could happen to me.
Getting AIDS is sure death sentence 3.54 1.52
Perceived Condoms are effective in preventing 2.66 1.34
Benefit HIV/AIDS infection.
Perceived It is better not to talk about HIV or AIDS 1.77 1.12
Barriers in my community.
HIV/AIDS prevention methods cost too 1.51 .67
much.
HIV/AIDS prevention methods are 1.85 1.00
inconvenient.
Self-Efficacy I am (would be) able to use condoms to 3.51 1.29
prevent HIV/AIDS infection.
I am able to talk to my partner about 4.11 .95
HIV/AIDS prevention methods to prevent
infection.
Using condoms to prevent HIV/AIDS is 3.48 1.32
(would be) easy for me.
Behavioral I intend to use condoms to prevent 2.86 1.41
Intention HIV/AIDS infection during the next 12

months.

Scale range: 1-strongly disagree to 5-strongly agree.

4O

CHAPTER 4

RESULTS

The analysis consisted of a two-part analysis. First, to test the hypotheses of this
study, zero-order relationships among exogenous and endogenous variables were initially
observed. Although bivariate correlation statistics commonly provide researchers with
useful information about the relationship between two variables, it should be indicated
that there is a clear distinction between a simple correlational relationship and a causal
relationship. Therefore, in order to test the hypothesized model based on the causal links

among variables, path analyses were conducted.

C orrelational Analyses

Hypothesis 1 was to test the relationship between cognitive engagement with the
narrative of the radio program and identification with a character. As shown in Table 4, a
significant positive association between cognitive engagement and identification with a
protagonist (Askale) was found (r = .25, p < .01 , two-tailed, n == 119). Thus, the result
suggests that the more listeners are cognitively engaged with the narrative of the E-E
program, the more likely they are to Show identification with the character of the program.
Hypothesis 1 was supported.

Hypothesis 2 was to test whether affective narrative engagement would be
positively related to identification with a character. The results Show that there was no
significant relationship between affective engagement and identification with a character

(r = -.04, us. two-tailed, n = 119). Hypothesis 2 was not supported.

41

Hypotheses 3—a through 3-d predicted the significant relationships between
identification with a character and five health belief variables. According to Table 4,
identification with a character was significantly related to both perceived severity (r = .28,
p < .01, two-tailed, n = 119) and perceived barriers (r = -.19, p < .05, two-tailed, n = 119).
Thus, the results suggest that the more listeners Show identification with the character of
the EB program that concerns HIV /AIDS prevention, the more likely they were to show
greater perceived severity of HIV /AIDS and less perceived barriers to safer sex behaviors.
Hypotheses 3-b and 3-d were supported. A relationship between identification with
character and perceived self-efficacy approached significance (r = .16, p < .10, two-tailed,
n = 118).5 Hypothesis 3-e -- that the more listeners show identification with the character
of the EB program that concerns HIV/AIDS prevention, the more they would show
perceived self-efficacy of safer sex behaviors -- was supported at the significance level
of . 10. In contrast, no significant relationships between identification with character and
perceived susceptibility and between identification and perceived benefit were found.
Hypotheses 3-a and 3-c were not supported.

Hypotheses 4-a through 4-e were to replicate the previous studies on the
relationships between health belief variables and behavioral intention related to future safer
sex. A relationship between perceived susceptibility and behavioral intention to adopt
condom use approached significance (r = .17, p < .10, two-tailed, n = 124). Thus, this
suggests that the more listeners show perceived susceptibility to HIV /AIDS, the more

they show an intention to use condoms to prevent HIV/AIDS infection in the future.

 

5 Significance level in this study was determined as less than p-value of .10. When considering that
statistical power is sensitive to sample size and the sample size in this study is relatively small, it is
reasonable that the criterion of the significance level is p-value of .10.

42

Hypothesis 4-a was supported at the significance level of .10. Hypothesis 4-b predicted a
positive relationship between perceived severity and behavioral intention. The results,
however, did not support this relationship. For perceived benefit and barriers, perceived
benefit of condom use for safer sex was positively related to behavioral intention at the
significance level of .10 (r = .16, p < .10 two-tailed, n = 124), whereas perceived barriers
to safer sex behavior was not found to be significantly related to behavioral intention.
Hypothesis 4-c was supported at the significance level of .10, but hypothesis 4-d was not. It
was also hypothesized that perceived self-efficacy would be positively related to behavioral
intention to use condoms. The results strongly support this hypothesis (r = .54, p < .001,
two-tailed, n = 123). Consistent with previous studies (e. g., Mattson, 1999; Polacsek et
a1., 1999; Steers et a1., 1996; Wong & Tang, 2001; Wulfert & Wan, 1993), hypothesis 4-e

was supported.

43

Table 4. Zero-Order Correlations among Exogenous and Endogenous

Variables in the Proposed Model

 

 

1 2 3 4 5
1. Cognitive Engagement
2. Affective Engagement .14
3. Identification with Askale .25" -.04
4. Perceived Susceptibility -.08 .05 -.14
5. Perceived Severity .24" .22* .28** -.25**
6. Perceived Benefit .02 .43*** .12 .00 .11
7. Perceived Barriers -.42*** -.06 -.19* .23* -.13
8. Perceived Self-Efficacy .18* .26** .16# .11 .01
9. Behavioral intention -.08 .16* .12 .17# -.04

 

# < .10, * p < .05, ** p < .01, and p < .001, two-tailed.

‘0

 

 

Cognitive Engagement

Affective Engagement

Identification with Askale

Perceived Susceptibility

Perceived Severity

Perceived Benefit

Perceived Barriers -.13

Perceived Self-Efficacy .34*** -.24**
Behavioral Intention .26** .00 .54***

SPP°NQQPSPNf

# p < .10, * p < .05, ** p < .01, and p < .001, two-tailed.

44

Path Analyses: Testing the Hypothesized Model

A path analysis, a statistical method of structural equation modeling (SEM), was
used to test the hypothesized model. The LISREL 8.30 SEM program was employed to
obtain maximum-likelihood (ML) estimates of the model’s parameters (J dreskog &

Sdrbom, 1999b).

Model Specification and Parameter Estimates. This study employed a classical
approach to path analysis, which consists of solving structural equations for the
dependent variables in terms of the independent and the random error terms C1, C2, ...., 4;,
(J dreskog & Sorbom, 1996a). According to Jdreskog and Sdrbom (1996a), estimating a
path analysis model for directly observed variables with LISREL rests on the
understanding that the model is a system of equations and estimates because it estimates
all equations rather than each equation separately.

Table 4 shows the correlation matrix employed for estimating the recursive model
hypothesized in this study. Cognitive engagement and affective engagement were treated as

exogenous variables, and the other seven variables were endogenous variables:

xi = Cognitive Engagement

)9 = Affective Engagement

y] = Identification with Askale
yz = Perceived Susceptibility
y3 = Perceived Severity

y4 = Perceived Benefit

y5 = Perceived Barriers

yo = Perceived Self-Efficacy
y7 = Behavioral Intention

45

Figure 2 represents a path diagram for the recursive model of this study. This diagram
illustrates that both cognitive engagement and affective engagement directly affect
identification that is a predictor of the five health belief variables (perceived
susceptibility to and perceived severity of HIV/AIDS, perceived benefit and perceived
barriers of HIV /AIDS, preventive behaviors, and self-efficacy toward safer sex behavior).
In the final block of the model, the five health belief variables directly affect individuals’

intention to use condoms in the future.

46

Figure 2. Path Diagram for the Recursive Model of the Effects of Entertainment-

Education Intervention on Safer Sex Behaviors

 

 

 

 

 

 

 

X2

 

 

in

Y1

    

 

61

47

KY4 l‘ I374

 

(f2

 

 

 

 

Y5

 

 

 

 

 

Yo

 

 

Y7

Two exogenous variables (cognitive engagement and affective engagement) were
allowed to freely covary with one another by definition. Normally, it is reasonable to
allow covariances among the exogenous variables, and forcing orthogonality (zero
correlation) between two exogenous variables should be undertaken only with strong
theoretical justification (Baer, 2002). The variance-covariance matrix of the exogenous

variables ((1) matrix) is as follows:

 

 

Two paths from each exogenous variable to y; (identification) were allowed in the
model, but all of the paths from the exogenous variables to the other six endogenous
variables (y; through y7) were constrained to equal zero. The matrix of the coefficients

linking exogenous variables with endogenous variables (F matrix) is:

I’ll 1’12
0 0
0 0
F= 0 0
0 0
0 0
_0 0 _

 

 

It is the case with any recursive model that the matrix of coefficients linking

endogenous variables with other endogenous variables (B matrix) is sub-diagonal (Baer,

48

2002). Parameters from y] (identification) to all five endogenous variables (Y2 through y6),
except for y-, (behavioral intention), were freely estimated. In turn, parameters from the

five endogenous variables (y; through y6) to y7 (behavioral intention) were estimated. The

beta matrix is:

I—

0

,321
P31
B: .541
.351
1661 0 0
0 ,372 .373 ,374

_

0
0
0
0
0

QQQDQQ

0
0
0
0
0
0

QQQQQQQ

 

0 0
0 0
0 0
0 0
0 0
0 0
fl75 fl76

 

Covariances among C2 through C6 (the random error terms of health belief
variables) were allowed to freely covary with one another by definition. Like the phi
matrix ((1) matrix), it is reasonable to allow covariances among the random error terms of
endogenous variables. The rest of the covariances (between C1 and (,2 through C6, between
C. and C7, and between C2 through Co and C7) were constrained to equal zero because the
connections between the second block (yi), the third block (Y2 through y6), and the fourth

block (y7) are unidirectional. As such, the variance-covariance matrix of the error terms

of endogenous variables (‘1’ matrix) is as follows:

 

 

Wu 0 0 0 0 0 07

0 W22 0 0 0 0 0

0 W32 W3 3 0 0 0 0
‘1’ = 0 W42 51,4 3 W44 0 0 0

0 Y’sz W53 W54 W55 0 0

0 W62 7’63 W64 W65 W66 0

0 0 0 0 0 0 W77

49

As such, the structural equations in this model are as follow:

Y1=711X1+712X2+C1

Y2=1321Y1+C2
Y3=1331M+C3
Y4=fl41YI+C4
YS=fl51 Y1+C5
Yo=floi YI+C6

Y7=,372 Y2+l373 Y3+1374 Y4+,BISYS+,376 Yo+C7

Multiple fit indices were used to test the model. Although the x2 goodness-of-fit
statistics and fit indexes are two of the most popular ways in which to evaluate model fit
(Hu & Bentler, 1999), there are problems with the x2 statistic as a fit index because it is
very sensitive to sample size (Kline, 1998). As a supplement to the x2 statistic, several fit
indexes have been employed, such as Goodness-of-Fit Index (GFI), the Adjusted
Goodness-of-Fit Index (AGFI), Gamma Hat, the Normed Fit Index (NF I), the Tucker-
Lewis Index (TLI), the Comparative Fit Index (CF I), SRMR, Root Mean Square Error of
Approximation (RMSEA), and so on. However, because a single index reflects only a
particular aspect of fit, a favorable value of that index does not by itself indicate good fit
(Kline, 1998). Although there is no solid consensus for good fit, SEM scholars have
recommended that researchers should use multiple indexes for assessing model fit. For
example, Hu and Bentler (1999) recommended that researchers employ a 2-index
presentation strategy, which includes use of the maximum likelihood (ML)-based SRMR

and its being supplementing with TLI, Bollen’s Fit Index (BL89), the Relative

50

Noncentrality Index (RNI), CFI, Gamma Hat, McDonald’s Centrality Index (Me), or
RMSEA. When sample sizes are less than 250, the recommended combinational rules
based on CFI, BL98, RNI, or Gamma Hat in combination with SRMR are more
preferable (Hu & Bentler, 1999).

This study considered SRMR and CFI as fit indices for the model. In general, for
the ML method, a value higher than .90 on the CPI was considered representative of a
well-fitting model, but a cutoff point value close to .95 has been recently recommended
as a good model fit (Hu & Bentler, 1999). A favorable value of the SRMR is close to .08
(Hu & Bentler, 1999), and an RMSEA value close to .06 or less (Browne & Cudeck,
1993). Unfortunately, the fit indices did not show a good model fit. (x2 (13) = 57.58;
SRMR = .l 1; CF I = .63). As such, this study attempted to respecify the original model

based on both theoretical and empirical considerations.

Model Respecification

To respecify a recursive or nonrecursive model, two ways in which to test a
model can often be used: model trimming and model building (Kline, 1998). Model
trimming refers to a strategy wherein the researcher begins a path analysis with a just-
identified model and simplifies it by eliminating paths. As paths are trimmed, the chi-
square increases, which means that the model fit becomes worse. In contrast, model
building is the opposite strategy to model trimming. The researcher adds paths to a null
model. As paths are added to the model, the chi-square decreases, which means that the
model fit becomes improved. In either strategy, the eliminating (or the adding) of paths

should be determined based only on theoretical considerations or face validity although it

51

is guided by empirical information obtained from a statistic program (i.e., a modification
index or the chi-square difference test).

In the early stages of model or theory development, it is useful to eliminate the
nonsignificant path coefficients and with another sample population, examine the
resulting respecified model against the full model (Kerlinger & Pedhazur, 1973). As
such, the model trimming strategy was employed to respecify the original model, and the

respecified models contain only the significant paths.6

Modified Model 1. Using a model trimming strategy, this study proposed a
respecified model. First, paths from the two exogenous variables to the five health belief
variables were allowed to be free. Conceptually, this means that individuals’ engagement
with narratives has a direct effect on their beliefs regarding HIV/AIDS as well as an
indirect effect on the beliefs. The findings of several experiments by Green and Brock
(2000) support this relationship. Green and Brock (2000) found that transportation in
persuasive narratives positively affects individuals’ story—consistent beliefs on both the
specific and the general items implied by the narrative. Thus, their findings showed that
highly transported participants reported beliefs more consistent with those implied in the
story. Sood (2002) also supports the association between audience involvement and
beliefs. Sood (2002) found that both the cognitive and the affective dimensions of

audience involvement significantly affect self-efficacy. In particular, the findings showed

 

6 A criterion for eliminating the nonsignificant path coefficients was a p-value higher than .10. In other
words, if the t-value of a path coefficient was less than 1.645, the path was eliminated from the model.
When considering that statistical power is sensitive to sample size and the sample size in this study is
relatively small, it is reasonable that the criterion of the significance level for determining model trimming
is p-value of .10.

52

that the affective dimension of audience involvement had a higher impact than the

cognitive dimension on self-efficacy.

Second, as noted above, all of the nonsignificant paths were eliminated from the
just-identified model. For the F matrix, paths from cognitive engagement to both
perceived benefit and perceived self-efficacy and paths from affective engagement to
both identification and perceived barriers were constrained to equal zero. Two paths from
two engagement variables to behavioral intention and two paths from two engagement

variables to perceived susceptibility were also constrained to equal zero. AS such, the F

matrix was changed into the following:

F711 0 q
0 0
1’31 1’32
F: 0 742
1’51 0
0 1’62
-0 0 .1

 

 

For the Beta matrix, all paths were eliminated from the model, except for the four
paths (identification 9 perceived severity, identification 9 perceived benefit,
identification 9 perceived self-efficacy, and perceived self-efficacy 9 behavioral
intention). As such, the Beta matrix was changed into the following:

I-

0
0

.531
B = 1341
0

Bo]
0

QQQQQQQ
QQQQQQQ

0
0
0
0
0
0
0

 

QQQQQQQ
EQQQQQQ

 

—

53

SEM was used to test the respecified model. Maximum-likelihood (ML) estimates
of the model’s parameters were obtained by using the LISREL 8.30 SEM program
(J dreskog & Sdrbom, 1999). As shown in Table 5, the indices for the model indicate a
good model fit ()6 (15) = 18.97, SRMR = .05, and CF I = .97). The fitted model accounted
for 24% of the variance in behavioral intention to use condoms. Figure 3 presents a
proposed respecified model including the standardized path coefficients with p-values.

The standardized path coefficients indicate that cognitive engagement positively
affects identification with character (y = .21 , p < .05), even when controlling for affective
engagement and socio-demographic variables. Thus, the more listeners are cognitively
engaged with the narrative of the EB program, the more likely they are to show
identification with the character of the program, supporting the finding of zero order
correlation analyses. In turn, identification positively affects perceived self-efficacy ([3
= .21, p < .05) that is a substantial positive predictor of behavioral intention to use
condoms ([3 = .50, p < .001). These results suggest that cognitive engagement -- through
identification and perceived self-efficacy -- indirectly affects behavioral intention to use
condoms.

Although no significant effect of affective engagement on identification with
character was found, perceived self-efficacy was a significant mediator of the effect of
affective engagement on behavioral intention to use condom. Consistent with the findings
of Sood (2002), affective engagement with the EB program positively affects perceived
self-efficacy of HIV/AIDS preventive behaviors (y = .21, p < .05). It suggests that
affective engagement has an indirect effect on future safer sex behavior through its effect

on perceived self-efficacy (perceived self-efficacy 9 future safer sex behavior: [3 = .50, p

54

< .001).

In the respecified model, it was also predicted that both cognitive and affective
engagement are significant predictors of health belief variables. As shown in Figure 3,
affective engagement positively affects perceived severity (y = .18, p < .05). Cognitive
engagement has a direct effect on perceived severity (y = .16, p < .10) as well as an
indirect effect, therein, through its effect on identification (identification 9 perceived
severity [3 = .24, p < .01). These results suggest that the more listeners were cognitively
and affectively engaged with the narrative of the E-E program, the more they tended to
perceive HIV/AIDS as a serious threat to them. Interestingly, cognitive engagement was a
strong negative predictor of perceived barriers (y = -.40, p < .001), whereas affective
engagement was a strong positive predictor of perceived benefit (y = .39, p < .001). An
indirect effect of cognitive engagement on perceived benefit through its effect on
identification was also found (identification 9 perceived benefit: [3 = .15, p < .10).
Figure 3 shows that neither perceived severity nor perceived benefit nor yet again
perceived barriers was a significant predictor of future safer sex behavior, when

controlling for other variables used in this model and socio-demographic variables.

55

Figure 3. Modified Recursive Path Model of the Effects of Entertainment-Education

Intervention on Safer Sex Behavior

 

 

 

  
 
  
   

 

Perceived
/ self-Efficacy
Z Perceived 50*“

Affective Severity
Engagement

. .24*/

Identification/ Future Safer

.1 571 Sex Behavror
_39m\ Perceived
Benefit
Cognitive
Engagement - 40“. > Perceived
Barriers

t p < .10, * p < .05, ** p < .01, and p < .001, two-tailed.

56

Modified Model 2. This study proposed another modified model in terms of the
relationships between perceived self-efficacy and health belief variables. As noted earlier,
to strengthen the explanatory power for preventive health behaviors, the HBM was
extended by including perceived self-efficacy in the model. Although most of the HBM
studies have considered self-efficacy as another predictor of preventive health behaviors,
like the other four health beliefs factors, several studies have suggested the causal
relationships between self-efficacy, perceived benefits, and perceived barriers (e.g.,
Basen-Engquist, 1992; Mattson, 1999; Winfield & Whaley, 2002). These studies have
found that self-efficacy is positively related to perceived benefits and negatively related
to perceived barriers. Underlying this study, as well, is the assumption that perceived self-
effrcacy would be a mediator between the effects of perceived benefit and the effects of
perceived barriers on future safer sex behavior. Modification indexes obtained from the
original model support this assumption that there will be causal relationships between
these three variables. As shown in the Modification Indices for Beta (see Appendix D), a
very high value was found in a path from perceived self-efficacy to perceived benefit
(15.71). In addition, the modification index value for a path from perceived self-efficacy
to perceived barriers is relatively high (3.38). These observations imply that if we release
the equality constraint of the perceived self-efficacy effect both on perceived benefit and
on perceived barriers, then it is to be expected that the chi-square of the model will
decrease, and, after all, the model fit will improve.

Therefore, this study proposed another modified model that included the causal
paths from perceived self-efficacy to perceived benefit and perceived barriers. By

employing the same procedures used to respecify Modified Model 1, Modified Model 2

57

was respecified. First, the parameters from cognitive engagement to perceived benefit
and to perceived self-efficacy and the parameters from affective engagement to
identification, to perceived self-efficacy, and to perceived barriers were constrained to
equal zero. All of the parameters from both the cognitive and the affective engagement to
both perceived susceptibility and behavioral intention were also constrained to equal zero.
As such, the 1" matrix was changed into the following:

Wu 0 -
0 0
1’31 1’32
F = 0 142
1’51 0

0
0

0
-0

 

 

_I

For the Beta matrix, two parameters from identification with character to perceived
susceptibility and to perceived barriers were constrained to equal zero. In addition,
parameters from health belief variables to behavioral intention were eliminated from the
model, except for a path from perceived self-efficacy to behavioral intention. Parameters
from both perceived severity and perceived susceptibility to perceived self-efficacy were

also constrained to be zero. As such, the Beta matrix was changed into the following:

 

0 0 0 0 0 0 0 1
0 0 0 0 0 0 0
.831 0 0 0 0 0 0
B = [341 0 0 0 0 0 0
0 0 0 0 0 0 0
1861 0 0 ,364 ,865 0 0
0 0 0 0 0 ,6’ 76 0

 

58

Since two parameters (perceived benefit 9 perceived self-efficacy, perceived
barriers 9 perceived self-efficacy) were released, the ‘1' matrix slightly changed. The
covariances between the error terms of perceived benefit, perceived barriers, and
perceived self-efficacy were constrained to equal zero. As such, the ‘1’ matrix is as

follows:

 

 

W11 0 0 0 0 0 0

0 W22 0 0 0 0 0

0 W3 2 W33 0 0 0 0
‘1” = 0 W42 W4 3 W44 0 0 0

0 W5 2 W53 0 W55 0 0

0 9’62 W63 W64 0 W66 0

0 0 0 0 0 0 W77

SEM with ML estimates was used to test the respecified Model 2, as well. Table 5
indicates that this model also shows a good model fit (x2 (13) = 22.60, SRMR = .06, and
CFI = .97). In order to compare this model with Model 1, a chi-square difference test was
conducted. As indicated in Table 5, a chi-square difference test shows that although the fit
of Model 2 was significantly improved compared to the original model (A);2 (5) = 36.77, p
< .001), there was no significant difference in model fit between Model 1 and Model 2
(Ax2 (3) = 1.42, p = .70). The fitted model accounted for 25% of the variance in
behavioral intention to use condoms, which means that the total variance explained
behavioral intention slightly increased compared to Model 1. Thus, to put it another way,

the overall fit of Model 2 was not worse than that of Model 1 although two degrees of

freedom increased.

59

Figure 4 presents a path diagram for Modified Model 2, including the standardized
path coefficients with p-values. Cognitive engagement was a significant positive predictor
of identification with character (y = .21, p < .05). A path from identification to perceived
benefit was significant at the significance level of .10 (B = .15, p < .10). Thus, these
suggest that cognitive engagement, through its effect on identification, indirectly affects
perceived benefit. Only the direct effects of cognitive engagement on both perceived
severity and perceived barriers were found. While a path from cognitive engagement to
perceived severity was significant at the significance level of .10 (y = .14, p < .10),
cognitive engagement appeared to have a strong negative effect on perceived barriers (y =
-.43, p < .001). Affective engagement was a positive predictor of both perceived severity
(y = .20, p < .05) and perceived benefit ()7 = .39, p < .001).

In the respecified Model 2, the mediating role of perceived self-efficacy in the
effects of both perceived benefit and perceived barriers on behavioral intention to use
condoms was examined. As predicted, both of the effects of both perceived benefit and
perceived barriers on behavioral intention were mediated by perceived self-efficacy.
Perceived benefit positively and substantially affects perceived self-efficacy (B = .28, p <
.001), and perceived barriers negatively influences perceived self-efficacy (B = .28, p <
.001) that was a strong positive predictor of behavioral intention (B = .50, p < .001).
These results support the reconceptualization of the HBM suggested by Mattson (1999),

which suggests that perceived benefits and barriers are the predictors of self-efficacy.

6O

Figure 4. Alternative Modified Recursive Path Model of the Effects of

Entertainment-Education Intervention on Safer Sex Behavior

Perceived
20* Severity

   
   
 

 

Affective .
Perceived
Engagement
/Self-Efficacy
Perceived/
. Benefit ...
Identif cation 1/V
. Future
Cognitive _'43... Safer Sex
Engagement + Perceived Behavior

Barriers

# p < .10, * p < .05, ** p < .01, and *** p < .001, two-tailed.

61

Table 5. Goodness-of-Fits Summary among Models

 

 

 

Contrast with
hypothesized
X2 df model GFI SRMR CFl

sz Adf
HypOtheSIZed 57.58*** 13 __ .. 91 _11.63
Model
Proposed m
Modified Model1 ”'39 15 38'19 1 '97 '05 '97
Pl°P°Sed 20.81 18 36.77*** 5 .96 .06 .97

Modified Model 2
Note: Chi-Square difference between Model 1 and Model 2 was not significant

(sz (3) = 1.42, p = .70).
p < .001.

62

CHAPTER 5

DISCUSSION

This study examined the persuasive effects of narrative content in an E-E program
on HIV/AIDS preventive behavior in Ethiopia. Utilizing causal modeling, this study
explored theoretical mechanism how E-E messages influence individuals’ beliefs and
behavior. In fact, in Spite of the abundant findings regarding the effects of E-E on
behavioral changes, there have been very few studies on the process that underlies the
workings of E-E. In terms of providing an explanatory model that sheds light on the
process of EB in the context of health-related media campaigns, this study might have

several implications for theoretical developments in E-E research.

The Roles of Narrative Engagement and Identification in E-E Intervention

As noted above, this study focused on the roles that both narrative engagement
and audience members’ identification with a character featured in an Ethiopian radio
serial drama, Journey of Life, played in the efficacy with which E-E gave rise to
behavioral change in adopting condom use for safer sex. A recursive model that includes
causal paths from the engagement of narrative content to behavioral intention to condom
use was hypothesized. This recursive model represents how the engagement of narrative
affects respondents’ identification with character, which is a predictor of individuals’
beliefs about HIV/ADDS and safer sex behavior (perceived susceptibility to and perceived
severity of HIV/AIDS, perceived benefits and perceived barriers of HIV/AIDS

preventive behaviors, and self-efficacy toward safer sex behavior). In turn, it was

63

hypothesized that these beliefs mediate the effects of narrative engagement and
identification on the behavioral intention to engage in safer sex.

Because the results of a path analysis showed, however, that the hypothesized
model did not fit the data, the model was respecified. Therefore, the current study
proposed two alternative models based on the results of the respecification of the original
model. In both of the alternative models, this study found that an audience member’s
identification with a character played a mediating role in the relationships between
narrative engagement and individuals’ beliefs about HIV/AIDS although this mediating
role was partially supported. The results of path analyses indicated that cognitive
engagement was significantly related to identification with the character. Thus, the listeners
who are cognitively engaged with the narrative of the radio program were more likely to
identify with the character of the program, supporting Slater and Rouner (2002)’s
argument that identification is conceived to be a mediator of the effect of engagement
with the narrative both on health beliefs and on behavioral intention to use condom. The
more the listeners cognitively engaged with the story world, which tells how the
characters of the program suffer from HIV/AIDS and how a protagonist fights against
ignorance of the disease and educates and persuades people to protect themselves from
the virus, the more they tended to identify with the protagonist, Askale. These findings
support the hypothesis that as engagement with the narrative content creates strong
cognitive responses to the characters in the narrative content, it promotes audience
members to identify with characters. The following notion of Oatley and Gholamain
(1997) explains the relationship between cognitive engagement with narrative content

and identification with a character. Since the audience takes on the goals and plans of a

64

character, an understanding of narrative involves “a specific mode of thinking with its
own characteristics related to actions and outcomes of actions to the goals and plans”
(Oatley and Gholamain, 1997, p. 269). Oatley and Gholamain (1997) illustrate an

example:

From the age of four or five, when they keen on playing games involving roles —
doctors and patients, cops and robbers, husbands and wives — children listening to
a story, or watching a movie, say spontaneously, “I will be that one.” They are
very clear about the role with which they want to identify. (p. 269)

A disappointing result of this study, however, is that the significant relationship
between affective engagement and identification was not found although previous studies
(e. g., Oatley, 2002; Slater & Rouner, 2002) have argued that affective engagement is
closely related to identification. A possible alternative explanation is that affective
engagement indirectly influences identification through its effect on cognitive
engagement. It suggested that “to feel” precedes and impacts “to think.” Oatley and
Gholamain (1997) argued that identification was governed by secondary processes such
as conscious, meaning-based thoughts. Several experiments of Murphy and Zajonc
(1993) have provided strong evidence that affect is processed early in the information-
processing chain. Thus, emotional reactions can precede subsequent cognitions (Murphy
& Zajonc, 1993). To explore the relationship among the three variables of affective
engagement, cognitive engagement, and identification, a simple recursive path model -—
in which affective engagement influences cognitive engagement that leads to

identification with character ——was tested. The results of the analysis indicated that the

65

data were consistent with the model (error = .005, z = 0.04, p = .97). Therefore, a future
study should explicate the causal relationships among these three variables.

Several hypotheses were to test the relationships between identification with
character and individuals’ beliefs related to HIV/AHDS. The results of a bivariate
correlation analysis indicted that identification was significantly related to three beliefs:
perceived severity, perceived barriers, and perceived self-efficacy. However, the
relationship between identification and perceived barriers was not significant in both of the
two modified models. According to the results of the path analyses, identification with
character significantly affects perceived severity and perceived benefit in both of the
models, whereas its effect on perceived self-efficacy was significant only in Modified
Model 1. In other words, in Modified Model 2 predicting a meditating role of self-efficacy
in the effects of perceived benefit and perceived barriers on future safer sex behavior,
identification was not a significant predictor of self-efficacy. AS such, the hypotheses on the
relationships between identification with character and individuals’ beliefs related to
HIV /AIDS were partially supported. These findings suggest that the more the individuals
identified with the protagonist, the more likely they were to perceive HIV/AIDS as serious
disease. In addition, it seems that through their identification with the protagonist,
listeners grew more aware that condom use is an effective method by which to prevent
HIV/AIDS infection.

In spite of the crucial role of identification in the E-E process, the “Archie Bunker
Effect” was also found in this study. The Archie Bunker Effect refers to “the degree to
which certain audience individuals identify with negative role models in the media”

(Singhal & Rogers, 1999, p. 158). To determine whether the Archie Bunker Effect appears

66

in this study, identifications with three types of characters—positive, transitional, and
negative role models—were compared. In so doing, three main characters of the program
were selected based on the story line of the drama (see Appendices A and B). Saba was
chosen as a transitional character. In general, transitional characters ofien develop in the
Show as they make mistakes, yet they are neither a protagonist nor a villain (Smith et a1.,
2003). Saba, who is Askale’s sister, enjoys drinking and socializing with many men. She
did not listen to her sister’s warnings to either abstain from sex or use protective methods.
Finally, she was suffering from discovering that she was HIV positive and had infected
others. The architect of this program created her as a transitional model that could convey
to the audience the foolishness of her ways and encourage the audience to learn from
them. Zeleke was selected as a negative role model or villain. Although Zeleke and his
wife agreed to take in Azeb’s son and find him a family to live with, they decided to raise
him themselves and to give the child to Alemu, who cripples children and forces them to
beg on the streets. Askale was the favorite Journey of Life character to whom subjects
best responded (65.5%), whereas Zeleke was the least favorite Journey of Life character
(50.4%). Respondents indicated that Saba was the second least favorite Journey of Life
character (36.5%) and the third most favorite character (9.2%). Not surprisingly, no one
indicated that Askale was his or her least favorite character or that Zeleke was his or her
most favorite character.

To see whether there are mean differences between audience members’
identification with the three characters, this study looked at the within-subject mean
differences between individuals’ identification with Askale, identification with Saba, and

identification with Zeleke via the General Linear Model (GLM) procedure. As shown in

67

Appendix E, significant mean differences in identifications across the three characters
were found (A= .26, F (2, 117) = 165.59, p < .001, n2= .74). Thus, respondents showed
significantly greater identification with Askale than with Saba or Zeleke. Pairwise
comparisons tested with the least significant difference (LSD) method shows that there
were also significant mean differences between each pair of identification (mean
difference between identification with Askale and identification with Saba = 4.06, SE

= .29, p < .001; mean difference between identification with Askale and identification
with Zeleke = 4.70, SE = .26, p < .001; mean difference between identification with Saba
and identification with Zeleke = .64, SE = .22, p < .01). It is suggested that individuals are
more likely to identify with a protagonist than with a transitional character or a villain. It
Should be noted, however, that 6.8% to 7.6% of the listeners identified with a negative
role model (Zeleke). Thus, the Archie Bunker Effect was found in this study although its
size was small, as is usually observed in many E-E programs. Given that the Archie
Bunker Effect is an audience-centered resistance that diminishes the effectiveness of EB
interventions (Singhal & Rogers, 2004), it should be suggested that firrther research should

investigate the factors that produce the Archie Bunker Effect in order to reduce the effect.

Influences of Engagement with Narrative on Individuals ’Beliefs

Two modified models hypothesized the direct effects of both affective
engagement and cognitive engagement on individuals’ health beliefs. In Modified Model
1, affective engagement was a significant predictor of perceived severity, perceived
benefit, and self-efficacy. These findings suggest that the more individuals affectively

engaged with the narrative of the program, the more they tended to perceive HIV/AIDS

68

as a serious disease, to have a perception that the threat from HIV /AIDS can be
substantially reduced through condom use, and to enhance self-efficacy perceptions of
condom use for preventing HIV /AIDS. On the contrary, cognitive engagement was
directly influenced perceived severity and perceived barriers. In particular, cognitive
engagement substantially diminished individuals’ perceived barriers to HIV /AIDS
preventive behavior. Thus, through their cognitive engagement with the program,
listeners might decrease their perception of the potential negative aspects of condom use
for the prevention of HIV/AHDS. These findings support an important assumption of the
E-E study, which is that the vicarious experiences that stem fi'om exposure to the E-E
program can change the values and beliefs of audience members and, thus, influence
audience members’ behaviors.

It should be noted that perceived self-efficacy was a strong mediator in both of the
models. In Model 1, self-efficacy was a significant mediator of the relationships between
affective engagement and behavioral intention to use condoms and between identification
with character and behavioral intention to use condoms. In Model 2, the effects of the
perceived benefit of and the perceived barriers on behavioral intention were mediated by
perceived self-efficacy. These findings support an important point—regarding the role of
self-efficacy in E-E interventions—that Bandura makes in his social cognitive theory.
According to Bandura (1995), vicarious experience is one of the sources that influence

people’s beliefs concerning their efficacy. Bandura (1995) made the following remark:

[An] influential way of creating and strengthening efficacy beliefs is through the
vicarious experiences provided by social models. Seeing people similar to
themselves succeed by perseverant effort raises observers’ beliefs that they, too,
possess the capabilities to master comparable activities. By the same token,
observing others fail despite high effort lowers observers’ judgments of their own

69

efficacy and undermines their level of motivation. (p. 3)

Given that self-eflicacy is a strong predictor of behavioral change, an implementation
model, as Bandura (2004) suggested, that involve creative translation of social cognitive
theory — specifically, self-efficacy - into practice is strongly recommended for the

successful E-E interventions.

Limitations of the Present Study

There are several potential limitations of the present study. First, it should be
indicated that, owing to the small sample size (n = 124) and large-sample technique (SEM)
in the present study, any interpretation of the results must be met with some caution. Kline
(1998) suggested two criteria by which to determine how much larger a sample is
required in path analysis in order for the results to be reasonably stable. First, sample
sizes less than 100 is considered small, between 100 and 200 is a medium sample size,
and more than 200 can be considered a large sample size. An additional consideration is
the complexity of the path model. In other words, the sample size can be considered the
ratio of the number of cases to the number of parameters in the model to be tested.
Suggesting 20:1 or 10:1 as a realistic ratio, Kline (1998) indicated that if the ratio is less
than 5:1, the statistical stability of the results might be doubtful although there are no
absolute standards in SEM literature. To achieve model parsimony, as noted above, socio-
demographic variables were residualized in the model. In spite of the residualization,
because the ratio of the number of cases to the number of parameters is still less than 5:1,

one should be cautious in interpreting the results of these analyses. Therefore, firture

70

replication using more substantial sample sizes should be required.

A second limitation centers on the measurement issues. Owing to the limitations
arise from the use of secondary data, most variables in this study employed a single-item
measure or a two-item measure. When considering the fact that the multi-item measure is
more reliable and valid than a single-item or a two-item measure (Snijders & Bosker,
1999), the weaknesses of the measure’s reliability and validity should be indicated as one
of the limitations of this study. Therefore, it can be suspected that the use of a single—item
for measuring perceived susceptibility might be a reason why the current study failed to
find its effect in the model. In addition, a considerable measurement issue concerns the
measurement of identification, which is one of the key variables in this study. As
mentioned above, this study employed only two items by which to measure identification.
The main reason for this was to distinguish identification from other types of reactions—
such as liking, similarity, and parasocial interaction—that audience members might have
to media characters. Given that identification with characters is a central mechanism by
which to explain the effectiveness of EB, future research on this particular topic should
include replication that benefits from a theoretically elaborated measure of identification
that, itself, enables researchers to examine with greater rigor, specificity, and validity the
role of identification in E-E interventions.

A third limitation concerns the fact that the current study did not consider a
hierarchy of effects (or stages of change) in E-E intervention. According to Slater (2002),
the cross-theoretical approach is useful because it provides researchers with an integrated
model that explains the persuasive influences of E-E. For example, the transtheoretical

model or the stages-of-change model (Prochaska & DiClemente, 1983; Prochaska,

71

DiClemente, & Norcross, 1992) provides a useful theoretical framework in which to
examine the hierarchy of effects in E-E interventions. The main concept underlying this
model is that different kinds of interventions based on different theoretical models are
appropriate, depending on the readiness with which individuals change their behavior
(Slater, 2002). According to Prochaska and others (1992), changes in behavior go
through five distinct stages: precontemplation, contemplation, preparation, action, and

maintenance.

Precontemplation is the stage at which there is no intention to change behavior in
the foreseeable future. Many individuals in this stage are unaware or underaware
of their problems... Contemplation is the stage in which people are aware that a
problem exists and are seriously thinking about overcoming it but have not yet
made a commitment to take action... Preparation is a stage that combines
intention and behavioral criteria. Individuals in this stage are intending to take
action in the next month and have unsuccessfully taken action in the past year...
Action is the stage in which individuals modify their behavior, experiences, or
environment in order to overcome their problems... Maintenance is the stage in
which people work to prevent relapse and consolidate the gains attained during
action. (Prochaska et a1., 1992, pp. 1103 - 1104)

For instance, exposure to E-E messages, including the processes of engagement and
identification, might influence the stages of precontemplation and contemplation, whereas
self-efficacy will affect the stage of action. Given that the stages models play an important
role for E-E practitioners in both defining where intended audiences are in the process of
change and designing messages to fit specific audience and cultural needs (Sood et a1.,

2004), further research should investigate the effects of E-E across the stages of change.

72

Sugestions for Future Directions of E-E : Issue Placement as an Alternative Strategy of
E-E

The final section features a discussion of “issue placement” in terms of future
directions of EB strategy. In fact, although most E-E research has been conducted in
developing countries of Latin America, Afiica, and Asia, such as India, Turkey, Peru,
Mexico, Tanzania, and Nigeria, there has been little evidence that E-E can be effective in
an industrialized nation like the United States. In highly saturated media environments
such as those common to the US, the effectiveness of E—E can be attenuated by various
other media contents (Sherry, 2002; Singhal & Rogers, 2002). For example, in highly
saturated media environments, because audiences receive “a barrage of competing
messages on many health issues” (i.e., anti-obesity health message vs. junk food
advertising), the effectiveness of campaign messages may be decreased (Greenberg,
Salmon, Patel, Beck, & Cole, 2004, p.199). Furthermore, in a country like the US,
which has a commercial media system, it is not easy to produce and disseminate whole
programs or shows that feature prosocial messages scripted by government or nonprofit
organizations. In this regard, as Greenberg and his colleagues (2004) indicated, the scope,
function, and form of E-E in the US. should be quite different from its counterpart in
developing countries. As a result, government and nonprofit organizations have made an
effort to embed educational messages in entertainment programming. For example, US.
sexual health agencies have been working with the commercial media to insert health
messages into existing entertainment programming (Keller & Brown, 2002). The
definition of the term issue placement can be derived from this activity. Thus, I define

issue placement as “the paid and intentional inclusion of a message or an issue in

73

entertainment programming by an individual or an organization that attempts to affect
individual and social change.”

In the area of advertising research, product placement has been considered a
popular advertising strategy that inserts branded products or brand identifiers into a
movie scene or television Show. Therefore, it is believed that a thoughtful discussion of
the major similarities and differences that characterize product placement in relation to
issue placement will provide a proper framework for developing and testing propositions
about issue placement.

Product placement, often called brand placement, is a growing marketing practice
(Balasubramanian, 1994; Gupta & Lord, 1998; Karrh, 1998; Snyder, 1992). Product
placement is defined as “the paid inclusion of branded products or brand identifiers,
through audio and/or visual means, within mass media programming” (Karrh, 1998).
According to Balasubrarnanian (1994), product placement is one of many hybrid message
forms, a growing genre of marketing communication that combines advertising and
publicity messages. Currently, it mostly occurs in movies and TV shows, but the usage
has been expanded to other media such as video games, Broadway shows, music videos,
novels, and so on (Balasubramanian, 1994; Gupta & Gould, 1997; Gupta & Lord, 1998;
Snyder, 1992).

Several empirical findings regarding the effectiveness of product placement
suggest that product placement in movies can increase the awareness of the placed brand
among audiences (Babin & Carder, 1996). Examining viewers’ recognition of brands
placed within a film, Babin and Carder (1996) found that viewers could recognize brands

placed in the film that they watched and could distinguish them from brands not placed in

74

the same film. Another contribution of their study is the fact that audiences will recognize
more placed brands than they will actually recall, suggesting that recognition is easier to
achieve than recall. Karrh (1998) found, for example, that identification of product
placements in a movie could enhance viewers’ recall and recognition, making the
represented brand more memorable in consumers’ minds. Furthermore, the findings of
product placement research has showed that product placement can increase short-term
purchase intention for the represented brands in movies (Karrh, 1998).

These findings from product placement literature can also be applied to
explanations of issue-placement effectiveness. For example, one can examine the
hypothesis that an issue placed in TV programs will increase the audience’s awareness of
the placed issue. Given that the awareness of the placed issue is a starting point from
which to change audiences’ attitudes toward the issue and their behaviors related to the
issue, then audience awareness of a specific issue is important for determining the
effectiveness of the EB intervention. In particular, this is the case in a competitive media
environment in which a variety of media messages compete because the message strategy
might be attenuated by competing messages from other sources. Types of exposure can
affect the effectiveness of placement: visual only, audio only, or a visual-audio
combination. Gupta and Lord (1998) found that a prominent placement generates greater
effectiveness than a subtle placement. With regard to the types of exposure, their findings
showed that a verbal only exposure results in better recall than a subtle visual only
exposure. Another study (Law & Braun, 2000) found that while the visual-audio
combination is best remembered, recognition is highest when using audio only placement.

The visual only type results in the least recall, but is the best choice. Prominence of

75

placement, which means a highly visible exposure by virtue of size and/or position on the
screen or its centrality to the action in the scene, also affects audiences’ responses in
terms of the effectiveness of product placement (Gupta & Lord, 1998). When products
are placed more prominently, viewers will find it easier to recognize the brands (Babin &
Carder, 1996). Prominent placement can also enhance the placed brand’s recognition,
recall, and memorability compared to subtle placement (Brennan, Dubas, & Babin, 1999;
Gupta & Lord, 1998; Law & Braun, 2000).

These findings from product placement research provide us with some empirically
grounded ideas that go far in explaining the issues raised above: namely, the effectiveness
of EB intervention. Touching on this concern, Greenberg and his colleagues (2004) have
suggested the importance of several research agenda related to message components: the
diet issue (e. g., what portion of each E is necessary and beneficial?)7, the amount issue
(e. g., the size or the length of the issue placed in a program), the order of presentation
issue (e. g., locating the strongest position of the entertainment component), the repetition
issue (e.g., how often the message can be repeated before tedium or even revulsion sets
in), the spread issue (e. g., is placement in a single scene or across scenes in a given
episode preferable?), genre comparison (e. g., sitcom vs. adventure), mode comparison
(e. g., audio vs. written), format comparison (e. g, text vs. graphics), and source

comparison (e.g., relative credibility and attractiveness).

 

7 The first “E” refers to entertainment, and second “E” refers to educational message.

76

Appendix A. Synopsis of Journey of Life

 

Askale, a female police investigator, and Bahiru, her husband, are an affectionate
married couple who actively educated and taught others about health topics including
HIV/AIDS and family planning. They are constantly trying, and eventually succeed, in
persuading their mother, the interfering and vocal Amelwork, that child -spacing is the best way
to raise a healthy and economically stable family. Saba is Askale‘s carefree sister who goes
out drinking and socializing with many men. She does not heed her sister's warnings to either
abstain from sex or use protection and Saba winds up discovering that she is HIV positive and
that she has infected others. Tedjie, their servant, frequently asks questions on health related
issues, adopts the use of condoms for himself, and develops healthy beliefs regarding the
benefits of child spacing and the birth control pill.

Elias, who is a judge, and his wife, Azeb, live next door to Askale and Bahiru. Elias and
Azeb are initially very happy in their marriage, but Azeb causes conflict, eventually destroying
the marriage. Azeb had a son that she had to give up for adoption because she had been
raped by an employer and needed to be childless to get a new job. She shares her concern
with Askale and they plot to investigate the crime, leaving Elias in the dark.

Meanwhile, another couple, Zeleke and Fikirte, agree to take in Azeb‘s son and find him a
family to live with. But they soon decide to raise him themselves. Zeleke plans to give Azeb‘s
son to the shady character, Alemu, who cripples children and forces them to beg on the
streets, extracting some of their money for profit. Fikirte catches wind of the plan and keeps
the boy from his life .In order to do so, she eventually turns her husband into Askale, and
Alemu and he are sent to jail. Unfortunately, the marriage of Azeb and Elias suffers, since he
is suspicious and she will not reveal what has been going on. Elias succumbs to Saba‘s
advances, contracts HIV, and infects his wife (Azeb). They both die. Saba decides to better
her life and teach others to avoid her path. lntermingled in these various plots are discussions
of different health-related topics like the problems of overpopulation, unwanted pregnancy,
different family planning methods, ways of HIV transmission, prevention of HIV/AIDS, crimes
committed against children, and so on, where the character Doctor Hailu is frequently

consulted for an expert opinion.

 

Source: Belete, Girgre, & Witte (2003)

77

Appendix B. Description of the Main Characters of Journey of Life

 

Character

Alemu

’ ArneIWork

' Askale A

N, 09891990" I. ML .. .__ W-
The VIIIaIn of the program who cripples children so they work as
street beggars for his profit. M _
She represents the "doubter"—the person who Is not
immediately willing to accept the new ideas about family
planning. She gradually learns that family planning is good and
helps herself and the country. She represents the large number

of older listeners who similarly react with doubt to advocates of

family planning.

The role model With regard to famIly planning and the person

. “who Is trying to be most helpful In Azeb s trouble.

”Azeb

Banird

‘1 Enos

"Fikirte A

' saoa ”

‘ Zeleke

learn from them.

The heroin who suffers the most and from Whom the audience I

will learn the most.

The male role model and husband of Askale These two (Askale M
and Bahiru) represent the family who decided, early on, to do

i “the right thing In terms of family size, among other things .
He Is a good man Who makes one foolish mistake. He was

created to be representative of people who (in anger or out of
hurt) fall into unsafe behaviors. The designers created him to
generate both sympathy and a realization of how easily one can
make a fatal mistake.

Zelekes wife Who suffers from Zeleke s mistakes despite her’

Innocence.

She is the young Woman Who is too eaSily "led astray' for the H

wrong reasons. The designers created her as a model from
which the audience can grasp the foolishness of her ways and

The Villain Who at firSt has "hoodwmked" his Wife, Fikirte, into
thinking he is a perfectly good man. He stands as a “warning"
that people are not always what they seem.

 

Source: Belete, Girgre, & Witte (2003)

78

Appendix C. Respondents’ Demographic Characteristics (N = 124)

 

 

Variable % or Mean (SD)
Sex
Male 22.6
Female 77.4
Age 22.04 (4.32)
Education
Literate, no schooling 1.6
1-4 years 1.6
5-8 years 21.0
9-12 years 62.9
Over 12 years or Diploma 11.3
Marital Status
Single 68.5
Living with partner 6.5
Married 18.5
Widowed .8
Separated .8
Divorced 4.8
Number of Sexual Partners during 12 months
0 53.2
1 45.2
2 .8
3 .8

 

79

Appendix D. Modification Indices for BETA

Modification Indices for BETA

ASKALE EFFICACY SEVERE SUSCEPT BARRIER
ASKALE - - 0.70 1 12 0 35 15.71
EFFICACY - - - - - - - - - -
SEVERE - - - - - - - - - -
SUSCEPT - - - - - - - - - -
BARRIER - - - - - - - - - -
BENEFIT - - - - - - - - - -
INTENTION 3.32 - - - - - - - -
INTENTION
ASKALE 4.87
EFFICACY 3.32
SEVERE 3.32
SUSCEPT 3 32
BARRIER 3.32
BENEFIT 3 32
INTENTION - -

80

BENEFIT

Appendix E. Mean Differences among Identifications with Three Main Characters

 

ASKALE SABA ZELEKE

 

M SD M SD M SD
7.80 2.00 3.74 2.35 3.10 1.92D

 

2
Wilks' Lambda = .26, F(2, 117) = 165.59, p < .001, n = .74.

81

Appendix F. Family Planning & HIV/AIDS Communication Survey, 2002

Johns Hopkins University/PCS and Development Studies Associates

Study Subjects: Males and Females — age 17 to 32 years at 2nd panel (15-30 at original

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

panel)
IDENTIFICATION #
Region Zone Region
Woreda Town
Kebele House Number
Respondent's Name (fill this
as you finish interviewing) CID
Male Female
Name of Household Head
Male _ Female
INTERVIEWER VISITS
1 2 3 FINAL VISIT

Date Montl|:|:|
lnterviewer’s Year
Name [:[3
Result"
*Result
Codes:
1= Completed 4: Refused 7= Dwelling not Interviewer
2: Not at 5: Partly found Result

home completed 8 = Dewelling [:13
3= Postponed 6= Incapacitated closed

9 = Others
(Specify)

 

 

 

 

 

 

 

82

 

 

LANGUAGE

 

 

 

 

 

 

 

 

 

 

 

 

 

Questionnaire Interview Respondent's Native Translator
Language Used During
Interview
Amharic = 01 Amharigna = 01 Amharigna = 01 Yes = 01 ED
Others = 02 Oromigna = 02 Oromigna = 02 N0 = 02
CE] Tigrigna = 03 ED Tigrigna = 03 D:
Sidamigna = 04 Sidamigna = 04
Wolaitigna = 05 Wolaitigna = 05
Guragigna = 06 Guragigna = 06
Others = 07 Others = 07
(specify) ------—- (Specify) -------
Supervisor Field Editor Keyed By
Name Name Name
Date Date Date

 

 

 

 

 

 

83

 

 

SECTION 1: DEMOGRAPHIC/RESPONDENT‘S BACKGROUND

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

No. Questions and Filters Coding Categories Skip
How old were you at your last Age in completed year
101 birthday?
In what month and year were you Month
102 born? _ _ _ _ Year
Don’t Know DK
Have you ever been married before?
103 g: fig ->
109
104 If yes, age at which first married: Years
What is your marital status now?
1: Single
2= Living with partner
3= Married
105 4= Widowed
5= Separated
6= Divorced
7: Other
105 In your opinion, what is the right age
a of marriage of boys & girls? BOYS
106 GIRLS
b
What is your daily work, if anything? 1: Wage labor
2: Work on own farm/cattle
3= Salaried job
107 4: Service job (maid, taxi)
5= Self-employed (tailor, etc.)
6: Household wor1<
7= Student
8: Others
In the house that you live (household
108 level), 1= yes 2: no
a Do you own a radio? 1= yes 2= no
~ Do you own a television? 1: yes 2: no
108 Do you own a car? 1= yes 2: no
d Do you have a cook? 1: yes 2: no
108 Do you have a gardener?
e

 

 

 

 

 

84

 

 

109

Educational Level

 

1= Illiterate
2= Literate, no schooling

 

 

 

3=1-4

4:5-8

5: 9 - 12

6= 12 + 8. Diploma
7: Degree

 

110

Can you read a letter or newspaper
easily?

1: No at all

 

 

2: With Difficulty

 

 

3= Average ability
4: Easily
5: Very Fluently

 

111

To which religion do you belong?

 

1= Orthodox

 

 

 

2= Muslim

3= Traditional

4: Protestant

5= Catholic

6= Seventh Day Adventist
7: Other:

 

112

How much influence would you say that
your religion has on the things you do?

1: No influence at all
2= Little influence

3: Average influence
4: Great influence
5: Guides everything I do
(total influence)

 

 

 

 

 

113

 

What is your ethnicity?
RECORD THE MAJOR ETHNIC
GROUP

 

01 = Amhara

02= Oromo

03= Tigre ED
04: Gurage

05= Sidama

06= Wolaita

07: Other (specify)

 

 

The next few questions are about alcohol and other drugs. When we mention alcohol, we

mean beer, wine, tej, wine coolers, or liquor—anything with alcohol in it.

-your answers are confidential.

Please be honest-

 

 

113
a

 

Have you ever had any alcohol to
drink, other than for religious
reasons?

 

 

1: Yes
2: No

 

 

 

 

no,
ski

p to
113]

 

85

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

113 Over your entire life, about how many 01= 0
b different days have you had alcohol to 02= 1-2 times ——
drink? 03= 3-5 times —--*
04: 6-9 times
05= 10-20 times
06= 21-40 times
07: More than 40 times
113 Over the past three months, about
c how many different days have you days
had alcohol to drink (TOTAL)?
113 The last time that you drank alcohol,
d how many drinks did you have? drinks
113 How old were you when you had your
e first alcoholic drink? years old
No Yes Yes Yes,
Have you ever: , .a a
One Fe Lot
e w
Tim
es
113f Gotten into a fight during or shortly 1 2 3 4
after drinking?
1139 Done things when drinking that you 1 2 3 4
normally would not do, and you later
regretted doing?
113h AftenNards, been unable to 1 2 3 4
remember things that happened
while you were drinking?
113i Been hassled by your friends or 1 2 3 4
family because of your drinking?
113 j Now let's talk about Chat. 1: Yes If no skip to
Over your entire life have you ever 2= No 201
used chat?
113k If yes. How frequently? 01= Daily
02= Every other day
03= Once a week
04= Every two weeks
05= Every month
06= Others (specify)
113 l After I use chat my sexual desire is 1= Yes [:3
aroused 2: No
113 After I use chat my sexual desire is 1: Yes [:3
m reduced 2: No
113 Chat has no influence on my sex life 1= Yes [:1
n 2= No

 

 

 

 

86

 

SECTION II.

RADIO LISTENERSHIP

 

Do you have the habit of listening to

1= Don’t have a working radio

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

called “Berhane?”

 

201 radio? 2: Don’t listen to the radio
(If yes ask how often they listen) 3: Listen at least once a
month
4: Listen at least once a
week
5: Listen several times a
week
6: Listen daily
202 What types of radio programs do you 1= News
a-h listen to? 2: Personal paid
(DO NOT READ LIST, CIRCLE ALL announcements
MENTIONED) 3= Commercials
4= Sports
5= Drama/soap opera
6= Religious Programs
7: Health Programs
8: Other, specify
203 What are the names of specific radio 1: Journey of Life
programs you listen to?
(DO NOT READ LIST, CIRCLE ALL 2: Ask Your Doctor (DKT
MENTIONED program)
3=Other
204 IF DON’T MENTION JOURNEY OF 1: Yes, heard of it but not
LIFE, Have you ever heard of a radio _
serial called “Journey of Life"? listened listened
to it? 2: Yes, listened to it
3: No
205 If yes listened to it, about how many of
the 26 episodes did you listen to?
206 IF DON’T MENTION Ask Your Doctor, 1: Yes, heard of it but not
Have you ever heard of a radio listen
program called “Ask Your Doctor”? , .
Listened to it? 2= Yes, listened to it
3: No
207 If yes listened to it, about how many of
the episodes have you listened to in
the last 6 months (out of 26 possible)?
208 Have you ever heard of a radio serial

1: Yes, heard of it but not
listen

2: Yes, listened to it
3: No

 

87

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Now I'm going to give several statements about your
opinions on the radio soap JOURNEY OF LIFE. I'm
interested in the extent to which you agree or disagree
with the following statements. Each of your answers will
be on this scale (show ruler) ranging from "Strongly
Disagree" to "Strongly Agree." You can simply point to or
say your answer in response to each statement.

*There are no right or wrong answers; these are simply
your opinions. Your answers will be kept confidential.
We appreciate your honesty.

If respondent has never listened to Journey of Life go to
301.

PRACTICE ITEM:

For example, in response to
the question,

"Doro wat is good".

I might point here, "Strongly
Agree"

How would you respond to the
statement? "Doro wat is
good." (MAKE SURE
PARTICIPANT GIVES A
VALID ANSWER.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

g i. i. g s ‘5 g
s .3 g 2‘ 3’ g <
m a D it)
209 Journey of Life influenced me to protect 1 2 3 4 5
myself against HIV/AIDS.
210 Journey of Life made me believe 1 2 3 4 5
HIV/AIDS is a serious health threat.
211 Journey of Life made me believe that it
is possible that I myself might be 1 2 3 4 5
susceptible to getting infected with
HIV/AIDS
212 Journey of Life made me feel hopeless
about fighting against AIDS 1 2 3 4 5
213 Journey of Life made me feel like there
was nothing I could do to avoid getting 1 2 3 4 5
HIV/AIDS
214 Journey of Life made me feel scared 1 2 3 4 5
about AIDS
215 Journey of Life influenced me to use
. . 1 2 3 4 5
family planning.
216 Journey of Life made me believe that
having children when I wasn’t ready 1 2 3 4 5
would be harmful.
217 Journey of Life made me believe that I
might have children when I didn’t want 1 2 3 4 5

them.

 

 

 

 

 

 

 

 

218 Journey of Life made me believe that
family planning worked in preventing
me from having children unless I really
wanted them.

 

 

 

 

 

 

 

 

 

 

 

 

88

 

 

219

Journey of Life made me believe that I
was able to use family planning

 

220

Journey of Life made me feel emotional.

 

221

Journey of Life had a strong emotional
impact on me.

 

222

Journey of Life was an emotional
program.

 

223

Journey of Life made me feel hopeful
about avoiding HIV/AIDS infection.

 

224

Journey of Life made me believe I could
personally do something to avoid
getting infected with HIV/AIDS.

 

225

Journey of Life made me believe there
were effective ways to avoid HIV/AIDS
infection.

 

226

Journey of Life was entertaining.

 

227

Journey of Life was boring.

 

228

I wish Journey of Life had lasted longer.

 

229

The quality of the Journey of Life story
was very good.

 

230

The quality of the production of Journey
of Life was very good (e.g., sound,
special effects).

 

231

Journey of Life was educational.

 

231

Journey of Life kept me in suspense
from episode to episode.

 

232

I learned a lot about HIV/AIDS from
Journey of Life.

 

233

I learned a lot about family planning
from Journey of Life.

 

234

Journey of Life provided a lot of good
information about HIV/AIDS.

 

235

Journey of Life provided a lot of good
information about family planning.

 

 

236

 

The characters on Journey of Life were
realistic Ethiopians.

 

89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

237 l have talked with my friends about
Journey of Life. 1 2
238 l have talked with my family about
. 1 2
Journey of Life.
239 I have talked with my
boyfriend/girlfriend/husband/wife about 1 2
Journey of Life.
240 After listening to Journey of Life, I 1 2
wanted to change my life for the better.
241 Please tell us who is your favorite
Journey of Life character. 1: Askale
2= Bahiru
3=Saba
4= Amelework
5= Elias
6= Azeb
7= Zeleke
242 Why is this character your favorite?
243 Please tell us who is your LEAST 1: Askale
favorite Journey of Life character. 2= Bahiru
3= Saba
4= Amelework
5= Elias
6= Azeb
7= Zeleke
244 Why is this character your LEAST

 

favorite?

 

 

 

 

 

90

 

 

 

Please tell us what you think
about each of these characters
- all answers should be on the
following scale:

 

 

 

 

 

 

 

 

‘ai
E 8
g n:'
E ii‘
a ., 3
Use this code: = 1: fi 9::
1 Strongly Disagree o:' m '5 m- I
ZDisa ree E I: 2": ¥0
9 2 i2 8 SE 3 3%
3Neutral m h E < 5 m 3
4Agree S: E 0 LE) 3 o o
SStroneg Agree (2'5 3 j g .2 €- : E
— 2
S2 é g I in .3 3';
3.— 039 Q.- 3 r
245 Askale 12345 12345 12345 12345
a-d
246 Bahiru 12345 12345 12345 12345
a-d
247 Saba 12345 12345 12345 12345
a-d
248 Amelework 12345 12345 12345 12345
a-d
249 Elias 12345 12345 12345 12345
a-d
250 Azeb 12345 12345 12345 12345
a-d
251 Zeleke 12345 12345 12345 12345
a-d

 

 

 

 

 

 

91

 

SECTION 3: SEXUAL HEALTH & HIV/AIDS

 

Now I’m going to ask some questions about sex and HIV/AIDS. It is important for us to get
complete answers to these questions because they will help us devise appropriate programs
for people in Ethiopia. Again, no one will ever see your answers and we really appreciate
your honesty.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Have you ever had sexual intercourse? 1: Yes
301 2: No
302 How old were you when you first had sexual Years
intercourse?
303 About how many sexual partners have you Partners
a had over your whole life?
303 During the last 12 months (the last year)? Partners
b
On average, how frequently do you have
sexual intercourse?
304 .
(e.g., a few times a year, once a year, once a
month, etc.)
On average, how frequently does the
AVERAGE ETHIOPIAN Male/Female
305 ,
[matched to respondents sex] have sexual
intercourse?
Do you have a regular sexual partner? 1: yes
306
2: no
Who is this person? 1: spouse
2: fiance
3= girlfriend/boyfriend
307 4: acquaintance
5= paid sex worker
6 other
The last time you had sex did you use a 1= Yes
308 contraceptive? 2: No

3: Can't remember
4: Don't know

 

92

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

309 What did you use? 01= pi"
02= Injections
03= IUD
04= Implants/Norplant
05= Condoms
06= Foam/jelly
07= Female sterilization
08= Male sterilization
09= Periodic abstinence
10 Withdrawal
11 Other
310 Under perfect circumstances, how frequently
would you LIKE to have sexual intercourse?
311 When comparing yourself to other Ethiopian 1: much lower than others just
[men/women] just like yourself, how does like me
your level of sexual desire compare? 2= lower than others just like me
3= about the same as others just
like me
4: higher than others just like me
5= much higher than others just
like me
3‘ 3.3 E is 2‘
a '5 a é’ 2’ a 5:”
312 I find it difficult to talk about sexual 1 2 3 4 5
issues with my parents
313 I find it difficult to talk about sexual 1 2 3 4 5
issues with my friends
314 I find it difficult to talk about sexual 1 2 3 4 5
issues with my romantic friend
315 My best friend tries to convince me to 1 2 3 4 5
be more sexually active
315 How often do you think about sex each day 1: Never
2: Once
3= Twice

 

 

4= Three times
5= Frequently

 

93

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

317 From where do you get most of your information 1: School
about HIV/AIDS 2: Friends
(DO NOT READ LIST, CIRCLE ALL _
MENTIONED) 3- Parents
4: Other relatives
5: Teachers
6: Others
313 Do you do anything to protect yourself against 1: Yes, always
HIV/AIDS infection? 2: Yes, sometimes
3: No, don’t need to
4: No, I don’t think about it
319 If yes, what have you done to protect yourself?
a-f 1= Become abstinent
For each spontaneous answer, follow up by _
asking, “How often do you use that strategy to 2" Use condoms
protect yourself against HIV/AIDS infection?” 3: Limit partners
1 Use all of the time 4: Get tested for HIV
2 Use most of the time _
3 Use some of the time 5- Screen partner carefully
4 Use rarely 5:
5 Use never Other
320a Sometimes people put themselves at risk of 1: never
getting HIV/AIDS by having sex with someone 2: once
they don’t know well, without using a condom. 3: 2-3 times
How frequently, if at all, have you ever done 4: 4-5 times
anything that might put you at risk for getting 5: 6-10 times
HIV/AIDS. 6= more than 10 times
We appreciate your honesty here. The
responses are “never,” “once," “2-3 times,” “4-5
times,” “6-10times,” “more than 10 times”
320b Do you have the virus that causes AIDS? 1: definitely no
2= may be no
3: not sure
4: maybe yes
5= definitely yes
321 Have you ever been tested for HIV/AIDS? 1= Yes
2= No
322 Have you ever talked with other people about 1= No never
HIV/AIDS? - 2: Yes, once
3= Yes, a few times
4: Yes, a lot

 

 

5= Yes, all of the time

 

94

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

323 Do you know where you could go to get an 1Yes,
AIDS test? If yes, where?
2 No
Please indicate the extent of your 11 m
agreement or disagreement with the 7: 9 9.’ a 7:
. i: 5’ 0’ 3:: 8 c m
followmg statements. 9 g g 1:, ‘6) 9 “a”
0'5 '5 '5 z < 6 <
324a HIV/AIDS test clinics in Ethiopia give 1 2 3 4 5
accurate results.
324b HIV/AIDS test clinics in Ethiopia provide 1 2 3 4 5
high quality services.
324c It is difficult to go to an HIV/AIDS test 1 2 3 4 5
clinic in Ethiopia.
324d I don’t trust HIV/AIDS test clinics in 1 2 3 4 5
Ethiopia.
325 By your own estimate, would you say that in 1. NOBODY HAS HIV/AIDS
your community or in your neighborhood: 2. ONLY FEW PEOPLE HAVE
HIV/AIDS
3. MANY PEOPLE HAVE
HIV/AIDS
4. OTHER
(SPECIFY)
8. DON'T KNOW/NOT SURE
326 If you had to describe the way people with 1. TREATED LIKE
AIDS are treated in your community would EVERYBODY ELSE
say they are: 2. TREATED WITH
SYMPATHY
3. ISOLATED FROM
OTHERS
4. MISTREATED BY
OTHERS
5. OTHER (SPECIFY)
8. DON'T KNOW/NOT SURE
327 Do you have a relative who has AIDS or the 1= definitely no

 

virus that causes AIDS?

 

2= may be no
3= not sure

4: maybe yes
5= definitely yes

 

95

 

 

 

 

 

 

 

328 If yes, if you had to describe the way 1. TREATED LIKE OTHER
your family and friends treat this person with FAMILY MEMBERS

AIDS would you say that he or she is: 2‘ TREATED WITH MORE

 

SYMPATHY

 

 

 

3. ISOLATED FROM FAMILY
MEMBERS

4. MISTREATED BY FAMILY
MEMBERS

5. OTHER
(SPECIFY)

8. DON'T KNOW/NOT SURE

 

 

329 Regarding medical treatment, do you think 1. LESS THAN

 

that the amount of medical treatment for 2. SAME AS

 

 

 

people with AIDS should be less than, the

same as or more than those with other 3' MORE THAN
illnesses? 4. DON'T KNOW/NOT SURE

 

 

330 If a person knows that he or she has AIDS or 1. SHOULD KEEP IT
the virus that causes AIDS do you think that pR|VATE

 

 

 

he/she'should keep it private or tell other 2. SHOULD TELL OTHERS
people in the community?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3. OTHER
(SPECIFY)
8. DON'T KNOW/NOT SURE
Would you be willing to personally take care No Yes DK
of a family member if he or she:
331a Had suffered an accident? 0 1 3
3315 Were sick with malaria? o 1 3
331s Were with tuberculosis? o 1 3
331d Were sick with AIDS? 0 1 3

 

 

 

 

 

 

 

 

 

Now I'm going to ask you several questions about your perceptions of AIDS and what
you think about ways to prevent AIDS. Each of your answers will be on this scale ranging from
"Strongly Disagree (1 )" to "Strongly Agree(2)".

* Remember, there are no right or wrong answers. Just give your honest opinions.

 

 

332 I am at risk for getting HIV/AIDS 1 2 3 4 5

 

SUSC

 

 

333 Getting AIDS is the worst thing that could 1 2 3 4 5

 

SEV happen to me.

 

 

334 Getting AIDS is a sure death sentence. 1 2 3 4 5

 

SEV

 

 

 

 

 

 

 

 

 

 

 

96

 

 

** Use “would be" if participant has never had sex.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

335 I am (would be) able to use condoms to
SE prevent HIV/AIDS infection.
336 I am able to be abstinent to prevent HIV/AIDS
SE infection.
337 I am (would be) able to be monogamous to
SE prevent HIV/AIDS infection.
338 I am able to talk to my partner about
SE HIV/AIDS prevention methods to prevent
infection
339 Using condoms to prevent HIV/AIDS is (would
SE be) easy for me.
340 Condoms are effective in preventing
RE HIV/AIDS infection.
341 Abstinence is effective in preventing
RE HIV/AIDS infection.
342 Being monogamous is effective in preventing
RE HIV/AIDS infection.
343 My using condoms to prevent HIV/AIDS
ATI' infection is (would be) good.
(USE 'WOULD BE" IF "Never" TO 2M1)
344 My being abstinent to prevent HIV/AIDS
ATT infection is (would be) good.
345 My being exclusively monogamous to prevent
ATT HIV/AIDS infection is (would be) good.
346 I try not to think about getting infected with
DA HIV/AIDS when having sex.
347 I intend to use condoms to prevent HIV/AIDS
INT infection during the next 12 months.
348 I intend to be abstinent to prevent HIV/AIDS
INT infection during the next 12 months.
349 I intend to be exclusively monogamous to
INT prevent HIV/AIDS infection during the next 12
months.
350 It is better not to talk about HIV or AIDS in my
BAR community.
351 HIV/AIDS prevention methods cost too much.
BAR
352 HIV/AIDS prevention methods are
BAR inconvenient.
353 Current campaigns on HIV/AIDS prevention
React are misleading.
354 My culture is against using condoms to
SN prevent HIV/AIDS

 

 

 

97

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

355
SN

My religion is opposed to using condoms to
prevent HIV/AIDS.

 

356
SN

My friends use condoms to prevent HIV/AIDS.

 

357
SN

My siblings use condoms to prevent
HIV/AIDS.

 

 

358
SN

 

Most Ethiopians just like me use condoms to
prevent HIV/AIDS.

 

98

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SECTION 4: CHILDREN AND FAMILY PLANNING

 

Now I would like to switch topics and ask about children

 

 

 

 

 

 

 

 

 

 

 

 

 

401 In your opinion what is the ideal age for a Years
woman to have her first child?
402 a If you have a child or children how old were Years
you at your first delivery?
402 b In terms of your biological children (children # of Sons Born
that you personally have had) have you
ever had sons? Daughters? IF NONE, # of Daughters Born
RECORD ‘00'
Total # of Children
403 How many of your sons or daughters are #of Sons Died
now not living (died)? IF NONE, RECORD
‘00’ # of Daughters Died
404 When YOUR LAST CHILD was conceived, Wanted then = 01
did you want to have a child m, did you —’ 141a
want to have a child but wanted to wait until Wanted later = 02
l_at_gr, or did you not want to have any Did not want at all = 03
(more) children at all?
405 How much longer would you like to have Months =
waited? Years = (convert
into months)
Undecided/don’t know 99
406
a- HAS LIVING CHILDREN Ideal Number of boys / girls
boys If you could go back to the time you did not
have any children and could choose exactly
b-girl the number of children to have in your
whole life, how many would that be?
MARK 00 IF NONE.
407

 

NO CHILDREN

If you could choose exactly the number of
children to have in your whole life, how
many would that be?

MARK 00 IF NONE.

 

Ideal Number of boys / girls

 

 

99

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

403 After how many births should a couple 1: before any birth
begin using a family planning method? 2: one birth
3= two births
4= three births
5= four or more births
6: after a son
7: after two sons
8= After a daughter
9: other
409 What should be the ideal age gap between 1: 1 year
two children be? 2: 1 ‘/2 years
3: 2 years
4: 2 V2 years
5: 3 years or more
6: Up to God's will
For the next few questions I'd like you to state your extent of agreement or
disagreement, as you did before, with 1 representing strongly disagree, 2
representing disagree, 3 representing neutral, 4 representing agree, and 5
representing strongly agree.
410 It is possible that I will HAVE MORE 1 2 3 4
SUSC CHILDREN THAN I REALLY WANT.
411a IF HAS NO CHILDREN: It is possible 1 2 3 4
SUSC that I will HAVE CHILDREN AT AN
EARLIER AGE THAN I REALLY WANT.
411b IF HAS CHILDREN: I had children at an 1 2 3 4
EARLIER AGE THAN I REALLY
WANTED.
412 HAVING MORE CHILDREN THAN I 1 2 3 4
SEV REALLY WANT leads to problems.
413 HAVING MORE CHLDREN THAN I WANT 1 2 3 4
SEV is harmful to my future.
414 HAVING CHLDREN AT AN EARLIER AGE 1 2 3 4
THAN I WANT (ED) is (WAS) harmful to
SEV
my future.
415 From where do you get most of your 1: School
information about Family Planning 2: Friends
3= Parents
4= Other Relatives
5: Teachers
6: Radio
7= Others

 

 

 

100

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Now I’m interested in how your personal knowledge and/or awareness of different ways to

prevent pregnancy.
prevention methods.

I'm going to ask you a series of questions about various pregnancy
(Please use the following codes to fill in this table.)

 

 

 

 

 

 

 

 

 

 

 

 

 

Awareness: Knowledge: Ever use: Current Use: Plan
Have you ever Ask them to Have you ever Do you currently (Intention):
heard of? completely describe used X method? use? Do you plan to
what it is and how use?
used and then code
(train coders)
1: Yes 1: Proper complete 1: Yes 1: Yes 1: Never
knowledge
2: No 2: Proper incomplete 2: No 2: No 2: May be
knowledge
3= Improper 3: Probably
knowledge
4: No knowledge 4: Yes, Definitely
No. Methods of Family Awareness Knowledge Whether Ever Plan to
Planning adopted by use
respondent (or
spouse)
PERMANENT
416 1) Tubal ligation
a-d
417a 2) Vasectomy
-d
MODERN Ever Current
418a 1) Female pills
-e Ask specifically for social
marketing brand
419a 2) Norplant
.9
420a 3) Injection
.9
421a 4) Condoms
-e Ask specifically for social
marketing brand
422a TRADITIONAL
-e Breast feeding

 

 

 

 

 

 

 

 

101

 

 

423a Rhythm abstinence

 

.9 Withdrawal

 

 

 

 

 

 

 

 

 

 

 

424a
-e
425 If yes to whether adopted any family planning 1= Health clinic
method, what was the most important source of 2= Another health clinic
information that convinced you to adopt a family 3: Radio
planning method? 4= TV
5: Print media
(spontaneous answers) 6: Spouse
7: Friend
8= Family Planning Field Worker
9= Other

 

 

 

 

 

426. The next few questions will ask you how you feel about these different family
planning methods.

Depending on whether or not the respondent is male or female, and whether or not the family
planning method is gender appropriate, begin the sentence with either “I” or “My partner”. For
instance, if asking a woman, the statement reads “I can use the pill to prevent pregnancy”.
Use this code: -

1 2 3 4 5
Strongly Disagree Neutral Agree Strongly
Disagree Agree

If the respondent has never had sex, use the “would easily be able to” stem instead of
“is easily able to” stem.

 

 

426a The l-lusband should be the one to 1 2 3 4 5
decide whether the couple should
use a family planning method

 

 

 

 

 

 

 

 

 

 

426b Using the pill to prevent pregnancy 1 2 3 4 5
SE is easy for me (or my partner) to do.

(IF NO SEX: would be easy. . .]
426c Getting Norplant/injections to 1 2 3 4 5

 

 

SE prevent pregnancy are easy for me
(or my partner) to do. (IF NO SEX:
would be easy. . .]

 

 

 

426d Using condoms to prevent 1 2 3 4 5
SE pregnancy is easy for me/my partner
to do (IF NO SEX: would be easy)

 

 

 

 

 

 

 

 

 

 

 

Now, we'd like to know the degree to which you think various family planning methods
are effective In preventing pregnancy

 

 

 

426a The pill is effective in preventing 1 2 3 4 5
RE pregnancy.

 

 

 

 

 

 

 

 

 

 

 

102

 

 

426f
RE

Condoms are effective in
preventing pregnancy.

 

4269
RE

Implant/injections are effective in
preventing pregnancy.

 

425h
ATT

Using the pill to prevent pregnancy
is (would be) good.

 

426i
ATT

Getting implant/injections to prevent
pregnancy is (would be) good.

 

426j
ATT

Using condoms to prevent
pregnancy is (would be) good

 

426k
DA

ONLY IF HAS HAD SEX: I try not
to think about getting pregnant
when having sex.

 

426I
REAC

Current campaigns promoting
family planning methods are
misleading.

 

426m
REAC

Current campaigns promoting
family planning methods are
untrustworthy.

 

426n
BAR

The pill has bad side effects.

 

4260
BAR

Norplant has bad side effects.

 

426p
BAR

Injections have bad side effects.

 

426q
BAR

Modern family planning methods
cost too much.

 

426r
BAR

Modern family planning services
are hard to get.

 

426s
BAR

It is embarrassing to use family
planning services

 

426t
SN

My culture is against family
planning

 

426u
SN

My religion is opposed to family
planning.

 

426v
SN

My friends use modern family
planning methods.

 

426w
SN

My siblings use modern family
planning methods.

 

 

426x
SN

 

Most Ethiopians just like me use
modern family planning methods.

 

 

103

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SECTION 5: MISC. Scales (GENDER ROLES, IMPULSIVITY, SENSATION

SEEKING, SELF-ESTEEM, SOCIAL CAPITAL, COLLECTIVE EFFICACY)

 

We’re almost done. This last section is like previous ones, where I asked you to give the extent
of your agreement or disagreement with each statement, from (show ruler) "Strongly Disagree"
to "Strongly Agree." You can simply point to or say your answer in response to each statement.

* Remember, there are no right or wrong answers; these are simply your opinions.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

501 GENDER ROLE ATT [TUDES scale a g g — E a E
E a a i .. § ..
en '5 E z T a) ‘

501a Males and females should have equal 1 2 3 4 5

rights.

501b Boys should help with housework the way

. 1 2 3 4 5
girls do.

501c A male child IS preferable to a female 1 2 3 4 5

child.

501d When resources are scarce only boys 1 2 3 4 5

should be sent to school.

501e It's okay for a man to beat his wife as a

sign of discipline if she does something 1 2 3 4 5
wrong.
501f Awoman should not question the authority
1 2 3 4 5
of a man.

5019 Women cannot make as good decisions

. 1 2 3 4 5
on important matters as men can

502 SELF-ESTEEM scale.

Now I am going to read you a list of E
statements dealing with your general 2- § § — g

- - g a 0’ E E >.
feelings about yourself. Tell me if you 2 3 3 o a. a
strongly agree, agree, disagree, or 5 5 E z T g
strongly disagree with each statement. ‘0
Order okay? Consistency

502a On the whole, I am satisfied with myself. 1 2 3 4 5

502b At times I think I am no good at all. 1 2 3 4 5

502c I feel that l have a number of good 1 2 3 4 5

qualities.

502d llike the way I look. 1 2 3 4 5

502e I like most things about myself. 1 2 3 4 5

 

 

 

 

 

 

 

 

 

 

 

 

 

104

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

502f I wish I were somebody else. 1 4 5
503 IMPULSIVITY scale. a >.
When I make a decision to do something: ,3 g g, g g 'g g
a g g 3 < i. <
503a I think about all of my choices very 1 4 5
carefully
503b 1mg}?j the first thing that comes into my 1 4 5
503e I don't even think about it; I just do it. ' 1 4 5
503d héake my time to carefully decide what to 1 4 5
503e I think things out step by step. 1 4 5
503f I consider if it will be good or bad for my 1 4 5
future.
5039 I do whatever feels good at the moment. 1 4 5
504 SENSATION SEEKING scale. > E >
§ é a g t s t
e .1 g a a g a
(7: E a z < m ‘
504a I like new and exciting experiences. 1 5
504b I like to do scary things. 1 5
5044:: I sometimes like to break the rules. 1 5
504d I prefer friends who are exciting and 1 5
unpredictable
Social Capital scale > > E
3’ § E. E g E g
5 E g 5 T m 3
505a People in this community are willing to help 1 2 3 4 5 ‘
their ngqhbors.
505b This is a close-knit community. 1 2 3 4 5 _
5050 People in this community can be trusted. 1 2 3 4 5
505d People in this community generally don't 1 2 3 4 5

 

get along with each other.

 

 

 

 

 

 

 

 

 

 

 

105

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

505e People in this community do not share the 1 2 3 4 5
same values
505f People in this community will intervenejf 1 2 3 4 5
someone’s children were engaging in
delinquent behavror.
5059 In this community, if you perform a favor for 1 2 3 4 5
one of your neighbors they will likely return
the favor at some future date.
505h One of the norms in this community is that 1 2 3 4 5
people help one another.
Individual and Collective Efficacy
e E t g g s E
5 g g: o a 5 .8;
53 a a 2 “ a“; a
5063 l have control over the decisions that affect 1 2 3 5
mWM.
506!) If someone opposes me, I can find means 1 2 3 5
and ways to get what I want.
5066 I can influence decisions that affect my 1 2 3 5
community.
506d By working together, people in my 1 2 3 5
community can influence decisions that
affect the community.
5069 People in my community can mobilize 1 2 3 5
resources to change things that are
bothersome.
506f People in my community can freely discuss 1 2 3 5
issuesthatafiectus.
5069 People in my community can organize 1 2 3 5
themselves to address a social problem.
50611 People in my community can work together 1 2 3 5
to influence decisions on the state or
national level.
507 Finally, this is the last question,
Overall, how honest would you say you were in answering this questionnaire?
Not Honest At Not Very Honest Fairly Honest Very Honest Completely
All 2 3 4 Honest
1 g 5

 

 

 

 

 

 

THANK YOU

We appreciate your help.
Your cooperation will enable Ethiopia to ensure that health programs meet your

needs.

106

 

BIBLIOGRAPHY

Abdulla, R. A. (2004). Entertainment-education in the Middle East: Lessons from the
Egyptian oral rehydration therapy campaign. In A. Singhal, M. J. Cody, E. M.
Rogers & M. Sabido (Eds), Entertainment-education and social change: History,
research, and practice (pp. 301-320). Mahwah, NJ: Lawrence Erlbaum Associates,
Publishers.

Allard, R. (1989). Beliefs about AIDS as determinants of preventive practices and of
support for coercive measures. American Journal of Public Health, 448-452.

Babin, L. A., & Carder, S. T. (1996). Viewers' Recognition of Brands Placed Within a
Film. International Journal of Advertising, 1 5, 140-151.

Bae, H., & Lee, B. (2004). Audience involvement and its antecedents: An analysis of the
electronic bulletin board messages about an entertainment-education drama on
divorce in Korea. Asian Journal of Communication, 14(1), 6-21.

Baer, D. (2002). Latent variable structural equation models. Unpublished manuscript,
ICPSR Summer Program.

Bandura, A. (1977). Social learning theory. Upper Saddle River, NJ: Prentice Hall.

Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ:
Prentice Hall.

Bandura, A. (1994). Social cognitive theory and exercise of control over HIV infection.
In R. J. DiClemente & J. L. Peterson (Eds.), Preventing AIDS theories and
methods of behavioral interventions (pp. 25-59). New York: Plenum Press.

Bandura, A. (1995). Exercise of personal and collective efficacy in changing societies. In
A. Bandura (Ed.), Self-eflicacy in changing societies (pp. 1-45). New York, NY:
Cambridge University Press

, Bandura, A. (1997). Self-eflicacy: The exercise of control. New York, NY: W. H. Freeman
and Company.

Bandura, A. (2004). Social cognitive theory for personal and social change by enabling
media. In A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido (Eds),
Entertainment-education and social change: History, research, and practice (pp.
75-96). Mahwah, NJ : Lawrence Erlbaum Associates, Publishers.

Baran, S. J ., Chase, L. J ., & Courtright, J. A. (1979). Television drama as a facilitator of
prosocial behavior: "The Waltons." Journal of Broadcasting, 23(3), 277-284.

107

Basen-Engquist, K. (1992). Psychosocial predictors of "safer sex" behaviors in young
adults. AIDS Education and Prevention, 4, 120-134.

Basil, M. D. (1996). Identification as a mediator of celebrity effects Journal of
Broadcasting & Electronic Media, 40(4), 478-495.

Basil, M. D., & Brown, W. J. (1997). Marketing AIDS prevention: The differential impact
hypothesis versus identification effects. Journal of Consumer Psychology, 6, 389-
411.

Belete, S., Girgre, A., & Witte, K. (2003). Summative evaluation of "Journey of Life ".'
The Ethiopia reproductive health communication project. Addis Ababa, Ethiopia:
J HU/CCP and Ethiopia National Office of Population.

Balasubramanian, S. K. (1994). Beyond Advertising and Publicity: Hybrid Messages and
Public. Journal of Advertising, 23(4), 29-46.

Brennan, 1., Dubas, K. M., & Babin, L. A. (1999). The Influence of Product- Placement
Type & Exposure Time on Product-Placement Recognition. International Journal
of Advertising, 18(3), 323-337.

Brock, T. C., Green, M. C., & Strange, J. J. (2002). Insights and research implications:
Epilogue to narrative impact. In M. C. Green, J. J. Strange & T. C. Brock (Eds),
Narrative impact: Social and cognitive foundations (pp. 343-354). Mahwah, NJ:
Lawrence Erlbaum Associates, Publishers.

Brock, T. C., Strange, J. J ., & Green, M. C. (2002). Power beyond reckoning: An
introduction to narrative impact. In M. C. Green, J. J. Strange & T. C. Brock
(Eds), Narrative impact: Social and cognitive foundations (pp. 1-15).

Brown, W. J ., Basil, M. D., & Bocamea, M. C. (2003). Social influence of an
international celebrity: Responses to the death of princess Diana. Journal of
Communication, 53(4), 587-605.

Brown, W. J ., & Fraser, B. P. (2004). Celebrity identification in entertainment-education.
In A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido (Eds), Entertainment-
education and social change: History, research, and practice (pp. 97-115).
Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In J. S.
Long (Ed.), Testing structural equation models (pp. 136-162). Thousand Oaks,
CA: Sage.

Bruner, J. S. (1990). Acts of meaning. Cambridge, MA: Harvard University Press.

108

Bryant, J ., & Miron, D. (2002). Entertainment as media effect. In J. Bryant & D.
Zillmann (Eds), Media eflects: Advances in theory and research (2nd ed., pp.
549-582). Mahwah: NJ: Lawrence Erlbaum Associates, Publishers.

Cohen, J. (2001). Defining identification: A theoretical look at the identification of
audiences with media characters. Mass Communication and Society, 4(3), 245-
264.

Cohen, J ., & Perse, E. M. (2003, May). Diflerent stroke for diflerent folks: An empirical
search for diflerent modes of viewer-character relatopnships. Paper presented at
the International Communication Association, San Diego, CA.

Cole, H. P. (1997). Stories to live by: A narrative approach to health behavior research
and injury prevention. In D. S. Gochman (Ed.), Handbook of health behavior
research IV: Relevance for professionals and issues for the future (Vol. 325-349).
New York: Plenum Press.

DeLorme, D. E., & Reid, L. N. (1999). Moviegoers' Experiences and Interpretations of
Brands in Films Revisited. Journal of Advertising, 28(2), 71-95.

Deighton, J ., Romer, D., & McQueen, J. (1989). Using drama to persuade. Journal of
Consumer Research, 16(3), 335-343.

DiClemente, R. J ., Durbin, M., Siegel, D., Krasnovsky, F., Lazarus, N., & Comacho, T.
(1992). Determinants of condom use among junior high school students in a
minority, inner-city school district. Pediatrics, 89(2), 197-202.

Dinoff, B. L., & Kowalski, R. M. (1999). Reducing AIDS risk behavior: The combined
efficacy of protection motivation theory and the elaboration likelihood model.
Journal of Social and Clinical Psychology, 18(2), 223-239.

Fazio, R. H., & Zanna, M. P. (1981). Direct experience and attitude-behavior consistency.
In L. Berkowitz (Ed), Advances in experimental social psychology (Vol. 14, pp.
161-202). San Diego, CA: Academic Press.

Forbes, D. (2000). Prime-time propaganda: How the White House secretly hooked
network TV on its anti-drug message. Retrieved June 6, 2003, from
http://wwwmapinc.org/drugnews/vOO/n043/309.html?126

Fraser, B. P., & Brown, W. J. (2002). Media, celebrity, and social influence: Identification
with Elvis Presley. Mass Communication and Society, 5(2), 183-206.

Freud, S. (1949). Group psychology and the analysis of the ego (J. Strachey, Trans). New
York: Boni and Liveright Publishers.

Gholamain, M. (1998). The attachment and personality dynamics of reader response.
Unpublished doctoral dissertation, University of Toronto, Toronto, Canada.

109

Green, M. C., & Brock, T. C. (2000). The role of transportation in the persuasiveness of
public narratives. Journal of Personality and Social Psychology, 79(5), 701-721.

Green, M. C., & Brock, T. C. (2002). In the mind's eye: Transportation-imagery model of
narrative persuasion. In M. C. Green, J. J. Strange & T. C. Brock (Eds), Narrative
impact: Social and cognitive foundations (pp. 315-341). Mahwah, NJ: Lawrence
Erlbaum Associates, Publishers.

Greenberg, B. S., Salmon, C. T., Patel, D., Beck, V., & Cole, G (2004). The evolution of
an E-E research agenda. In A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido
(Eds), Entertainment-education and social change: History, research, and
practice (pp. 191-206). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Gupta, P. B., Balasubramanian, S. K., & Klassen, M. L. (2000). Viewers' Evaluations of
Product Placements in Movies: Public Policy Issues and managerial Implications.
Journal of Current Issues and Research in Advertising, 22(2), 41-53.

Gupta, P. B., & Gould, S. J. (1997). Consumers' Perceptions of the Ethics and
Acceptability of Product Placements in Movies: Product Category and Individual
Differences. Journal of Current Issues and Research in Advertising, 19(1), 37-50.

Gupta, P. B., & Lord, K. R. (1998). Product Placement in Movies: The Effect of
Prominence and Mode on Audience Recall. Journal of Current Issues and
Research in Advertising, 20(1), 47-59.

Hingson, R., Strunin, L., Berlin, B., & Heeren, T. (1990). Beliefs about AIDS, use of
alcohol and drugs, and unprotected sex among Massachusettes adolescents.
American Journal of Public Health, 80, 295-299.

Hoffner, C., & Cantor, J. (1991). Perceiving and responding to mass media characters in
responding to the screen. In J. Bryant & D. Zillmann (Eds), Responding to the
screen : reception and reaction processes (pp. 63-101). Hillsdale, NJ: Lawrence
Erlbaum Associates.

Horton, D., & Wohl, R. R. (1956). Mass communication and para-social interaction:
Observation on intimacy at a distance. Psychiatry, 19(3), 188-211.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure

analysis: Conventional criteria versus new alternatives. Structural Equation
Modeling, 6, 1-55.

J anz, N. K., & Becker, M. H. (1984). The Health Belief Model: A decade later. Health
Education Quarterly, 11(1-47).

Jdreskog, K. & Sdrbom, D. (1996a). LISREL 8: User is reference guide. Scientific
Software International Inc.

110

J oreskog, K. & Sdrbom, D. (1999b). LISREL 8.30 and PRELIS 2.30. Scientific Software
International Inc.

Karrh, J. A. (1998). Brand placement: A review. Journal of Current Issues & Research in
Advertising, 20(2), 31-49.

Keller, S. N., & Brown, J. D. (2002). Media interventions to promote responsible sexual
behavior. The Journal of Sex Research, 39(1), 67-72.

Kelloway, E. K. (1998). Using LISREL for structural equation modeling: A researcher 3‘
guide. Thousand Oaks, CA: SAGE Publications, Inc.

Kelman, H. C. (1961). Processes of Opinion Change. Public Opinion Quarterly, 25(1),
57-78.

Kerlinger, F. N., & Pedhazur, F. (1973). Multiple regression in behavioral research. New
York: Holt, Rinehart, & Winston.

Kincaid, D. L. (2002). Drama, emotion, and cultural convergence. Communication
Theory, 12(2), 136-152.

Kincaid, D. L., Yun, S. H., Piotrow, P. T., & Yaser, Y. (1993). Turkey's mass media family
planning campaign. In T. E. Backer & E. M. Rogers (Eds), Organizational
aspects of health communication campaigns: What works? (pp. 68-92). Newbury
Park, CA: Sage.

Kline, R. B. (1998). Principles and practice of structural equation modeling. New York: -
NY, The Guilford Press.

Kopfrnan, J. E., Smith, S. W., Yun, J. K. A., & Hodges, A. (1998). Affective and cognitive
reactions to narrative versus statistical evidence organ donation messages. Journal
of Applied Communication Research, 26, 279-300.

Krugman, H. E. (1966). The measurement of advertising involvement. Public Opinion
Quarterly, 30, 583-596.

Langley, D. J. (1995). Examining the personal experience of student skill Ieaming: A
narrative perspective. Research Quarterly for Exercise and Sport, 66(2), 116-128.

Law, 8., & Braun, K. A. (2000). I'll Have What She's Having: Gauging the Impact of
Product Placements on Viewers. Psychology & Marketing, 1 7(12), 1059-1075.

Levy, M. R. (1979). Watching TV news as para-social interaction. Journal of
Broadcasting, 23(1), 69-80.

111

Levy, M. R., & Windahl, S. (1984). Audience activity and gratifications: A conceptual
clarification and exploration. Communication Research, 11(1), 51-78.

Liebes, T., & Katz, E. (1990). The export of meaning: Cross-cultural readings of
"Dallas”. New York: Oxford University Press.

Livingstone, S. M. (1991). Audience reception: The role of the viewer in retelling
romantic drama. In J. Curran & M. Gurevitch (Eds), Mass media and society (pp.
285-306). London: Great Britain: Edward Arnold.

Mattson, M. (1999). Toward a reconceptualization of communication cues to action in the
health belief model: HIV test counseling. Communication Monographs, 66, 240-
265.

Meissner, W. W. (1972). Notes on identification. Psychoanalytic quarterly, 41, 224-259.

Montgomery, 8., Joseph, J ., Becker, M., Ostrow, D., Kessler, R., & Kirscht, J. (1989).
The health belief model in understanding compliance with preventive
recommendations for AIDS: How useful? AIDS Education and Prevention, 1 ,
303-323.

Murphy, S. T., & Zajonc, R. B. (1993). Affect, cognition, and awareness: Affective
priming with optimal and suboptimal stimulus exposures. Journal of Personality
and Social Psychology, 64(5), 723-739.

Oatley, K. (2002). Emotions and the story worlds of fiction. In M. C. Green, J. J. Strange
& T. C. Brock (Eds), Narrative impact: Social and cognitive foundations (pp. 39-
69). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Oatley, K., & Gholamain, M. (1997). Emotions and identification: Connections between
redaers and fiction. In M. Hjort & S. Laver (Eds), Emotions and the arts (pp.
263-281). New York: Oxford University Press.

Papa, M. J ., Singhal, A., Law, 8., Pant, S., Sood, S., Rogers, E. M., et a1. (2000).
Entertainment-education and social change: An analysis of parasocial interaction,
social learning, collective efficacy, and paradoxical communication. Journal of
Communication, 50(4), 31-55.

Pastina, A. C. L., Patel, D. S., & Schiavo, M. (2004). Social merchandizing in Brazilian
telenovelas. In A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido (Eds),
Entertainment-education and social change: History, research, and practice (pp.
261-279). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Petty, R. E., Cacioppo, J. T., & Schmann, D. (1983). Central and peripheral routes to

advertising effectiveness: The moderating role of involvement. Journal of
Consumer Research, 1 0, l 35- 145.

112

 

Piotrow, P. T., & F ossard, E. d. (2004). Entertainment-education as a pulic health
intervention. In A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido (Eds),
Entertainment-education and social change: History, research, and practice (pp.
39-60). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers

Piotrow, P. T., Kincaid, D. L., Rimon II, J. G, & Rinehart, W. (1997). Health
communication: Lessons from family planning and reproductive health. Westport,
CT: Praeger.

Polacsek, M., Celentano, D. D., O'Campo, P., & Santelli, J. (1999). Correlates of condom
use stage of change: Implications for intervention. AIDS Education and
Prevention, 11(1), 38-52.

Prentice, D. A., Gerrig, R. J ., & Bailis, D. S. (1997). What readers bring to the processing
of fictional texts. Psychonomic Bulletin and Review, 4, 416-420.

Prochaska, J. O., & DiClemente, C. C. (1983). Stages and processes of self-change of
smokinngoward an integrative model of change. Journal of consulting and
clinical psychology, 51(3), 390-395.

Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992). In search of how people
change: Applications to addictive behaviors. American Psychologist, 47(9), 1102-
1114.

Rosenstock, I. M. (1974). Historical Origin of the Health Belief Model. In M. H. Becker
(Ed.), The Health Belief Model and personal health behavior (pp. 1-8). Thorofare,
NJ: Charles B. Slack, Inc.

Rosenstock, I. M., Strecher, V. J ., & Becker, M., H. (1988). Social learning theory and the
Health Belief Model. Health Education Quarterly, 15(2), 175-183.

Rosenstock, I. M., Strecher, V. J ., & Becker, M. H. (1994). The health belief model and
HIV risk behavior change. In R. J. DiClemente & J. L. Peterson (Eds),
Preventing AIDS theories and methods of behavioral interventions (pp. 5-24).
New York: Plenum Press.

Rotheram-Borus, M. J ., Reid, H., Rosario, M., & Kasen, S. (1995). Determinants of safer
sex patterns among gay\bisexual male adolescents. Journal of Adolescence, 18(1),
3-15.

Rubin, A. M., & Step, M. M. (2000). Impact of motivation, attraction, and parasocial
interaction on talk radio listening. Journal of Broadcasting & Electronic Media,
44(4), 635-654.

Rubin, R. B., & McHugh, M. P. (1987). Development of parasocial interaction
relationships. Journal of Broadcasting & Electronic Media, 31(3), 279-292.

113

Schwarzer, R., & Fuchs, R. (1995). Changing risk behaviors and adopting health
behaviors: The role of self-efficacy beliefs. In A. Bandura (Ed.), Self-eflicacy in
changing societies (pp. 1-45). New York, NY.

Shankar, A., & Goulding, C. (2001). Interpretive consumer research: Two more
contributions to theory and practice. Qualitative Market Research, 4(1), 7-16.

Sherry, J. L. (2002). Media saturation and entertainment-education. Communication
Theory, 12(2), 206-224.

Singhal, A., & Rogers, E. M. (2004). The status of entertainment-education worldwide. In
A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido (Eds), Entertainment-
education and social change: History, research, and practice (pp. 3-20). Mahwah,
NJ: Lawrence Erlbaum Associates, Publishers.

Singhal, A., & Rogers, E. W. (1999). Entertainment-Education: A communication
strategy for social change. Mahwah, NJ: Erlbaum.

Singhal, A., & Rogers, E. W. (2002). A theoretical agenda for entertainment-education.
Communication Theory, 12(2), 117-135.

Slater, M. D. (2002). Entertainment education and the persuasive impact of narratives. In
M. C. Green, J. J. Strange & T. C. Brock (Eds), Narrative impact: Social and
cognitive foundations (pp. 157-181). Mahwha, NJ: Lawrence Erlbaum Associates,
Publishers.

Slater, M. D., & Rouner, D. (1996). Value affirmative and value protective processing of
alcohol education messages that include statistics or anecdotes. Communication
Research, 23, 210-235.

Slater, M. D., & Rouner, D. (1997, May). The processing of narrative fiction from the
persuasion perspective: An exploratory study. Paper presented at the International
Communication Association, Montreal, Canada.

Slater, M. D., & Rouner, D. (2002). Entertainment-education and elaboration likelihood:
Understanding the processing of narrative persuasion. Communication Theory,
12(2),173-191.

Smith, R. A., Witte, K., & Downs, E. (2003). 1 want to be like you: Entertainment
education, EPPM, and health education. Unpublished manuscript, Michigan State
University, East Lansing, MI.

Snijders, T. A. B., & Bosker, R. J. (1999). Multilevel analysis: An introduction to basic &
advanced multilevel modeling. Thousand Oaks, CA: Sage Publications.

Snyder, L. B., & Rouse, R. A. (1992). Targeting the audience for AIDS messages by
actual and perceived risk. AIDS Education and Prevention, 4, 143-159.

114

Sood, S. (2002). Audience involvement and entertainment-education. Communication
Theory, 12(2), 153-172.

Sood, S., Menard, T., & Witte, K. (2004). The theory behind entertainment-education. In
A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido (Eds), Entertainment-
education and social change: History, research, and practice (pp. 117-149).
Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Sood, S., & Rogers, E. M. (2000). Dimensions of parasocial interaction by letter-writers
to a popular entertainment-education soap opera in India. Journal of Broadcasting
& Electronic Media, 44(3), 386-414.

Steers, W. N., Elliott, E., Nemiro, J ., Ditman, D., & Oskamp, S. (1996). Health beliefs as
predictors of HIV-preventive behavior and ethnic differences in prediction. The
Journal of Social Psychology, 136(1), 99-110.

Stephenson, M. T., & Palmgreen, P. (2001). Sensation seeking, perceived message
sensation value, personal involvement, and processing of anti-marijuana PSAs.

Communication Monographs, 68(1), 49-71.

Strange, J ., & Leung, C. C. (1999). How anecdotal accounts in news and in fiction can
influence judgments of a social problem's urgency, causes, and cures Personality
and Social Psychology Bulletin, 25(4), 436-449.

Sypher, B. D., McKinley, M., Ventsam, S., & Valdeavellano, E. E. (2002). Fostering
reproductive health through entertainment-education in the Peruvian Amazon:
The social construction of Bienvenida Salud! Communication Theory, 12(2), 192—

205.

Usdin, S., & Singhal, A. (2004). No short cuts in entertainment-education: Designing
Soul City step-by-step. In A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido
(Eds), Entertainment-education and social change: History, research, and
practice (pp. 153-175). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Valente, T. W., Kim, Y. M., Lettenmaier, C., Glass, W., & Dibba, Y. (1994). Radio
promotion of family planning in the Gambia. International Family Planning
Perspectives, 20(3), 96-100.

Volk, J. E., & Koopman, C. (2001). Factors associated with condom use in Kenya: A
test of the health belief model. AIDS Education and Prevention, 13(6), 495.

Weeler, S. C., Green, M. C., & Brock, T. C. (1999). Fictional narratives change beliefs:
Replications of Prentice, Gerrig, & Bailis (1997) with mixed corroboration.
Psychonomic Bulletin and Review, 6(1), 136-141.

115

Wiggers, L. C. W., de Wit, J. B. F., Gras, M. J ., Coutinho, R. A., & van den Hoek, A.
(2003). Risk behavior and social-cognitive determinants of condom use among
ethnic minority communities in Amsterdam. AIDS Education and Prevention,
15(5), 430-447.

Wilson, D., Manual, A., & Lavelle, S. (1991). Psychological predictors of condom use to
prevent HIV transmission among Zimbabwean students. International Journal of
Psychology, 26(6), 705-721.

Winfield, E. B., & Whaley, A. L. (2002). A comprehensive test of the health belief model
in the prediction of condom use among Afi‘ican American college students.
Journal of Black Psychologz, 28(4), 330.

Witte, K., Girma, B., & Girgre, A. (2001). Ethiopia reproductive health communication
project: Family planning HI V/AIDS prevention formative and baseline study.
Addis Ababa, Ethiopia: JHU/CCP and Ethiopia National Office of Population.

Wollheim, R. (1974). Identification and imagination. In R. Wollheim (Ed.), Freud: a
collection of critical essays (pp. 172-195). Garden City, N.Y.: Anchor Books.

Wong, C.-t., & Tang, C. S.-k. (2001). Understanding heterosexual Chinese college
student's intention to adopt safer sex behaviors. Journal of Sex Research, 38(2),
118-126.

Wulfert, E., & Wan, C. K. (1993). Condom use: A self-efficacy model. Health Psychology,
12(5), 346-353.

Yaser, Y. (2004). The Turkish Family Health and Planning Foundation's entertainment-

education campaign. In A. Singhal, M. J. Cody, E. M. Rogers & M. Sabido (Eds),
Entertainment-education and social change: History, research, and practice (pp.
321-329). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers.

Yep, G A. (1993). HIV prevention among Asian-American college students: Does the
Health Belief Model work? Journal of College Health, 41, 199-205.

Zillmann, D. (1994). Mechanisms of emotional involvement with drama. Poetics, 23, 33-
51.

Zimmerman, R. S., & Olson, K. (1994). AIDS-related risk behavior and behavior change
in a sexually active, heterosexual sample: A test of three models of prevention.
AIDS Education and Prevention, 6, 189-204.

116

   

lllllllllllllllllllll