How to Persuade Adolescents to Use Nutrition Labels:

Effects of Health Consciousness, Argument Quality

and Source Credibility

DONG, Zhuowen

A Thesis Submitted in Partial Fulfilment

of the Requirements for the Degree of

Master of Philosophy

in

Communication

The Chinese University of Hong Kong

April 2013

Thesis/Assessment Committee

Professor Louis W. C. Leung (Chair)

Professor Annisa C. H. Lai Lee (Thesis Supervisor)

Professor Francis L. F. Lee (Committee Member)

i

Abstract of thesis entitled:

How to Persuade Adolescents to Use Nutrition Labels: Effects of Health

Consciousness, Argument Quality and Source Credibility

Submitted by DONG Zhuowen

for the degree of Master of Philosophy in Communication

at The Chinese University of Hong Kong in (April 2013)

This study tested the utility of the Elaboration Likelihood Model in creating

effective Public Service Announcements (PSAs) of nutrition label use. Students

(N=169) from a secondary school in Hong Kong participated in a 2x2x3 (Health

consciousness: High vs. Low x Argument: Informational vs. Testimonial x Source:

Expert vs. Celebrity vs. Ordinary Person) factorial design experiment. Hypotheses

were partially supported. Main effect was located for health consciousness.

Adolescents with high health consciousness were more persuaded by the PSAs. There

was an interaction effect between health consciousness and source. A dietitian as the

source was most persuasive to people with high health consciousness, while low

health conscious people were most influenced by a celebrity. The information

processing route was biased by the match of argument and source. Informative

argument made by an ordinary person was more likely to motivate adolescents to use

nutrition label, while testimonial argument made by a celebrity was the most effective.

The findings are discussed in light of ELM and match-up hypotheses.

Recommendations are also offered for public health communication campaigns.

ii

中文摘要

本研究探討深思的可能性模型（Elaboration Likelihood Model）在設計推廣

使用營養標籤的廣告中的適用性。香港的一所中學的 169 位學生參與了一個 2 x 2

x 3（健康意識：高 vs. 低 x 論據：數據 vs. 經驗 x 代言人：專家 vs. 名人 vs.

普通人）的實驗。 部分假設得到支持。 健康意識的主效果明顯。 具有較高健

康意識的青少年更易被廣告說服。 健康意識和代言人具有交叉效應。對於健康

意識較高的人來說，營養師更有說服力， 而對於健康意識較低的人來說，名人

更有影響力。論據和代言人的搭配影響了資訊的處理路徑。如果以數據為主的論

據是由普通人提供的，則更易推動青少年使用營養標籤。而如果以個人經驗為主

的論據是由名人提供的，則更有效。統計結果將在深思的可能性模型下得以展開

討論。 本文還會對開展針對青少年的公共健康傳播運動提出建議。

iii

Acknowledgements

This thesis would not be possible without the guidance, encouragement and

support of my teachers, friends, and my family. They helped me reach a goal I did not

originally think was possible.

First and foremost, I would like to express sincere gratitude to my committee

members. Professor Annisa Lee, thank you for being my supervisor, mentor and

friend throughout my study at CUHK. Your expertise and rich experience in health

communication have inspired me to explore this field further and further. I am grateful

for the countless hours you spent discussing the theoretical framework and details of

my work. Your unwavering support has always given me strength to move forward.

Professor Louis Leung, thank you for having guided me on the journey of

communication research since the first day of the M.Phil program. Your pursuit of

perfection is a reminder of the rigorous attitude I should embrace as a researcher.

Professor Francis Lee, your passion for teaching and doing research is infectious.

Thank you for always being ready to help and answering my questions patiently, no

matter how tedious they were.

I am grateful to other faculty members at the school. The varied research

interests of them have expanded my understanding of the field. I enjoyed attending

the classes, seminars, study groups and workshops they held, where I got exposed to

elementary and advanced theories and research methods, and most importantly, I

gained insights of society and humanity from different perspectives.

Professor Anthony Fung, thank you for evaluating my performance at the

iv

interview, which opened a new page in my life. It was in your class where I learned

the theories applicable to my research.

Professor Paul Lee, I didn’t know the production of knowledge carries so many

philosophic meanings until I took your class. Thank you for reading my proposals

carefully even though I changed the topic three times.

Professor Lo, Ven-hwei, thank you for making statistics interesting. Without the

methods taught by you, it would have been impossible for me to analyze the research

data for any term paper or my thesis.

Professor Eric Ma, thank you for teaching me skills to communicate with people

and observe our life. Everything is there for a reason, and the process of searching for

it is much more exciting than I imagined.

I am indebted to many people for their helps with the experiment, including

Professor Donna Chu from CUHK, Dr. Teresa Choi from Center for Food Safety, Mr.

Polo Lo and Principal Chau from CUHKFAA Chan Chun Ha Secondary School,

Principal Au-yang and Principal Lau from Po Leung Kuk Mrs Ma Kam Ming-Cheung

Fook Sien College, and Ms. Agnes Lee from Good Hope School.

In the past two and a half years, I feel very fortunate to develop friendships with

so many brilliant and hard-working M.Phil and Ph.D students. Lisa Tam, it is you

who made our papers qualified to be accepted by IAMCR and NCA, which exposed

me to a bigger world. Pei Zheng, thank you for reminding me, with your own practice,

that the more we work, the more we are able to achieve. Allan Au, thank you for

sharing your perspectives in various channels. Your example as a professional media

v

person and traveler has motivated me to live a more meaningful and productive life.

Other classmates, Peiyi Huang, Mengmeng Zhao, Weiwei Zhang, Hongzhe Wang,

Yue Wang, Oiyan Ng, Zhifei Mao Ni-chen Sung and Frankie Cheung, I miss the

thought-provoking class discussions we had together, and thank you for sharing your

knowledge with great passion. Yin Zhang, Chen Lu, Haiyan Wang, Cindy Chong, Qi

Ling, Xueting Liao, Xin Zhang and Yang Liu, it’s my great pleasure to have your

company in my campus life and conference trips. Our life journey is long, but how

wonderful it is that we have once enjoyed the beautiful scenes together. Miao Li and

Fang Luo, thank you for allowing me to look for accommodation with fewer burdens.

Special thanks go to Xiaoxiao Zhang, who is a sister figure to me, giving me the

biggest encouragement whenever I feel frustrated and doubtful about myself. Her

smiles, with calmness and happiness, can melt the snow.

My sincere gratitude also goes to all general clerks at the school. Heidi, thank

you for the ongoing support in almost every aspect. Most of the things you have done

may seem simple, but just as the air and the sunshine, they are vital. I wish in the

future I could become as patient, kind-hearted and helpful as you.

I cannot be spare in my appreciation of Ms. Ping Ping Yu and Mr. Hong Huang,

who have been taking care of me in Hong Kong. You not only offer me a place to live,

but also give me security, warmth, and endless love.

Lastly, I would like to thank my family, who are always cheering me up and

standing by me through the good times and bad. I am endlessly thankful for the love

and belief they have in me.

vi

Table of Contents

Abstract i

Acknowledgements iii

Table of Content vi

Chapter 1: Introduction 1

Publicity Campaign of Nutrition Labelling Scheme in Hong Kong 1

Adolescent and Nutrition Label (NL) 3

Significance of the Research 4

Organization of Thesis 7

Chapter 2: Literature Review 9

Adolescents and Preventive Health Behavior 9

Health Consciousness 10

Information Processing of Adolescents 12

Informative or Testimonial Argument 13

Expert vs. Celebrity vs. Ordinary person 15

Source credibility 20

Initial attitude 24

Self-efficacy 25

Chapter 3: Theoretical Framework 28

The Elaboration Likelihood Model of Persuasion 28

Chapter 4: Research questions and Hypothesis 34

34

Chapter 5: Methodology 39

Subjects and Design 40

Procedure 40

Materials and Manipulations 41

Independent variables 44

Dependent variables 46

Chapter 6: Results 49

Health Consciousness Assessment 49

vii

General results 49

Covariate 56

Predictors of Behavioral Intention 56

Hypothesis Testing and Answers to Research Questions 57

Chapter 7: Discussion 59

Chapter 8: Implication 66

Chapter 9: Limitation and Direction for Future research 69

References 72

Lists of Appendices

Appendix A: Experiment Questionnaire 85

Appendix B: Experiment Masterial--- PSA 87

Appendix C: Experiment Masterial--- PSA 88

Appendix D: Experiment Masterial--- PSA 89

Appendix E: Experiment Masterial--- PSA 90

Appendix F: Experiment Masterial--- PSA 91

Appendix G: A Proposed PSA of Nutrition Labels 92

List of Tables:

Table 6.1: MANCOVA Results--- Wilks’ Lambda Statistics 50

Table 6.2: Standardized Multivariate Means and Standard Deviation of the Influence of

Health Consciousness, Argument and Source on Perceived knowledge gain, Post

Attitude and Behavioral Intention 51

Table 6.3: Multivariate Analysis of Covariance ---The Influence of Initial attitude, Health

Consciousness, Argument and Source on Perceived knowledge gain, Post Attitude

and Behavioral Intention 57

Table 6.4: Multiple Regressions to Predict Behavioral Intention to Use Nutrition Labels 57

Lists of Figures:

Figure 1.1: Tabular Format of Nutrition Labelling in both English and Chinese 3

Figure 3.1: Central and Peripheral routes to persuasion. 33

Figure 4.1: Modification of the ELM in the Processing of PSAs of Nutrition Label Use 38

Figure 6.1: Interactive effect of health consciousness and source on perceived knowledge gain

53

Figure 6.2: Interactive effect of argument and source on behavioral intention 53

Figure 6.3: Marginal interactive effect of argument and source on perceived knowledge gain

55

1

Chapter1:

Introduction

Publicity Campaign of Nutrition Labelling Scheme in Hong Kong

Nutrition information on food labels is an important tool to promote a balanced

diet. It is particularly urgent to regulate the provision of this tool in Hong Kong,

where 60% of deaths are caused by cancers, heart diseases, cerebrovascular diseases

and diabetes, four of the top ten killers attributed largely to imbalanced diets,

especially those high in fat, sugars and sodium (Center for Food Safety, 2010).

To assist consumers to make informed food choices, a mandatory nutrition

labelling scheme was enforced on 1 July, 2010. It requires information on energy and

seven specified nutrients, namely, protein, carbohydrates, total fat, saturated fat, trans

fat, sodium and sugars, to be listed on food labels (See Figure 1.1 for the tabular

format of nutrition labeling in Hong Kong). With this information, consumers can

compare nutrient contents among different food for a healthier choice and choose

food that is lower in fat, sodium (or salt) and sugars.

However, many consumers haven’t made full use of the nutrition labels to make

food choices. Before the enforcement of the nutrition labelling scheme, a survey

conducted by Center for Food Safety (CFS) showed that more than 80% of the 1,200

respondents believed that nutritional information was useful, but only half of them

had the habit of reading nutrition labels when buying prepackaged food products for

the first time (CFS, 2009).

To tie in with the implementation of the scheme, the Government has been

2

actively engaging in publicity and public education, aiming to drive home the

advantages of the nutrition labelling scheme and enhance public understanding on

how to read the information on nutrition labels. A media campaign “Live it, use it”

was launched, with the utilization of carious channels, such as brief TV programme,

TV& radio announcements and advertisements in public transport system, etc. In

addition, through collaboration with retailers, education stakeholders and community

organizations, the CFS has organized serial talks, workshops and roving exhibitions to

reach out to the public.

While the campaign has succeeded in raising awareness and enhancing

understanding of the benefits and knowledge of nutrition labels, it has not effectively

motivated people to use nutrition labels in their daily life. In a survey conducted by

the University of Hong Kong's Public Opinion Programme in 2011, only 13 percent of

the respondents said they read nutrition labels. And many of them thought it difficult

and time-consuming to process the information on nutrition labels (HKU POP, 2011).

The finding indicates that attitude change and subsequent behavioral change require

continuing and sustained public education effort.

3

Figure 1.1 Tabular Format of Nutrition Labelling in both English and Chinese (CFS,

2008)

Adolescent and Nutrition Label

Adolescents are a target group in the promotion of nutrition label. According to

the latest government data, children in Hong Kong are putting on fat: General obesity

rates among secondary school students climbed to 17% in 2009-2010, from 13% in

1996-1997. For primary-school students, the figure saw a significant increase from

16.7% to 22.2% (Shay, 2011). More and more young people suffer from diabetes, and

the youngest patient is only eight years old.

Eating habits are established early in the life cycle and tend to be carried through

to adulthood. The use of nutrition information can help adolescents to develop and

consolidate healthy eating habits, which can make for a healthy body in later life.

4

Given the benefit of promoting nutrition labels among students, emphasis has been

placed on youth participation in executing the nutrition labelling publicity strategy.

For example, nutrition labels and relevant topics have been integrated into the school

curriculum at different levels. In the 2010/11 academic year, the CFS and the

Education Bureau jointly organized the “Live It, Use It” Nutrition Labelling

Promotion Award Scheme, under which a total of 35 teams from 21 secondary schools

designed and implemented creative activities to promote nutrition label among

adolescents’ peers, teachers, family and communities.

There have been no evaluations of the effectiveness of the past activities targeted

at youth. The sample of previous surveys investigating nutrition label use was the

general population, instead of a specific age group. In other countries where the

advocacy of nutrition label use was initiated earlier than that in Hong Kong, some

scholars had deeper findings. For example, Nørgaard and Brunsø (2009) found that

children aged under 12 use nutritional information rarely or not at all. Parents seldom

use nutritional information when they had problems processing an overflow of

information, or information that is too technical. They preferred food labels with less

information and more visual cues, like a colorful panel.

Significance of the Research

This research has practical and theoretical significance.

Practically, it aims to give inspiration to the health practitioners about how to

promote nutrition labels via PSAs targeted at adolescents. While many health

education studies have been concerned about the adolescents in Hong Kong, the

5

health behaviors they investigate focus on either risk behaviors such as substance

abuse (e.g. Lee, Tsang, & Healthy School Research Group, 2004), smoking and

suicide (Stewart et al., 1999), or health preventive behaviors such as physical activity

(e.g. Ho & Lee, 2001) and eating fruits and vegetables. Some other studies are

conducted as an evaluation or a report of a general health promotion program (e.g.

Lee et al., 2003). Only a few studies investigate the nutritional diets or nutrition label

promotion among adolescents, but they are based on school-based intervention or

community intervention, instead of PSAs. Mass media is influential to the adolescents,

but the research investigating the impact of media on adolescent health pay more

attention to the negative social influence of TV commercials, TV programs or films,

rather than the positive effect of PSAs, even though PSAs are widely used by the

governments and non-governmental organizations.

Unlike the research on the patterns of health behaviors of adolescents, this study

aims to figure out how adolescents process and respond to the messages that advocate

using nutrition labels when choosing prepackaged food. Communication theories will

be applied to examine what kinds of messages in PSAs have potential power to affect

adolescents.

In terms of theoretical significance, this study attempts to integrate the factors

that affect the PSAs processing of adolescents into a traditional cognitive model to

increase the effectiveness of health communication targeted at this age group. In

addition, this study examines concurrently components of an information processing

model and a construct which is included in many health behavior theories, that is, the

6

construct of self-efficacy to adopt the desired behavior.

According to previous survey by CFS, many consumers don’t read nutrition

labels because they are not aware of the usefulness of them or they are lacking in

ability to read and make use of them. Therefore, an effective PSA should increase the

motivation and ability of adolescents to read nutrition labels. But above all, a PSA

should be processed in an effective way at first, which means motivating the target

audience to read the messages, making them understand and learn from the content,

agree with it, store it and retrieve it later, and make decision based on it (McGuire,

1989). This study will explore the psychological mechanism of adolescents when they

are exposed to a PSA.

A PSA comprises many factors, such as argument, communicator, graphic design,

logos and so on. In the existing PSAs of nutrition labels, the salient components with

stronger persuasive power are argument and communicator. Regarding arguments,

some of them are supported by research or survey data indicating the severity of

health problems or by specific instructions of using nutrition information, while others

are supported by testimonial evidence of the communicator who claims to have

benefited from the nutrition labels. Regarding communicators, some of them are

health experts, such as dietitians and doctors, some are celebrities or famous local

artists, and the others are ordinary people at different age groups. This current study

will focus on how different arguments and communicators influence adolescents’

processing of the PSA.

Since this study concerns about how motivation to process information, and

7

messages and communicators in the information affect the processing of a PSA, the

Elaboration Likelihood Model (ELM, Petty & Cacioppo, 1981) will be adopted. ELM

has important implications for advertising in that the effectiveness of different kinds

of message appeals is affected by receivers’ individual difference (e.g. involvement in

the issue) and situational conditions (e.g. distraction) (Petty et al., 1983). In addition,

ELM illustrates the roles of argument and communicator in either route of

information processing.

Apart from the cognitive processing of the PSA, this study moves one step

further to investigate the predictive power of self-efficacy to motivate behavioral

change.

Many cognitive or behavior theories have a component representing individuals’

ability to adopt the desired behaviors, for example, self-efficacy in Social Cognitive

Theory (Bandura, 1977), perceived efficacy in Health Belief Model (Rosenstock,

1974; Rosenstock et al., 1988) and self-efficacy in Extended Parallel Process Model

(EPPM, Witte 1992). This study will inject such a component into the cognitive

processing model to understand the process of behavior change of adolescents.

Organization of theThesis

This thesis is divided into 8 chapters. Following this introductory chapter is

Chapter 2, which presents background information about the findings of preventive

health behavior of adolescents, and indicates the significance of further research.

Several elements which influence the information processing of adolescents are

detailed. It is argued that these elements can be integrated into the Elaboration

8

Likelihood Model to explain adolescents’ processing modes of the messages about

nutrition label use. Chapter 3 illustrates the theoretical framework of this thesis.

Chapter 4 presents the research questions and hypothesis according to the estimated

framework, which is a modification of the Elaboration Likelihood Model to adapt to

the promotion of nutrition label use among adolescents. Chapter 5 demonstrates the

procedure of the experiment and presents the measurements of the variables being

tested. Chapter 6 shows the results of data analysis, with a focus on the main effects

and interaction effects of some major variables. Hypotheses are tested, and research

questions are answered. Chapter 7 provides possible interpretation of the statistical

results and discusses the theoretical insights. Chapter 8 summarizes the thesis and put

forwards several ideas for health interventions. Chapter 9 reflects on the limitations of

the research and suggests some directions for future work. References and a few

appendices are followed, including the questionnaire and stimulus for the

experiments.

9

Chapter2:

LiteratureReview

Adolescents and Preventive Health Behavior

The increasing rate of obesity among adolescents is attributed to both

environmental factors and individual factors. Regarding the environment, the great

amount of food-relating advertising is to blame the most. According to Powell et al.

(2007), 26% of the advertisements viewed by American adolescents are related to

food, among which the most commonly promoted products are fast food, sweets, and

beverage products, which are well in reach of adolescents’ purchasing power. For

individual factors, as children grow up from childhood through adolescence, they

have more considerable autonomy in food choices and more likely to buy food for

themselves. Their preference for taste and enjoyment may result in more consumption

of food with high calories, sugar and fat, putting them at higher risk for health

problems.

Newmark-Sztainer et al. (1999) argued that adolescents do not perceive any

urgency to keep a healthy diet because the risk seems so distant to them that they are

not concerned about health and nutrition issues. Contento, Michela and Goldberg

(1988) divided the food choice motivation of adolescents into three groups: the

hedonistic group, for whom taste and convenience were paramount; the socially

controlled group, who regard friends’ opinions as the most important; and the health-

oriented groups, who put the most weight to personal health. However, even for the

last group, the perceived beneficial health outcomes are limited to short-term

10

outcomes, such as more energy, better athletic performance and better-looking skin.

They may especially feel anxious about physical changes of puberty. However, they

seldom consider long-term outcomes like disease prevention.

Compared to preschool children, adolescents have meals in more settings, but

their home remains to be the major one (Guthrie, Lin, & Frazao, 2002). On one hand,

food choices made by the family matters to adolescents’ health, on the other,

adolescents can influence what to serve at home (Contento et al., 2006). Nørgaard et

al. (2007) found that adolescents have more influence on the family’s purchase of

unhealthy products (e.g. sweets) than healthy ones (e.g. fruit, vegetables and fish).

Given the impact of adolescents’ preference on the family menu, it is meaningful

to motivate adolescents to use nutrition labels to make informed choices, so as to

`change the dietary habit of themselves and their family.

Health Consciousness

As mentioned above, a difference between adolescents and adults is that

adolescents are less alert to unhealthy behaviors, because they haven’t realized the

severity of the consequence. Therefore, they are less motivated to change. That is to

say, the health consciousness of adolescents is low.

Dutta-Bergman (2004a) defined health consciousness as “an indicator of the

consumer’s intrinsic motivation to maintain good health” as well as “a reflection of

his or her responsibility toward health” (p. 398). Jayanti and Burns (1998, p. 10)

referred this concept to “the extent to which health concerns are integrated into a

person’s daily activities”. Hong (2009) concluded five major dimensions that have

11

comprised the definitions of health consciousness in previous research: (1)

engagement in health behaviors, (2) psychological attention to one’s health, (3) health

information seeking and usage, (4) personal responsibility, and (5) health motivation.

Based on the five dimensions, health conscious people feel responsible for their health,

incorporate healthy behaviors in their daily life actively, keep attentive to their health

conditions, engage in health information seeking from diverse sources actively and

apply the knowledge to future behaviors.

In the current study, health consciousness is defined as the extent to which

people are conscious about their overall health status and take actions accordingly.

Exposed to the same health message, people with different levels of health

consciousness evaluate the relevance of the messages to them differently. It is by

influencing perceived self-relevance that health consciousness affects the modes of

information processing. Highly health conscious people considered health messages

to be more personally relevant, and paid more attention to them, so systematic

processing of the arguments occurred (Rosen, 2000). As Wright (1997) pointed out,

people who are generally health conscious are more likely to perceive information of

ingredients on the packaged food relevant to them.

According to Blum and Stark (1985), young adolescents cannot truly

comprehend the abstract concept of “health” or project the outcomes of their current

risk behavior into the future. They are less conscious of the importance of a healthy

body. For example, in his study about substance abuse, Scott (1996) found that

adolescents thought substance abuse only had negative impact on older people.

12

Likewise, the majority of adolescents have not lived long enough to have experienced

long-term physiological, psychological, and social problems caused by an unhealthy

diet. Therefore, they may ignore the benefits of a healthy diet, and show little interest

in the functions of the nutrients on food labels. As a result, they are less motivated and

feel less competent in applying their knowledge and experience to process such

information.

Most research used health behaviors as proxy for health consciousness. However,

what is of interest here is not only adolescents’ consciousness of a healthy and

nutritional diet, but also their comprehensive mental and behavioral orientation

towards health.

Information Processing of Adolescents

Before designing persuasive messages targeted at adolescents, it is worthy to first

identify the way they process information.

During preadolescence and adolescence, changes in cognitive processes occur.

Pre- and young adolescents are concrete thinkers who need to experience something

by seeing, touching, or hearing it, rather than just relying on abstract descriptions.

Formal operational or abstract thinking is possible by middle or late adolescence

(Blum & Stark, 1985).

According to John’s (1999) consumer socialization model, adolescents aged 11

-16 years begin to think reflectively and critically, and develop ability to formulate

hypotheses to explain occurrences. Regarding advertising messages, they understand

not only the persuasive intent but also the specific advertising tactics and appeals.

13

Most theorists contend that the heuristic system of younger people have more

various characteristics, such as the impact or involvement of affect and images in

decision-making, as well as a tendency towards superficial processing (Gibbons,

Houlihan, & Gerrard, 2009).

The mechanism that motivation and ability affects the process of attitude change

develops through adulthood. The young child, with relatively little motivation and less

ability to think about true merits of issues, may establish an attitude based on his good

or bad feeling. As children grow, the motivation to express correct opinions gets

stimulated, but the lack of knowledge hinders their ability to scrutinize issue-relevant

arguments. As a result, they are more susceptible to appeals based on behavioral cues

and self-perceptions, and rely more on cognitive rules based on personal experience

(Petty & Cacioppo, 1986).

Informative or Testimonial Argument

Based on the processing modes of different arguments and the appeals in existing

PSAs, this study designs two kinds of messages with either informative arguments or

testimonial arguments.

An informative argument is based on factual assertion or aggregated reports of

statistical evidence, while a testimonial argument comprises a vivid exemplar or a

fist-person account of someone who experienced a condition that may also affect the

message recipient.

In previous research, informative arguments, with scientific or empirical data, are

perceived to be valid and more persuasive. However, some scholars point out that

14

testimonials are more advantageous than purely informative messages in three major

aspects: enjoyment of the messages, ease of reading and vividness. All of them might

serve as heuristic cues increasing persuasiveness when the message is peripherally

processed (Braverman, 2008). Tversky and Kahneman (1973) provided an

explanation for this persuasive effect with the notion of availability heuristic.

According to this concept, the testimonials with vividly presented information, such

as a personal case history, comes to mind faster, and assist recipients to construct a

scenario more easily. The ease of imagination increases individual’s estimate of the

likelihood of encountering the same problem (De Wit, Das, & Vet, 2008; Tversky &

Kahneman, 1974). To the contrary, people have more difficulties in processing

informative messages with abstract data, like percentages and probabilities (Brosius &

Bathelt, 1994), so their risk perception is less likely to increase. In a word, the

different degrees of enjoyment, ease of reading and vividness make informative

arguments and testimonials vary in their comprehensibility and persuasiveness.

However, according to ELM, the persuasive effects of informative messages and

testimonials depend on other variables, for example, motivation or personal

involvement.

Containing a variety of peripheral and heuristic information, like vivid examples

and emotional appeals, testimonial messages have more direct impact on individuals

who are less motivated to elaborate the content in depth (Petty et al., 2002). In

contrast, to people who are concerned about their current state and are more likely to

scrutinize the messages, a testimonial story with a single source rather than

15

aggregated statistical data is less convincing. Braverman (2008) found that

participants who reported high readiness to change their diet were more persuaded by

the informational message, whereas those who did not care about their diet were more

convinced by the testimonial message. In addition, informational message about

alcohol drinking was more convincing for the highly involved participants, who

regarded it important to change their drinking habit, than for the low-involved

participants. In contrast, the testimonial message was more convincing to the low

involved participants than to the high-involved participants.

Moreover, initial attitudes might interact with message types to affect recipients’

perception of the quality or validity of the persuasive messages. For example, Slater

and Rouner (1996) discovered that alcohol education messages with statistical

evidence are evaluated to have higher quality by respondents whose initial values

were congruent with the message content. However, to those whose initial values

were incongruent with the message, messages with anecdotal evidence were perceived

to have higher quality. Therefore, testimonials are especially effective at reducing

counterarguments overcoming resistance to the persuasive messages (Braverman,

2008).

Expert vs. Celebrity vs. Ordinary person

In the transitional stage between childhood and adulthood, adolescents attempt to

seek autonomy from their parents and explore a new identity. As they spend more

time outside the family, the opinions and advice of others become more influential.

Adolescents imitate the attitudes, values and life styles of the role models in their

16

social environment (Chan et al., 2009). Modeling is one of the most powerful means

of transmitting patterns of thought and behavior (Bandura, 1986).

Then who is the most influential role model for adolescents? Whose opinion is

the most trustworthy to adolescents? According to Steinberg (1999), adolescents trust

different people in different situations. On issues requiring specific objective

information, teachers and other adults who are likely to have the necessary knowledge

are perceived to be the authorities. When it comes to social matters, such as styles of

dress and choices of leisure activities, adolescents are more likely to conform to their

peers’ opinions. Parents are the advisers of questions of values, ethics, and educational

plans. However, it is worthy to note that the perception of adult authority change over

the course of adolescence. Preadolescents are more likely to rely on parents and

teachers on a wide range of decisions. This adult authority is questioned over time

when adolescents begin to see some issues as “personal” and out of “jurisdiction” of

adults (Smetana & Bitz, 1996).

Mass media presents different kinds of role models or authorities in social issues.

The current PSAs of nutrition label mainly use three kinds of spokespersons: one is an

expert, like a dietitian, another is a celebrity, like a pop singer in the youth culture, the

other is an ordinary person, like a mother in a family or an office lady. Therefore, this

study primarily examines the effects of these three sources.

An expert has been traditionally defined as a source who knows the correct stand

on an issue (McGuire, 1969), and makes statements which have been verified

empirically (Birnbaum & Stegner, 1979). The perceived validity of the information

17

provided by an expert produces greater agreement with the subject than the same

information attributed to a non-expert (Tedeschi, 1972).

The effectiveness of an expert spokesperson is generated through internalization,

a process that individuals conform to the attitudes or behavior of the spokesperson. It

occurs when the message is credible, self-relevant and congruent with the recipients’

own value system.

Although adolescents perceive the messages from adults as correct, they tend to

resist them. For example, a health advice given by a physician, who is considered an

authority, might cause young people and minorities to avoid the message because they

preconceive the message to be boring and contain complicated information they do

not relate to (Salmon & Atkin, 2003).

In public service announcement, celebrity spokespersons often take on a

testimonial role to increase adolescents’ intention to quit high-risk behaviors, such as

drunk-driving, drug use and unprotected sex (Shead et al., 2011). Their persuasive

power rests on some characteristics, such as attractiveness, likeability, familiarity and

trustworthiness (Freiden, 1984; Shead et al., 2011; Till & Shimp 1998). McCracken

(1989) pointed out that a celebrity endorser brings a distinguishing feature in terms of

personality and lifestyle meanings to an endorsement process. In health campaign,

audience’s affect of the preconceived image of the celebrity can be transferred to the

theme of the key message. In this case, the attitude change of audience occurs through

the process of identification that an individual attempts to establish or maintain the

identity associated with a celebrity, so as to enhance his own self-image (Kelman

18

1961).

In health campaign, a celebrity with a high level of connection to a health issue is

more likely to influence people with lower involvement (Petty & Cacioppo, 1986).

For example, in a study of youth gambling (Shead et al., 2011), a celebrity-endorsed

gambling prevention message was more effective to youth with relatively low

gambling involvement.

However, some adolescents may suspect that celebrities are paid to deliver the

message, so they are not trustworthy.

Compared to an expert or a celebrity, an ordinary person would be more likely to

increase the self-relevance of a message to adolescents. For example, since family is a

major food provider for adolescents and parents usually give health advices, a visual

picture with a spokesperson that looks like a mother delivering a message about

nutritious diets may be more familiar to adolescents. Adolescents who conform to

parents’ opinion may accept the appeal, while those who strongly doubt parents’

authority may reject the message. If the ordinary person is similar to adolescents in

one way or another, for example, the age, adolescents tend to process the message,

because they think the message is relevant to their peer group and is cared about by

other adolescents (Brin˜ol & Petty, 2009). According to Mackie, Gastardo-Conaco,

and Skelly (1992), a communication from a same-category source might be seen as

more reflecting, defining, and informing about social reality for people similar to the

recipient, so it is more persuasive.

Among the three kinds of sources mentioned above, who is the most effective?

19

According to the match-up hypothesis (Kahle & Homer 1985), the effectiveness of an

endorser is contingent on the type of product advertised or the issue discussed. For

example, attractive endorsers are more effective when promoting beauty products.

This strong match serves as a source of information regarding the value and utility of

the product. Likewise, it is better to use a spokesperson who can be readily identified

with the issue being presented. A female celebrity with an attractive figure may be

effective to promote physical activity when the audience agrees that doing sports

contributes to a good-looking body.

Biswas et al. (2006) discovered that for high technology–oriented products (for

example, a computer in their study), an expert as the endorser is more effective in

reducing perceived risk than a celebrity or an ordinary person. This effect is further

magnified on consumers who are highly knowledgeable about the product. For low

technology–oriented products (for example, a treadmill in their study), the differential

effects of the three kinds of endorsers do not exist. From the perspective of ELM,

high-knowledge consumers who have basic knowledge of the attribute information

relate the endorser’s messages to their original knowledge framework. The identity of

the endorser and the congruence between the endorser and the product serve as a

dimension of the argument. Consumers are actually going through the central routes

to scrutinize the advertisement and make a decision. In contrast, consumers with less

product knowledge rely on whatever form of the peripheral cues as diagnostic signals

to analyze the benefits associated with a product (Rao & Monroe 1988).

Based on the match-up hypothesis, not only the match of spokesperson and

20

product or issue images has been tested, but also the match of spokesperson and

message appeals. Perse, Nathanson and McLeod (1996) found that the safe-sex PSA

using a female spokesperson in the rational appeal generated the most positive

emotional attitudes, while a female spokesperson in the emotional appeal generated

the least favorable emotional attitudes.

With ELM as the major framework, this study attempts to examine if issue

involvement, which is indicated by health consciousness in the current context,

interacts with the match of spokesperson and argument to affect attitude.

Source credibility

The early learning theory paradigm illustrates that communicators’ influence

derives from the incentives they provide, which are based on some source

characteristics, such as credibility, attractiveness and similarity. These characteristics

motivate individuals to process the message (Kruglanski, 2005). The current

investigation focuses on one of the most widely studied characteristics: source

credibility.

The discussion of source credibility is guided by the source-credibility model,

dating back to a landmark study by Hovland, Janis and Kelley (1953), which contends

that source credibility is underscored by two factors: expertise and trustworthiness.

Expertise refers to “the extent to which a communicator is perceived to be a

source of valid assertions” (Hovland, Janis, & Kelley, 1953). It is topic-specific. An

expert must possess special knowledge or experience on an issue, but he doesn’t

necessarily have expertise in another field. In nutrition education, an accredited

21

dietitian is likely to be regarded as an expert, because the title of “accredited dietitian”

indicate that he or she must have fulfilled the minimum academic and professional

requirements to qualify for the credential. According to a meta-analysis of 114 articles

about source effect (Wilson & Sherrell, 1993), expertise had a stronger effect on

persuasion than other types of source manipulations. It may be explained by the

methodological artifact that expertise is often established on some objective basis and

may be easier to assess compared to other dimension of source credibility, such as

similarity or trustworthiness.

Trustworthiness is “the degree of confidence in the communicator’s intent to

communicate the assertions he considers most valid” (Hovland, Janis, & Kelley,

1953). As a clear cue to validity, trustworthiness influences the extent of elaboration.

A source high in trustworthiness engenders a nonthoughtful acceptance of the

message, because the recipient trusts the source in providing valid information. In

contrast, a source with low or questionable trustworthiness increases the amount of

attention to the message, because the recipient is unsure whether the source is

providing accurate information and thinks it necessary to scrutinize the message.

(Priester & Petty, 2003).

Apart from the above two dimensions, source credibility has been

operationalized as including more dimensions, such as attractiveness, likability,

sociability, and so on. Ohanian (1990) developed a valid instrument to measure

celebrity endorsers’ credibility, encompassing the dimensions of trustworthiness,

expertise and attractiveness. The dimension of attractiveness refers to physical

22

attractiveness. This scale has been widely adopted and will be modified and used to

evaluate the credibility of the communicators in the current study.

Early empirical studies demonstrated that high-credibility sources tend to be

more likely to let recipients accept the message (e.g., Hovland & Weiss, 1951).

However, in the dual-mode paradigm, accumulated contemporary research suggests

that the effect of credibility is not completely straightforward, but depends on other

factors, which can be divided into two general categories: factors influencing the

magnitude of effects, and factors influencing the direction of effects (O’ Keefe, 1990).

The magnitude of the credibility’s effect is affected by receivers’ involvement

with the issue. When exposed to issues of little personal relevance, receivers may just

let the source’s superficial credibility shape their opinions, rather than making efforts

to process the arguments (Wilson & Sherrell, 1993). To the contrary, for highly

involving issues, receivers tend to scrutinize the source’s arguments and evidence, and

thus become more engaged in the messages.

In terms of the direction of credibility effect, increasing credibility may increase

or decrease persuasiveness. It depends on whether the advocated position is initially

opposed or agreed with by the receiver. For a proattitudinal message, a low-credibility

source appears to have a persuasive advantage over the high-credibility one, because

receivers believe that the high-credibility source would do a good job to defend their

views, so they don’t need to think closely about the messages. The other way around,

they back up the low-credibility source by scrutinizing the supporting arguments.

However, for a counterattitudinal message, it is still the highly credible source who

23

achieves a greater success (O’ Keefe, 1990).

According to ELM, credibility may interact with argument quality to influence

attitudes, and the interaction effect depends on the perceived relevance of the message

to the audience. For example, when unambiguous strong or weak arguments were

presented, source expertise had a greater impact on attitudes towards an unimportant

topic rather than an important topic. In contrast, when arguments are not clearly

strong or weak, an expert source was more persuasive than an inexpert source under

both high and low importance conditions (Chaiken & Maheswaran, 1994).

The effect of source credibility is based on the validity of credibility heuristic

that highly credible source present strong arguments and low credible source present

weak argument. However, when the validity of credibility heuristic is undermined by

message content, for example, expert sources presenting weak arguments, or

non-expert sources presenting strong arguments, systematic processing, or central

processing would attenuate the judgmental impact of heuristic or peripheral

processing (Chaiken & Maheswaran, 1994).

Source credibility is not the intrinsic property of a source, but the perception of

message recipients, so it is influenced by many recipient variables, for example, age.

People in different lifespan developmental phases may perceived different kinds of

source as credible in specific domains. For example, Bar-Tal et al. (1991) found that

as people grow up, they perceive their friends as a more reliable source of knowledge

of physical appearance and personal feelings. Meanwhile, expert and media are

important source of science knowledge to people aged 17 to 18. It is argued that

24

individuals in their late childhood and early adolescence begin to separate from their

family and establish self-identity and independence when engaging in youth culture. A

study of pop singers (Raviv et al., 1996) showed that adolescents aged 10-11, 13-14,

and 16-17 relied on singers’ opinions more in personal matters (including future

panning, social relations, physical appearance, etc) than overall knowledge and values

(including politics, values, etc). This reliance declined with age.

Initial attitude

This study is interested in the attitude change of adolescents after exposure to the

PSA. Therefore, initial attitude is measured.

Initial attitudes influence resistance to attitude change, biasing cognitive

processing in a manner that favors attitudinally congruent information (Eagly &

Chaiken, 1998). Information supporting one’s initial attitudes is easier to be learned

and remembered than information contradicting his attitudes (for a review, see Eagly

& Chaiken, 1992). Initial attitudes towards an object or issue may contribute to

selective exposure and attention, selective perception and judgment, and selective

memory. Some people ignore discrepant information and pay more attention to

congruent information, because the former one is less connected to stored knowledge

about the object and is more difficult to be retrieved from memory. In contrast, other

people process discrepant information more actively and try to refute it, and thus

remember more about it (e.g., Cacioppo & Petty, 1979). It may be attributed to the

attitudinal ambivalence triggered by the messages with evaluatively inconsistent

aspects. Such ambivalence decreases perceived confidence in one’s own attitude,

25

creates a resultant gap between desired and actual confidence and evokes an increase

in systematic processing.

The effectiveness of persuasive messages is contingent on the relation between

the message itself and recipients’ initial attitude. Slater and Rouner (1996) suggested

that a statistical message was more persuasive when it was consistent to initial attitude.

In contrast, when the message was inconsistent to the initial position of the receiver,

narrative evidence was more effective.

This study explores the extent to which initial attitude can predict the level of

elaboration of persuasive messages and the intention to conduct health behaviors.

Self-efficacy

According to a survey done by CFS, more than 50% of respondents think

nutrition labels are useful, but only 15% of them read the labels when choosing food.

Many consumers are confused with the format or figures on the nutrition (CFS, 2010).

In spite of the perceived benefits of nutrition labels, they are not sure of the reference

daily intake of nutrients and think it difficult to understand the information on food

labels. As a result, they preferred not to waste time to read the label when choosing

food. In those countries where mandatory nutrition labelling has a longer history,

many consumers also find it difficult to comprehend the quantitative information

presented on labels, especially the recommended daily amounts, percentage of daily

values, and serving sizes (Campos, Doxey, & Hammond, 2011).

Previous research has identified other predictors of intention to execute a healthy

behavior, such as subjective norms, perceived behavioral control, and risk perception,

26

etc. Based on the finding of the administrative research mentioned above, this study

focuses on the effect of self-efficacy on adolescents’ intention to use nutrition labels.

According to Bandura’s Social Cognitive Theory (1977), self-efficacy is the

belief about one’s own capacities to produce desired changes by taking particular

action. It is a focal determinant of personal change because it affects individuals’

goals and aspirations. People with stronger perceived self-efficacy set higher goals for

their behaviors and have firmer commitment to them. Self-efficacy beliefs also

determine how obstacles and impediments are perceived. In the face of difficulties,

people of low efficacy are easily convinced of the futility of effort and give up trying,

while those of high efficacy make perseverant effort to improve self-management

skills to overcome the difficulties (Bandura, 2004). Perceived self-efficacy can

influence all aspects of a behavior, including the acquisition of new behaviors,

inhibition of existing behaviors, and disinhibition of behaviors. For example, Ewart et

al. (1983) discovered that changes in efficacy scores after treadmill exercise testing

predicted both the duration and intensity of subsequent self-reported home activity. In

Strachan and Brawley ‘s (2009) observational study, participants who had a higher

level of self-efficacy belief to consume more fruits, vegetables and food of high

nutritional value were more likely to keep a healthy eating diet.

The concept of self-efficacy has been integrated into other preventive health

behavior models. For example, the Extended Parallel Process Model (EPPM, Witte,

1992) postulates that individuals process the threatening messages by appraising the

threat of the message and by evaluating the perceived efficacy of the recommended

27

response. For efficacy, the EPPM emphasizes two different types: response efficacy

and self-efficacy. Response efficacy pertains to beliefs about the effectiveness of the

recommended response in deterring the threat. Self-efficacy is the perception of one’s

own ability to perform the recommended response (Gore & Bracken, 2005, Witte et

al., 2001). According to EPPM, the messages processing results in one of the three

outcomes: no response; or acceptance; or rejection of the message (Witte et al., 2001).

When adopting the model to study the processing of television health news stories,

Hong (2011) found that the level of perceived efficacy to take the preventive actions

determined how much individuals accept or reject the messages when they realized

the unhealthy condition was a severe health problem.

The concept of self-efficacy belief overlaps with some determinants in other

psychosocial models of health behavior, but under different names. For example, the

construct of perceived behavioral control in the Theory of Planned Behavior (TPB) is

compatible with self-efficacy. It refers to one’s perception of the ease or difficulty of

performing the behavior (Ajzen, 1991). In the framework of TPB, perceived

behavioral control operates together with attitudes and subjective norms in the

regulation of human intention to execute causes of action.

In a word, both practical and theoretical research have specified the strong effect

of self-efficacy on behavioral change, so this study places this construct within a more

general framework of the relations among information processing, knowledge gain,

attitudes and intention.

28

Chapter3:

TheoreticalFramework

The Elaboration Likelihood Model of Persuasion

Focusing on the information processing route of adolescents, which is influenced

by such factors as health consciousness, argument, source and initial attitude, this

study is guided by the Elaboration Likelihood Model of Persuasion (Petty & Cacioppo,

1981, 1986). Elaboration, in a persuasion context, refers to the extent to which a

person thinks about the issue-relevant arguments contained in a message. Petty and

Cacioppo (1981, 1986) proposed a continuum of message elaboration, along which

different attitude change processes may operate. By the high end of the continuum,

“central route” to persuasion occurs, indicating that people are likely to scrutinize and

elaborate upon the message arguments in light of the associations available from

memory, and draw inferences about the merits of the arguments for a recommendation

based upon their analyses. Consequently, people derive on overall attitude toward the

recommendation.

By the low end of the continuum, in contrast, “peripheral route” is taken, under

which people change their attitude by associating the issue or object with positive or

negative cues or by making a simple inference about the merits of the advocated

position based on various simple cues (Petty & Cacioppo, 1981, 1983, 1986). The

simple cues, such as the attractiveness or credibility of the source, provide means of

maximizing the likelihood that the communicator’s position is correct while

minimizing the cognitive efforts to achieve this position (Scott, 1995). In a word, the

29

central route activates long term message processing with extensive thought and

reflection, while the peripheral route activates shortcuts with very little thought.

The two different routes of processing are not mutually exclusive. The degree of

elaboration varies along the continuum. O’ Keefe (1990) pointed out that there is a

trade-off between peripheral cues and elaboration in affecting the persuasive

outcomes. That is to say, the relative impact of elaboration and peripheral cues will

vary as elaboration likelihood varies. As elaboration likelihood increases, the impact

of peripheral cues declines (but doesn’t disappear) and the impact of the receiver’s

issue-relevant thinking increases. With lower elaboration likelihood, persuasive

effects depend more and more on peripheral cues.

The amount of issue-relevant thinking or elaboration about an object determines

the strength of an attitude towards it. Central route processing generates attitudes that

are postulated to be stronger, more resistant to change and more predictive of behavior

compared to those generated via peripheral route (See Figure3.1).

While central route is more predictive of attitudinal persistence, resistance, and

attitude-behavior consistency, it should be noted that the persuasiveness of different

routes is based on elaboration likelihood. The basic assertion is when the elaboration

likelihood is high, central route predicts persuasion the best; when the elaboration

likelihood is low, peripheral route predicts persuasion the best (Petty, Haugtvedt, &

Smith, 1995). Therefore, in the publicity of nutrition label, practitioners should first

have an insight of the likelihood that audience would elaborate on this issue, and then

design the messages with specific components in either processing route. In order to

30

encourage persistent use of labels, central route processing should be induced.

The likelihood of elaboration is determined by two factors: motivation and ability.

When people are highly motivated and able to process the messages, central route

persuasion is triggered. Many situational and individual variables can increase either

motivation or ability. Regarding motivation, personal relevance, need for cognition,

and accountability for one’s attitude judgments can predict high motivation. And

variables such as distraction, message repetition and affective state, are predictors of

ability.

Among the variables, issue involvement, or personal involvement, has long been

recognized as a major influence on the cognitive strategy to process persuasive

communication. Based on different definitions of this construct, Salmon (1986)

concluded that involvement encompasses a family of constructs, including salience,

relevance, perceived risk, attention, elaboration and audience activity.

As a research on preventive health behavior interventions, this study uses a factor

called health consciousness as a proxy for issue involvement, or involvement in health

issue.

The other predictor of elaboration routes is ability to process the information,

which is affected by the message’s nature such as comprehensibility and repetition,

situational factors like environmental noise, and personal differences like prior

knowledge (Cacioppo & Petty, 1989; Petty & Cacioppo, 1983, 1984). Generally

speaking, high ability predicts central processing, while low ability predicts peripheral

processing.

31

According to ELM, variables can affect the amount and direction of attitude

change in four fundamental processes: (a) serving as persuasive arguments, (b)

biasing the ongoing thinking when processing is high, (c) serving as peripheral cues

when thinking is low, (d) affecting the level or direction of argument-relevant

elaboration when thinking is unconstrained to be particularly high or low (Petty &

Cacioppo, 1986; Petty & Wegener, 1998).

Central to the ELM is the construct of argument quality, referring to a recipient’s

perception that a message’s arguments are strong and cogent as opposed to weak and

specious (Eagly & Chaiken, 1993). According to Petty and Cacioppo (1986), strong

arguments give rise to predominantly favorable thoughts about an issue, while weak

arguments evoke predominantly unfavorable reactions. Therefore, many studies

defined argument quality empirically, on the basis of observed effects under

conditions of high elaboration. However, this effect-based definition of argument

quality doesn’t specify what message properties make an argument strong and

persuasive.

Eagly and Chaiken (1993) pointed out that the latent variable of argument quality

is actually the inherent strength, plausibility, or believability of the form and content

of an argument.

Petty et al. (1981) described the strong messages as those providing persuasive

evidence (statistics, data, etc.), and the weak ones as those relying on quotations,

personal opinion and examples. According to this clarification, the two arguments

under investigation here, which are informative argument and testimonial argument,

32

are supposed to be strong message and weak message respectively.

Another variable that has been tested in many ELM research and manipulated in

most health campaign is source characteristic. When people are not motivated or not

able to process the message, the source or spokesperson can attract people’s attention

to read the message and associate the issue with the source and make a simple

inference about the merits of the advocated position. In addition to serving as a simple

rejection or acceptance cue, the source variable can be processed as arguments and

analyzed for its evidentiary. When a person is highly motivated and able to think

carefully about the merits of an appeal, an attractive source, for example, would be

analyzed for its relevance and cogency (Petty & Brin˜ol, 2008).

The ELM also identifies another role of source variable: affecting the amount of

thinking when thinking is not constrained to be high or low by motivation or ability.

For example, when people perceive a source to be similar to them, they evaluate the

message to be more self-relevant, and think it more carefully. In addition, when the

source is regarded as untrustworthy or surprising, he or she would induce a sense of

doubt or uncertainty, and thus increase careful thinking (Brin˜ol & Petty, 2009).

33

Figure 3.1.Central and peripheral routes to persuasion. This figure depicts the two anchoring endpoints on the elaboration likelihood continuum (adapted from Petty & Cacioppo, 1981, 1986).

NATURE OF COGNITIVE PROCESSING (initial attitude, argument quality, etc.)

FAVORABLE THOUGHTS PREDOMINATE

UNFAVORABLE THOUGHTS PREDOMINATE

NEITHER OR NEUTRAL PREDOMINATE

CENTRAL POSITIVE ATTITUDE CHANGE

CENTRAL NEGATIVE ATTITUDE CHANGE

Yes (Unfavorable)

Yes

Yes (Favorable)

No

PERSUASIVE COMMUNICATION

MOTIVATED TO PROCESS (personal relevance; need for cognition,

personal responsibility, etc.)

ABILITY TO PROCESS (distraction; repetition; prior

knowledge, message comprehensibility, etc.)

PERIPHERAL ATTITUDE SHIFT

Attitude is relatively temporary, susceptible, and unpredictive of

behavior

COGNITIVE STRUCTURAL CHANGE:

Are new cognition adopted and stored in memory? Are different responses

made salient than previously?

Attitude is relatively enduring, resistant, and predictive of behavior

PERIPHERAL CUE PRESENT?

Positive/Negative Affect; attractive/negative sources; number of arguments, etc.

RETAIN OR REGAIN INITIAL

ATTITUDE

Yes No

No

Yes

No

34

Chapter4:

ResearchquestionsandHypothesis

ELM postulates that individuals are more likely to engage in central route

processing if they have a high level of personal involvement in an issue. In that case,

they will consider the true merits of the issue-relevant information in the message. In

contrast to a testimonial argument with only a case of another person that may not be

generalized to everyone, an informative argument with scientific evidence seems

more convincing to them. Therefore, people with a high involvement in healthy eating,

who are thus regarded to have high health consciousness, are expected to elaborate on

informative arguments. And their careful scrutinizing of the benefits of nutrition

labels is expected to have a greater effect on positive attitude change. Adolescents

who are less health conscious don’t have the motivation to read the substantive

arguments carefully, so they only rely on simple cues to determine how much to think

about the message. Source credibility is one of the cues. Dietitians as nutrition experts

may be perceived to have higher expertise and trustworthiness than celebrities or

ordinary people. Therefore, a dietitian is more influential to low health conscious

adolescents.

H1: Only under high health consciousness conditions, an informative argument

generates significantly more: (a) perceived knowledge gain, (b) positive attitudes

towards nutrition label use and (c) behavioral intention, than a testimonial argument.

H2: Only under low health consciousness conditions, a dietitian as the

peripheral cue generates more: (a) perceived knowledge gain, (b) positive attitudes

35

towards nutrition label use and (c) behavioral intention than either a celebrity or an

ordinary person.

The current research intended to test the effect of ability to process the created

PSA. However, the manipulation check showed no significant difference in subjects’

ability to read the PSA across the experimental conditions. There was not a significant

variation in the comprehension of the message. Therefore, the primary determinant of

the processing route that was actually tested was health consciousness.

Experts are expected to give scientific evidence to prove the importance of

reading nutrition labels. As an authority of nutrition, when they share their

professional knowledge, the message becomes more credible. If they share their own

experience instead of knowledge, it may be less persuasive, because adolescents may

think the experience inapplicable to them. After all, dietitians are supposed to have

healthier eating habits, but ordinary people are not necessarily able to follow their

example. In contrast, celebrities in this study, who are young singers popular among

teenagers, share more similarities with adolescents and may be perceived to have the

similar preference for food. If they claimed to keep a healthy eating habit, adolescents

would be more confident in their own ability to adopt the same health behavior.

Compared to an ordinary person, a celebrity can be more effective because she brings

a distinguishing feature in terms of personality and lifestyle meanings to the

recommended behavior (Biswas et al., 2006). Adolescents have a preconceived

impression on the celebrity and this image affect is transferred to the health behavior.

H3: An informative argument is more persuasive when it is made by a dietitian,

36

while a testimonial argument is more persuasive when it is made by a celebrity.

ELM postulates that under central route, people pay attention to arguments, and

are not affected by the characteristics of the source. But previous studies suggested

that source credibility could also become one component of the argument and change

the persuasiveness of the message. Therefore, I come up with the question:

RQ1: To adolescents with either high or low health consciousness, is there a

significant difference in the persuasiveness of the match of argument and source?

The ELM model holds that attitude formed via central route is more predictive of

consistent behavior.

H4: After exposure to the PSA, adolescents with a more favorable attitude to

nutrition labels are more likely to use them.

It is argued that one single traditional model is not enough to understand the

complexity of the mechanism that influences preventive health behavior (Jayanti &

Burns, 1998). ELM doesn’t point out if there are other variables to facilitate the

attitude-behavior consistency. Practical survey demonstrates that not all people who

have a positive attitude towards nutrition label take actual action to use it. One of the

barriers is their self-efficacy to understand the label. A nutrition label shows the

quantity of several nutrients in each serving size. But usually the actual quantity of the

nutrient needs to be calculated according to the weight of the product. However, many

consumers are not able to do the calculation, neither are they willing to spend time

doing it.

To examine how adolescents’ self-efficacy to understand nutrition labels and to

37

make use of the nutrition information predicts their intention, this study extends the

ELM model by integrating the construct of self-efficacy, which was extracted from

Social Cognitive Theory and EPPM. It is assumed that self-efficacy added

significantly to the prediction of intention to use nutrition label.

H5: Adolescents who have a higher level of self-efficacy to make use of nutrition

labels are more likely to use them.

Previous research found that information congruent with initial attitudes is easier

to be processed and remembered. In the current study, information in all PSAs is

advocating nutrition label use. I will investigate the effect of adolescents’ initial

attitudes on the persuasiveness of the message, to see if this effect is even stronger

than the effect of the PSA itself.

RQ2: To what extend can initial attitude affect the effectiveness of the PSAs?

Figure 4.1 depicts the modification and application of ELM in the current study.

38

Figure 4.1. Modification of the ELM in the Processing of PSAs of Nutrition Label Use. Constructs and arrows in dashed line are not tested in this study.

DIETITIAN

CELEBRITY

ORDINARY PERSON

INFORMATIVE ARGUMENTS

TESTIMONIAL ARGUMENTS

H2

Low High

Yes

Yes

No

PUBLIC SERVICE ANNOUNCEMENT OF NUTITION LABLE USE

HEALTH CONSCIOUSNESS

(motivated to process)

ABILITY TO PROCESS (distraction; repetition; prior

knowledge, message comprehensibility, etc.)

PERIPHERAL ATTITUDE SHIFT

Attitude is relatively temporary, susceptible, and unpredictive of

behavior

PEICEIVED KNOWLEDGE GAIN Are new cognition adopted and stored in memory? Is the information useful?

PERIPHERAL CUE: SOURCE CREDIBILITY

Yes

No

SELF-EFFICACY TO USE

NUTRITION LABELS.

BEHAVIORAL INTENTION

(INTENTION TO USE NUTRITION

LABELS)

H1

H5 H4

INITIAL ATTITUDE

H3

ATTITTDE IS RELATITIVELY

ENDURING, RESISTANT, AND

PREDICTIVE OF BEHAVIOR

39

Chapter5:

Methodology

This chapter includes detailed description of the design, procedure, materials and

measurements of this experimental research.

This study aims to examine what kind of argument (informational or testimonial)

and source (dietitian or celebrity) in a PSA has a stronger power to persuade

adolescents to use nutrition labels. A 2 (argument: informative vs. testimonial) x 3

(source: expert vs. celebrity vs. ordinary person) factorial experiment was conducted.

Before the experiment, a pilot test had been undertaken to develop the PSAs for the

experiment.

A total of 134 students from CUHKFAA Chan Chun Ha Secondary School

participated in the pilot tests. After that, one informative statement and one

testimonial statement, both with the highest rating of persuasiveness were selected.

According to the rating of source credibility, two dietitians and two celebrities were

chosen as the sources of the statement. The description of the PSAs is given in the

part of “Materials and Manipulations”

Pilot tests data indicated a significant difference in the perceived persuasiveness

of the arguments (informative: M=3.27, SD=.86; testimonial: M=2.91, SD=.94) on a

5-point scale, t (132) =2.25, p<.03. No differences were shown in comprehensibility

of the message. The dietitian (M=3.59, SD=.79) was rated as more credible than the

celebrity (M=2.72, SD=.51), t (132) =6.43, p< .001.

40

Subjects and Design

On April 2, 2012, a total of 169 students at the Po Leung Kuk Mrs Ma Kam

Ming-Cheung Fook Sien College participated in the experiment as an exercise of a

General Education course about social science research. 76 boys (45%) and 93 girls

(55%), aged from 12 years to 15 years old (M=13.67, SD=.65) participated. They

were randomly assigned to each of the cells in a 2 (argument type: informative vs.

testimonial) x 3 (source: expert vs. celebrity vs. ordinary person) factorial design.

Procedure

Three booklets were prepared for the study. The first contained the pre-test

measures regarding participants’ initial attitudes towards nutrition label, their health

consciousness and perceived ability to process the label and demographic information.

The second contained the PSA stimuli. And the third booklet contained the evaluation

of the credibility and attractiveness of the source and dependent measures: perceived

knowledge gain, attitude, and behavioral intention.

Upon arrival, participants were told that the current research was about the habit

of nutrition label use among the public. There was no sample answer to each question

and no identifiable information would be asked for. Therefore, participants were

encouraged to provide their own real opinions. The entire research consists of three

parts, lasting for about 20 minutes.

After the brief introduction, the first one-page questionnaires for pre-test were

distributed. The participants spent three minutes filling out the questionnaire and then

returned it to the experimenter. Then each of the participants received one of the six

41

versions of the second booklet, a color-print PSA. They were asked to spend one to

two minutes reading the PSA. After all the print PSAs were returned, the third booklet,

a questionnaire for post-test was given to each participant. Upon completion, the

participants were debriefed, thanked with a small gift and dismissed.

Materials and Manipulations

Arguments. Two versions of the persuasive arguments were created. The

informative version was written in an impersonal manner, with scientific and survey

data.

Nutrition label lists out the energy content and values of seven nutrients in per

serving of food, which can help consumer make informed choices to meet

individual need. For example, there is 30g to 43g of sugar in a 350ml soft drink,

but for most adolescents, the daily intake upper limit of sugar is 50g. In Hong

Kong, there have been 17% of secondary school students suffering from obesity,

which imposes problem to their study and life. It is beneficial for them to use

nutrition labels to choose low sugar food, so as to reduce the risk of developing

chronic diseases and stay healthy to study more efficiently.

Much content of the message, including the data was taken from the official

website of Centre for Food Safety (http://www.cfs.gov.hk/). Precautions were taken to

make sure the information was objective, without deceptive and misleading content.

The testimonial version was written in the tone of first person, with personal

experience.

42

When choosing prepackaged food, I take note of the nutrition labels on the

package, to make comparison among the content of energy and other nutrients of

various products. Some of my relatives and friends suffered from chronic diseases,

overweight or dental caries. Thanks to nutrition labels, we are able to choose food

lower in sugar, total fat and calcium, so as to reduce the risk of developing

diseases and stay healthy. See, now we have a fit body to work energetically and

efficiently.

Source. In the “expert” condition, a picture of a real dietitian was attached next to

the argument, with a caption indicating her identity. Two accredited dietitians were

used in the study, so there were two versions of the captions: one is “Lam See Way,

accredited dietitians of Hong Kong Dietitian Association, Master in Nutrition Therapy,

University of Sydney University.” And the other one is “Chan King Chi, accredited

dietitians of Hong Kong Dietitian Association, accredited dietitians of the United

States.” The personal information was taken from the website of Hong Kong Dietitian

Association (http://www.hkda.com.hk/) and verified with media coverage. The

pictures were carefully chosen according to the pixel, dress code and facial expression

of the dietitian. In either picture, the dietitian is smiling and having eye contact with

the viewer. Miss Lam is dressed in a black suit, while Miss Chan in a white doctor

gown. The preference for formal dress aims at highlighting the profession and

expertise of the source. To catch attention, the headline above the picture read “The

dietitian tells you…”.

43

In the “celebrity” condition, the PSA featured a picture of a real celebrity, with a

caption indicating her identity. Two local emerging singers, G.E.M Tang and Candy

Wong were showed as the source in either version. Both of them were active in media

publicity and participated in many activities that took place on campus. In addition,

their images were perceived to be relatively healthy and pretty. The pilot test also

indicated that they were popular among the secondary school students. In either

picture, G.E.M or Candy is dressed in casual clothes, smiling at the viewers. To catch

attention, the headline above the picture read “G.E.M tells you…” or “Candy tells

you…”.

In the “ordinary person” condition, another picture of the dietitian Miss Lam was

shown, with no caption or headline. In the picture, Lam is dressed in casual clothes

with a plate full of salad in her hand. With average looking unfamiliar to the

participants, she is likely to be perceived as an ordinary person without expertise in

nutrition.

In addition to the arguments and a picture of the source, the slogan of the “Live it,

use it” campaign, “Make better choices, use nutrition label”, was shown at the top of

each print PSA. Besides, a sample of nutrition label was put in the middle, above the

main arguments. And the bottom of the PSA carried the logos of the campaign, Food

and Environmental Hygiene Department, and Center for Food Safety, together with

the website of campaign (www. nutritionlabel.gov.hk). The format of the PSA was

similar to the actual one released by the current campaign.

44

Independent variables

Health consciousness was measured with five items, which were derived from

previous studies, regarding individuals’ concern about health in general and

specifically food consumption. The five modified items included:

Are you aware of the state of your health? (Gould, 1988)

Are you interested in the health information about how to avoid health hazards

and keep healthy? (Hong, 2009)

Do you worry about getting overweight or obese? (Dutta-Bergman, 2004)

Do you take actions (e.g. keep a balanced diet, exercise, etc.) actively to stay

healthy? (Dutta-Bergman, 2004, 2006)

Do you monitor the daily intake of energy and some other nutrients (e.g. calcium,

sugar, total fat, etc.) in your food? (Furnham & Forey, 1994)

Participants were asked to indicate how much they agreed with each statement

on a 5-point Likert scale, from 1 (not at all) to 5 (extremely). Scores were averaged to

calculate an index of the health consciousness variable (Cronbach’s alpha=.658). A

median split divided the sample into the high (M=3.44, n=94) and low (M=2.35, n=74)

health-consciousness conditions, F (l, 194) = 302.65, p< .001.

Based on the classification, the technique of blocking was applied to transform

the 2x3 experimental design into a 2x2x3 design, with health consciousness as the

third factor.

45

Initial attitude towards nutrition label use was measured by three items

(Cronbach’s alpha =.768)

Is it necessary to attach nutrition labels to the package of food products?

Is the nutrition label useful for you to keep a balanced diet?

Is the nutrition label useful for you to avoid health hazards and stay healthy?

Three 5-point bipolar adjective scales (unnecessary-necessary, useless-useful,

useless-useful) were used to measure the items above.

Self-efficacy to process the information on nutrition labels was assessed by

asking respondents to rate the extent to which they understand the content of nutrition

labels and their patience in processing information on labels. The measurement can

also reflect the level of prior knowledge. The items are as follows (Cronbach’s alpha

= .739):

Do you know how to use nutrition labels to compare nutrient contents among

different foods to make a healthier choice?

Is it easy for you to understand the information on nutrition labels?

Do you feel annoying when reading nutrition labels?

The first items were rated on a 5-point scale ranging from 1(not at all) to 5

(extremely). And the second and third item was measured with two 5-point bipolar

adjective scales: 1(difficult) to 5 (easy), and 1(annoying) to 5 (ease).

Given the task and behavior specificity of self-efficacy, it is suggested that the

46

instrument of self-efficacy should be created for a specific health behavior studied

(AbuSabha & Achterberg, 1997). In the current research, the three items measuring

perceived ability to use nutrition label were created and tested for reliability and

validity in the pilot-test before it was used in the experiment.

Dependent variables

Manipulation checks

As a check on the effectiveness of the source manipulation, a set of questions

regarding credibility was asked. The questions were derived from Ohanian’s (1990)

scale and modified according to the context of nutrition label.

The ratings of source credibility included the following: (a) the level of her

knowledge about nutrition labels, (b) her execution of reading nutrition label, (c)the

extent to which she makes full use of nutrition label to make a healthy choice, (d)

expertise, (e) trustworthiness, (f) attractiveness of her looking, (g) beauty, (h) elegance.

All the eight questions were rated on 5-point scales ranging from 1 (not at all) to

5(extremely). The mean of the eight items formed an internally consistent index of

source credibility (Cronbach’s alpha =.877).

The dietitian (M=3.54, SD=.82) was rated as the most credible, followed by the

ordinary person (M=3.06, SD=.74) and the celebrity (M=2.92, SD=.87), F (2, 166)

=10.03, p<.001.

Perceived knowledge gain. Many investigators of ELM believe that one

determinant of whether or not attitude change would occur is individual’s cognitive

response to the information. It is measured by thought-listing technique, asking

47

participants to list all thoughts that occurred to them while they attended to the

stimulus (e.g., Cacioppo & Petty, 1979). The responses are classified into one of the

three dimensions: (1) polarity--- “favorable”, “neutral/irrelevant” or “unfavorable”

thoughts of the stimulus; (2) origin---thoughts originated from the information, acting

as reactions to the stimulus, or not traceable directly to the stimulus; (3) target---

thoughts pertaining to the issue concerned, the source, the recipients, or the channel

through which the stimulus is transmitted (Cacioppo & Petty, 1981).

The pilot test adopted the thought-listing method to measure cognitive response,

by asking participants to list whatever thoughts that come to their mind while reading

the PSAs. However, participants didn’t write down informative statements to indicate

their thoughts. Most of them just wrote a word “nutrition”, or recalled the names of a

few nutrients. One reason why adolescents didn’t elicit much thought may be they

didn’t feel personally relevant to the topic (Petty & Cacioppo, 1979). Another reason

is that adolescents haven’t developed a high ability to think about their own thought.

Steinberg (1999) pointed out that knowledge about one’s own thinking processes

improves during adolescence. Older adolescents are better able to explain the thinking

process they are using and to introspect their emotions.

The outcome of the pilot test indicated it may not be appropriate to measure

cognitive response of adolescents with thought-listing technique. Since early scholars

argued that comprehension of the argument in the message is also an important

determinant of attitude change (Hovland, Janis, & Kelley, 1953), in the experiment,

three items were used to measure the “perceived knowledge gain” after exposure to

48

the PSA, instead of cognitive response. This variable was equally assessed by three

items (Cronbach’s alpha =.747):

I learn a lot about nutrition labels from the PSA.

I understand the information on the PSA.

I find the information useful.

All the three questions were rated on 5-point scales ranging from 1 (strongly

disagree) to 5(strongly agree).

Post-attitude towards nutrition label use was assessed by the same three items

and scales which had been used to measure initial attitude.

Behavioral intention to use nutrition label

Behavioral intention was equally assessed by three items (Cronbach’s alpha

=.871):

Will you encourage more people to read nutrition labels when choosing

prepackage food?

Will you read nutrition labels when you choose prepackage food next time?

Will you remind your family to read nutrition labels?

The three items were rated on 5-point scales ranging from 1 (extremely unlikely)

to 5 (extremely likely).

49

Chapter6:

Results

Health Consciousness Assessment

The median split (median 3.00) of health consciousness (M=2.96, SD=.68) was

taken to define “high” and “low” health consciousness, resulting in 95 participants in

the group of high health consciousness and 74 participants in the group of low health

consciousness.

Since the three dependent variables were significantly correlated (ranging from r

= .46 to r = .74), and the three factors, health consciousness, argument and source

have more than two levels, multivariate analysis of covariance (MANCOVA) in SPSS

was conducted for statistical analyses. Bartlett’s test of sphericity (χ² (5) =100.82, p

< .001) indicated that a MANCOVA was appropriate.

General results

The results of the factorial MANCOVA revealed two significant multivariate

interaction effects on the effectiveness of the PSA. One is the interaction effect

between health consciousness and source, Wilks’ λ = .92, F (6,304) = 2.24, p < .05, η2

= .04, the other is between source and argument, Wilks’ λ = .89, F (6,304) = 3.12, p

< .01, η2 = .06. There is also a significant multivariate main effect for health

consciousness, Wilks’ λ = .80, F (3,152) = 12.98, p < .001, η2 = .20. Table 6.1

displays the result of the multivariate analysis. Standardized multivariate means for

two levels of health consciousness, two levels of argument across three levels of

50

source are presented in Table 6.2. To aid in the interpretation of these effects, three

factorial ANOVAs were obtained using health consciousness, argument and source as

independent variables, and perceived knowledge gain, post attitude and behavioral

intention as separate dependent variables.

Table 6.1 MANCOVA Results--- Wilks’ Lambda Statistics Wilks’ λ F Approx p-value partial eta2 Gender .15 .93 .003 Age .94 .43 .02 Initial attitude 16.67 <.001 .25 Health Consciousness (HC)

12.98 <.001 .20

Argument (A) 1.82 .15 .04 Source (S) .91 .49 .02 HC x A 1.27 .29 .02 HC x S 2.24 .04 .04 A x S 3.16 .01 .06 HC x A x S .94 .47 .02

51

Table 6.2 Standardized Multivariate Means and Standard Deviation of the Influence of Health Consciousness, Argument and Source on

Perceived knowledge gain, Post Attitude and Behavioral Intention

Dependent variables

Health consciousness

Informative Argument Testimonial Argument

Expert Celebrity Ordinary Person Expert Celebrity Ordinary Person

Perceived knowledge gain

High 3.59(.77) 3.19(.54) 2.81(.88) 3.50(.50) 3.56(.92) 3.44(1.10)

Low 3.00(.64) 3.00(.94) 2.95(.64) 2.71(.93) 2.50(.90) 3.25 (.82)

Post Attitude High 3.59(.53) 3.56(.55) 4.06(.78) 3.65(.45) 3.72(.91) 3.61(.82)

Low 2.94(.84) 3.55(.72) 2.83(.71) 2.77(.96) 2.76(1.17) 3.18(.77)

Behavioral Intention

High 3.35(.80) 3.04(.58) 3.92(.68) 3.06(.73) 3.31(.85) 3.07(1.02)

Low 2.46(.83) 2.70(.75) 2.37(.55) 2.02(1.03) 2.51(1.09) 2.17(.61)

52

Perceived knowledge gain

The results of the first factorial ANOVA revealed a significant interaction effect

between health consciousness and source on perceived knowledge gain, F (2, 154) =

3.63, η2 = .05, p < .05. Highly health conscious adolescents were most influenced by

the dietitian (M=3.54, SE=.11), followed by the celebrity (M=3.43, SE=.13) and the

ordinary person (M= 2.95, SE= .19). In contrast, low health conscious adolescents

were most influenced by the ordinary person (M=3.12, SE=.17), and the celebrity is

the least influential to them (M=2.79, SE=.15) (see Figure 6.1).

A marginal interaction effect of argument and source emerged on perceived

knowledge gain, F (2, 154) = 2.90, η2 = .04, p < .06. For informative argument,

dietitian as the source was the most persuasive (dietitian M=3.34, SE=.13 vs. celebrity

M=3.05, SE= .16 vs. ordinary person M=2.75, SE=.19). For testimonial argument, the

dietitian was more persuasive than the celebrity (dietitian M=3.12, SE=.13 vs.

celebrity M=3.08, SE= .13), but the ordinary person was the most persuasive (M=3.30,

SE=.17) (see Figure 6.2).

A main effect of health consciousness was also found on perceived knowledge

gain, F (1, 154) = 6.86, η2 = .04, p < .01, indicating adolescents who are highly

conscious about their health find the PSAs more useful, and have a stronger

perception of knowledge gain (M=3.28, SE=.08) than those with low health

consciousness (M=2.94, SE=.09).

53

Figure 6.1. Interactive effect of health consciousness and source on perceived knowledge gain. Covariate appearing in the model is evaluated at the following value: age=13.67, gender= .45, initial attitude= 3.43.

Figure 6.2. Marginal interactive effect of argument and source on perceived knowledge gain. Covariate appearing in the model is evaluated at the following value: age=13.67, gender= .45, initial attitude= 3.43.

54

Post Attitude

The second univariate ANOVA revealed that only health consciousness had

significant main effect on post attitude towards nutrition labels, F (1, 154) = 20.93, η2

= .12, p < .001, indicating highly health conscious adolescents have a more positive

attitude towards nutrition label (M=3.60, SE=.08) than the low health conscious ones

(M=3.06, SE=.08) after exposure to the PSA.

Behavioral Intention

The third univariate ANOVA showed an interaction effect between argument and

source on behavioral intention, F (2, 154) = 3.24, η2 = .04, p < .05. If the argument

was informative, a dietitian as the source made the PSA more likely to motivate

people to use nutrition label (M=2.97, SE=.13), compared to an ordinary person

(M=2.96, SE=.18) and a celebrity (M=2.80, SE=.15). If the argument was testimonial,

celebrity as the source was the most effective (celebrity M=3.00, SE=.12 vs. ordinary

person M=2.57, SE=.17 vs. dietitian M=2.54, SE=.12) (see Figure 6.3).

A main effects of health consciousness emerged on behavioral intention, F (1,

154) = 38.14, η2 = .20, p < .001, indicating highly health conscious adolescents are

more likely to use nutrition label (M=3.18, SE=.08) than low health conscious

adolescents (M=2.43, SE=.09).

55

Figure 6.3. Marginal interactive effect of argument and source on behavioral intention. Covariate appearing in the model is evaluated at the following value: age=13.67, gender= .45, initial attitude= 3.43.

56

Covariate

It is shown that initial attitude as a covariate in the analysis, had significant effect

on perceived knowledge gain (F (1, 154) = 16.85, η2 = .10, p < .001), post attitude (F

(1, 154) = 36.49, η2 = .20, p < .001), and behavioral intention (F (1, 154) = 41.24, η2

= .21, p < .001).

Table 6.3 shows the effect of the covariate and independent variables on the

three dependent variables.

Predictors of Behavioral Intention

A hierarchical multiple regression analysis was performed to access the effects of

initial attitudes, health consciousness, source credibility, argument, perceived

knowledge gain, post attitude and self-efficacy on intention to use nutrition labels (see

Table 6.4). Results show that the whole model could explain 59.6% of the variance in

the dependent variable, F (8, 159) =31.79, p<.001. After controlling for gender and

age, three independent variables accounted for a significant portion of the variance in

adolescents’ behavioral intention: health consciousness (β=.16, t=2.61, p<.01), post

attitude (β=.36, t=4.99, p<.001), and self-efficacy (β=.33, t=5.56, p<.001).

57

Table 6.3 Multivariate Analysis of Covariance ---The Influence of Initial attitude, Health Consciousness, Argument and Source on Perceived knowledge gain, Post Attitude and Behavioral Intention Perceived

Knowledge Gain Post Attitude Behavioral Intention

F p η2 F p η2 F p η2

Gender .15 .70 001 .41 .52 .003 .07 .79 <.001Age .11 .74 .001 1.46 .23 .01 2.67 .10 .02Initial attitude 16.85 <.001 .10 36.49 <.001 .19 41.24 <.001 .21Health Consciousness (HC)

6.87 .01 .04 20.93 <.001 .12 38.14 <.001 .20

Argument (A) .93 .34 .01 .34 .56 .002 2.96 .09 .02Source (S) 1.04 .36 .01 .57 .57 .01 .62 .54 .01HC x A 2.83 .10 .02 .05 .82 <.001 .18 .67 .001HC x S 3.63 .03 .05 .71 .49 .01 1.74 .18 .02A x S 2.90 .06 .04 .42 .66 .01 3.24 .04 .04HC x A x S .07 .93 .001 1.54 .22 .02 .70 .50 .01

Table 6.4 Hierarchical Multiple Regressions to Predict Behavioral Intention to Use Nutrition Labels (n=169) Variable β t Adjusted R2 R2 change (%)Step 1 Gender -.04 -.77 Age .05 1.05 .02 .03 Step 2 Health consciousness .16** 2.61 .32 .30 Step 3 Source credibility .03 .50 .35 .03 Step 4 Argument .10 1.99 .35 .01 Step 5 Perceived knowledge gain .06 .91 .40 .05 Step 6 Post Attitude .36*** 4.99 .52 .12 Step 7 Self-efficacy to use nutrition labels

.33*** 5.56 .60 .08

*p<.05, **p<.01, *** p<.001

Hypothesis Testing and Answers to Research Questions

The following hypothesis testing and solutions to research questions were given

based on the statistical findings.

58

No interaction effect of health consciousness and argument emerged, so

Hypothesis 1 was rejected, indicating that even high health conscious adolescents,

who were expected to scrutinize the arguments of PSA, didn’t have significant

difference in the processing of informative argument and testimonial argument.

Hypothesis 2 was partially supported. A significant interaction effect of health

consciousness and source emerged on perceived knowledge gain. For low health

conscious adolescents, an expert as the source generated more positive attitudes

towards nutrition label use than a celebrity, regardless of the types of arguments. It is

noteworthy that under high health consciousness condition, the variation of source

credibility also contributes significantly to persuasion.

The interaction effect of argument and source only existed in behavioral

intention, so Hypothesis 3 was partially supported. An informative argument made by

an expert was more effective to incur positive behavioral intention than the same

argument made by a celebrity, while a testimonial argument made by a celebrity was

more effective than that made by an expert. However, the match of argument and

source didn’t show significant difference in attitude and perceived knowledge gain.

In accordance with Hypothesis 4 and hypothesis 5, post attitude and self-efficacy

appeared to be strong predictors of behavioral intention to use nutrition labels.

To answer Research Question 1, for adolescents, the match of testimonial

argument and a celebrity is significantly more effective than other matches to incur

intention to use nutrition labels. The informative argument made by either source was

less effective.

Research Question 2 asked about the effect of initial attitude. It is shown that the

observed difference in the three dependent variables is largely attributed to the

pre-existing differences in adolescents’ initial attitudes towards nutrition labels.

59

Chapter7:

Discussion

The goal of this study is to explore the effectiveness of the Elaboration

Likelihood Model in creating effective PSAs to advocate nutrition label use among

adolescents. It was hypothesized that highly health conscious adolescents processed

the PSA via central route, so the informative arguments would be significantly more

persuasive to them. However, this hypothesis was not supported. Highly health

conscious adolescents and low health conscious adolescents paid similar levels of

attention to either argument. It may be attributed to the moderate level of high health

consciousness. Even though there was a significant difference between high and low

health consciousness (t=40.811, p<.001), the score of high health consciousness was

actually moderate (M=3.44, SD= .36). In other words, the topic of health problems or

healthy lifestyles was not perceived to be so important that deserved high involvement

and careful deliberation.

Adolescents with low health consciousness were hypothesized to rely on source

credibility to judge the usefulness of PSAs and change their attitudes accordingly.

This hypothesis was supported for perceived knowledge gain. Although the expert

was rated as the most credible source, followed by the ordinary person and celebrity,

the ordinary person was the most useful one for low health conscious adolescents to

learn about nutrition labels. This is partially incongruent to the view that a source with

higher credibility is more persuasive.

Actually, previous empirical research has found that source credibility may

inhibit or have no effect on the communicator’s persuasiveness under some

circumstances like when the messages are defending receivers’ initial attitude.

60

According to O’ Keefe (1990), when a high credible source is making proattitudinal

arguments, recipients will sit back and let the communicator present the perceived

adequate and cogent argument, and thus they will make less effort to think about the

arguments. In contrast, if it is a source low in credibility who delivers the

proattitudinal message, recipients might feel a greater need to ensure that their

position is being adequately represented, and thus they will be more stimulated to

think intensively about the argument to defend its viewpoints. In that case, they will

generate more supportive arguments and will be more persuaded (Sternthal et al.,

1978). Before the experiments in this study, adolescents had acknowledged the

benefits of nutrition labels. Then the messages presented to them were advocating the

use of nutrition labels, compatible with their initial attitudes. Therefore, they may find

it unnecessary to scrutinize the arguments made by an expert who was perceived to be

credible to provide adequate evidence. In contrast, the argument made by an ordinary

person with lower credibility should be supported by more evidence. Therefore,

adolescents elaborated on the messages to defend their own views.

Another finding regarding health consciousness and source credibility is that

highly health conscious adolescents were more inclined to find the PSA useful and

meaningful when they were exposed to an expert as the source. Here, the high

credibility of the source may serve as evidence supporting the validity of the

arguments. In this case, highly health conscious adolescents still processed the PSA

via the central route, because they view the source as a part of the argument

influencing the amount of thinking, rather than a simply peripheral cue. According to

Petty and Wegener (1998), source characteristics have effects across the entire

elaboration continuum by taking up different roles. When the elaboration likelihood is

moderate and people are unsure how much effort they should devote to processing the

61

message, sources give guidance as to whether the message is worth thinking about

(Heesacker, Petty, & Cacioppo, 1983). In my study, adolescents with comparatively

higher health consciousness actually didn’t have a strong motivation to process the

PSA, so they relied on credible source to evaluate the message. Likewise, adolescents

interested in popular culture were more likely to be attracted by a celebrity and

curious about what a celebrity said in the PSA.

The finding of an interaction effect between argument and source indicates the

importance of appropriate message-spokesperson matches to create effective PSAs,

regardless of health consciousness of the adolescents. It is shown that a dietitian made

the informative arguments more effective to arouse intention to use nutrition labels,

while a celebrity made testimonial argument more persuasive.

Bohner et al. (2002) argued that when ambiguous arguments were presented by

an expert, recipients evaluated these arguments very critically because the expert was

supposed to make cogent arguments. However, when the same arguments were

presented by a non-expert, recipients’ thoughts about the message were more

favorable. According to the manipulation check of the credibility of the sources, the

dietitians were perceived to have the most expertise in nutrition labels and were most

trustworthy to provide valid information. Therefore, they were expected to show

scientific evidence to prove the benefits of nutrition labels. When they made

testimonial arguments, recipients may consider the nutrition labels as not that

beneficial because even the dietitians were not able to provide strong evidence to

support the argument. In addition, they may not have confidence in themselves to

imitate the expert to be self-disciplined and conscious to conduct specific behaviors.

However, it is interesting to find that an ordinary person, despite of his or her

lower credibility, made the informative argument as persuasive as the one made by an

62

expert. It is understandable given that some adolescents, at their age of rebelliousness,

tend to resist authority. In this study, soda drink was taken for an example to

recommend an appropriate amount of nutrients intake. Soda drink is one of

adolescents’ favorite food, so when they were told to drink less or drink under some

restrictions, they were likely to become argumentative and to try to assert their

freedom by avoiding the message processing (Wolburg, 2006). In terms of ordinary

persons, their persuasiveness is partly attributed to adolescents’ perception of

similarity to the source. As Atkin (2001) argued, ordinary people as a messenger can

improve self-efficacy. If adolescents think they share similar characteristics and

abilities with the ordinary person in the PSA, they may tend to believe that they are

able to do the same, acquire the same knowledge and get the same benefits.

With the lowest perceived credibility, a celebrity became the most persuasive

source when making a testimonial argument. The celebrities in the experiment had

high exposure in advertisements and campus road shows, so they had built up a

healthy image among the adolescents, which made them suitable to advocate healthy

eating. In addition, they were in their twenties and look young, so it is easy for

adolescents to identify with them. According to social learning theory (Bandura,

1977), individuals are more likely to model a behavior when they perceive themselves

to be similar to the model. Therefore, although celebrities have less expertise in

nutrition labels, their claim of conducting a behavior would be more effective to

establish a connection with the adolescents, improving their confidence in developing

the same habit. When either of the celebrities says “my family and friends,” the

receivers would readily connect the subjects to their own family and peers.

Public health research applying ELM seldom examines the effect of health

consciousness on the information processing strategies. Given the wide application of

63

this concept to other research, this study tested its power in the ELM framework, by

regarding it as a reflection of the original construct of issue involvement. It was found

that health consciousness had significant main effects on perceived knowledge gain,

attitude and behavioral intention. One possible explanation is that adolescents who

care about their health and have the habit of seeking health information and applying

it to their life may have acquired greater amount of health knowledge, including the

knowledge of nutrition labels. Therefore, they are more willing to spend time

analyzing the claims in the PSA and evaluating their reliability based on their initial

knowledge.

Consistent with expectations, initial attitude caused significant variation in

perceived knowledge gain, post attitude and behavioral intention. Adolescents who

had a more positive initial attitude towards nutrition labels were more likely to be

persuaded by the PSA. This initial attitude explained a large part of the persuasive

effect. According to Lavine et al (2000), initial attitude is cognitively accessible to

guide information processing and influence judgment and decision-making. But the

attitude accessibility, which refers to how quickly a person can retrieve an attitude

from memory and use it in making a judgment (Fazio, 1986), is determined by a few

factors, including the extent to which the attitude is linked to other constructs (e.g.,

beliefs, values). The finding showed that initial attitudes towards nutrition labels were

highly positively correlated with health consciousness (λ=.36, p<.001). That partly

explained why initial attitude influenced the elaboration of persuasive messages.

However, results show that post attitude was not significantly different from

initial attitude (t=.30, p>.70, ns). We identified no other health consciousness x source

interaction effect or argument x source interaction effect on post attitudes. It may be

the result of peripheral processing. Although adolescents comprehended the PSAs and

64

restructured their cognition about nutrition labels, they didn’t engage in the cognitive

elaboration to internalize the positions represented by the messages (Scott, 1995).

Another reason may be the manipulation of the PSA is not strong enough to

change adolescents’ attitudes. The information provided in the PSA is not attractive to

the audience, and lacks persuasiveness. In addition, as a characteristic of their age

group, adolescents are resistant to change their attitudes.

For the sake of health, adolescents in Hong Kong should be alert to their

excessive consumption of total fat, sugar and sodium. However, they may not realize

the negative effects of the unhealthy eating diet. Those who are generally more health

conscious than their peers may ignore potential health problems and pursue good taste,

too. After all, the chronic disease is perceived to be far away from them. As Atkin

(2001) mentioned, people who perform risky behaviors are highly resistant during

early phases, especially during the teenage years. The Transtheoretical Model

(Prochaska, Diclemente, & Norcross, 1992) proposes five stages in the process of

behavior change: precontemplation, contemplation, preparation, action and

maintenance. People in the precontemplation stage are not aware of the risk of their

current behavior, and thus have no intention to change. In the contemplation stage,

with awareness of the problem, people are considering changing in the near future but

have not taken action. Adolescents with an unhealthy diet are probably in the stage of

precontemplation or contemplation. However, they may not be receptive to the PSA

that advocates nutrition label use to regulate their diet, not to mention enhancing their

attitudes to the benefits of using labels or conducting actual behavior.

In addition, according to reactance theory, when people are threatened by any

perceived restrictions to their freedom, an attempt to restore sense of freedom is

aroused, so the restricted behavior becomes more attractive to them (Brehm, 1966;

65

Brehm & Brehm, 1981). With their great need for independence, adolescents might

especially respond with reactance to advocacy of change in their unhealthy eating

habits (Czyzewska & Ginsburg, 2007).

The other explanation for the unchanged attitudes is that participants in the

experiment only had a short time to report their attitudes after reading the PSA, while

the change of attitudes may emerge after a longer delay. People may first change their

behaviors and then change attitudes in order to bring them in line with the behaviors

to reduce cognitive dissonance (e.g., Axsom & Cooper, 1985; Bem, 1967; Festinger,

1962; Seligman, Fazio, & Zanna, 1980).

Despite of the reasons above, it is important to note that attitudes changed

thoughtfully via central route processing are not necessarily stronger than one’s initial

attitude. What ELM emphasizes is the thoughtfully formed attitude is stronger than an

unthoughtfully attitude formed by peripheral processing. Initial attitude can exist

along with the new attitude to guide behavior.

The publicity campaign of nutrition label has reached a large number of

individuals, families and schools. However, the impact on actual adoption of nutrition

labels is unsatisfactory. Aiming to refine the messages delivered in the campaign to

promote behavior change, I referred to Social Cognitive Theory, Theory of Planned

Behavior and Extended Parallel Model, and focused on the effect of attitude and

self-efficacy on behavioral intention. As expected, both attitude towards nutrition

labels and self-efficacy to use them were distinct and powerful predictors of the

intention to read nutrition labels when choosing food. Adolescents evaluate their

efficacy to read nutrition labels in the face of different obstacles: time constraint, too

many names of nutrients, and complicated calculations, etc. Those with high efficacy

think the impediments are surmountable, so the behavior is easy to perform.

66

Chapter8:

Implication The findings bear implications for health interventions.

First, adolescents’ health consciousness greatly influenced the effectiveness of

the PSA, which implies that in the publicity campaign of nutrition label use,

additional messages should be delivered to enhance adolescents’ awareness of the

importance of a healthy body and a balanced diet. But these messages should not be

regarded as a threat to deprive individual freedom to do what is being done; otherwise

adolescents would show reactance to the messages. Hence, the messages should

emphasize adolescents’ freedom to accept or to reject any recommendations.

Another effective strategy is community-based intervention. People who are

moderately or highly health conscious are more willing to participate in community

activities (Dutta-Bergman, 2004a), such as attending church and volunteering in a

community organization. Campaign planners can design specific health promotion

activities in such settings. For people with low health consciousness, campaign

planners should target the specific at-risk populations in the context of the specific

health behavior, and address individual-level factors that limit their action.

Second, most adolescents are not aware of the importance of a healthy diet, and

seldom take initiatives to seek health information. With a moderate level of health

consciousness, they prefer to rely on simple cues, like source credibility to process the

information in PSA. Dietitians, who are perceived to have more expertise in nutrition

labels and be more trustworthy, can be effective communicator to the adolescents.

Third, the persuasiveness of the PSA also depends on the match-up of the

message and the source. A testimonial argument, with vivid examples and a

first-person account of individual experience delivered by a young and healthy

67

celebrity, is more effective than other argument-source matches. Therefore, to design

the most persuasive PSA, campaign planners should investigate the popular culture to

see who is appealing to adolescents.

Fourth, the power of self-efficacy implies that practitioners should enhance

adolescents’ confidence in understanding nutrition labels, and improve their perceived

efficacy to make use of them. Bandura (1977) points out four ways to gain efficacy

information, one of which is vicarious experience, including learning by observing

others’ behaviors. The school-based health campaign can invite a few students as peer

role models to share their experience of learning and adopting nutrition labels.

Strecher et al. (1986) suggested the role model should be viewed as overcoming

difficulties with determined efforts rather than with ease. In addition, a clear

rewarding outcome can make the modeled behaviors more attractive. It is also

necessary to improve the ways that nutrition information is presented on the labels, to

make it less complex and clearer, so that consumers can process the information in a

quick and meaningful manner.

The key findings were applied to design a new PSA (see Appendix G). Apart

from the testimonial argument with personal benefits from using nutrition labels, there

are additional messages emphasizing the ease of the usage behavior and the way to do

it.

In sum, this study explores the applicability of the Elaboration Likelihood Model

to the mechanism that adolescents process PSAs which promote nutrition label on

food products. I focus on the effect of issue involvement, argument and source

credibility in the model. Although few strong effects of the argument and source

credibility were found, issue involvement, which was indicated by health

consciousness in this study, showed strong power to influence the effectiveness of

68

PSA. Therefore, it is suggested that, to promote nutrition labels to adolescents, who

haven’t realized the risk of unhealthy diets and had no motivation to make positive

change, the initial focus should be on increasing their general health consciousness

before matching a spokesperson with specific messages.

69

Chapter9:

LimitationandDirectionforFutureresearch

According to ELM, many other variables can affect the routes of information

processing, such as need for cognition, distraction in the message environment, and

message comprehensibility, etc. Since participants in the experiment were randomly

assigned to each group, it controlled for the effects of those extraneous variables

across all treatment conditions. Therefore, any significant differences between

conditions can fairly be attributed to the difference of argument and source. However,

the limited effects of the PSAs warrant additional attention to these variables, for

example, need for cognition, which refers to “individual’s tendency to engage in and

enjoy effortful cognitive endeavors.” (Cacioppo, Petty, & Kao, 1984). It is developed

as an individual difference variable affecting the motivation to process persuasive

messages. Attitudes of individuals with high need for cognition are primarily

influenced by strong message arguments, while attitudes of those low in need for

cognition were primarily influenced by attractiveness of the communicators (Batra &

Stayman 1990; Haugtvedt et al., 1988; Haugtvedt, Petty, & Cacioppo, 1992).

This research adopted the source credibility model to study source effect,

focusing on the difference in perceived credibility of different sources. However,

spokespersons in PSAs differ in other aspects, for example, attractiveness. Source

attractiveness model is another widely adopted model. It identifies four aspects of the

attractiveness of a source: “familiarity”, “likability”, “similarity” (McGuire, 1969)

and “physical attractiveness” (Mills & Harvey, 1972). Attractiveness affects

persuasion when audience feel satisfied in believing they are similar to the

spokesperson and thus identify with or conform to his or her attitudes and behaviors

(Friedman & Friedman, 1979).

70

In this study, I didn’t manipulate the attractiveness of the dietitians, celebrities or

ordinary person, and didn’t ask participants to rate their perception of this

characteristic of the sources. It is possible that information processing routes are

biased by source attractiveness. And it is also argued that argumentation, source

expertise, and source attractiveness have independent and additive effects on

persuasion (Maddux & Rogers, 1980). For example, Norman (1976) manipulated

source credibility and source attractiveness orthogonally to interact with argument,

and found that female undergraduates agreed with the conclusion made by an

attractive nonexpert, no matter supporting arguments were provided or not. However,

an unattractive expert was more effective in influencing agreement if he provided

justifying arguments for his position. Future research on the PSA of nutrition labels

can manipulate source credibility and attractiveness at the same time.

The stimuli in this study are print PSAs, so the results cannot be generalized to

the PSAs in other media channels, such as radio and TV. Previous research has tested

the interaction effect of message delivery mode and message type. For example,

Braveman (2008) found that testimonial messages about drinking water to lose weight

or moderating alcohol drinking are more persuasive when presented through the audio

mode, while informational messages were more persuasive when presented through

the written mode. Adolescents have a high exposure to television and Internet, and

spend less time on traditional print media. Future research can explore the influence

of PSAs in different media channels on adolescents’ information processing mode.

This study was interested in the predictors of behavioral intention and tested the

effects of initial attitude, perceived knowledge gain, post attitude and self-efficacy.

But there was 40.4% of the variance in behavioral intention unexplained. According

to other behavior models, some other variables can influence intention to undertake

71

health behavior, for example, subjective norms (Theory of Planned Behavior, Ajzen,

1985), For the behavior of reading nutrition label when buying food, other studies

indicate that weight control, knowledge of diet-disease relationships, grocery

shopping habits, eating habits and concern about price are associated with nutrition

label use (Campos, Doxey, & Hammond, 2011). The effects of these variables on

adolescents can be examined in the future.

The execution of the experiment has limitations that may decrease the internal

validity. Firstly, all six groups of participants were presented the PSA at the same

auditorium at the same time. When the questionnaires and PSAs were handed out or

collected, some students talked to each other despite of the early reminder of

no-talking. It may affect the performance of some participants. Secondly, the

experiment was taken at the last class of a day around 5:00pm. Some students filled

out the questionnaires carelessly, because they wanted to leave early to play or go

back home. One student complained that he was hungry. The negative mood could

influence the amount of thinking a person does about the PSA. According to Blaney

(1986), a positive mood can bias ongoing information processing by selectively

facilitating the retrieval of positive material and/or inhibiting the retrieval of negative

material from memory. In addition, a good mood makes positive consequences more

desirable than when the same consequence is considered in a neutral mood (Wegener,

Petty, & Smith, 1995).

In order to increase the external validity, there are two different dietitians for the

expert condition, and two different celebrities for the celebrity condition. For the same

condition, there is no significant difference between the two versions. Since all the

spokespersons in the study are females, future research can examine if the gender of

the source has an effect on the persuasiveness of the PSAs.

72

References

AbuSabha, R., & Achterberg, C. (1997). Review of self-efficacy and locus of control

for nutrition and health-related behaviour. Journal of the American Dietetic

Association, 97, 1122-1132.

Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl

& J. K. J.Beckmann (Eds.), Action control: From cognition to behavior, 11-39.

Berlin: Springer-Verlag.

Ajzen, I. (1991). The Theory of Planned Behavior. Organizational Behavior and

Human Decision Processes, 50, 179-211.

Atkin, C. (2001). Impact of public service advertising: Research evidence and

effective strategies. Menlo Park, CA: Kaiser Family Foundation.

Axsom, D., & Cooper, J. (1985). Cognitive dissonance and psychotherapy: The role

of effort justification in inducing weight loss. Journal of Experimental Social

Psychology, 21, 149-160.

Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change,

Psychological Review, 84, 191-215.

Bandura, A. (2004). Health Promotion by Social Cognitive Means. Health Education

and Behavior, 31, 143-164.

Bar-Tal, D., Raviv, A., & Brosh, M. E. (1991). Perception of epistemic authority and

attribution for its choice as a function of knowledge area and age. European

Journal of Social Psychology, 21, 477-492.

Batra, R., & Stayman, D.M. (1990). The Role of Mood in Advertising Effectiveness.

Journal of Consumer Research, 17, 203-214.

Bern, D. J. (1967). Self-perception: An alternative interpretation of cognitive

dissonance phenomena. Psychological Review, 74, 183-200.

73

Birnbaum, M. H., & Stegner, S. E. (1979). Source credibility in social judgment: Bias,

expertise, and the judge’s point of view. Journal of Experimental and Social

Psychology, 37,48–74.

Biswas, D., Biswas, A., & Das, N. (2006). The differential effects of celebrity and

expert endorsements on consumer risk perceptions: The role of consumer

knowledge, perceived congruency, and product technology orientation.

Journal of Advertising, 35, 17-31.

Blaney, P, H.(1986). Affect and memory: A review. Psychological Bulletin, 99(2),

229-246.

Blum, R.W., & Stark T. (1985). Cognitive development in adolescence. Seminars in

Adolescent Medicine, 1(1), 25–32 .

Bohner, G., Ruder, M., & Erb, H. P. (2002). When expertise backWres: Contrast and

assimilation eVects in persuasion. British Journal of Social Psychology, 41,

495–519.

Braverman, J. (2008). Testimonials Versus Informational Persuasive Messages: The

Moderating Effect of Delivery Mode and Personal Involvement.

Communication Research, 35(5), 666-694.

Brehm, J. W. (1966). A theory of psychological reactance. New York, NY: Academic

Press.

Brehm, S. S., & Brehm, J. W. (1981). Psychological reactance: A theory of freedom

and control. New York, NY: Academic Press.

Brin˜ol, P., & Petty, R. E. (2009). Source factors in persuasion: A selfvalidation

approach. European Review of Social Psychology, 20, 49-96.

Brosius, H. B., & Bathelt, A. (1994). The Utility of Exemplars in Persuasive

Communications. Communication Research, 21(1), 48-78.

74

Cacioppo, J. T., & Petty, R. E. (1979). Effects of message repetition and position on

cognitive responses, recall, and persuasion. Journal of Personality and

Social Psychology, 37, 97-109.

Cacioppo, J. T., & Petty, R. E. (1981). Social psychological procedures for cognitive

response assessment: The thought listing technique. In T. Merluzzi, C. Glass,

& M. Genest (Ed.), Cognitive assessment (pp. 309-342). New York: Guilford.

Cacioppo, J. T., & Petty, R. E. (1983). Social psychophysiology : a sourcebook. New

York: Guilford Press.

Cacioppo, J. T., & Petty, R. E. (1989). Effects of message repetition on argument

processing, recall, and persuasion. Basic and Applied Social Psychology, 10,

3-12.

Cacioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The Efficient Assessment of Need

for Cognition. Journal of Personality-Assessment, 48, 306-307.

Campos, S., Doxey, J., & Hammond, D. (2011). Nutrition labels on pre-packaged

foods: a systematic review. Public health nutrition, 14(8), 1496-1506.

Center for Food Safety. (2010). Food Safety Focus, 47, 3.

Center for Food Safety. (2009). Nutrition labelling publicity and education campaign

launched. Retrieved from

http://www.cfs.gov.hk/english/press/2009_03_12_1_e.html.

Chaiken, S., & Eagly, A. H. (1983). Communication Modality as a Determinant of

Persuasion - the Role of Communicator Salience. Journal of Personality and

Social Psychology, 45(2), 241-256.

Chaiken, S., & Maheswaran, D. (1994). Heuristic processing can bias systematic

processing: Effects of source credibility, argument ambiguity, and task

importance on attitude judgment. Journal of Personality and Social

75

Psychology, 66, 460-473.

Chan, K., Prendergast, G., Gronhoj, A., & Bech-Larsen, T. (2009). Communicating

healthy eating to adolescents. Journal of Consumer Marketing, 26(1), 6-14.

Contento, I.R., Michela, J.L., & Goldberg, C.J. (1988). Food choice among

adolescents: population segmentation by motivations. Journal of Nutrition

Education, 20 (6), 289-298.

Contento, I.R, Williams, S.S, Michela, J.L., & Franklin, A.B. (2006). Understanding

the food choice process of adolescents in the context of family and friends.

Journal of Adolescent Health, 38(5), 575–82.

Czyzewska, M., & Ginsburg, H. J. (2007). Explicit and implicit effects of

anti-marijuana and anti-tobacco TV advertisements. Addictive Behaviors, 32,

114-127.

De Wit, J. B. F., Das, E., & Vet, R. (2008). What works best: Objective statistics or a

personal testimonial? An assessment of the persuasive effects of different

types of message evidence on risk perception. Health Psychology, 27(1),

110-115.

Dutta-Bergman, M. J. (2004a). An Alternative Approach to Social Capital: Exploring

the Linkage Between Health Consciousness and Community Participation.

Health Communication, 16(4), 393-409.

Dutta-Bergman, M. J. (2004b). The Impact of Completeness and Web Use Motivation

on the Credibility of e-Health Information. Journal of Communication , 54(2),

253-270.

Dutta-Bergman, M. J. (2006). The Demographic and Psychographic Antecedents of

Attitude Toward Advertising. Journal of Advertising Research, 46(1), 102-112.

Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Fort Worth, TX:

76

Harcourt Brace Jovanovich.

Eagly, A.H., & Chaiken, S. (1998). Attitude structure and function. In D. T. Gilbert, S.

T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th Ed., Vol.

1, pp. 269-322). Boston: McGraw-Hill.

Ewart, C. K., Taylor, C. B., Reese, L. B., & DeBusk, R. F. (1983). Effects of early

postmyocardial infarction exercise testing on self-perception and subsequent

physical activity. American Journal of Cardiology, 51, 1076-1080.

Fazio, R. H. (1986). How do attitudes guide behavior? In R. M. Sorrentino & E. T.

Higgins (Eds.), The handbook of motivation and cognition: Foundation of

social behavior (pp. 204-243). New York: Guilford Press.

Festinger, L. (1962). Cognitive dissonance. Scientific American, 20, 93-102.

Freiden, Jon B. (1984). Advertising Spokesperson Effects: An Examination of

Endorser Type and Gender on Two Audiences. Journal of Advertising

Research, 24, 33-41.

Friedman, H., & Friedman, I. (1976). Whom Do Students Trust? Journal of

Communication, 26 (1), 48-49.

Furnham, A., & Forey J.(1994). The attitudes, behaviors, and beliefs of patients of

conventional vs complementary alternative medicine. Journal of Clinical

Psychology, 50, 458-469.

Gibbons, F. X., Houlihan, A. E., & Gerrard, M. (2009). Reason and reaction: The

utility of a dual-focus, dual-processing perspective on promotion and

prevention of adolescent health risk behaviour. British Journal of Health

Psychology, 14, 231-248.

Gore, T.D., & Bracken, C.C. (2005). Testing the theoretical design of a health risk

message: reexamining the major tenets of the extended parallel process model.

77

Health Education & Behavior. 32, 27-41.

Gould, S. J.(1988). Consumer attitudes toward health and health care: a different

perspective. Journal of Consumer Affairs , 22, 96–118.

Guthrie, J.F, Lin, B-H., & Frazao, E. (2002). Role of Food Prepared Away From

Home in the American Diet, 1977–78 versus 1994–96: Changes and

Consequences. Journal of Nutrition Education and Behaviors, 34, 140-150.

Haugtvedt, C. P., Petty, R. E. & Cacioppo, J. T. (1992). Need for Cognition in

Advertising: Understanding the Role of Personality Variables in Consumer

Behavior. Journal of Consumer Psychology, 1(3), 239-260.

Haugtvedt, C., Petty, R. E., Cacioppo, J. T., & Steidley, T.(1988). Personality and Ad

Effectiveness: Exploring the Utility of Need for Cognition. Advances in

Consumer Research, 15, 209-212.

Heesacker, M., Petty, R. E., & Cacioppo, J. T. (1983). Field dependence and attitude

change: Source credibility can alter persuasion by affecting message-relevant

thinking. Journal of Personality, 51, 653-666.

HKU POP. (2011). Public Opinion Survey on Nutrition Label. Retrieved from

http://hkupop.hku.hk/english/report/nutrition11/index.html.

Ho, S.M, & Lee, T. M. (2001). Computer usage and its relationship with adolescent

lifestyle in Hong Kong. Journal of Adolescent Health, 2001, 29, 258-266.

Hong, H. (2009). Scale Development for Measuring Health Consciousness:

Reconceptualization. 12th Annual International Public Relations Research

Conference, Holiday Inn University of Miami Coral Gables, Florida.

Hong, H. (2011). An extension of the Extended Parallel Process Model (EPPM) in

television health news: The influence of health consciousness on individual

message processing and acceptance. Health Communication, 26 (4), 343-353.

78

Hovland, C. I., & Weiss, W. (1951). The Influence of Source Credibility on

Communication Effectiveness. Public Opinion Quarterly, 15 (4), 635-650.

Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion.

New Haven, CT: Yale University Press.

Jayanti, R. K., & Burns, A. C. (1998). The antecedents of preventive health care

behavior: An empirical study. Journal of the Academy of Marketing Science,

26(1), 6-15.

John, D.R. (1999). Consumer socialization of children: A retro perspective look at

twenty five years of research. Journal of Consumer Research, 26,183-213.

Kelman, H. C. (1961). Processes of opinion change. Public Opinion Quarterly, 25,

57-78.

Kruglanski, A. W. (2005). Says who?: Epistemic authority effects in social judgment.

Advances in Experimental Social Psychology, 37, 345-392.

Lavine, H, Borgida, E., & Sullivan J, L. (2000). On the Relationship between Attitude

Involvement and Attitude Accessibility: Toward a Cognitive-Motivational

Model of Political Information Processing. Political Psychology, 21(1),

81-106.

Lee, A., Lee, S. H., Tsang, K. K. & To, C. Y. (2003) A comprehensive ‘Healthy

Schools Programme’ to promote school health: the Hong Kong experience in

joining the efforts of Health and Education Sectors. Journal of Epidemiology

and Community Health, 573, 174–177.

Lee, A., Tsang, K. K. & Healthy School Research Group. (2004) Youth risk

behaviours in a Chinese population: a territory wide Youth Risk Behavioural

Surveillance in Hong Kong. Public Health, 118, 88–95.

Mackie, D. M., Gastardo-Conaco, M. C., & Skelly, J. J. (1992). Knowledge of the

79

advocated position and the processing of in-group and out-group persuasive

messages. Personality and Social Psychology Bulletin, 18, 145-151.

Maddux, J.E., & Rogers, R. W. (1980). Effects of Source, Expertness, Physical

Attractiveness, and Supporting Argumentson Persuasion: A Case of Brains

over Beauty. Journal of Personality and Social Psychology, 39 (2), 235-244.

McCracken, G. (1989). Who Is the Celebrity Endorser? Cultural Foundations of the

Endorsement Process. Journal of Consumer Research, 16, 310-321.

McGuire, W. J. (1969). Theory-oriented research in natural settings: The best of both

worlds for social psychology. In M. Sherif & C. Sherif (Eds.),

Interdisciplinary relationships in the social sciences.Chicago: Aldine.

McGuire, W. J. (1989). Theoretical foundations of campaigns. In R. E. Rice & C. K.

Atkin (Eds.), Public communication campaigns (pp. 43-65). Newbury Park,

CA: Sage

Mills, J., & Harvey, J. (1972). Opinion change as a function of when information

about the communicator is received and whether he is attractive or expert.

Journal of Personality and Social Psychology, 21, 52-55.

Neumark-Sztainer, D, Story, M, Perry, C., & Casey, M.A. (1999). Factors influencing

food choices of adolescents: findings from focus-group discussions with

adolescents. Journal of the American Dietetic Association: 99, 929-934.

Nørgaard, M. K., & Brunsø, K. (2009). Families' use of nutritional information on

food labels. Food Quality and Preference, 20(8), 597-606.

Nørgaard, M. K., Brunsø, K., Christensen, P. H., & Mikkelsen, M. R. (2007).

Children's influence on and participation in the family decision process during

food buying.Young Consumers: Insight and Ideas for Responsible Marketers,

8(3), 197-216.

80

Norman, R (1976). When what is said is important: a comparison of expert and

attractive sources. Journal of Experimental Psychology, 294-300.

Ohanian, R. (1990). Construction and validation of a scale to measure celebrity

endorsers’ perceived expertise, trustworthiness, and attractiveness. Journal of

Advertising, 19(3), 39-52.

O’Keefe, D.J. (1990). Persuasion: Theory and Research. Newbury Park, CA: Sage

Perse, E.M., Nathanson, A.I., & McLeod, D.M. (1996). Effects of spokesperson sex,

public service announcement appeal, and involvement on evaluations of

safe-sex PSAs. Health Communication, 8, 171-189.

Petty, R. E., Brin˜ol, P., & Tormala, Z. L. (2002). Thought confidence as a determinant

of persuasion: The self-validation hypothesis. Journal of Personality and

Social Psychology, 82, 722-741.

Petty, R. E., & Brin˜ol, P. (2008). Persuasion: From single to multiple to

metacognitive processes. Perspectives on Psychological Science, 3, 137-147.

Petty, R. E., & Cacioppo, J. T. (1979). Issue-involvement can increase or decrease

persuasion by enhancing message-relevant cognitive responses. Journal of

Personality and Social Psychology, 37, 1915-1926.

Petty, R. E., & Cacioppo, J. T. (1981). Attitudes and persuasion: classic and

contemporary approaches. Dubuque, Iowa: W.C. Brown Co. Publishers.

Petty, R. E., & Cacioppo, J. T. (1983). The role of bodily responses in attitude

measurement and change. In J. T. Cacioppo, & R. E. Petty (Eds.), Social

psychophysiology: A sourcebook (pp. 51-101). New York: Guilford.

Petty, R. E., Cacioppo, J. T. & Schumann, D. (1983). Central and Peripheral Routes

to Advertising Effectiveness: The Moderating Role of Involvement. Journal of

Consumer Research, 10, 135-146.

81

Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion : central and

peripheral routes to attitude change. New York: Springer-Verlag.

Petty, R. E., & Cacioppo, J. T. (1996). Attitudes and persuasion: classic and

contemporary approaches. Boulder, Colo.: Westview Press.

Petty, R. E., Fazio, R. H., & Briänol, P. (2009). Attitudes: insights from the new

implicit measures. New York: Psychology Press.

Petty, R. E., Goldman, R., & Cacioppo, J. T. (1981). Personal Involvement as a

Determinant of Argument-Based Persuasion. Journal of Personality and

Social Psychology, 41(5), 847-855.

Petty, R. E., Haugtvedt, C. P., & Smith, S.M. (1995). Elaboration as a determinant of

attitude strength. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength:

Antecedents and consequences (pp. 93-130). Mahwah, NJ: Erlbaum.

Petty, R. E., Ostrom, T. M., & Brock, T. C. (1981). Cognitive responses in persuasion.

Hillsdale, N.J.: L. Erlbaum Associates.

Petty, R. E., & Wegener, D. T. (1998). Matching versus mismatching attitude

functions: Implications for scrutiny of persuasive messages. Personality and

Social Psychology Bulletin, 24(3), 227-240.

Powell, L. M., Szczpka, G., Chaloupka, F.J., & Braunschweig, C. L. (2007).

Nutritional content of television food advertisements seen by children and

adolescents. Pediatrics, 120, 576-583.

Priester, J. R., & Petty, R. E. (2003). The influence of spokesperson trustworthiness on

message elaboration, attitude strength, and advertising effectiveness. Journal

of Consumer Psychology, 13(4), 408-421.

Prochaska, J. O., DiClemente, C. C., & Norcross, J. C. (1992). In search of how

people change: Applications to addictive behaviors. American Psychologist,

82

47, 1102-1114.

Rao, Akshay, R., & Monroe, K. B. (1988). The Modernizing Effect of Prior

Knowledge on Cue Utilization in Product Evaluations. Joumal of Consumer

Research,15, 253-264.

Raviv, A., Bar-Tal, D., Raviv, A., & Ben-Horin, A. (1996). Adolescent Idolization of

Pop Singers: Causes,Expressions, and Reliance, Journal of Youth and

Adolescence, 25(5), 631-650.

Rosen, C. S. (2000). Is the sequencing of change processes by stage consistent across

health problems? A meta-analysis. Health Psychology, 19, 593–604.

Salmon, C.T. (1986). Perspectives on Involvement in Consumer and Communication

Research. In Progress in Communication Sciences, B. Dervin & M. J. Voight

(Eds). Ablex Publishing Corporation, Norwood, New Jersey. 243-269.

Salmon, C., & Atkin, C. (2003). Using media campaigns for health promotion. In T. L.

Thompson, A. M. Dorsey, K. I. Miller & R. Parrott (Eds.), Handbook of health

communication. Hillsdale, NJ: Lawrence Erlbaum, 449-472.

Seligman, C, Fazio, R. H., & Zanna, M. P. (1980). Effects of salience of extrinsic

rewards on liking and loving. Journal of Personality and Social Psychology,

38, 453-460.

Shay, C. (2011). Hong Kong’s Expanding Obesity Problem. Retrieved from

http://blogs.wsj.com/hong-kong/2011/01/10/hong-kongs-expanding-obesity-pr

oblem/

Shead, N. W., Walsh, K., Taylor, A., Derevensky, J. L. & Gupta, R.(2011). Youth

Gambling Prevention: Can Public Service Announcements Featuring Celebrity

Spokespersons be Effective?. International Journal of Mental Health

Addiction, 9, 165-179.

83

Slater, M. D., & Rouner, D. (1996). How message evaluation and source attributes

may influence credibility assessment and belief change. Journalism & Mass

Communication Quarterly, 73(4), 974-991.

Smetana, J. G., & Bitz, B. (1996). Adolescents’ conceptions of teachers’ authority and

their relations to rule violations in school. Child Development, 67, 1153-1172.

Steinberg, L. Adolescence. New York, NY; McGraw-Hill, Inc.: 1999.

Sternthal, B., Phillips, L., & Dholakia, E. (1978). The Persuasive Effects of

Source Credibility: A Situational Analysis. Public Opinion Quarterly,

285-314.

Strachan, S.M., & Brawley, L.R. (2009). Healthy-eater identity and self-efficacy

predict healthy eating behavior: A prospective view. Journal of Health

Psychology, 14, 684-695.

Strecher, V. J, DeVellis, B.M, Becker, M.H, & Rosenstock, I.M. (1986). The role of

self-efficacy in achieving health behavior change. Health Education Quarterly,

13, 73-92.

Stewart, S. M., Lam, T. H., Betson, C., & Chung, S. F. (1999). Suicide ideation and its

relationship to depressed mood in a community sample of adolescents in Hong

Kong. Suicide & Life-Threatening Behavior, 29, 227-240.

Tedeschi, J. T. (1972). The Social Influence Processes, Chicago: Aldine-Atherton.

Till, B., & Shimp, T. (1998). Endorses in advertising: the case of negative celebrity

information. Journal of Advertising, 27(1), 67-82.

Tversky, A., & Kahneman, D. (1973). Availability - Heuristic for Judging Frequency

and Probability. Cognitive Psychology, 5(2), 207-232.

Tversky, A., & Kahneman, D. (1974). Judgment under Uncertainty - Heuristics and

Biases. Science, 185(4157), 1124-1131.

84

Wegener, D., Petty, R., & Smith, S. (1995). Positive mood can increase or decrease

message scrutiny: the hedonic contingency view of mood and message

processing. Journal of Personality and Social Psychology, 69(1), 5-15.

Wilson, E.J., & Sherrell, D.L. (1993). Source effects in communication

and persuasion research: A meta-analysis of effect size. Journal of The

Academy of Marketing Science 21(2), 101-112.

Witte, K. (1992). Putting the fear back into fear appeal: The extended parallel process

model. Communication Monographs, 59, 329-349.

Witte, K., Meyer, G., & Martell, D. (2001). Effective health risk messages: A step-by

step guide .Thousand Oaks, CA: Sage.

Wolburg, J. (2006). College students’ responses to antismoking messages: Denial,

defiance, and other boomerang effects. The Journal of Consumer Affairs, 40(2),

294-323.

Wright, L.T. (1997) Exploring the need for extended research: an investigation of

consumer attitudes to product. Journal of Product and Brand Management, 66,

417-427.

85

APPENDIXA:EXPERIMENTQUESTIONNAIRE 關於食物營養標籤的調查

第一部分：個人基本資料:

1. 性別： 男____ 女______

2. 年齡：_________歲

3. 體重：_________kg； 身高：________m 或 cm

4. 你平時會自己購買預先包裝好的食物（例如汽水，牛奶，餅乾，薯片，等等）嗎？

從來不買 1 2 3 4 5 經常買

5. 你留意到食物包裝上有右圖所示的這一類營養標籤嗎？ 有______ 無_______

6. 你懂得如何根據這類營養標籤上的資料來選擇食物嗎？ 完全不懂得 1 2 3 4 5 非常懂得

7. 在選擇預先包裝食物時，你會閱讀這類營養標籤嗎？ 從來不會 1 2 3 4 5 經常會

8. 你覺得在食物包裝上加上營養標籤是多餘還是必要的呢？

非常多餘 1 2 3 4 5 非常有必要

9. 你認為營養標籤對你保持飲食均衡有用嗎？ 完全無用 1 2 3 4 5 非常有用

10. 你覺得營養標籤上的資訊對你維持健康，減少患病風險有用

嗎？ 完全無用 1 2 3 4 5 非常有用

11. 你有興趣了解更多關於營養標籤的資訊嗎？ 完全無興趣 1 2 3 4 5 非常有興趣

12. 你有沒有控制每天攝取的能量或某些營養素（例如，鈉，糖，總脂肪）的份量呢？ 完全沒有控制 1 2 3 4 5 非常嚴格控制

13. 你擔心自己超重或者患肥胖症嗎？

從來不擔心 1 2 3 4 5 經常擔心

14. 對於那些告訴你如何降低患病風險，保持身體健康的資訊，你感興趣嗎？

非常不感興趣 1 2 3 4 5 非常感興趣

15. 你關注自己的健康狀況嗎？

完全不關注 1 2 3 4 5 非常關注

16. 日常生活中，你會為了保持身體健康而採取積極行動（例如：調節飲食，運動，等等）

嗎？

從來不會 1 2 3 4 5 經常會

營養資料 Nutrition Information

每 100 克/Per 100g

能量/Energy 380 千卡/kcal

蛋白質/Protein 6 克/g

總脂肪/Total fat 3 克/g

-飽和脂肪/Saturated fat 1.5 克/g

-反式脂肪/Trans fat 0 克/g

碳水化合物/Carbohydrates 82 克/g

-糖/Sugars \ 5 克/g

鈉/Sodium 120 毫克/ mg

親愛的同學： 您好！我們是香港中文大學新聞與傳播學院的學生，現正進行一項關於食物營養標籤

的學術研究，希望您能提供寶貴的意見。本研究所有資料僅供學術研究之用，所有問題皆

無標準答案，您只需按照個人意見作答。非常感謝您的合作！

86

第二部分： 現在，研究者向你展示一則廣告，請閱讀廣告上的資訊，並繼續回答問題。

17. 看完廣告後，你覺得在食物包裝上加上營養標籤是多餘還是必要的呢？ 非常多餘 1 2 3 4 5 非常有必要

18. 對於剛才看的廣告，你記得些什麽呢？請寫下你記得的文字，人物，LOGO 等內容。

19. 剛才看到那個廣告時，你第一時間想到什麼?

_______________________________________________________________________________

_____ 20. 請為這則廣告上的幾句話打分：

（1）非常難理解 1 2 3 4 5 非常容易理解

（2）說服力非常弱 1 2 3 4 5 說服力非常強

（3）提供的資訊完全無用 1 2 3 4 5 提供的資訊非常有用

（4）廣告內容和我完全無關 1 2 3 4 5 廣告內容和我非常有關

21. 你覺得這則廣告的代言人在以下表述中應該獲得多少分： （1）她其實完全不了解營養標籤 1 2 3 4 5 她確實非常了解營養標籤

（2）她其實從來不閱讀營養標籤 1 2 3 4 5 她確實經常閱讀營養標籤

（3）她其實完全不懂得利用營養

標籤作出健康的選擇

1 2 3 4 5 她確實非常懂得利用營養

標籤作出健康的選擇

（4）她絕對不是營養專家 1 2 3 4 5 她絕對是營養專家

（5）她完全不值得信賴 1 2 3 4 5 她非常值得信賴

（6）她外貌完全不吸引 1 2 3 4 5 她外貌非常吸引

（7）她完全不美麗 1 2 3 4 5 她非常美麗

（8）她外表完全不優雅 1 2 3 4 5 她外表非常優雅

22. 看完廣告後，你覺得營養標籤對你保持均衡飲食有用嗎?

完全無用 1 2 3 4 5 非常有用

23. 看完廣告後，你覺得營養標籤對你維持健康，減少患病風險有用嗎？

24. 閱讀營養標籤會令你厭煩嗎？

非常煩 1 2 3 4 5 完全不煩

25. 你覺得容易讀懂營養標籤上的資料嗎？

非常難讀懂 1 2 3 4 5 非常容易讀懂

26. 看完廣告後，你會呼籲更多人利用營養標籤選擇食物嗎?

27. 下次挑選預先包裝食物時，你會閱讀包裝上的營養標籤嗎？

一定不會 1 2 3 4 5 一定會

28. 看完廣告後，你會提醒家人在挑選預先包裝食物時閱讀包裝上的營養標籤嗎？

一定不會 1 2 3 4 5 一定會

問卷結束，非常感謝你的合作！

完全無用 1 2 3 4 5 非常有用

一定不會 1 2 3 4 5 一定會

87

APPENDIX B: EXPERIMENTMASTERIAL‐‐‐ INFORMATIVE ARGUMENT X DIETITIAN A;TESTIMONIALARGUMENTXDIETITIANA

88

APPENDIX C: EXPERIMENT MASTERIAL‐‐‐ INFORMATIVE ARGUMENT X DIETITIAN B;TESTIMONIALARGUMENTXDIETITIANB

89

APPENDIXD:EXPERIMENTMASTERIAL‐‐‐INFORMATIVEARGUMENTXCELEBRITYA; TESTIMONIALARGUMENTXCELEBRITYA

90

APPENDIXE:EXPERIMENTMASTERIAL‐‐‐INFORMATIVEARGUMENTXCELEBRITYB; TESTIMONIALARGUMENTXCELEBRITYB

91

APPENDIXF:EXPERIMENTMASTERIAL‐‐‐INFORMATIVEARGUMENTXORDINARYPERSON;TESTIMONIALARGUMENTXORDINARYPERSON

92

APPENDIXG:APROPOSEDPSAOFNUTRITIONLABELS