"u

an

(r-rvxi a

.31.. V}
5.. ‘24

3.... ,.... . .C
. . . ‘ . ...‘.u.......n. .
I .5 if . x.
.3; :5} . . \.:I

3
.. 1
A .v ..:«.v

.1. .....

.r.r iv.

a" a

'LO

v-Lw

A.

W‘Iﬂﬂvlﬁ‘l

I I .4 («up

Me:

wad:

a ‘ . o‘otow'rvO-r.0‘0

. P

v. 1.: .. ...

 

 

This is to certify that the
dissertation entitled

Americans’ Intentions to Vacation in East Asia:
An Empirical Investigation of the Relationships among
Information Source Usage, Destination Image,
Perceived Risks, and Intention to VISit

presented by

Jeonghee Noh

has been accepted towards fulﬁllment
of the requirements for the

Park, Recreation, and
Tourism Resources

Ph.D degree in

 

 

WM

Major Professor’s Signature

Date

MSU is an Al‘ﬁnnativo ActiorVEquaI Opportunity Institution

iiiii —> 7 v

LIBRARY
Michigan State
University

 

 

 

 

a.-.-n-Con-O.-.-o--a-o-o--aO0-:-n-o-o---o-c-n-n---o--‘--.-n----uno-v-o-o-u-s-u-u-u-u-o-u-o-

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/C|RC/Dale0ue.indd-p.1

 

AMERICANS’ INTENTIONS TO VACATION IN EAST ASIA:
AN EMPIRICAL INVESTIGATION OF THE RELATIONSHIPS AMONG
INFORMATION SOURCE USAGE, DESTINATION IMAGE, PERCEIVED RISKS,
AND INTENTION TO VISIT

By

Jeonghee Noh

A DISSERTATION

Submitted to
Michigan State University
In partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Park, Recreation, and Tourism Resources

2006

ABSTRACT

AMERICANS’ INTENTIONS TO VACATION IN EAST ASIA:
AN EMPIRICAL INVESTIGATION OF THE RELATIONSHIPS AMONG
INFORMATION SOURCE USAGE, DESTINATION IMAGE, PERCEIVED RISKS,
AND INTENTION TO VISIT

By
Jeonghee Noh

In 2004, 22.3 million Americans vacationed overseas, increasing by 11 percent
since 2003 (TIA, 2005). However, only 10 percent of these tourists visited China, Japan,
or South Korea. To attract more American tourists to these destinations, East-Asia-based
tourism marketers must better understand which factors contribute to Americans’
decisions to vacation for the ﬁrst time in China, Japan, and South Korea.

The aim of this research was to deveIOp, and to empirically test, a model
explaining Americans’ intentions to vacation for the ﬁrst time in China, Japan, and South
Korea. Based on a review of the relevant literature, information source usage, cognitive
and affective destination image, and perceived risks were identiﬁed as key factors

inﬂuencing travelers’ intentions to visit East Asia.

Data were collected using a self-administered, on-site survey. Only American
citizens, 18 years of age or older, who had never visited‘China, Japan, or South Korea,
were included in the participant sample. Structural equation modeling (SEM) was used to

test the proposed conceptual model relative to these data.

The results of the SEM analyses showed that respondents reporting both a high
(emotionally neutral) cognitive destination image and a high (emotionally positive)
affective destination image of China, Japan, and South Korea, as well as low perceived

risks of vacationing in these three countries, had greater intention to visit them. Similarly,

respondents reporting a high cognitive image of China, Japan, and South Korea and low
perceived risks of vacationing in these countries also reported having high affective image
of these destinations. Additionally, respondents who frequently obtained information on
China, Japan, and South Korea were found to have a high cognitive image of these three
countries and low perceived risks of vacationing there. However, information source
usage was not found to directly influence affective image. SEM analysis showed that
information source usage indirectly inﬂuenced affective image through cognitive image
and perceived risks. Cultural experiences was the information source most strongly
correlated with Americans’ intentions to visit China, Japan, and South Korea. Similarly,
for all three countries the strongest cognitive image items were interesting historical
attractions, many things to see and do, and unique cultural resources.

Despite these similarities in participants’ responses for China, Japan, and South
Korea, some differences were identiﬁed. One key difference was that, depending on the
country, a different factor was shown to be most strongly correlated with intention to visit;
for China, it was affective image; for Japan, it was cognitive image; and, for South Korea,
it was perceived risks.

These ﬁndings suggest that tourism marketers should jointly sponsor cultural
events in the US. to promote vacations in China, Japan, and South Korea. Joint trips, in
which Americans would visit similar attractions in all three countries could also be
marketed. Not only were cultural experiences identiﬁed as an information source to be
studied in ﬁrture research, but study results also demonstrated the value of studying
cognitive and affective destination image as being distinct but interrelated. A conceptual
model is presented that might be tested in the future regarding other underperforrning

international tourism regions.

Copyright by
J eonghee Noh
2006

First and foremost, this dissertation is dedicated to my parents,
my father MyungSun Noh and my mother ChoonJa Lim,
for their unconditional love and support,
which has given me the conﬁdence to face the world.

ACKNOWLEDGEMENTS

My journey to completing my doctoral program has been long and arduous.
Without the considerable support of the people mentioned below, it would not have been
possible. In this space, I express my sincere appreciation.

I wish to thank the members of my dissertation committee for their constant
encouragement in helping me to complete this dissertation.

I am extremely grateful for the invaluable support that I have received from my
dissertation director and committee chair, Dr. Donald Holecek. Providing the right
balance of challenge and support as I drafted my dissertation, Dr. Holecek was a true
mentor. His wisdom will continue to guide me long after I have left Michigan State
University.

Dr. Carl Brochegrevink, a professor in Hospitality Business, served as my
academic advisor throughout my Master’s and doctoral programs, as well as one of my
dissertation committee members. Dr. Brochegrevink has always been there to listen, to
counsel, and to inspire. I have greatly appreciated his advice on navigating the Ph.D. and
academic life beyond. Dr. Christine Vogt generously shared her insights on the research
methods employed in this study. Dr. Vogt has been my role model as a researcher, as a
teacher, and as a woman. Dr. Sarah Nicholls also offered useful advice, especially during
the writing process.

My friends and colleagues also deserve my sincere thanks for their support.

‘ Soyeon Ahn, Jaemin Cha, Sejung Choi, KaiLonnie Dunsmore, Chang Huh, Hyunsoon
Park, SungHee Park, MiKyung Kim, Miran Kim, SeungHyun Kim, Eunsil Lee, Michael

vi

Morris, Lori Langone, Dottie Schmidt, Charles Jinyen Shih, Kyungoh Song, and Mark
VerNooy. I will never forget the happy moments that we have shared together and the
tough times that I would not have overcome without them.

I am especially grateful to J inyoung Choi for her endless encouragement, smile,
and warmth. Together, we have celebrated many milestones in the United States.
Whenever I needed true understanding from a friend, J inyoung was always there. My
special thanks also go to Ann Lawrence, a friend and mentor throughout all of the
dissertation process, whose help enabled me to ﬁnish this hard process. I cannot articulate
sufﬁciently how much I thank J inyoung Choi and Ann Lawrence.

Additionally, I would like to thank Hyunjoo Choi for assisting me with collecting
data for the dissertation study. Without Hyunjoo’s enormous charisma, the sample size for
this study would no doubt have been much smaller and the ultimate ﬁndings, less
interesting.

Most importantly, I would like to dedicate this dissertation to my parents,
MyungSun Noh and ChoonJ a Lim, and to my brother Kookjin and sisters Yoonhee and
Heejung. I thank them for loving, supporting, and praying for me. Despite traveling a
great distance, their love has always reached me. To my family, I owe all that I have

become.

vii

TABLE OF CONTENTS

LIST OF TABLES .......................................................................................................... xi

LIST OF FIGURES ........................................................................................................ xiii
CHAPTER 1

INTRODUCTION .......................................................................................................... 1

Purpose of the study ................................................................................................ 4

Organization of the Dissertation ............................................................................. 5
CHAPTER 2

LITERATURE REVIEW ............................................................................................... 6

Destination Image ................................................................................................... 6

Components of Destination Image .................................................................. 11

Destination Image Formation Processes and Information Sources ................ 12

Information source usage ...................................................................... 15

Direct experiences with the destination ................................................ 17

Proximity to the destination .................................................................. 19

Inﬂuence of Destination Image on Travelers’ Visit Intentions ...................... 20

Perceived Risks of Vacationing at a Destination .................................................... 21

Types of Risks Associated with the Destination and Their Inﬂuences on

Travel Behavior .............................................................................................. 22

Role of Information Sources in Perceived Risks ............................................ 25

Proposed Conceptual Model ................................................................................... 27

Hypothesized Relationships ............................................................................ 28
CHAPTER 3

RESEARCH DESIGN AND METHODOLOGY .......................................................... 29

Data Collection and Analysis Procedures ............................................................... 29

Data Collection ............................................................................................... 29

Sub-sample Size .............................................................................................. 30

Proﬁle of Respondents .................................................................................... 31

Survey Instrument and Scales ................................................................................. 33

Destination Image ........................................................................................... 33

Perceived Risks ............................................................................................... 34

Information Source Usage .............................................................................. 34

Intention to Visit ............................................................................................. 34

viii

Data Analysis .......................................................................................................... 35

CHAPTER 4
RESULTS ....................................................................................................................... 40
Preliminary Analyses .............................................................................................. 41
Data Screening ................................................................................................ 41
Description of the Variables ........................................................................... 43
Analysis of Mean Differences among the Three Countries ............................ 51
Testing the Measurement Model ............................................................................ 55
The China Model ............................................................................................ 58
Assessment of overall model ﬁt ............................................................ 58
Assessment of reliability and validity ................................................... 59
The Japan Model ..................................................... . ....................................... 61
Assessment of overall model ﬁt ............................................................ 61
Assessment of reliability and validity ................................................... 62
The South Korea Model .................................................................................. 64
Assessment of overall model ﬁt ............................................................ 64
Assessment of reliability and validity ................................................... 65
Comparison of the Measurement Model for the Three Countries. ................. 67
Testing the Structural Equation Model ................................................................... 67
The China Model ............................................................................................ 69
The Japan Model ............................................................................................. 71
The South Korea Model .................................................................................. 73
Comparison of the Structural Equation Model for the Three Countries ......... 75
Hypotheses Testing ................................................................................................. 75
The China Model ............................................................................................ 76
Direct inﬂuence related to study hypotheses ........................................ 77
Multiple squared correlation ................................................................. 79
Indirect and total inﬂuences of the factors on visit intention ................ 79
The Japan Model ............................................................................................. 80
Direct inﬂuence related to study hypotheses ........................................ 82
Multiple squared correlation ....................................................... 84
Indirect and total inﬂuences of the factors on visit intention ................ 84
The South Korea Model .................................................................................. 86
Direct inﬂuence related to study hypotheses ........................................ 87
Multiple squared correlation ................................................................. 89

ix

Indirect and total inﬂuences of the factors on visit intention ................ 89
Comparison of the results of hypotheses testing for the three East Asian

countries .......................................................................................................... 90
Additional Analysis:
The Comparative Inﬂuence of Different Types of Information Source on
Destination Image and Perceived Risks .......................................................... 93
CHAPTER 5
DISCUSSION, IMPLICATIONS, AND CONCLUSIONS ........................................... 95
Summary and Discussion of the Findings .............................................................. 95
Mean Differences of the Variables among Three East Asian Countries ........ 96
Results of Hypotheses Testing ........................................................................ 99
Overall Pattern of Relationships among the Factors across the Three
Countries ......................................................................................................... 104
Relationships between Types of Information Source Used and
Destination Image and Perceived Risks .......................................................... 105
Implications ............................................................................................................. 106
Theoretical Implications ................................................................................. 106
Practical Implications ...................................................................................... 108
Limitations of this Study and Suggestions for Future Research ............................. 111
Conclusions ............................................................................................................. 1 12
APPENDIX: Survey Questions ...................................................................................... 114
REFERENCES ............................................................................................................... 1 19

LIST OF TABLES

Table 1. Overview of previous tourism research on destination image ........................... 7
Table 2. Distribution of respondents who have and have not visited China, Japan, ....... 31
Table 3. Characteristics of respondents who have never visited China, Japan, and
South Korea. ...................................................................................................... 32

Table 4. Constructs and observed variables in the proposed model. ............................... 37
Table 5. Normalitya test results for variables included in the proposed model ............... 42
Table 6. Mean and SD of intention to visit the three countries. ...................................... 43
Table 7. Mean and SD of cognitive destination image of the three countries. ................ 44
Table 8. Mean and SD of affective image of the three countries. ................................... 46
Table 9. Mean and SD of perceived risks of visiting the three countries. ....................... 48
Table 10. Mean and SD of information source usage (type and frequency) for the

three countries. ................................................................................................. 50
Table 11. Results of Repeated Measure ANOVA. .......................................................... 52
Table 12. The recommended goodness-of-ﬁt indices ...................................................... 57
Table 13. Comparison of the proposed and modiﬁed full measurement models for

China. ............................................................................................................... 59
Table 14. Standardized factor loadings for the modiﬁed measurement model for

China. ............................................................................................................... 60
Table 15. Correlation analysis of factors for China. ........................................................ 61
Table 16. Comparison of the proposed and modiﬁed full measurement models for

Japan. ............................................................................................................... 62
Table 17. Standardized factor loadings for the modiﬁed measurement model for

Japan. ............................................................................................................... 63
Table 18. Correlation analysis of factors for Japan. ........................................................ 64
Table 19. Comparison of the initial and modiﬁed full measurement models for South

Korea ....... 65
Table 20. Standardized factor loadings for the modiﬁed measurement model for

South Korea. .................................................................................................... 66
Table 21. Correlation analysis among factors for South Korea. ...................................... 67
Table 22. Comparison of the initial and modiﬁed structural models for China. ............. 71
Table 23. Comparison of the initial and modiﬁed structural models for Japan. .............. 73
Table 24. Comparison of the initial and modiﬁed structural models for South Korea... 75
Table 25. Results of hypotheses testing for China ........................................................... 76

Table 26. Standardized estimates for direct, indirect, and total inﬂuences for China. 80

xi

Table 27. Results of hypotheses testing for Japan. .......................................................... 81

Table 28. Standardized estimates for direct, indirect, and total inﬂuences for Japan ...... 84
Table 29. Results of hypotheses testing for South Korea. ............................................... 86
Table 30. Standardized estimates for direct, indirect, and total inﬂuences for South

Korea. ............................................................................................................... 89
Table 31. Results of hypotheses testing across three East Asian countries. .................... 91
Table 32. Comparison of total inﬂuence .......................................................................... 92

Table 33. Correlation between types of information source used and cognitive image
and perceived risk. ........................................................................................... 93

xii

LIST OF FIGURES

Figure l. A model of a tourist’s image formation process ............................................... 14
Figure 2. Theoretical Model of Intention to Visit Three East Asia countries .................. 27
Figure 3. The proposed measurement model for the three countries. .............................. 56
Figure 4. The pr0posed structural equation model. ......................................................... 68
Figure 5. Modiﬁed SEM for China. ................................................................................. 70
Figure 6. Results of testing the proposed structural model with standardized path
coefficients and squared multiple correlations (R2 ) for China. ....................... 77
Figure 7. Results of testing the proposed structural model with standardized path
coefﬁcients and squared multiple correlations (R2) for Japan. ........................ 82
Figure 8. Results of testing the proposed structural model with standardized path
coefﬁcients and squared multiple correlations (R2) for South Korea. ............. 87
Figure 9. Proposed conceptual model .............................................................................. 99

xiii

CHAPTER 1
INTRODUCTION

International leisure travel is an important market for the international tourism
industry in terms of its size and economic importance. According to the World Tourism
Organization’s Tourism 2020 Vision (2006), international tourist arrivals are expected to
reach over 1.6 billion in the year 2020. This is more than twice the number in 2000. The
number of American overseas leisure travelers has also increased from 2003 to 2004.
Approximately 22.3 million American visited overseas destinations in 2004: an increase
of 11.2 percent since 2003 (Travel Industry Association of America, 2005). Of the 7.6
percent of Americans who vacationed abroad in 2004, 17 percent visited Asian countries,
while only 9.8 percent visited China, Japan, or South Korea. A much larger portion of
American travelers vacationed in other destinations, such as Europe (43 percent of the
total) and the Caribbean (20 percent of the total). These American travelers are a potential
market for East-Asia-based tourism marketers.

Despite the promise of the East Asian tourism market, most tourism research has
been conducted on the largest international tourism markets (America and Europe) by
primarily American and European tourism researchers. To capitalize on Americans’
interest in vacationing in East Asian countries, East Asian tourism marketers must
understand why relatively few Americans have decided to vacation in China, Japan, and
South Korea, and how to attract larger numbers of potential American tourists to East
Asian countries. However, tourism researchers must ﬁrst provide tourism marketers with

this information. This study, conducted by a South Korean tourism researcher, was an

attempt to identify the factors inﬂuencing Americans’ intentions to vacation in China,
Japan, and South Korea for both international tourism researchers and East-Asia-based
tourism marketers.

For several decades, research in travel and tourism has demonstrated that
destination image is a critical factor inﬂuencing travelers’ decision-making behavior
(Court & Lupton, 1997; Milman & Pizam, 1995; Rittichainuwat, Qu, & Brown, 2001;
Ross, 1993; Urn & Crompton, 1992; Woodside & Lysonski, 1989). Positive images of a
destination encourage travelers to choose it over other possible destinations. For this
reason, destination marketers have competed progressively more intensely to produce
positive destination images in travelers’ minds, and have tried to develop positions that
differentiate their destinations from other attractive sites (Baloglu & McCleary, 1999b).
Based on research reporting the strengths and weaknesses of their destinations, tourism
marketers have developed marketing programs to create and enhance favorable images, or
correct negative images, of their destination. For East-Asia-based tourism marketers to
design programs that entice prospective American tourists to China, Japan, and South
Korea, a thorough understanding of Americans’ images of these three countries is crucial.

Numerous tourism researchers have examined prospective and actual tourists’
destination images of cities, regions, states, and countries in order to help tourism
marketers design effective positioning strategies. Destination image studies have
compared travelers’ evaluations of tourism-related attributes across particular destinations
(Goodrich, 1978; Hunt, 1975; Javalgi, Thomas, & Rao, 1992) and identiﬁed which
attributes are strong or weak, given the destination (Ahmed, 1991; Chou, 1991; Court &

Lupton, 1997; Gartner & Hunt, 1987). However, past tourism research has tended to

center on cognitive destination image (knowledge about a place), largely overlooking
affective destination image (feelings about a place).

Contrary to this trend, recent studies (Baloglu & Brinberg, 1997; Gartner, 1993)
have proposed that both travelers’ cognitive and affective destination images deserve
consideration. The present study attempted to extend this recent body of knowledge on
destination image by employing a two-dimensional framework (cognitive and affective)
to assess Americans’ images of China, Japan, and South Korea.

Tourism research has also shown that travelers consider their perceived risks of
vacationing in a destination, as well as their image of the destination in general, before
making travel decisions (Coshall, 2003; Lam & Hsu, 2006). Even though travelers may
perceive a destination as having many attractive features (e. g., interesting cultural
resources and good night life), if they consider vacationing in this destination to be a risk,
they may decide not to visit the destination (Crompton, Fakeye, & Lue, 1992).

Perceived risks of vacationing in a destination have become especially inﬂuential
on travelers’ decision-making behavior in recent years, given international awareness of
tragic events occurring around the globe (Coshall, 2003; Faulkner, 2001; Floyd, Gibson,
Pennington-Gray, & Thapa, 2003; Mazzocchi & Montini, 2001 ). For example, travel in
the Asia-Paciﬁc region has been devastated by Severe Acute Respiratory Syndrome
(SARS), bird ﬂu, and the tsunami that struck Southeastern Asia. Because of the
considerable inﬂuence of tourists’ perceived risks of vacationing in a destination on their
travel decisions, East-Asia-based tourism marketers must understand American travelers’
reasons for avoiding China, Japan, and South Korea, as well as their reasons for visiting

these countries.

 

 

Different types of perceived risk affect tourists’ travel decisions (Roehl &
Fesenmaier, 1992). However, previous tourism studies have isolated travelers’ perceived
risks (e. g., terrorism, natural disasters, and disease) in considering their effects on travel
behavior (Coshall, 2003; Faulkner, 2001; Floyd et a1., 2003; Gartner & Shen, 1992;
Mazzocchi & Montini, 2001; Sonmez & Graefe, 1998a, 1998b). These studies have not
considered the combined effect of these individual perceived risks on intention to visit.
Such comprehensive information would be useful to East-Asia—based tourism marketers
seeking to attract ﬁrst-time American tourists to China, Japan, and South Korea.

Travelers’ image of a destination in general and perceived risks of vacationing in
a destination are determined by the types of information source travelers consult on a
destination (Baloglu, 1999; Beerli & Martin, 2004b; F akeye & Crompton, 1991; Gunn,
1972). Although several tourism studies have compared the differential effects on travel
behavior of direct experience and indirect information source usage, no study before the
present dissertation has investigated the inﬂuence of type and frequency of indirect
information source usage on destination image and perceived risks regarding China,
Japan, and South Korea, ﬁndings which would greatly beneﬁt East-Asia—based tourism

marketers.

Purpose of the Study
Given these gaps in previous tourism research, the purpose of the present study
was to develop and empirically test a conceptual model based on the ﬁndings of previous
tourism studies that would explain prospective American tourists’ intentions to visit

China, Japan, and South Korea for the ﬁrst time. To this end, the present study sought to

determine the interrelationships among information source usage (type and frequency),
cognitive and affective destination image, perceived risks, and visit intention, in order to
offer both theoretical and practical recommendations to international tourism researchers

and East-Asia—based tourism marketers.

Organization of the Dissertation
A general background for the present study and discussion of its purpose and

signiﬁcance are provided in Chapter One. Relevant literature is reviewed in Chapter Two.

 

The methodological speciﬁcations of this study, including procedures relating to study
design, data collection, and data analysis are detailed in Chapter Three. The ﬁndings of
this study are presented in Chapter Four. Finally, theoretical and practical implications

derived from these ﬁndings are outlined in Chapter Five.

CHAPTER 2
LITERATURE REVIEW

Previous studies were reviewed in order to develop a conceptual model that
explains American tourists’ intentions to visit China, Japan, and South Korea. The results
of these studies are summarized in the three sections of this chapter. In the ﬁrst section,
previous research on destination image—its components, its formation processes as they
relate to information source usage on a destination, and its inﬂuence on tourists’
intentions to visit a destination——is presented. In the second section, previous studies on
travelers’ perceived risks of vacationing in a destination are discussed; speciﬁcally, types
of perceived risk, the relationship between perceived risks and information source usage,
and the inﬂuence of perceived risks on visit intention. In the ﬁnal section, previous
research is synthesized to create a conceptual model that would potentially clarify the
interrelationships among destination image, perceived risks, and information source usage,

and explain American tourists’ intentions to visit China, Japan, and South Korea.

Destination Image
In tourism research, destination image has been considered as an important
concept in explaining travelers’ decision-making processes and tourism behavior.
Assessing destination image helps destination marketers to design effective positioning
strategies, which create, correct, or enhance target travelers’ destination image (Pike &
Ryan, 2004). Researchers in the tourism ﬁeld began to conceptualize destination image in

the 19703. Although a number of researchers have attempted to conceptualize the

construct, a precise deﬁnition of this term is problematic. Generally, destination image
refers to the sum of beliefs, ideas, and impressions that people have about a destination
(Baloglu & Brinberg, 1997; Crompton, 1979). The image-related literature has focused
primarily on the following issues: (1) the components of destination image, (2)
destination image formation processes in relation to information sources, and (3) the
inﬂuence of destination image on travelers’ behavior. Highlights from destination image
studies conducted since 1975 on American tourists are grouped according to destination
and presented in Table 1.

Table 1. Overview of previous tourism research on destination image.

1) In the US.

 

 

Author(s) Destination Destination Image(s) 1m! age(s)l
Hunt (1975) Colorado, Montana, Utah, Disparities between the summer and Cognitive
and Wyoming winter climates at home and the

destination caused respondents to
have a negative image of the

 

destination.
Goodrich California, Florida, and California was perceived to possess Cognitive
(1978) Hawaii suitable accommodations,

entertainment, and cuisine. Florida
and Hawaii were perceived to
possess water sports, golﬁng, and
suitable accommodations.

 

Fridgen (1987) Michigan Respondents familiar with the Cognitive
destination had a more favorable
image of the destination.

 

Gartner & Hunt Utah Respondents living closer to the Cognitive
(1987) destination were more impressed by
the state than respondents living
farther away. Those who had
previously visited the destination
had a generally more favorable
image of the state than those who
had never visited it.

 

 

 

 

 

Table 1: Overview of previous tourism research on destination image (cont’d).

 

 

 

 

 

 

 

 

 

Respondents living in nearby states
answered that the destination was
more friendly and less unique as a
vacation destination, and that it had
many more places of interest to
visit, than did respondents living
farther away.

 

1) In the US.
Author(s) Destination Destination Image(s) Ilmage(s)l
Reilly (1990) Montana and Montana Respondents living closer to Cognitive
ski area destination had a stronger and more
positive image of the state. Potential
visitors from distant areas lacked a
___W _ _ W ,, vivid image of the state. .- _______-__...
Ahmed (1991) Utah Perceptions were generally more cognitive
favorable for those who had
, .. W WW - W previously visited destination. m... -__ W -
Chon, Weaver Norfolk, Virginia The three most satisfying features Cognitive
& Kim (1991) were suitable accommodations, easy
access, and the variety and quality

. - _ _W __W WW _ W of attractions. -_.- __ _._____ . W
Fakeye & Lower Rio Grande Valley Repeat visitors are well aware of the Cognitive
Crompton in Texas destination’ 3 attributes.

(1991) The more frequent the visit to the
destination, the more recognizable
and appreciated its social
__ _, ______,...-_._ ,_ . opportunities and attractions. _ , _W_ W __
Milman & Central Florida Respondents familiar with the Cognitive
Pizam (1995) destination had a more positive
image of it and were more interested
in the destination and more likely to
_. ____ WW -- revisit. , . , .__
Court & Lupton New Mexico A positive relationShip existed Cognitive
(1997) between prior experiences with the
destination and intention to visit:

___-.._._... _ __ W ._ . W Climate was not inﬂuential. -_.._ W_____WW
Vogt & Arizona Whereas cognitive image improved Cognitive
Anderecek during a trip, affective Image, when And
(2003) positive before a trip, remained Affective

___ _ ___ . . stable. ,. _...__ -_-
Hsu, Wolfe & Kansas Visitors had a more positive image Cognitive
Kang (2004) of the destination than non-visitors.

 

 

 

Table 1: Overview of previous tourism research on destination image (cont’d).

2) Near the US.

 

Author(s)

Destination

Destination Image(s)

Image(s)
Assessed

 

Goodrich
(1978)

Mexico and ﬁve West
Indian Islands (Bahamas,
Barbados, Jamaica,
Puerto Rico, and Virgin
Islands)

Five West Indian Islands were
perceived as being similar. These
areas were perceived as having
rest, relaxation, and water sports.
Mexico was perceived as being
an historical and cultural

interesting site.

Cognitive

 

Crompton
( l 979)

Mexico

There were signiﬁcant
differences between ideal images
and actual images of the
destination as a tourism site.
Respondents living farther away
from the destination had a more
favorable image of it. The two
most positive attributes were
climate and low cost.

Cognitive

 

Joppe, Martin
&Waalen
(2001)

Toronto, Canada

American visitors were
concerned about personal safety,
value for money, and the
hospitality of local people.

The three most satisfying
features were attractions,
transportation, and shopping
facilities.

Cognitive

 

3) In Eurom

 

Author(s)

Destination

Destination Image(s)

Image(s)
Assessed

 

Phelps (1986)

Minorca

First-time visitors were more
concerned about violence and
poverty, and had a vague
impression of the destination.
Retum visitors had a detailed
impression of the destination and
were more interested in natural
environment and traditional
cultural sites.

Cognitive

 

Pizam(l991)

U.S.S.R.

Post- trip respondents showed
more positive images than did
pre-trip respondents.

Cognitive and
Affective

 

Javalgi,
Thomsa & Rao
( 1 992)

 

British Isles, Central
Europe, Scandinavian
countries, and Southern
Europe

 

Better-educated, employed and
retired travelers with a higher
annual income chose this
destination. Younger travelers
(aged 24 and under) were more
attracted to Southern Europe than
to the other three regions.

 

Cognitive

 

Table 1: Overview of previous tourism research on destination image (cont’d)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3) In Europe
. . . . Image(s)
Author(s) Destlnatlon Dest1nat10n Image(s) Assessed

Baloglu & Four Mediterranean Most cognitive images were Cognitive and

McCleary destinations (Egypt, similar between visitors and Affective

(1999b) Greece, Italy, Turkey,) non-visitors. Egypt was
perceived less positively than
Turkey, Italy, and Greece.

Baloglu (2001) Turkey Age and education positively Cognitive and
correlated with familiarity and Affective
familiarity inﬂuenced
destination image.

Respondents more familiar with
the destination had a more
positive image of the
destination.
4) Asia
. . . . Image(s)
Author(s) Destlnation 1 Destmatlon Image(s) Assessed
Chon (1991) South Korea Y Post trip respondents showed Cognitive
more positive images than pre-
trip respondents.
Gartner & Shen China Aﬁer the Tiananmen Square Cognitive
(1992) conflict, tourists showed more
negative images of China
.. _- chanaIedto before the event.
Choi, Chan & Hong Kong The destination was considered Cognitive
Wu (1999) as a “shopping paradise” but also
as a crowded, busy, and stressful
urban environment.
Rittichainuwat Thailand The more positive and the fewer Cognitive
(2001) negative images of a destination,
the more likely travelers were to
return to that destination.
Suh & Seoul, South Korea Business travelers’ were more Cognitive and
Gartner (2004) likely to focus on security, Affective
opportunities to increase
knowledge, and overall
atmosphere; pleasure travelers
were more likely to focus on
opportunities for adventure and
overall cost.

Kim & South Korea Respondents having either a Cognitive

Morrsion negative image or no image of

(2005) the destination had a positive

 

 

image after watching matches on
TV hosted by the destination.

 

 

10

 

Componenm of Destination Image

The most recent studies (Baloglu & Brinberg, 1997; Gartner, 1993) tend to agree
that destination image is composed of two components: cognitive and affective. Cognitive
image includes beliefs and knowledge about destination attributes, whereas affective
image comprises feelings, moods, and emotions toward a destination. Cognitive
destination image is, thus, emotionally neutral, while affective destination image is either
emotionally positive (e. g., arousing, pleasant, exciting, and relaxing) or negative (e.g.,
sleepy, unpleasant, gloomy, and distressing).

As shown in Table 1, the earliest destination image studies focused only on
cognitive destination image (Ahmed, 1991; Chon, Weaver, & Kim, 1991; Gartner & Hunt,
1987; Goodrich, 1978; Hunt, 1975; Phelps, 1986; Reilly, 1990). Some of this research
compared travelers’ evaluations of cognitive destination attributes across particular
destinations (Goodrich, 1978; Hunt, 1975; Javalgi et al., 1992) in order to design
positioning strategies for these destinations. For example, in their study of American
tourists’ images of European destinations (the British Isles, Central Europe, Scandinavia,
and Southern Europe), Javalgi, Thomas, and Rao (1992) found that travelers had different
knowledge of each destination. On the other hand, some tourism researchers sought to
identify which cognitive image attributes were strong and weak for a target destination in
order to develop more effective tourism-marketing strategies (Ahmed, 1991; Chon et al.,
1991; Court & Lupton, 1997; Gartner & Hunt, 1987).

Affective destination image has traditionally been overlooked in tourism research.

Reviewing the 142 studies on destination image from 1973 to 2000, Pike (2002) found

11

 

that only six studies examined affective image attributes. Of the studies that focused on
Americans’ destination images, only 4.2 percent examined their affective component.

Recently, some researchers have argued that total destination image cannot be
understood without considering both its cognitive and affective dimensions (Baloglu &
Brinberg, 1997; Gartner, 1993). Baloglu and Brinberg (1997) noted that cognitive
destination image is insufﬁcient in explaining travelers’ complex of impressions about a
place. Using Russell, Ward, and Pratt’s (1981) binary framework (e.g.,
unpleasant/pleasant, and sleepy/arousing) to measure affective destination image, Baloglu
and Brinberg (1997) sought to develop a more complete understanding of travelers’
destination image. Their study provided empirical evidence supporting the applicability
of this framework to tourism contexts.

Baloglu and McCleary (1999a) later developed a path analytic model of
Americans’ destination image formation regarding four Mediterranean countries, a model
which addressed both its cognitive and affective components, employing the term overall
image to encompass these two destination image components. Baloglu and McCleary
found that destination image formation proceeded sequentially: While cognitive and
affective image directly inﬂuenced travelers’ overall image, cognitive image also
inﬂuenced affective image. Baloglu and McCleary’s study demonstrated that cognitive

and affective image components are distinct but interrelated.

Destination Image Formation Processes and Information Source Usage
The existing literature suggests that a tourist’s image of a destination is formed through a

multi-stage process. According to Gunn (1972), the initial image formation is largely a

12

function of information source usage before the ﬁrst visit to the destination. Gunn noted
that destination image formation proceeds in two stages: organic image followed by
induced image. Gunn asserts that organic image is formed by non-touristic and non-
commercial information sources, such as popular culture, education, news reports in print
and on TV, word of mouth from family and friends, and other non-tourism-speciﬁc
information sources. However, Gunn states that induced image is a function of destination
marketing. Adapting Gunn’s model, F akeye and Crompton (1991) added complex image
to their model of a tourist’s destination image formation process. In their model,
destination image formation proceeds in three stages: (1) organic image, (2) induced
image, and (3) complex image. Figure 1 shows a model of the image formation process
developed by Fakeye and Crompton. (Italicized portions have been added for

clariﬁcation.)

l3

 

 

Exposure to non-touristic
information sources

p——_————-—-———-

 

 

 

 

Organic Image

 

 

 

 

 

Active information
search and processing of
commercial materials

 

 

 

‘mm-m--------------

 

 

 

Evaluation of
alternative
destinations’ beneﬁts
and images

Induced Image

 

 

 

 

 

 

 

 

 

 

Destination selection
and visitation

 

 

 

 

 

 

 

Complex Image

 

 

 

 

 

 

Figure 1. A model of a tourist’s image formation process. Modiﬁed from Fakeye and
Crompton (1991).

According to Fakeye and Crompton’s model, ﬁrst, organic image is developed in
travel customers through exposure to non-touristic and non-commercial information
sources. This information familiarizes potential tourists, to some extent, with a destination
and provides them with some knowledge of a destination. Second, induced image is
produced when travel customers, having decided to take a trip, seek out promotional
materials from tourism suppliers. This information is often called persuasive promotion

and is most effective once prospective tourists have already developed an organic image

14

of a destination. Third, travel customers achieve a complex image of a destination once
they have visited this place and had the opportunity to integrate their own impressions of
the destination with their previous organic and induced images of the place. Complex
image may be reactivated through promotional materials, once travel customers have
returned home, and positive complex image has been shown to increase repeat visits and
favorable word of mouth.

A precursor to Fakeye and Crompton (1991), Phelps (1986) divided destination
image into primary and secondary destination image: primary, being the image formed
before the ﬁrst visit to the destination; and secondary, being the modiﬁed image resulting
from personal experience with the destination. Phelps’s and F akeye and Crompton’s
models suggest that travelers’ image of a destination may change after visitation. Thus, it
would seem that both information sources about a destination and direct experiences with
a destination are critical factors in shaping travelers’ destination images.

Information source usage. Information sources refer to all of the resources from
which prospective tourists gather impressions of a destination before their (next) actual
visit. Information sources can be categorized as being either internal or external. Internal
information sources include personal experiences, either with a speciﬁc destination or
with a similar destination, and the knowledge accumulated through an ongoing
information search (Fodness & Murray, 1997; Fodness & Murray, 1998; Gursoy, 2003;
Vogt & F esenmaier, 1998). External information sources include the following four
resources: (1) family and friends (i.e., word of mouth advice from friends and relatives),

(2) destination-speciﬁc literature, (3) media, and (4) travel consultants (Snepenger &

15

Snepenger, 1993). Additionally, many tourism researchers consider the Internet as a ﬁfth
external information source (Lang, 2000).

It is important to understand the role of information source usage in shaping
tourists’ destination image because destination image has been shown to be a critical
factor in tourists’ intentions to visit a destination. Several studies have demonstrated how
information source usage inﬂuences travelers’ destination image, which, in turn, affects
their visit intention (Baloglu, 1999; Beerli & Martin, 2004a; F akeye & Crompton, 1991;
Gunn, 1972; Hanlan & Kelly, 2005; Litvin & Ling, 2001). Information source usage
inﬂuences travelers’ destination-choice behavior because knowledge about a destination
reduces prospective travelers’ uncertainty about vacationing there. New information
about a destination can create new positive images in prospective travelers’ minds or
correct previous negative images.

According to Gartner (1993), not only the quantity of information prospective
tourists obtain on a destination, but also the type of information sources consulted,
inﬂuences cognitive destination image. Similarly, Stern and Krakover (1993) studied the
effects of the type and amount of information tourists had on the Israeli city Beer-Sheva
on their overall image of the place. Stern and Krakover categorized types of information
source as either direct (ﬁrsthand) experience with the destination or indirect (secondhand)
experience via print or digital media, and word of mouth. However, Stern and Krakover
used an imprecise instrument to measure participants’ amount of information on Beer-
Sheva: Respondents (residents of Haifa, Jerusalem, and Tel-Aviv) were asked to rate their
familiarity with “public institutions and services” in Beer-Sheva (p.142) on a Likert scale

ranging from 0 to 10; however, their responses were inconclusive in that Stern and

16

Krakover, in reporting their results, did not indicate the exact amounts of information to
which these rating numbers referred.

Similarly, Baloglu (2001) found that both the type and amount of information on a
destination inﬂuenced travelers’ visit intentions indirectly through their cognitive image
of the destination. In their study of American non-visitors’ images of Turkey, Baloglu
identiﬁed four types of information source (nine total sources): professional advice (tour
operators, travel agents, and airlines), word of mouth (friends, relatives, and social clubs),
advertisements (print or broadcast media), and non-tourism sources (books, movies, and
news). However, as in Stern and Krakover’s (1993) study, Baloglu (2001) imprecisely
deﬁned “amount” of information on a destination: Respondents were asked merely to
check whether they had received information on Turkey from the nine information
sources; the exact amount of information was not speciﬁed by participants, who
effectively gave “yes-or-no” responses.

On the whole, previous research suggests that both qualitative (type) and
quantitative (amount) measures of information on a destination are important given the
inﬂuence of information source usage on cognitive destination image and, through it,
affective destination image and visit intention. However, limitations in previous studies,
regarding the speciﬁcation of “amount” of information on a destination, suggest that
“frequency” of information source usage may be a more accurate quantitative measure.

Direct experiences with the destination. A number of previous studies, as
presented in Table 1, found that direct experience with the destination seems to positively
affect Americans’ destination image. Generally, respondents who have previously visited

a given destination have a more positive image of this destination than do non-visitors

17

(Ahmed, 1991; Baloglu & McCleary, 1999b; Chon, 1991; Court & Lupton, 1997; Hsu,
Wolfe, & Kang, 2004; Milman & Pizam, 1995; Pizam, Jafari, & Milman, 1991). In his
study of tourists’ image of Utah, Ahmed (1991) found that perceptions were generally
more favorable for those who had previously visited the state. Similarly, in their study of
Americans’ images of the USSR, Pizam et a1. (1991) conﬁrmed that post-trip
respondents had more positive images than pre-trip respondents.

Moreover, previous research suggests that American repeat visitors, in general,
have a more positive image of destinations in the US. than do ﬁrst-time visitors. Milman
and Pizam (1995) compared the images of Central Florida held by those whohad visited
the state and by those who were merely aware of it. Milman and Pizam found that repeat
visitors had a more positive image, and more accurate knowledge, of Central Florida than
did respondents who were only vaguely familiar with the region. Fakeye and Crompton
(1991) extended Milman and Pizam’s ﬁndings. In their study of differences in destination
image of prospective ﬁrst-time and repeat visitors to the Lower Rio Grande Valley in
Texas, Fakeye and Crompton discovered that respondents who had not previously visited
the destination had a more negative image of the region than did those who had visited it.
They also noted that the more frequently a tourist visited the destination, the more social
opportunities and attractions he or she recognized and appreciated. Similarly, regarding
Americans’ image of Asian destinations, Chon (1991) discovered that post-trip
respondents had a more favorable image of South Korea than did pre-trip respondents.
This review of previous studies suggests that repeated direct experience with a destination

improves Americans’ image of the place.

18

However, visitors’ image of a destination is not always more positive than non-
visitors’ image. Baloglu and McCleary (1999b) observed that visitors’ and non-visitors’
images of the Mediterranean countries Egypt, Greece, Italy, and Turkey were largely the
same. American tourists may already have a positive image of a destination before their
ﬁrst visit, and this image may remain static, despite actual experiences in the destination
(V ogt & Andereck, 2003). Given that Baloglu and McCleary were focusing on
understudied American tourist destinations, it would seem that differences in destination
image between visitors and non-visitors may depend on the destination in question.

On the whole, previous tourism research suggests that repeat visitors have a more
positive image of a destination than do ﬁrst-time visitors, who, in turn, have a more
positive image of a destination than do non-visitors. It would seem, then, that destination
marketers should prepare different advertising strategies for visitors and non-visitors.
This study examines only the destination images of non-visitors because this population
represents a larger potential market for East Asian tourism marketers.

Proximity to the destination. Another pattern emerging in Table 1, which relates
to the ﬁrst (direct experience), is that proximity to the destination seems to positively
affect Americans’ destination image. Generally, visitors living near a destination have a
more positive image of the place than do those living farther away (Crompton, 1979;
Hunt, 1975; Hsu, Wolfe & Kang, 2004). In his study of 1,225 out-of-state Americans’
images of Montana, Hunt (1975) observed that proximity to a region enhanced
destination image. Respondents living in nearby states, compared to those living further
away, more often answered that Montana was friendly and familiar as a vacation

destination and that it had many places of interest to visit. Reilly (1990) conﬁrmed Hunt’s

19

ﬁndings, establishing that respondents living closer to Montana had a stronger and more
positive image of the state than did those living farther away. Hsu, Wolfe and Kang
(2004) supported both Hunt’s and Reilly’s ﬁndings, in their study of visitors’ and
residents’ images of Kansas. However, in his study of American tourists’ images of
Mexico, Crompton (1979) found a different result: Respondents living closer to the
destination had more negative images of the place than did those living farther away.
Crompton explained that Americans living near the border with Mexico associated the
country with crowding, poverty, and political unrest. These studies suggest that proximity

to the destination reinforces American’s destination image, whether positive or negative.

Inﬂuence of Destination Image on Travelers ’ Visit Intentions

Intention to visit refers to travelers’ perceived likelihood of visiting a speciﬁc
destination within a speciﬁc time period (Woodside & Lysonski, 1989). Intention to visit
is important as an outcome variable because it has been shown to be signiﬁcantly and
substantially correlated with travel behavior (Um & Crompton, 1992; Woodside &
Lysonski, 1989).

Generally, visitors who hold a positive image of a destination are likely to visit
that destination (Court & Lupton, 1997; Milman & Pizam, 1995; Rittichainuwat et al.,
2001; Ross, 1993). In their image study of Central Florida, Milman and Pizam (1995)
found that respondents who were familiar with the destination had a more positive image
of it and as a result, they were more interested in revisiting the destination. Court and
Lupton (1997) conﬁrmed Milman and Pizam’s ﬁnding that a positive relationship existed

between destination image and intention to visit. Similarly, in her study of Americans’

20

images of Thailand, Rittichainuwat (2001) conﬁrmed that the more positive the image (or
the less negative the image) of a destination, the more likely it was that travelers would
plan to revisit the destination. However, Crompton, et a1. (1992) found that, even if
visitors had a positive image of a destination, they might not have any intention to visit

the place, that perceived risks might outweigh positive destination image.

In sum, this review of previous research on destination image indicates that both
cognitive and affective destination image should be studied as distinct factors in order to
better understand their interrelationship and their independent effects on other factors.
Additionally, as illustrated by Table 1, much of the previous research on Americans’
destination image has focused on images of speciﬁc states and regions in the US, of
countries neighboring the US, or of other Western countries. Few studies have examined
Americans’ destination images of East Asian countries. This trend signals the need for
destination image research on China, Japan, and South Korea. The present dissertation
study examined both Americans’ cognitive and affective destination images of these three

countries in an effort to explain prospective tourists’ intentions to visit East Asia.

Perceived Risks of Vacationing at a Destination
Perceived risk is another factor possibly explaining Americans’ intentions to visit,
or not to visit, a destination. Surveying 500 international tourists, Sonmez and Graefe
(1998a, 1998b) found that perceived risks were a stronger predictor of tourists’ avoiding a

particular region than of their planning to visit it. In this section, the types of risks

21

associated with tourism, the inﬂuence of perceived risks on travel behavior, and, the role

of information source usage in the development of tourists’ perceived risks are discussed.

Types of Risks Associated with the Destination and Their Inﬂuence on Travel Behavior

In 1992, the tourism researchers Roehl and Fesenmaier borrowed consumer
behavior researchers’ deﬁnition of perceived risks in an attempt to investigate tourism
behavior. The seven types of perceived risks included in their study were: (1) equipment
risk (the possibility of mechanical problems while on vacation), (2) financial risk (the
possibility that a vacation will not provide value for the money spent), (3) physical risk
(the possibility of physical danger, injury, or sickness while on vacation), (4)
psychological risk (the possibility that a trip to the destination will not reﬂect one’s
personality or self-image), (5) satisfaction risk (the possibility that the vacation will not
provide personal satisfaction), (6) social risk (the possibility that the vacation will affect
others’ opinions of oneself), and (7) time risk (the possibility that the vacation will take
too much time or be a waste of time).

In 1998, three more speciﬁc risks were added by Sonmez and Graefe: (8) health
risk (the possibility of becoming sick while traveling to, or while staying at, the
destination), (9) terrorism risk (the possibility of being involved in a terrorist actwhile on
vacation), and (10) political instability (the possibility of becoming involved in the
political turmoil of the host country while on vacation). In recent years, natural disaster
(1 1) has also been studied (Faulkner, 2001; Mazzocchi & Montini, 2001). Dolnicar
(2005) noted that a number of previous studies had attempted to establish a complete

range of possible perceived risks exhibited by tourists. In contrast, she attempted to

22

determine which travel-related fears were most salient to tourists living under the global
market conditions of 2004. Using open-ended questions, Dolnicar identiﬁed ﬁve risk
categories: (1) political risk, such as terrorism and political instability, (2) environmental
risk, such as natural disasters, (3) health risk, such as lack of access to clean food and
water, (4) planning risk, such as an unreliable airline, and (5) property risk, such as theft
and loss of luggage.

Among all of the perceived risks identiﬁed above, terrorism and political turmoil
have been identiﬁed as having the most negative effect on visit intention (Coshall, 2003;
Gartner & Shen, 1992; Sonmez & Graefe, 1998a, 1998b). In their study of Americans’
reaction to worldwide coverage of the 1989 Tiananmen Square conﬂict, Gartner and Shen
(1992) observed that China’s newly developing tourism industry suffered: Hotel
occupancy in Beijing fell below 30 percent; 300 groups (about 11,500 individuals)
canceled travel plans; and tourism earnings declined by $430 million in 1989.

As did Gartner and Shen (1992) and Sonmez and Graefe (1998a, 1998b), Coshall
(2003) found that people perceiving terrorism as a risk were more likely to avoid the
Middle East. Coshall demonstrated that the Gulf War had a signiﬁcant impact on British
air travel to almost all destinations. Similarly, British air trips to the Near East (Israel,
Jordan, Lebanon, and Syria) and the Middle East (Iran, Iraq, Kuwait, the Persian Gulf
states, North and South Yemen, Saudi Arabia, and the United Arab Emirates) began to
fall in 1990, due to the Iraqi invasion of Kuwait and the seizure of Kuwaiti oil ﬁelds. The
number of British air passengers declined signiﬁcantly below expected levels: by 15.4
percent to the Near East and by 9.9 percent to the Middle East. Regarding terrorism,

Floyd et a1. (2003) investigated how perceived risks impacted New Yorkers’ intentions to

23

travel during the period of aftershock following September 11, 2001. Floyd et al. found
four statistically signiﬁcant effects on intention to take a pleasure trip in the next 12
months. Two of these effects were shown to be attributed to perceived risks: physical
risks and social risks.

Natural disaster is another formidable barrier to international tourism (Mazzocchi
& Montini, 2001). In their study of the inﬂuence of natural disasters on tourism demand,
Mazzocchi and Montini (2001) observed that tourist ﬂows decreased after the earthquake
that hit Umbria in Central Italy on September 26, 1997. Tourist arrivals fell drastically the
month following the main shock, compared to the same month of the previous year. The
total loss estimate for the period immediately following the earthquake (October 1997 to
June 1998) was about $71 million or a 40 percent reduction in estimated expenditure for
the same period under normal conditions.

Recently, health risk has become a major factor in choosing a vacation
destination (Dolnicar, 2005; KTO, 2004; Richter, 2003). International travel to the Asia-
Paciﬁc region has recently been devastated by tragic health-related events, such as Severe
Acute Respiratory Syndrome (SARS) and bird ﬂu. According to a Korean Tourism
Organization (KTO) report, in 2004, entry tourist numbers decreased noticeably, from 5.3
million in 2002 to 4.8 million in 2003: a 10 percent decrease, attributed mostly to three
causes, SARS (even though, no SARS patients were ultimately recorded in South Korea),
the Iraq war, and the North Korean nuclear controversy. Richter (2003) supported KTO’s
report in her study of West Nile fever in New York and Mad Cow Disease in Europe,
which also had a signiﬁcant negative impact on travel ﬂows to these areas. Dolnicar

conﬁrmed both KTO’s and Richter’s ﬁndings. In her study comparing 373 Australian

24

students’ patterns of perceived risks regarding domestic and international travel, Dolnicar
(2005) found that health risk emerged much more frequently in the context of
international travel than in the context of planned domestic travel.

Language risk is also a factor possibly inﬂuencing destination choice (Basala &
Klenosky, 2001). In their study of travel preferences, Balsa and Klenosky (2001)
discovered that tourists’ lack of ﬂuency in the native language(s) of the destination was a
barrier to international travel to that destination. It seems that language difﬁculty is a
twelfth perceived risk. However, Bales and Klenosky suggest that language risk is the
least studied by tourism researchers. Language risk should be explored in the future. Few
studies in the ﬁeld have been conducted on the precise impact of language difficulty on
destination choice, even though communication risk has been shown to affect
international travel.

Given Dolnicar’s (2005) identiﬁcation of tourists’ most salient travel-related fears,
and given recent political instability, terrorism, natural disaster, and disease, as well as the
ongoing language barrier for Americans, in East Asia, the present dissertation study
selected the following perceived risks to compose the perceived risks factor in the
proposed conceptual model: physical risk, political instability risk, terrorism risk,

environmental risk, health risk, and language risk.

Role of Information Source Usage in Perceived Risks
Several researchers (Fodness & Murray, 1999; Maser & Weiermair, 1998;
Sonmez & Graefe, 1998b) have indicated that travelers’ search for information on a

destination can reduce their perceived risks of vacationing there and increase their visit

25

 

intention. For example, Gemunden (1985) found that, when people decide on complex
products, such as travel holidays, information sources are one of several risk-reducing
factors that facilitate their decision making. Similarly, Roehl and Fesenmaier (1992)
explored the relationship between risk perceptions and leisure travel, ﬁnding that the
more uncertain or dissatisﬁed prospective tourists became about the consequences of a
destination choice, the more perceived risks they had regarding this destination. This
study identiﬁed information search behavior as a common risk-reduction strategy.
However, despite the important role of information source usage in reducing travelers’
perceived risks of vacationing in a destination, no research, before the present dissertation
study, has been conducted on the comparative inﬂuence of different types of information
source, and different frequencies of I information source usage, on travelers’ perceived
risks of vacationing in China, Japan, and South Korea.

In sum, the desire for risk-free travel is a key factor inﬂuencing tourists’ intentions
to visit a destination. Despite the importance of perceived risks in travelers’ decision-
making processes, there has been little research on perceived risks, before the present
study. Some destination image studies have included safety as a cognitive image attribute;
however, other important risks were excluded. Additionally, these studies did not
adequately distinguish prospective tourists’ perceived risks of vacationing in a destination
from their general image of a destination. The present study compared destination image
and perceived risks as two possible explanations for American tourists’ intentions to visit

China, Japan, and South Korea.

26

Proposed Conceptual Model

This literature review suggests a conceptual model of the relationships between
intention to visit and destination image, perceived risks, and information source usage, as
presented in Figure 2. The present dissertation study tested the appropriateness of this

model regarding Americans’ intentions to vacation for the ﬁrst time in three East Asian

countries (China, Japan, and South Korea).

 

 
      
     

Cognitive
Image

 

 

   

Perceived

Risks

 

 

H6
Information Intention
Source - Affective H7 : To.
Usage Image Vrsrt
H3 ‘. H 5 H8

 

Figure 2. Theoretical model of intention to visit three East Asian countries.

27

 

Hypothesized Relationships

Derived from the literature review, eight hypotheses were tested to achieve the

purposes of this study.

H1:

H2:

H3:

H4:

H5:

H6:

H7:

H8:

Increased usage of information sources increases Americans’ (emotionally
neutral) cognitive image of the destination in general.

Increased usage of information sources increases Americans’ (emotionally
positive) affective image of the destination in general.

Increased usage of information sources decreases Americans’ perceived risks of
vacationing in the destination.

Increased (emotionally neutral) cognitive destination image increases Americans’
(emotionally positive) affective destination image.

Increased perceived risks of vacationing in the destination decreases Americans’
(emotionally positive) affective image of the destination in general.

Increased (emotionally neutral) cognitive destination image increases Americans’
intention to visit the destination.

Increased (emotionally positive) affective destination image increases Americans’
intention to visit the destination.

Increased perceived risks of vacationing in a destination decreases American’s

intention to visit the destination.

The literature review presented in Chapter Two informed the eight hypotheses

listed here and the resulting conceptual model, as well as the study design presented in

Chapter Three.

28

CHAPTER 3
RESEARCH DESIGN AND METHODOLOGY

In Chapter Three, the design and research methodology of this study are detailed.
In the ﬁrst section, the theoretical model suggested by the literature review and the related
hypotheses are discussed. In the second section, the sampling and survey methods
employed in this study are described. In the third section, the measures and scaling
utilized in data collection are discussed. Finally, in the fourth section, the statistical

methods used in data analyses are presented.

Data Collection and Analysis Procedures

Data Collection

A self-administered survey was used to collect the data for this study.
Respondents were chosen based on a convenience sampling method, given the scope of
the present dissertation. Bookstore and coffeehouse frequenters were targeted, as the
sociodemographics of this group match that of the population targeted in this study.

With no particular pattern in mind during the time of data collection, potential
respondents were approached about participating in the study by two colleagues and me
in bookstores and coffeehouses at three Michigan-area shopping malls and in downtown
Chicago. Once respondents agreed to participate in the survey, they were screened. Only
American citizens, 18 years of age or older, were invited to complete the questionnaire.

The purpose of this survey was explained before a copy was distributed for completion.

29

The questionnaire required approximately 15 minutes to complete, and about one third of
those approached refused to participate. Questionnaires were collected on site, but some
respondents could not complete the survey, due to lack of time; in this case, a pre-paid
return envelope was provided. Seventy-two such envelopes were distributed; only 23
questionnaires were returned by mail. Out of 895 collected surveys, 73 were either
partially, or not at all, completed. Data collection was completed between May 20 and

June 20, 2006 (210-240 hours in total).

Sub-Sample Size

One purpose of this study was to test a model explaining prospective American
tourists’ intentions to visit China, Japan, and South Korea for the ﬁrst time. Therefore,
previous visitation to one of these countries was a criterion employed to exclude
respondents. Based on this criterion, a sub-sample was drawn from the entire sample of
822 respondents. The distribution of respondents in terms of China, Japan, and South
Korea is provided in Table 2. Of the 822 completed surveys, 61 were completed by
respondents who had visited China in the past 10 years. These 61 were dropped from the
analysis because the study’s intention was to examine the perceptions of potential, not
repeat, travelers. Similarly, 70 completed surveys were dropped regarding Japan, and 41
regarding South Korea. A total of 709 respondents had not previously visited any of the
three East Asian countries of focus in this study; therefore, the ﬁnal number of subjects

used for SEM analysis was n = 709.

30

Table 2. Distribution of respondents who had, or had not, visited China, Japan,
or South Korea before the survey.

 

 

 

Destinations VlSlted Total
Yes No
China 61(7.4%) 761(92.6%) 822
Japan 70(8.5°/o) 752(91 5%) 822
South Korea 41 (5 .0°/o) 78 l (95 .0%) 822
Had visited any of the 3 countries ll3(14%) - 709(860%)W 822

 

Profile of Respondents
The majority of the participants were male (60.1%), with an average age of 38.

The majority of participants (80.0 %) were European—American. The largest percentage

 

of participants had obtained a Bachelor’s degree (48.0%) as their highest level of
education, followed by those with a high-school degree (23.0%), those with a Master’s
degree (20.0%), and those with a doctorate (5.9%). A majority of respondents were
married (50.8%), while 39.5% were single (never married). Of those who answered the
income question (90.7%), 20.4% reported an annual household income in excess of
$100,000; 66%, between $40,000 and $59,999; 13.7%, between $60,000 and $79,999;

and 10.1% between $80,000 and $99,999.

 

31

Table 3. Characteristics of respondents who had never visited China, Japan, or South

 

 

 

 

 

 

 

 

 

 

Korea.
Category % n
Gender Male 60. 1%
Female , 39.9 % 691‘
Age 18-24 years 25.1%
Mean age: 25-34 years 23.3%
38 years 35-44 years 19.3%
45-54 years 16.4%
55-64 years 9.7%
65-74 years 4.9%
75 years or older 1.3% 678‘
Education Middle/Junior High School 0.6%
High School 23.0%
College(Undergraduate) 48.0%
Master’s Studies 20.0%
Doctoral Studies 5.9%
Post-doctoral Studies 2.5% 690‘
Income Less than $20,000 19.3%
$20,000 - $39,999 19.9%
$40,000 - $59,999 16.6%
$60,000 - $79,999 13.7%
$80,000 - $99,999 10.1%
$100,000 - $119,999 9.2%
$120,000 - $139,999 3.1%
$140,000 or more 8.1% 643‘
Ethnic African American/Black 9.5%
Asian American 3.1%
Native American/Native Alaskan 2.5%
European American/White 80.0%
Latino/Latina 4. 1 %
Native Hawaiian or Other Paciﬁc 0.3%
Islander
Other 0.4% 681‘
Marital Status Married/Committed relationship 50.8%
Widowed/Widower/
. 1.9%
Never remarrred
Separated 1.3%
Divorced 6.5%
Single, never married 39.5% 679.

 

I . o
Note: The value of “n” vanes, due to questions not answered.

32

 

Survey Instrument and Scales

An on-site survey was used to collect the data. The survey instrument consisted of
four printed pages, each divided into ﬁve sections. The ﬁrst section evaluated
participants’ general image of each country (China, Japan, and South Korea). The second
section targeted participants’ perceived risks of vacationing in each country. The third
section registered the information sources on which participants had based their
impressions of each country. The fourth section recorded participants’ intention to visit
each country within the next ﬁve years. Finally, the ﬁfth section ascertained selected

demographic characteristics of the participants.

Destination Image

Cognitive destination image attributes were derived from previous destination
image studies (Crompton, 1979; Echtner & Ritchie, 1993; Fakeye & Crompton, 1991;
Goodrich, 1978; Hunt, 1975). Respondents were asked to indicate their level of
agreement with these attributes, regarding each country, on a 5-point Likert scale, ranging
from 1 (strongly disagree) to 5 (strongly agree).

Affective destination image attributes were adapted from Russell’s (1981) bipolar
scales. Those used in this study were Amusing-Sleepy, Pleasant-Unpleasant, Exciting-
Gloomy, and Relaxing-Distressing. A 5-point scale was used for all four bipolar scales in

which the positive end of each spectrum was assigned the highest value.

33

 

Perceived Risks

The construct perceived risks comprised seven items modiﬁed from previous
studies (Floyd et al., 2003; Lepp & Gibson, 2003; Roehl & Fesenmaier, 1992; Sonmez &
Graefe, 1998b), and was measured on a ﬁve-point Likert scale (ranging from 1 = strongly
disagree to 5 = strongly agree). However, this study replicated Yamamoto and Gill’s
(1999) technique by adding language risk items, such as “I will not be able to understand
social expectations or read body language during my vacation here,” a modiﬁcation of
Yamamoto and Gill’s (1999) item “It is important that the people I encounter on a

vacation trip speak my language.” All perceived risk items were reverse coded.

Information Source Usage

Data on information sources were collected by asking survey respondents to
answer, on a 5-point Likert scale, ranging from 1 (“never”) to 5 (“frequently”), how often
they obtained information on each country from seven information sources: TV/radio,
word of mouth (family, friends and native people), books and movies, the Internet,
cultural experiences (in art museums, theaters, concert halls, restaurants, courses, etc.),
newspapers/magazines, and professional advice (travel consultants, brochures). These
seven sources had been identiﬁed in previous studies (Chen, 2000; Lo, Cheung, & Law,

2004; Money & Crotts, 2003).

Intention to Visit
To ensure construct validity, or the survey’s alignment with the proposed

conceptual model, several intention to visit items were developed. Intention to visit was

34

deﬁned as interest in, and likelihood of, visiting each country within the next ﬁve years.
Interest was measured using a 5-point Likert scale (1 = completely uninterested, 3 =
neither interested nor uninterested, 5 = strongly interested), and likelihood of visiting was
measured on a similar scale (1 = very unlikely, 3 = no opinion, 5 = very likely). These
items had been successfully used by previous researchers (Court & Lupton, 1997;

Milman & Pizam, 1995; Rittichainuwat, 2001).

Data Analysis

Survey data were analyzed in three steps. First, preliminary statistics were
obtained using the Statistical Package for the Social Sciences (SPSS). Descriptive
statistics were obtained to determine the distributional characteristics of each variable
including the means, standard deviation, skewness, and kurtosis.

Second, a one-way repeated measure ANOVA was used to examine the mean
differences in intention to visit, cognitive destination image, affective destination image,
perceived risks, and information sources across the three countries. After the null
hypothesis was rejected, a post hoc test was performed to determine where the differences
lay (Hair, Anderson, Tatham, & Black, 1998).

The third step in the analysis was to engage in structural equation modeling, using
the Analysis of Moment Structures (AMOS) to test the adequacy of the hypothesized
model. The structural equation model enabled testing of the eight hypotheses, which
speciﬁed direct and indirect relationships among the latent, as well as the observed,
variables (Kline, 1998). In general, structural equation models consist of two sub-models:

(1) a measurement model and (2) a structural model. The measurement model deﬁnes

35

relations between the variables as indicators of underlying constructs (the observed
variables) and the underlying constructs that the observed variables were designed to
measure (unobserved latent variables). The structural model represents relations among
the latent variables. The constructs, observed variables, and measurement scales in the

proposed model are presented in Table 4.

36

Table 4. Constructs and observed variables in the proposed model.

Observed

 

 

 

 

Constructs . Survey questions Scale
varrables
Infoa Infol TV/Radio 1: never to
Info2 Word of mouth (Family, Friends, Native 5: frequently
peeple)
Info3 Books and Movies
Info4 Internet
Inf05 Cultural experiences (Art Museums,
Theaters, Restaurants, Concert Halls,
Classes, etc.)
Info6 Newspapers/Magazines
Info7 Professional advice (Travel consultants,
Brochures)
Cog“ Cogl Many things to see and do 1: strongly disagree
Cog2 Natural scenic beauty to
Cog3 Easy to navigate 5: strongly agree
Cog4 Plenty of quality accommodations
Cog5 Friendly local people
Cog6 Appealing local food and drink
Cog7 Interesting historical attractions
Cog8 Unique cultural resources
Affec Affel Unpleasant vs. Pleasant 1: negative emotion
Affe2 Sleepy vs. Arousing to
Affe3 Gloomy vs. Exciting 5: positive emotion
Affe4 Distressing vs. Relaxing
PRd PR1 I will get sick from eating food or 1: strongly disagree
drinking water here to
PR2 Political instability will keep me from 5: strongly agree
vacationing here
PR3 I will be concerned about potential health
problems here
PR4 I will worry about terrorism while
vacationing here
PR5 I will be vulnerable to violence if I
vacation here
PR6 I will encounter a natural disaster if I
vacation here
PR7 I will not be able to understand social

expectations or read body language
during my vacation here

 

Note: a Information Source Usage regarding the destination; 5 Cognitive Destination
Image based on knowledge gathered about the place; ° Affective Destination Image based
on emotional evaluations of the place; d Perceived Risks of vacationing at the destination;
° Interest and perceived likelihood of visiting the destination within the next ﬁve years.

37

Table 4. Constructs and observed variables in the proposed model @ont’d).

 

 

IV° IVl How interested are you in visiting the 1: completely
following vacation destinations in the next uninterested
5 years? 5: strongly interested
1V2 How likely is it that you will visit the 1: very unlikely
following vacation destinations in the next 5: very likely
5 years?

 

Note: “ Information Source Usage regarding the destination; b Cognitive Destination
Image based on knowledge gathered about the place; ° Affective Destination Image based
on emotional evaluations of the place; d Perceived Risks of vacationing in the destination;
‘ Interest and perceived likelihood of visiting the destination within the next ﬁve years.

Generally, two-step modeling is used to test the structural equation model (Kline,
1998). The ﬁrst step is to ﬁnd an acceptable measurement model. A model is ﬁrst
speciﬁed as a conﬁrmatory analysis (CF A) measurement. The CFA model is then
analyzed to determine how well it ﬁts the data. Overall ﬁt indices are then reported. Chi-
square statistics are commonly used to evaluate model ﬁt, but these statistics are known
to be too sensitive to sample size (Kline, 1998). Thus, a number of ﬁt indices were
considered in this study to evaluate the measurement model for each country, including
chi-square statistics adjusted by degrees of freedom (xZ/df), the Tucker-Lewis index (TLI),
the Comparative Fit Index (CPI), the Incremental Fit Index (IFI), and the Root Mean
Square Error of Approximation (RMSEA). If the ratio of x2 to degrees of freedom
(CMH‘I/DF) is less than 5, an acceptable ﬁt is indicated; a CMIN/DF of less than 2 or 3
also denotes a good ﬁt (Garson, 2006; Tabachnick & Fidell., 2001). An RMSEA value of
less than .05 similarly indicates a good ﬁt, though an RMSEA value of less than .08
signiﬁes an acceptable ﬁt (Brown, Mowen, Donavan, & Licata, 2002; Byme, 2001). TLI,
CPI, and IF I values of more than .90 are generally considered as indicating an acceptable

model ﬁt (Kline, 1998). In this study, if the initial measurement model suggested

38

inadequate goodness of ﬁt, the model was re-speciﬁed, based both on theories presented
in the literature review and residual matrixes. After modiﬁcation, differences in xz/df
statistics were assessed to determine the signiﬁcance of any improvement to the modiﬁed
models.

If the overall ﬁt of the CFA model was deemed acceptable, the second stage of
two-step modeling was initiated: the testing of the structural model. Testing a structural
model includes testing the structural paths between the latent variables as well as testing
the overall ﬁt of the hypothesized model. In this study, because the initial model
suggested an inadequate overall ﬁt, the initial model was modiﬁed, based both on theories
presented in the literature review and residual matrixes. After modiﬁcation, differences in
xz/df statistics were assessed to determine the signiﬁcance of any improvement to the
modiﬁed models.

Chapter Three explained the data collection procedures used in the present
dissertation study, including details on survey design and research methodology. Data
collection from the on-site survey resulted in 709 usable responses from Americans, 18
years of age or older, who had never before visited China, Japan, or South Korea. This
chapter also discussed model ﬁt assessments for both the measurement and structural
equation models developed for this study. Chapter Four presents the results of these

model ﬁt analyses and of hypotheses testing.

39

CHAPTER 4
RESULTS

Findings on the relationships between Americans’ intentions to vacation for the
ﬁrst time in East Asia (China, Japan, and South Korea) and their general destination
image of these countries (both affective and cognitive), their perceived risks of
vacationing in these countries, and their information source usage regarding these
countries are presented in Chapter Four. The results of preliminary analyses, including
data screening, description of the variables, and analyses of the mean differences among
the variables across China, Japan, and South Korea are discussed in the ﬁrst section. The
results of tests conducted on the measurement models for each of these three East Asian
countries, including assessments of overall model ﬁt, reliability, and validity are reported
in the second section. The results of tests conducted on the structural equation models
(SEM) for China, Japan, and South Korea are presented in the third section. The results of
hypotheses testing, including evaluations of the direct, indirect, and total inﬂuences
among the factors identiﬁed in the models, are discussed in the fourth section. Finally,
additional analyses of the inﬂuence of information source usage on cognitive destination
image and perceived risks, from which implications for tourism researchers and marketers

will be derived in Chapter Five, are provided in Chapter Four.

40

Preliminary Analyses
In this section, data screening procedures, descriptive statistics for the variables
included in this study, the mean differences among the variables across the three countries,
and results from tests identifying statistically signiﬁcant mean differences among the

variables across the three countries, are described.

Data Screening

The means and ranges of all of the variables of interest were inspected ﬁrst for
signs of data entry errors and outliers. Then, normality (i.e., the assumption that each
variable should be normally distributed) was assessed by evaluating the skewness and
kurtosis of each variable in the study. These tests indicated that all values for univariate
skewness and kurtosis were inside the acceptable range (-3 to 3 for skewness and -8 to 8
for kurtosis), as established by Kline (1998). Thus, the results of the normality test

showed no extreme departure ﬁom normality. These results are presented in Table 5.

41

 

Table 5. Normality a test results for variables included in the proposed model.

 

 

 

 

 

 

 

 

Variable Skewnessb Kurtosisc : Skewnessh Kurtosisc Skewnessb Kurtosisc
Constructs a -
Names ; H __m_t
T China ...__._--.,____J3123!!.- _ South Korea
Information Info] .12 g -.57 ‘ .14 l -.51 .68 l -.16
source Info2 .38 -.45 .44 ‘ -.51 f 1.98 1 .53
usage lnfo3 —.00 I -.32 ' .08 -.38 3 .65 -.06
lnfo4 .57 -.52 ' .53 1 -.66 .95 .07
Inf05 .16 g -.62 .22 . -.51 i .81 .04
lnfo6 .24 -.32 : .24 —.33 Q .66 -.14
Info? :52. __ _A “:67, . - .47. 82 87 2.;
Cognitive Cogl -1.22 3 1.34 ’ -1.38 1.86 -.26 -.49
destination Cog2 -1.04 I .70 -.95 .45 l -.33 -.35
image Cog3 -.02 3 .60 -.()2 .39 -.01 .88
Cog4 -.68 g .143 -.63 -.09 j -.26 -.12
CogS -120 1.26 . -102 .63 -.37 -.43
Cog6 -.76 .31 5 -.69 . .29 -.35 ’ -.16
Cog7 -.78 .31 -.82 z .42 -.51 1 -.18
C083 ..._..__’-_79 . ...._-46.ﬂ _ ‘35 L _.-7.l. 48......9.5..
Affective Affel -.35 5 .85 -.12 -.66 T -.44 - 54
destination Affe2 ; -.22 ‘ -.98 -.02 -1.17 -.55 -.53
image Affe3 -.28 - -1.12 -.42 -1.01 L -.43 g -.78
Affe4 51.. s18; .50. .54 80 _-__._-_2.§_
Perceived PR1 i .78 -.03 ,3 .94 . .44 .53 -.33
risks PR2 .04 -.75 ; .25 -.62 .04 -.47
PR3 -.05 -.71 i_ .45 -44 -.08 -.63
PR4 ; .10 i -.70 , .49 -.55 .04 -.48
PR5 .30 -' -.72 .36 -.64 ‘ .16 -.71
PR6 -.10 . -27 -.06 -27 ' -.11 -.06
PR7 - .13 . . . -.63 _ .15 . 5.78” .03___ _‘ --...._:-.5._‘L
Intention INTI , -.35 g -1.11 -.49 -.98 ' .10 —1.10
to visit INT2 1 .95 -.56_ : .79 -.79 ; 1.12 g .59

 

 

Note: aNormality was examined in terms ofﬂsk‘ewness andmkurtosis. b Skewness refers to

the symmetry of the distribution. Skewness with a value above 3 is conventionally

considered as extremely skewed. ° Kurtosis indicates a relative excess of cases in the tails
of a distribution relative to a normal distribution. A kurtosis value of 10 is a conventional
criterion indicating normality distribution in terms of its peakedness. A value above 10 is
considered extremely peaked.

Description of the Variables

Americans ’ intentions to visit three East Asian countries. Visit intentions were
measured by two items: (a) interest in visiting and (2) likelihood of visiting. The
descriptive statistics for the variables measuring intention to visit are presented in Table 6.
The mean scores can be considered as study participants’ average evaluations of intention

to visit the three East Asian countries.

Table 6. Mean and SD of intention to visit the three countries.

 

 

 

. China Japan South Korea
Variable
Mean (SD c) Mean (SD c) Mean (SD c)
Interest in visiting a 3.10(.81) 3.30(.83) 2.61(.75)
Likelihood of visiting b 1.95(.66) 2.08(.72) 1.72(.51)

 

Note: a All items were on 5-point likert scales (1 = completely uninterested and 5 =
strongly interested), b All items were on 5-point likert scales (1 = very unlikely and 5 =
very likely). c the number in the brackets is standard deviation.

While the mean scores of interest in visiting the three East Asian countries ranged
from 2.61 to 3.30, the mean scores of likelihood of visiting China, Japan, and South
Korea were low, ranging from 1.72 to 2.08. Study participants’ interest in visiting the
countries was higher than their likelihood of visiting. Among the three countries, Japan
had the highest mean scores for both interest in visiting and likelihood of visiting, while

South Korea had the lowest for both visit intention items. '

Americans ’ cognitive images of three East Asian countries. The means and
standard deviations of items measuring cognitive image of the three East Asian countries
are reported in Table 7. The mean scores can be considered as study participants’ average

evaluations of their cognitive image of China, Japan, and South Korea.

43

Table 7. Mean and SD of cognitive destination image of the three countries.

 

 

 

. a China Japan South Korea
Vanable

Mean (so I’) Mean (so I)) Mean (SD b)

Many things to see and do 4.25(.48) 4.31(.46) 3.48(.59)
Natural scenic beauty 4.17(.51) 4.13(.53) 3.65(.43)
Easy to navigate 2.46(.51) 2.67(.57) 2.51(.49)
Quality accommodations 3.17(.48) 3.59(.41) 3.01(.42)
Friendly local people 3.29(.42) 3.37(.44) 3.21(.43)
Appealing local food and drink 3.68(.51) 3.58(.56) 3.20(.53)
Interesting historical attractions 4.28(.46) 4.20(.58) 3.70(.52)
Unique cultural resources 4.09(.48) 4.13(.55) 3.79(.49)

 

Note: a All items were on 5-point likert scales (1 = strongly disagree and 5 = strongly
agree), b The number in the brackets is standard deviation

The mean scores for China ranged widely from 2.46 to 4.28. The cognitive image
attributes with mean scores of greater than 4.00 were many things to see and do, natural
scenic beauty, interesting historical attractions, and unique cultural resources, indicating
that China’s major strengths were its vacation attractions. (Study participants gave the
highest mean ratings to China’s interesting historical attractions.) The three attributes,
quality accommodations, ﬁiendly local people, and appealing local food and drink, were,
on average, given rather high mean scores, ranging from 3.17 to 3.68. Only one cognitive
image attribute was, on average, given a low mean score. This attribute was easy to
navigate, with a mean score of 2.46.

The mean scores for Japan widely ranged from 2.67 to 4.31. The cognitive image
attribute with the highest mean score was many things to see and do. Three more
cognitive image attributes, natural scenic beauty, interesting historical attractions, and
unique cultural resources, had mean scores of greater than 4.00, indicating that Japan’s
major strengths were its vacation attractions. The three attributes, ﬁiendly local people,

appealing local food and drink, and quality of accommodations were on average

44

perceived to be positive, though not strongly so, with mean scores ranging from 3.37 to
3.59. As in the case of China, only one cognitive image attribute for Japan was, on
average, perceived to be negative. This attribute was easy to navigate, with a mean score
of 2.67.

The mean scores for South Korea ranged from 2.51 to 3.79. Most of the cognitive
image attributes had mean scores of more than 3.00, indicating that they were, on average,
perceived to be rather positive. While cognitive image attributes for South Korea were not
strongly positive, two cognitive image items, unique cultural resources and interesting
historical attractions, seemed to be South Korea’s strength as a vacation destination, with
mean scores of 3.79 and 3.70, respectively. As for both China and Japan, only one
cognitive image attribute for South Korea, easy to navigate, was perceived to be negative,
with a mean score of 2.51.

Examination of cognitive image attributes for the three East Asian countries
shows that historical and cultural attractions seemed to be the strength of all three, even
though the mean scores for these attributes differed across China, Japan, and South Korea.
Easy to navigate seemed to be the weakness of all three countries, suggesting that tourism
marketers working in China, Japan, and South Korea must improve this cognitive image
attribute if they are to attract more American vacationers. When the mean scores of
cognitive image attributes are compared across the three East Asian countries,

Americans’ cognitive images of China and Japan were found to be more positive than
those of South Korea. China had higher mean ratings on natural scenic beauty, appealing

local food, and interesting historical attractions than did Japan, while Japan had higher

45

scores for ﬁiendly local people, quality accommodations, many things to see and do, and
unique cultural resources than did China.

Americans ’ affective images of three East Asian countries. The means and
standard deviations of items measuring study participants’ affective image of the three
East Asian countries are reported in Table 8. The mean scores can be considered as study
participants’ average evaluations of their affective image of China, Japan, and South

Korea.

Table 8. Mean and SD of affective image of the three countries.

 

 

 

China Japan South Korea
Variableal
Mean (so I’) Mean (SD b) Mean (SD 1’)
Pleasant 2.94(.69) 3.37(.64) 2.61(.64)
Arousing 2.77(.77) 3 .01 (.84) 2.56(.67)
Exciting 3.37(.79) 3.51(.77) 2.77(.69)
Relaxing 2.43(.73) 2.51(.66) 2.35(.53)

 

Note: a All items were on 5-point Likert scales(l = negative emotion and 5 = positive
emotion), b The number in the brackets is the standard deviation.

The mean scores for China ranged from 2.43 to 3.37. While the mean score for
exciting was 3.37, the mean scores for other affective image items, pleasant, arousing,
and relaxing, were less than 3.00, the midpoint of the scale. Study participants gave the
lowest mean rating to relaxing. Whereas the mean score for only one affective image item
for China was greater than 3.00, the mean scores for three affective image items for Japan
were all greater than 3.00. This indicates that study participants had a more favorable
affective image of Japan than of China.

Study participants gave the highest mean rating to Japan for its being an exciting

vacation destination and the second highest mean rating for its being a pleasant vacation

46

destination. As in the case of China, study participants gave the lowest mean rating to
Japan for its being a relaxing destination.

The mean scores for South Korea ranged from 2.35 to 2.77, indicating that study
participants held relatively unfavorable affective image of this country. Study participants
gave the highest mean rating to South Korea for its being an exciting vacation destination,
but this mean score was much lower than it was for China and Japan. As for China and
Japan, South Korea received its lowest mean score for being a relaxing vacation
destination.

Examination of the mean scores for affective image of the three East Asian
countries shows similarities and differences among China, Japan, and South Korea.
Americans perceived all three of these countries as being relatively exciting, though not
relaxing, vacation destinations. However, study participants held a more favorable

affective image of Japan than of China and South Korea.

Americans ’ perceived risks of vacationing in three East Asian countries. Study
participants were asked to rate their level of agreement with seven statements regarding
their perceived risks of vacationing in China, Japan, and South Korea. The means and

standard deviations of items measuring these perceived risks are presented in Table 9.

47

Table 9. Mean and SD of perceived risks of visiting the three countries.

 

 

 

China Japan South Korea
Variable a

Mean (so I)) Mean (sob) Mean (SD b)
Likelihood of illness 2.60(.57) 2.37(.52) 2.64(.54)
Political instability 2.92(.67) 2.27(.57) 2.94(.67)
Health problems 2.39(.80) 2.15(.78) 2.40(.79)
Terrorism 2.38(.59) 2.31(.57) 2.59(.65)
Violence 2.46(.55) 2.31(.51) 2.59(.57)
Natural disaster 2.08(.44) 2.14(.48) 2.12(.47)
Language barrier 3.14(.56) 3.08(.59) 3.13(.56)

 

Note: Higher mean scores indicated higher perceived risks. a All items were evaluated on
a 5-point Likert scale (1 = strongly disagree and 5 = strongly agree), b The number in the
brackets is the standard deviation.

The mean scores for perceived risks of vacationing in China ranged from 2.08 to
3.14. Higher mean scores indicate that study participants perceived higher risks. Study
participants were most likely to agree that they would have a language barrier while
vacationing in China: a mean score of 3.14. The item that scored the second highest mean
was political instability (mean score of 2.92). Study participants were least likely to agree
that they would encounter a natural disaster while vacationing there (mean score of 2.08).

The mean scores for perceived risks of vacationing in Japan ranged from 2.14 to
3.08. As in the case of China, study participants were most likely to agree that they would
experience a language barrier while vacationing in Japan (mean score 3.08). Also
similar to responses for China, the items that had the lowest mean scores were natural
disaster and potential health problems (mean scores of 2. 14 and 2.15, respectively).
However, though six of seven items measuring Americans’ perceived risks of vacationing
in Japan had the lowest mean scores among the three countries, natural disaster

registered the highest mean score.

48

The mean scores for perceived risks of vacationing in South Korea ranged from
2.12 to 3.13. As in the cases of China and Japan, study participants were most likely to
agree that they would experience a language barrier while vacationing in South Korea,
with a mean score of 3.13. The item with the second highest mean score was political
instability (mean score of 2.94). Study participants were least likely to agree that they
would encounter a natural disaster while vacationing in South Korea. The mean scores
for two items, terrorism and violence, were highest for South Korea of the three East
Asian countries.

Examination of mean scores for Americans’ perceived risks of vacationing in
China, Japan, and South Korea shows similarities and differences among the three
countries. While study participants were most likely to agree that they would experience a
language barrier while vacationing in China, Japan, and South Korea, they were least
likely to agree that they would encounter a natural disaster there. Among the three East
Asian countries, Japan appeared to evoke the weakest perceived risks, and South Korea

appeared to evoke the strongest perceived risks.

Americans ’ information usage regarding the three East Asian countries. Study
participants were asked to rate how often they obtained information on China, Japan, and
South Korea, using a ﬁve-point Likert scale (1 = never, 5 = frequently). Seven different
information sources were included: TV/radio, word of mouth, books and movies, Internet,
cultural experiences, newspapers/magazines and professional advice. The means and

standard deviations for these items are reported in Table 10.

49

Table 10. Mean and SD of information source usage (type and frequency a) for the three

 

 

 

countries.
China Japan South Korea
Variable
Mean (sob) Mean (so I)) Mean (sot)
TV/Radio 2.64(.60) 2.65(.62) 2.25(.54)
Word of mouth 2.33(.53) 2.37(.59) 1.97(.57)
Books and Movies 2.71(.49) 2.70(.54) 2.12(.51)
Internet 2.24(.63) 2.29(.66) 1.95(.56)
Cultural experiences 2.62(.58) 2.58(.58) 2.06(.54)
Newspapers/Magazines 2.56( .54) 2.57(.55) 2.16(.55)
Professional advice 2.07(.67) 2.10(.65) 1.90(.62)

 

Note: a Frequency for all items was measured on a ﬁve-point Likert scale (1= never and 5
= frequently), b The number in the brackets is the standard deviation.

Regarding China, the mean scores for information sources ranged from 2.07 to
2.71, indicating that study participants appeared to lack information on China. Among the
seven information sources identiﬁed in this study, books and movies were used most often
by study participants, with a mean score of 2.71, followed by TV/radio and cultural
experiences, with mean scores of 2.64 and 2.62, respectively. The least used information
source was professional advice, with a mean score of 2.07.

Regarding Japan, the mean scores for information sources ranged from 2.10 to
2.70, indicating that study participants also appeared to lack information on Japan. Books
and movies on Japan were most frequently consulted by study participants (mean score of
2.70). TV/radio followed, with a mean score of 2.65. As in the case of China, the least
used information source for Japan was professional advice, with a mean score of 2.10.

The mean scores obtained for South Korea ranged from 1.90 to 2.25. When
compared to China’s and Japan’s mean scores for information source usage, South

Korea’s are much lower. Unlike for China and Japan, study participants most frequently

50

consulted TV/radio for information on South Korea (a mean score of 2.25), then
newspapers/ magazines (a mean score of 2.16). However, as in the cases of China and
Japan, the least used information source regarding South Korea was professional advice,
with a mean score of 1.90.

The mean scores for information source usage regarding all three East Asian
countries were less than 3.00, showing that study participants did not have enough
information about the countries. While the mean scores for books and movies were
highest for China and Japan, the mean score for TV/radio was highest for South Korea.

The mean scores for each of the seven information sources were lowest for South Korea.

Analysis of Mean Differences among the Three Countries

As mentioned above, across China, Japan, and South Korea, there were
differences in mean scores for visit intention variables, cognitive image variables,
affective image variables, perceived risks variables, and information source usage
variables. In order to determine the statistical signiﬁcance of these differences, Repeated
Measure Analysis of Variance (ANOVA) was employed. As shown in Table 11, results
of Repeated Measure ANOVA indicated statistically signiﬁcant differences across the

three countries for most variables.

51

com a 5 33ch .850 88m avenge mm 9on 533 @982 3 wow: on :8 $3 0338 < .3558

025 05 38% Boat? :03 E coconut? 508 Emu—maﬁa a 33065 3.8 cc: ﬁom Qmm >83? . dea 1. £9on ... a $32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$04 :86 men: a? m a .4 a: 86382 eases 8%:

0540 toad 2.. am can one ”we 288% serene assess

$540 .486 m2: 8.». ”2 88 are 2a use was. wanna/s.

umon :86 Ram 8m tum as 6.8% :32 been;

$04 :86 mm. E :2 a? a a .m 22388888 £20

magma :86 :2 EN SN Sn sense: 3 5m

$540 :85 $.42 as 2+ 2.4 beans sees Baez swan:
Mm .94 .53 8.28 was 5. m3. 8 e5 as e ass as: 6%:an
v5. .04 .23 o3 mom EN m3 weave

V504 :80 as: in BM Em magnum

V504 :86 3% e3 5.... EN 3862 ass
5.04 :03 as: 38. RM 38 Bass 6386?
V504 tag :8 N: was ac means a6 82:85 am;
V504 :85 8.92 Ga com o; wages a 2225 2 assess

vaeeex
.. ash mega 2.33 58m 5 5%.. 6%an 6525» seen
com “men
6325» as can:

 

.< >0 Z< 0.832: 830%.. .«o $33M .2 2an

52

.58 a 5 3:on .850 :85 “accommu mm 95% :03? @552 8 com: on :8 68 8558a < .8558 085

05 $88 035...? 5.3 5 oocoeobmo :38 Heme—mama a 58855 68 8: 58m Dmm woo—db. .8.on .1. .modvq _. a $32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MonA :80 emu: 8; 88 SN 823 massage
V546 .3? cm. _ 2 2 .N EN 62 85882 €88...an
V540 toad 93.8 88 was. med 88885 .833
M84 :86 ~33 was an an 5585
V546 teed mesa NZ EN EN use: Be seen swam:
v7.4.0 toes 8.8 8.2 EN m2 589 as 263 85%
van? :85 £82 an men 43 see: 8:283
anTo e3 91. m _ .m 3m 2 .m sense eweaeﬁ
on Mm. a .86 :6 NZ 3 .N a: sesame 3:32
709m .52 ea. _ m an an eves. 85.65
7on :85 2% an :56. mom Sweeney
7% .o 2.86 3. E can a _ .N as asezea exam
7% .u :86 $82 an Rd 8:6. 353% 823 3.5
7mm .0 2.8.0 3% 88 SN 88 was: as 865$: 852$
A 88
pumoh. said/AH 03.0 >1 m Vmnwuvzom M AC Swamp. aUngU oBmv—dxr uoaomm
83 Ken
ezeea> as see:

 

A5288 < >0 Z< 0.332: 50803.“ mo 338% ”2033.

53

As displayed in Table 11, signiﬁcant differences were found in mean scores for all
variables, except one: the perceived risks variable language barrier. This ﬁnding is not
surprising, given that none of the three countries claims English as a national language.
Further ﬁndings are discussed below.

First, signiﬁcant differences were found across the three countries regarding study
participants’ responses to the two items representing their intentions to visit China, Japan,
and South Korea. As shown in Table 11, the mean scores of visit intention items were
highest for Japan and lowest for South Korea.

Second, signiﬁcant differences were found across the three countries regarding
study participants’ responses to the eight items representing cognitive image. As shown in
Table 11, Japan was ranked ﬁrst, in terms of mean score, as being easy to navigate, as
having friendly local people, and unique cultural resources, and as offering many things
to see and do and quality accommodations. However, Japan ranked second for natural
scenic beauty, appealing local food and drink, and interesting historical attractions, all
for which China ranked ﬁrst. South Korea ranked third for all of the cognitive image
items.

Third, signiﬁcant differences were found among the three East Asian countries
regarding study participants’ responses to the four items representing affective image. As
shown in Table 11, the mean scores for affective image items were consistently highest
for Japan and lowest for South Korea.

Fourth, across the three countries, signiﬁcant differences were found in six of the
seven perceived risks items. (The mean differences for language barrier were
insigniﬁcant.) The mean scores for terrorism and violence were highest for South Korea,

while the mean scores of likelihood of illness, political instability, and potential health

54

problems were highest for China. The mean score for natural disaster was highest for
Japan, while the mean scores for other perceived risk items regarding Japan were equal to,
or lower than, those regarding China and South Korea.

Finally, signiﬁcant differences were observed across the three countries regarding
information source usage items. The mean scores for each of the seven information
sources identiﬁed in this study were lowest for South Korea. The mean scores for
TV/radio, word of mouth, and Internet were highest for Japan, while the mean scores for
books and movies and cultural experiences were highest for China.

In sum, results indicated that study participants’ responses regarding the variables
for the three East Asian countries were different. For this reason, it was appropriate to
isolate responses for China, Japan, and South Korea in order to test the hypothesized

model.

Testing the Measurement Model

The measurement model deﬁnes the relations between observed variables and the
underlying constructs that the observed variables are presumed to measure. A
conﬁrmatory factor analysis (CFA) was conducted to conﬁrm the hypothesized relations
between observed variables and their underlying constructs to assess the degree to which
the data ﬁt the proposed measurement model. The measurement model was estimated
using the maximum likelihood method. As shown in Figure 3, the proposed measurement
model consisted of ﬁve constructs and 28 observed variables. The ﬁve constructs were

Information Source Usage, Affective Image, Cognitive Image, Perceived Risks, and

55

Intention to Visit. Each construct (e. g., Affective Image) was measured by a number of

observed variables (e.g., Aﬂ‘el, Affe2, Affe3, and Affe4).

3'3eesss

loom cooscooo oo

 

 

 

Cognitive
Image @ ®
5’ M
o Imm
6’ 1m
6 .
6‘ lnb3 lnfmtion ® Intention 0
Affective Aim @ To
Source I mage
A (3‘

\mo

9

@0639
£553.

Perceived
Risks

 

l PR1 PR4 PR6 PR7

dggédéé

 

 

 

 

 

 

Figure 3. The proposed measurement model for the three countries.

Note: Affel: Pleasant, Affe2: Arousing, Affe3: Exciting, Affe4: Relaxing;

Cogl: Many things to see and do, Cog2: Natural scenic beauty, Cog3: Easy to navigate,
Cog4: Plenty of quality accommodations, CogS: Friendly local people,

Cog6: Appealing local food and drink, Cog7: Interesting historical attractions,

Cog 8: Unique cultural resources;

Infol: TV/Radio, Info2: Word of mouth, Info3: Books and Movies, Info4: Internet,
Inf05: Cultural experiences, Info6: Newspapers/Magazines, Info7: Professional advice;
1V1: Interest, IV 2: Likelihood;

PR1: Likelihood of illness, PR2: Political instability, PR3: Health problems,

PR4: Terrorism, PR5: Violence, PR6: Natural disaster, PR7: Language barrier

Error Terms: e1-e27, rl-r3.

56

The proposed measurement model was ﬁrst examined for overall model ﬁt using
AMOS (Analysis of Moment Structures) to determine the degree to which the model ﬁts
the data. A number of commonly used goodness-of-ﬁt indices were utilized to assess the
model ﬁt. The goodness-of-ﬁt indices are generally categorized into three types: (1)
absolute ﬁt indices, (2) parsimonious ﬁt indices, and (3) incremental ﬁt measures (Hair et
al, 1998). For this study, six goodness-of-ﬁt indices from three categories were chosen. In
Table 12, the goodness-of-ﬁt indices chosen for this study and recommended range of

good ﬁt are presented.

Table 12. The recommended goodness-of-ﬁt indices

 

 

 

 

Category Goodness-ﬁt indices Recommended Range
Used for the Study of Good Model Fit
Absolute Fit Indices o Chi-Square Statistics (x2 ) Insigniﬁcant p-value (p>.01)
0 Root Mean Square Error of < 05
Approximation (RMSEA) '
Parsirnonious Fit Indices 0 Chi-Square/Degree of < 3
Freedom (x2 /df)
Incremental Fit Indices 0 Comparative Fit Index (CPI) > .90
0 Tucker -Lewis Index (TLl) > .90
O Incremental Fit Index (IFI) > .90

 

Sources: Byme, 2001; Garson, 2006; Kline, 1998; Tabachnick & Fidell, 2001.

After evaluating the overall ﬁt of the model, its validity and reliability were
assessed. A set of conﬁrmatory factor analyses (CFAs) were conducted for China, Japan,

and South Korea, the results of which are separately reported below.

57

The China Model

Assessment of overall model ﬁt. Results of conﬁrmatory factor analyses for the
China model suggested that the proposed measurement model was not acceptable ()8
(340) = 1054.26, p = .01, CMIN/DF = 3.10, CFI = .88, TLI =.87, IFI = .88, RMSEA
= .05) as displayed in Table 13. While xz/dfand RMSEA results were within the
recommended range for a good model, other ﬁt indices indicated a poor ﬁt to the data.
Thus, it was decided to modify the proposed model for all three countries, beginning with
China.

AMOS provides the modiﬁcation indices for identifying which variables
contributed to this poor ﬁt. In order to modify the proposed model, signiﬁcance tests of
parameters were ﬁrst examined. Inspection of the standardized parameter estimates
showed that all observed variables loaded signiﬁcantly on their respective constructs,
indicating that the observed variables were good measures of the underlying constructs.
Then, modiﬁcation indices were examined. Inspection of modiﬁcation indices indicated
that the overall model ﬁt could be improved by removing three items (easy to navigate,
relaxing, and potential health problems).

After modifying the proposed model, another CFA was conducted for China. The
modiﬁed model had a better ﬁt to the data. The x2 was equal to 712.68, with 265 degrees
of freedom and a p-value of less than .01. While an insigniﬁcant value of x2 is preferable,
since 12 may be signiﬁcant if the sample size is large (Kline, 1998), other ﬁt indices were
examined to determine the overall model ﬁt of the modiﬁed measurement model. As
shown in Table 13, xz/df, CFI, TLI, IFI and RMSEA results were within the
recommended ranges of a good model ﬁt. Thus, it was concluded that the modiﬁed model

was acceptable. The x2 difference test, comparing the proposed and the modiﬁed

58

measurement models, showed that the )8 difference was 341.58 and statistically
signiﬁcant (p < .01). This means that the proposed model was signiﬁcantly improved

through modiﬁcation.

Table 13. Comparison of the proposed and modiﬁed full measurement models for China.

 

Model x2 Df xz/df CFI TLI IFI RMSEA sz Adf

 

Initial Model 1054.26 340 3.10 .88 .87 .88 .05

MOdiﬁid 712.68 265 2.69 .92 .90 .92 .04 341.58’ 75
Model

 

Note: CF I = Comparative Fit Index, TLI = Tucker-Lewis Index, IFI = Incremental Fit
Index, RMSEA = Root Mean Square Error of Approximation. Criteria to accept model:
xz/df should be smaller than 3; CFI, TLI, and IFI should be greater than .90; RMSEA
should be smaller than .05. signiﬁcant at p < 0.01

aModiﬁed model = Initial model — (Cog3 (Easy to navigate), Affe4 (Relaxing), and PR3
(Potential health problems»

Assessment of reliability and validity. The modiﬁed measurement model
provided support for both validity and reliability. As displayed in Table 14, all factor
loadings were statistically signiﬁcant with t-values ranging from 10.15 to 17.05 (p < .01)
and standardized factor loadings ranging from .44 to .83, both of which provide evidence
supporting convergent validity. As can be seen in Table 15, estimated correlations
between the constructs were not excessively high (less than .85, according to Kline
(1998)) and thus offered evidence supporting discriminant validity. Reliability analysis
using Cronbach’s alpha (or) indicates the internal consistency of the observed variables
measuring each factor. As shown in Table 15, all ﬁve factors had composite reliability
scores of greater or equal to .70. Since Cronbach’s or exceeded 0.60, it was deemed

acceptable (Burnkrant & Thomas, 1982), indicating that the observed variables reliably

measured their respective constructs. Given the results of these tests of model ﬁt, validity,

59

and reliability, it was concluded that the modiﬁed measurement model was acceptable as

the ﬁnal measurement model for China.

Table 14. Standardized factor loadings for the modiﬁed measurement model for China.

 

Standardized

 

 

 

 

 

 

 

Factors Paths . t-value
Estimates

Information Source Usage Info] .58 " 13.39
Inf02 .44 " 10.15

Inf03 .66 ” 13.88

Info4 .63 " 13.82

Inf05 .64 " 13.95

Info6 .69 " 14.77

lnfo7 .56 " 12.27

Cognitive Image Cogl .76 " 16.57
Cog2 .67 " 15.06

Cog4 .46 " 10.75

Cog5 .49 ” 11.40

Cog6 .52 " 12.11

Cog7 .76 " 17.05

Cog8 .66“ 15.12

Affective Image Affel .78 " 14.96
Affe2 .62 " 14.26

Affe3 ,3 .71 ” 14.73

Perceived Risks PR1 .55 " 11.28
PR2 .62 " 12.03

PR3 .67 ” 13.33

PR4 .76 " 14.03

PR5 .58 ” 11.80

PR6 .52 " 10.92

Intention to Visit V11 .83 " 14.43
V12 .62 " 12.95

 

Note:

"Factor loadings are all signiﬁcant at p < .01.

60

Table 15. Correlation analysis of factors for China.
N of Cronbach's

 

1 2 3 4 5 Variables a
1. Information Source 1 7 .82
Usage
2. Cognitive Image .18M 1 7 .75
3. Affective Image .22M .54" l 3 .81
4 .Perceived Risks - .24M -.39" -.46M 1 6 .71
5. Intention to Visit .33M .42" .49M -.45M 1 2 .78

 

Note: ”Correlations were all signiﬁcant at p < .01.

The Japan Model

Assessment of overall model ﬁt. Results of conﬁrmatory factor analyses for the
Japan model suggested that the proposed measurement model was not acceptable (x2
(340) = 931.11, p = .01, CMIN/DF = 2.74, CFI = .90, TLI = .89, IFI = .91, RMSEA
= .05) as displayed in Table 16. While xz/df and RMSEA results were within the
recommended range for a model with good ﬁt to the data, other ﬁt indices indicated a
poor ﬁt of the proposed model to the data. Thus, it was decided to modify the proposed
model.

AMOS provides the modiﬁcation indices for identifying which variables
contributed to this poor ﬁt. In order to modify the proposed model, signiﬁcance tests of
parameters were examined ﬁrst. Inspection of the standardized parameter estimates
showed that all observed variables loaded signiﬁcantly on their respective constructs,
indicating that the observed variables were good measures of the underlying constructs.

Aﬁer modifying the proposed model, another CFA was run for Japan. The
modiﬁed model had a better ﬁt to the data. The x2 was equal to 649.07, with 265 degrees
of freedom and a p-value of less than .01. While an insigniﬁcant value of x2 is preferable,

since )8 may be signiﬁcant if the sample size is large (Kline, 1998), other ﬁt indices were

61

examined to determine the overall model ﬁt of the modiﬁed measurement model. As
shown in Table 16, xz/df, CF I, TLI, and RMSEA were within the recommended range for
a model with good ﬁt to the data. Thus, given these positive model ﬁt test results, it was
concluded that the modiﬁed model was acceptable. The 12 difference test, comparing the
pr0posed and the modiﬁed measurement models, showed that the x2 difference was
282.15 and statistically signiﬁcant (p < .01 ). This means that the proposed model was

signiﬁcantly improved through modiﬁcation.

Table 16. Comparison of the proposed and modiﬁed full measurement models for Japan.

 

 

Model x2 Df {my CFI TLI IFI RMSEA 8x2 Adf
Initial Model 931.22 340 2.74 .90 .89 .91 .05

MOdiﬁfd 649.07 265 2.44 .93 .92 .93 .04 282.15‘ 75
Model

 

Note: CFI = Comparative Fit Index, TLI = Tucker-Lewis Index, IFI = Incremental Fit
Index, RMSEA = Root Mean Square Error of Approximation. Criteria to accept model:
ledf should be smaller than 3; CFI, TLI, and IFI should be greater than .90; RMSEA
should be smaller than .05. signiﬁcant at p < 0.01

aModiﬁed model = Initial model — (Cog3 (Easy to navigate), Affe4 (Relaxing), and PR3
(Health problems))

Assessment of reliability and validity. The modiﬁed measurement model provided
support for both validity and reliability. As displayed in Table 17, all factor loadings were
statistically signiﬁcant with t-value ranging from 10.21 to .1614 (p < .01) and
standardized factor loadings ranging from .49 to .82, both of which provided evidence
supporting convergent validity. As shown in Table 18, estimated correlations between the
constructs were not excessively high (less than .85, according to Kline (1998)) and thus
offered evidence supporting discriminant validity. Reliability analysis using Cronbach’s

or indicated the internal consistency of the observed variables measuring each factor. As

shown in Table 18, all ﬁve factors had composite reliability scores greater than or equal

62

to .70. If Cronbach’s or exceeds 0.60, it is deemed acceptable (Bumkrant & Thomas,
1982). Thus, the reliability test showed that observed variables reliably measured their
respective constructs. Given the results of these tests of model ﬁt, validity, and reliability,
it was concluded that the modiﬁed measurement model was acceptable as the ﬁnal

measurement model for Japan.

Table 17. Standardized factor loadings for the modiﬁed measurement model for Japan.

 

 

 

 

 

 

 

 

 

 

Factors Paths Standardized t—value
..- Estimates

Information Source Usage Info] .57" 13.60
Inf02 .50" 11.53

Info3 .67" 14.90

Info4 .65“ 14.48

Inf05 .70” 15.60

Info6 .80” 16.22

Info7 __,___W___ .69" 15.12

Cognitive Image Cogl .75" 16.14
Cog2 .68” 15.01

Cog4 .50" 11.67

CogS .49” 1 1.40

Cog6 .51 " 11.65

Cog7 .72” 15.82

cogs ., .65” 14.77

Affective Image Affel .82" 15.84
Affe2 .62” 14.46

Affe3 .81" 15.63

Perceived Risks PR1 .59" 11.80
PR2 .60” 11.69

PR3 .72” 13.73

PR4 .74" 13.89

PR5 .58” 11.27

PR6W________W W .48” W 10.21

Intention to Visit V11 .80" 14.34
V12 .65" 13.01

 

Note: "Factor loadings are all signiﬁcant at p < .01.

63

Table 18. Correlation analysis of factors for J eman.
N of Cronbach's

 

l 2 3 4 5 Variables a
l. Inforrnatron 1 7 .79
Source Usage
2. Cognitive Image .27” 1 7 .75
3. Affective Image .25M .54” l 3 .81
4 .Perceived Risks -22" -38" -.39** 1 6 .73
5. Intention to Visit .37" .47" .43" -38" 1 2 .78

 

Note: ”Correlations were all signiﬁcant at p < .01.

The South Korea Model
Assessment of the overall model ﬁt. Results of conﬁrmatory analyses for the

South Korea model suggested that the proposed measurement model was not acceptable
()8 (340) = 1144.51, p = .01, CMlN/DF = 3.37, CFI = .90, TLI = .88, IFI = .90, RMSEA
= .06) as displayed in Table 19. While xZ/df and RMSEA results were within the
recommended range for a model with good ﬁt to the data, other ﬁt indices indicate a poor
ﬁt of the proposed to the data. Thus, it was decided to modify the proposed model.

AMOS provides the modiﬁcation indices for identifying which variables
contributed to a poor ﬁt of the proposed model to the data. In order to modify the
proposed model, signiﬁcance tests of parameters were examined ﬁrst. Inspection of the
standardized parameter estimates showed that all observed variables loaded signiﬁcantly
on their respective constructs, showing that the observed variables were good measures of
the underlying constructs.

After modifying the proposed model, another CF A was run for South Korea. The
modiﬁed model had a better ﬁt to the data. The x2 was equal to 750.30, with 265 degrees
of freedom and a p-value of less than .01. While the insigniﬁcant value of x2 is preferable,

since 752 may be signiﬁcant if the sample size is large (Kline, 1998), other ﬁt indices were

64

examined to determine the overall model ﬁt of the modiﬁed measurement model. As
shown in Table 19, xz/df, CF I, TLI, and RMSEA were within the recommended ranges
for a model with good ﬁt to the data. Thus, given these positive model ﬁt test results, it
was concluded that the modiﬁed model was acceptable. The x2 difference test, comparing
the proposed and the modiﬁed measurement models, showed that the )8 difference was
394.21 and statistically signiﬁcant (p < .01). This means that the proposed model was

signiﬁcantly improved through modiﬁcation.

Table 19. Comparison of the initial and modiﬁed full measurement models for South

 

 

Korea.

Model x2 df xz/df CFI TLI IFI RMSEA sz Adf
initial Model 1144.51 340 3.37 .90 .88 .90 .06
M°diﬁfd 750.30 265 2.83 .93 .92 .93 .04 394.21' 75
Model

 

Note: CFI = Comparative Fit Index, TLI = Tucker-Lewis Index, IFI = Incremental Fit
Index, RMSEA = Root Mean Square Error of Approximation. Criteria to accept model:

xz/df should be smaller than 3; CF I, TLI, and IFI should be greater than .90; RMSEA
should be smaller than .05. signiﬁcant at p < 0.01

aModiﬁed model = Initial model — (Cog3 (Easy to navigate), Affe4 (Relaxing), and PR3
(Health problems))

Assessment of reliability and validity. The modiﬁed measurement model provided
support for both validity and reliability. As displayed in Table 20, all factor loadings
were statistically signiﬁcant with t-values ranging from 9.91 to 17.07 (p < .01) and
standardized factor loadings ranging from .44 to .83, both of which provided evidence
supporting convergent validity. As can be seen in Table 21, estimated correlations
between the constructs were not excessively high (less than .85, according to Kline

(1998)), and thus offered evidence supporting discriminant validity. Reliability analysis

using Cronbach’s or indicates the internal consistency of the observed variables measuring

65

each factor. As shown in Table 21, all ﬁve factors had composite reliability scores of
greater than or equal to .70. If Cronbach’s a exceeds 0.60, it is deemed acceptable
(Bumkrant & Thomas, 1982). Thus, the reliability test shows that observed variables
reliably measured their respective constructs. Given the results of these tests of model ﬁt,
validity, and reliability, it was concluded that the modiﬁed measurement model was

acceptable as the ﬁnal measurement model for South Korea.

Table 20. Standardized factor loadings for the modiﬁed measurement model for S. Korea.

 

 

 

 

 

 

 

 

Factors Paths Standardlzed t-value
EstImates

Information Source Usage Infol .64" 13.49

Info2 .44" 9.91

Inf03 .69" 13.69

Info4 .70” 13.40

Inf05 .75" 13.73

Info6 .77” 13.99

. Info7 .57” 13.11

Cognitive Image Cogl .77" 16.58
Cog2 .74” 15.07

Cog4 .56" 10.75

Cog5 .59“ 11.40

Cog6 .76” 12.12

Cog7 .68" 17.07

Cog8 ., .66” 16.89

Affective Image Affel .78" 14.95
Affe2 .61" 13.26

Affe3 .71" 14.45

Perceived Risks PR1 .63" 11.28
PR2 .71" 12.03

PR3 .62” ' 13.33

PR4 .71" 14.04

PR5 .59” 10.80

PR6 .79” 14.97

Intention to Visit V11 .83" 15.01
V12 .62” 12.87

 

Note: "F actor loadings are all signiﬁcant at p < .01.

66

Table 21. Correlation analysis among factors for South Korea.
N of Cronbach's

 

1 2 3 4 5 Variables a
1. Information 1 .7 .80
Source Usage
2. Cognitive Image .25" 1 7 .75
3. Affective Image .18M .55" 1 3 .81
4 .Perceived Risks -.21** -45“ -.55** 1 6 .74
5. Intention to Visit .32“ .44" .41“ -48“ 1 2 .78

 

Note: ”Correlations were all signiﬁcant at p < .01.

Comparison of the Measurement Model for the Three Countries.

The proposed measurement model was tested for China, Japan, and South Korea.
The overall model ﬁt of the proposed model was poor across all three countries. The
proposed model was modiﬁed for each of the three countries: Inspection of modiﬁcation
indices indicated that the overall model ﬁt could be improved by removing three items
(easy to navigate, relaxing, and potential health problems). This modiﬁed model was
used for China, Japan, and South Korea. Since the modiﬁed model had an adequate ﬁt to
the data across all three countries, it was accepted as the ﬁnal measurement model for

testing the structural equation model for China, Japan, and South Korea.

Testing the Structural Equation Model
Since the ﬁnal measurement model ﬁt was satisfactory, the structural equation
model (SEM), combining both the ﬁnal measurement model and the structural model,
was tested. Testing the structural equation model involves assessment of both the
measurement model and the hypothesized relationships among the factors identiﬁed in

this model. The proposed structural equation model is presented in Figure 4. Employing

67

AMOS, a set of SEM analyses was conducted for the three countries, the results of which

are separately reported below.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9‘ Infol
92 mm
°3 Inf03
°‘ lnfo4
Inf05
66 Inf06
07 Info7

 

 

 

ea 1

6i

@‘1

 

 

 

 

 

 

 

 

6? l

 

 

 

 

 

 

 

 

 

cool cooz coo4 coos (3007 C003
Cognitive
Image
rl
‘2 Affel 1
Ingmon Affective Affez e1
Usage Image Mk3 1
r3
Perceived
Risks
p15 PR4 PR5 PR6 PR7

 

 

 

 

 

 

 

 

 

 

 

 

 

arid

 

$88

 

lVl IV2

Intention '4

To
Visit

 

Figure 4. The proposed structural equation model.

Note: Affel: Pleasant, Affe2: Arousing, Affe3: Exciting;
Cogl: Many things to see and do, Cog2: Natural scenic beauty, Cog4: Plenty of quality
accommodations, Cog5: Friendly local people, Cog6: Appealing local food and drink,

Cog7: Interesting historical attractions, Cog 8: Unique cultural resources;

Infol: TV/radio, Inf02: Word of mouth, Info3: Books and movies, Info4: Internet,

Inf05: Cultural experiences, Info6: Newspapers/magazines, Info7: Professional advice;
1V1: Interest, 1V2: Likelihood;
PR1: Likelihood of illness, PR2: Political instability, PR4: Terrorism, PR5: Violence,
PR6: Natural disaster, PR7: Language barrier
Error Terms: el-e27, rl-r3

68

 

The China Model

The goodness-of-ﬁt indices used to assess the measurement model were also
utilized to determine whether the proposed structural equation model ﬁt the data for
China. As displayed in Table 22, the )6 statistic was equal to 746.03, with 245 degrees of
ﬁ'eedom and a p-value of less than .01. When the )8 statistic is insigniﬁcant, the model is
deemed acceptable. However, since chi-square statistics based on a large sample size may
result in a signiﬁcant x2 statistic, other ﬁt indices were utilized as alternatives to determine
the adequacy of the model ﬁt. CFI, IFI, and TL] values of over .90, RMSEA values of
less than .05, and xz/df values of less than three indicate an acceptable ﬁt. Based on these
criteria, the results of the ﬁt indices displayed in Table 22 indicated a poor ﬁt of the
proposed structural equation model to the data, suggesting that the proposed model
needed to be modiﬁed.

Inspection of modiﬁcation indices (MIs) showed that the MI value for the error
variable parameter, bearing on the association between the error term of cognitive images
(r1) and the error term of perceived risks (r3), was relatively high, with a value of 59.81.
This suggested that a better ﬁt could be achieved by correlating the errors of the factors.
In order to link the error terms, it must be supported by a substantive and/or empirical
rationale (Byme, 2001). Several authors have suggested that there is an association
between cognitive image of a destination and perceived risks of vacationing there (Suh
and Gartner, 2004). When travelers perceive very low risks of vacationing in a destination,
their evaluations of the destination tend to be very high. Also, travelers’ negative
evaluations may be inﬂuenced by perceptions of high risks of vacationing in the
destination. Since these suppositions are plausible, it was decided to link the error terms,

a decision which resulted in the model for China depicted in Figure 5.

69

 

“9863896?

ICOGI C002 C004 005 006 007 008

 

Cognitive
Image

‘
Affective

Perceived
Risks

    
 

 

eoooooo
EEEEEEE

 

I???"

PR4 PR5 PR6

added.

 

 

 

 

 

Figure 5. Modiﬁed SEM for China.

Note: Affel: Pleasant, Affe2: Arousing, Affe3: Exciting;

Cogl: Many things to see and do, Cog2: Natural scenic beauty, Cog4: Plenty of quality
accommodations, Cog5: Friendly local people, Cog6: Appealing local food and drink,
Cog7: Interesting historical attractions, Cog 8: Unique cultural resources;

Infol: TV/radio, InfoZ: Word of mouth, Info3: Books and movies, Info4: Internet,
Inf05: Cultural experiences, Info6: Newspapers/magazines, Info7: Professional advice;
IVl: Interest, 1V2: Likelihood;

PR1: Likelihood of illness, PR2: Political instability, PR4: Terrorism, PR5: Violence,
PR6: Natural disaster, PR7: Language barrier

Error Terms: el-eZ7, rl-r3

70

The modiﬁed model had a better ﬁt to the data. The x2 statistic was equal to
681.01, with 244 degrees of freedom and a p-value of less than .01. While an insigniﬁcant
value of x2 is preferable, since 752 may be signiﬁcant if the sample size is large (Kline,
1998), other ﬁt indices were examined to determine the overall model ﬁt of the modiﬁed
measurement model. As shown in Table 22, xz/df, CFI, TLI, and RMSEA were within the
recommended ranges for a model with good ﬁt to the data. Thus, it was concluded that
the modiﬁed structural equation model was acceptable. The x2 difference test, comparing
the proposed and the modiﬁed measurement models, showed that the x2 difference was
65.02 and statistically signiﬁcant (p < .01). This means that the proposed model was

signiﬁcantly improved through modiﬁcation.

Table 22. Comparison of the initial and modiﬁed structural models for China.

 

 

Model 1 df ledf CFI TLI IFI RMSEA sz Adi
Initial Model 746.03 245 3.04 .90 .89 .90 .05
Modiﬁed Model 681.01 244 2.79 .91 .90 .91 .04 65.02‘ 1

 

Note: CFI = Comparative Fit Index, TLI = Tucker-Lewis Index, IFI = Incremental Fit
Index, RMSEA = Root Mean Square Error of Approximation. Criteria to accept model:
xz/df should be smaller than 3; CFl, TLI, and IFI should be greater than .90; RMSEA
should be smaller than .05. ~‘signiﬁcant at p < 0.01

The Japan Model

The goodness-of-ﬁt indices used to assess the measurement model were also
utilized to determine whether the proposed structural equation model ﬁt the data for
Japan. As displayed in Table 23, the x2 statistic was equal to 687.63, with 245 degrees of
freedom and a p—value of less than .01. When the )8 statistics is insigniﬁcant, the model is
accepted. However, since chi-square statistics based on a large sample size may result in a

signiﬁcant )8 statistic, other ﬁt indices were utilized as alternatives to determine the

71

adequacy of the model ﬁt. CFI, IFI, and TLI values of over .90, RMSEA values of less
than .05, and ledf values of less than three indicate an acceptable ﬁt. Based on these
criteria, the results of the ﬁt indices displayed in Table 23 indicated a poor ﬁt of the
proposed structural equation model to the data, suggesting that the proposed model
needed to be modiﬁed.

Inspection of modiﬁcation indices (Mls) showed that the MI value for the error
variable parameter, bearing on the association between the error term of cognitive image
(r1) and the error term of perceived risks (r3), was relatively high, with a value of 61 .41.
This suggested that a better ﬁt could be achieved by correlating the errors of the factors.
In order to link the error terms, it must be supported by a substantive and/or empirical
rationale (Byme, 2001). Several authors have suggested that there is an association
between cognitive image of a destination and perceived risks of vacationing there (Suh &
Gartner, 2004). When travelers perceive very low risks of vacationing in a destination,
their evaluations of the destination tend to be very high. Also, travelers’ negative
evaluations may be inﬂuenced by their perceptions of high risks of vacationing in the
destination. Since these suppositions are plausible, it was decided to link the error terms.
The modiﬁed structural equation model for Japan was the same as that for China as
shown in Figure 5.

The modiﬁed model had a better ﬁt to the data. The x2 was equal to 620.64, with
244 degrees of freedom and a p—value of less than .01. While an insigniﬁcant value of x2
is preferable, since x2 may be signiﬁcant if the sample size is large (Kline, 1998), other ﬁt
indices were examined to determine the overall model ﬁt of the modiﬁed measurement
model. As shown in Table 23, xz/df, CFI, TLI, and RMSEA were within the

recommended ranges for a model with good ﬁt to the data. Thus, it was concluded that

72

the modiﬁed structural equation model was acceptable. The )6 difference test, comparing
the proposed and the modiﬁed measurement models, showed that the )8 difference was
66.99 and statistically signiﬁcant (p < .01). This means that the proposed model was

signiﬁcantly improved through modiﬁcation.

Table 23. Comparison of the initial and modiﬁed structural models for Japan.

 

 

Model x2 df ﬂay CFI TLI IFI RMSEA Ax2 Adi
urinal Model 687.63 245 2.81 .92 .89 .92 .05
Modiﬁed Model 620.64 244 2.54 .93 .92 .93 .04 66.99‘ 1

 

Note: CFI = Comparative Fit Index, TLI = Tucker-Lewis Index, IF I = Incremental Fit
Index, RMSEA = Root Mean Square Error of Approximation. Criteria to accept model:
xz/df should be smaller than 3; CF I, TLI, and IFI should be greater than .90; RMSEA
should be smaller than .05. ‘signiﬁcant at p < 0.01

The South Korea Model

The goodness-of-ﬁt indices used to assess the measurement model were also
utilized to determine whether the proposed structural equation model ﬁt the data for
South Korea. As displayed in Table 24, the x2 statistic was equal to 837.97, with 245
degrees of freedom and a p-value of less than .01. When the )8 statistic is insigniﬁcant,
the model is accepted. However, since chi-square statistics based on a large sample size
may result in a signiﬁcant )8 statistic, other ﬁt indices were utilized as alternatives to
determine the adequacy of the model ﬁt. CFI, IFI, and TLI values of over .90, RMSEA
values of less than .05, and xz/df values of less than three indicate an acceptable ﬁt. Based
on these criteria, the results of the ﬁt indices displayed in Table 24 indicated a poor ﬁt of
the proposed structural equation model to the data, suggesting that the proposed model

needed to be modiﬁed.

73

Inspection of modiﬁcation indices (MIs) showed that the MI value for the error
variable parameter, bearing on the association between the error term of cognitive image
(r1) and the error term of perceived risks (r3), was relatively high, with a value of 95.01.
This suggested that a better ﬁt could be achieved by correlating the errors of the factors.
In order to link the error terms, it must be supported by a substantive and/or empirical
rationale (Byme, 2001). Several authors have suggested that there is an association
between cognitive image of a destination and perceived risks of vacationing there (Suh
and Gartner, 2004). When travelers perceive very low risks of vacationing in a destination,
their evaluations of the destination tend to be very high. Also, travelers’ negative
evaluations may be inﬂuenced by perceptions of high risks of vacationing in the
destination. Since these suppositions are plausible, it was decided to link the error terms.
The modiﬁed structural equation model for South Korea was the same as that for China
and Japan, as shown in Figure 5.

The modiﬁed model had a better ﬁt to the data. The x2 was equal to 733.13, with
244 degrees of freedom and a p-value of less than .01. While an insigniﬁcant value of x2
is preferable, since x2 may be signiﬁcant if the sample size is large (Kline, 1998), other ﬁt
indices were examined to determine the overall model ﬁt of the modiﬁed measurement
model. As shown in Table 24, xz/df, CF I, TLI, and RMSEA were within the
recommended ranges of a model with good ﬁt to the data. Thus, it was concluded that the
modiﬁed structural equation model was acceptable. The x2 difference test, comparing the
proposed and the modiﬁed measurement models, showed that the 12 difference was
104.84 and statistically signiﬁcant (p < .01). This means that the proposed model was

signiﬁcantly improved through modiﬁcation.

74

Table 24. Comparison of the initial and modiﬁed structural models for South Korea.

 

Model x2 df xz/df CFI TLI IFI RMSEA Ax’ Adi
Initial Model 837.97 245 3.42 .91 .90 .91 .06
Modiﬁed Model 733.13 244 2.98 .93 .92 .93 .05 104.84‘ 1

 

Note: CFI = Comparative Fit Index, TLI = Tucker-Lewis Index, IF I = Incremental Fit
Index, RMSEA = Root Mean Square Error of Approximation. Criteria to accept model:
x2/df should be smaller than}; CFI, TLI, and IFI should be greater than .90; RMSEA
should be smaller than .05. signiﬁcant at p < 0.01
Comparison of the Structural Equation Model for the Three Countries

The proposed structural equation model, combining both the measurement and
structural models, was tested for China, Japan, and South Korea. The overall model ﬁt of
the proposed structural equation model was poor across all three countries. The proposed
model was modiﬁed for each of the three countries, and this modiﬁed model was used for
China, Japan, and South Korea. Since the modiﬁed structural equation model had an
adequate ﬁt to the data across all three countries, it was accepted as the ﬁnal measurement

model for examining the hypothesized relationships among the factors. In the next section,

the results of these hypothesis tests are presented.

Hypotheses Testing
The eight hypotheses regarding the relationships among the factors were tested in
the structural equation model. The results of these hypothesis tests are separately reported
for China, Japan, and South Korea. In this section, ﬁndings regarding the direct inﬂuence
of one factor on another, as they relate to the hypotheses of this study, are reported ﬁrst.
Then, squared multiple correlations are reported to show the total variance in the factor
(endogenous variables: dependent variables) as explained by the direct inﬂuences of the

factor(s) (exogenous variables: independent variables or predictors). One advantage of

75

structural equation modeling is that it allows one to test not only direct inﬂuence, but also
indirect and total inﬂuence between factors (Joreskog & Sorbom, 1993). Findings
regarding the indirect and the total inﬂuence of one factor on another, as they relate to the

hypotheses of this study, are presented at the end of this section.

The China Model

The results of hypotheses testing are presented in Table 25, including standardized
path coefﬁcients, critical ratio, and p values. A path diagram is also presented in Figure 6.
The path diagram shows standardized path coefﬁcients, representing the direction and
strength of the direct inﬂuence of one factor on another, and squared multiple correlations

indicating the total variance in a factor explained by the factor(s).

Table 25. Results of hypotheses testing for China.

 

 

Paths Cofffrtgient Critical Ratio P-value Supporta
“$2.3 x
$213523; sage 9 .04 .98 0.32 Y
afrigaoﬁgﬁsmggsage 9 -.26 -5.19 0.00 x
gligetlttilt‘l: 13:5: (:4) -56 11.36 0.00 x
Arise?) x
ﬁlﬂldhetrlrnihift (:16) -19 3.00 0.00 x
31:23:21: tbm;§;t()H7) -41 5.38 0.00 X
firetre‘hii‘ofldt?\s’li(ssit_()H8) ~25 -4.86 0.00 x

 

Note: 3X = Supported hypothesis, Y = Did not support hypothesis

76

 

 

 

 

 

R-square=.l 1
Cognitive
Image
.56 .19
‘ R-square=.53
Affective .41 Intentlon
Image * To
Visit
R-square=.56
-.28 -.25
Perceived
Risks
— signiﬁcant
R-square=.07 ........ non signiﬁ cant

 

 

 

Figure 6. Results of testing the proposed structural model with standardized path
coefficients and squared multiple correlations (R2) for China.

Note: Squared multiple correlations indicate the total variance explained by the direct
inﬂuence of one factor on another.

Direct inﬂuence related to study hypotheses. Regarding China, seven of eight
hypotheses were supported, as shown in Figure 6. Information source usage on China had
signiﬁcant direct inﬂuences on cognitive image of China and perceived risks of
vacationing in China (support for H1 and H3). The direct inﬂuence of information source
usage on cognitive image ([3 = .33, p < .01) was larger than that on perceived risks ([3 = -
.26, p < .01). The results of the hypotheses testing indicate that respondents who had

more information on China appeared to have more positive cognitive image of the

country and lower perceived risks of vacationing there. Only the hypothesis on the

77

relationship between information source usage and affective image (H2) was unsupported.
No signiﬁcant direct inﬂuence of information source usage on affective image was found
(B = .04, p > .01).

Cognitive image of China and perceived risks of vacationing in China had a
signiﬁcant direct inﬂuence on affective image of China (support for H4 and H5). This
indicated that respondents who had a more positive cognitive image, and lower perceived
risks, of China appeared to have a more positive affective image of China. The direct
inﬂuence of cognitive image on affective image was very large (B = .56, p < .01), which
supported previous studies showing the strong association between cognitive and
affective image. Perceived risks of vacationing in China had a signiﬁcant and negative
inﬂuence on affective image of China (B = - .28, p < .01). Respondents who perceived
higher risks of vacationing in China had a more negative affective image of the country.

Cognitive image of China, affective image of China, and perceived risks of
vacationing in China had signiﬁcant direct inﬂuences on respondents’ intentions to visit
China (support for H6, H7, and H8). This indicated that respondents who had more
positive cognitive and affective images of China and lower perceived risks of vacationing
in China appeared to have stronger intentions to visit China. Among the three factors
inﬂuencing Americans’ intentions to visit China, affective image of China had the
strongest direct inﬂuence on visit intention (B = .41 , p < .01), while perceived risks of
vacationing in China had the second strongest direct inﬂuence on Visit intention (B = -.25,
p < .01). Respondents who had more positive cognitive images of China showed greater
intention to visit China ((B = .19, p < .01), even though the inﬂuence of cognitive image

on Visit intention was weaker than those of affective image and perceived risks.

78

Multiple squared correlations. As shown in Figure 6, 53 percent of variance in
intention to visit China was explained by the direct inﬂuences of cognitive image,
affective image, and perceived risks. Information sources explained 11 percent of
variance in cognitive image of China and 7 percent in perceived risks of vacationing in
China. The large portion of unexplained variance in cognitive image and perceived risks
suggests the possibility of including other factors which may inﬂuence cognitive image '
and perceived risks. Fifty-six percent of variance in affective image was explained by

cognitive image and perceived risks.

Indirect and total inﬂuences of the factors on visit intention. The standardized
indirect inﬂuences of the factors on intention to visit China are presented in Table 26.
Information source usage on China had a positive indirect inﬂuence on intention to visit
China through perceived risks and cognitive image (B = .25). In other words, information
source usage on China had a positive direct inﬂuence on cognitive image, which, in turn,
had a positive direct inﬂuence on visit intention. Cognitive image and perceived risks also
had indirect inﬂuences on visit intention through affective image. While the indirect
inﬂuence of cognitive image on visit intention (B = .23) was larger than the direct
inﬂuence (B = .19), the indirect inﬂuence of perceived risks on Visit intention (B = - .12)
was smaller than the direct inﬂuence (B = - .25).

Examination of the standardized total inﬂuence combining both direct and indirect
inﬂuences, as presented in Table 26, showed that cognitive image had the largest
standardized total inﬂuence (B = .42) and affective image had the second largest (B = .41).
The total inﬂuence of perceived risks (B = .37) was smaller than that of the two

destination image factors.

79

Table 26. Standardized estimates for direct, indirect, and total inﬂuences for China.

 

Standardized Standardized Standardized

 

 

 

 

 

 

 

Exogenous Endogenous Direct Indirect Total
variables variables . . .
1nﬂuence Inﬂuence Inﬂuence
Affective Image .04 .26 .30
Information Cognitive Image .33 --- .33
30”“ Perceived Risks -.26 -.26
Usage

Intention to Visit --- .25 .25
Affective Image Intention to Visit .41 --- .41
Cognitive Image Affective Image .56 --- .56
Intention to Visit ._ .19 .23 .42
Perceived risks Affective Image -.28 --- -.28
Intention to Visit -.25 -.12 -.37

 

The Japan Model

The results of hypotheses testing are presented in Table 27, including standardized
path coefﬁcients, critical ratio, and p values. A path diagram is also presented in Figure 7.
The path diagram shows standardized path coefﬁcients, representing the direction and
strength of the direct inﬂuences of one factor on another, and squared multiple

correlations indicating the total variance explained by direct inﬂuences between factors.

80

Table 27. Results of hypotheses testing for Japan.

 

 

Paths Coffaﬁtfien t Critical Ratio P-value Supporta
g$3;0?mi°gf:§gsage ‘) .38 7.85 0.00 x
$233201?“ Egrfﬁztisage 9 .06 1.41 0.16 Y
gigggoﬁi 51555516335835 9 —.29 -5.98 0.00 X
gig??? 11:21:53“ .57 1 1.15 0.00 x
5.555551513152551 "’2 ‘4'” 0'00 x
giggfletin‘l/aliiet (:16) '34 5' 16 000 X
855 55555317) 28 “5 0'00 X
5152:5555? 1815515318) "20 5'89 0'00 x

 

Note: 5X = Supported hypothesis, Y = Did not support hypothesis

81

 

 

 

 

 

 

R-square=. 14
Cognitive
Image
.57 .34
‘ R-Square=.49
Affeh .28 Intention
Image To
Visit
R-square=.52
-.22 -.20
Perceived
Risks
— signiﬁcant
R-square=.08 ........ non Signiﬁcant

 

 

Figure 7. Results of testing the proposed structural model with standardized path
coefficients and squared multiple correlations (R2) for Japan.
Note: Squared multiple correlations indicate the total variance explained by the direct

inﬂuence of one factor on another.

Direct inﬂuences related to study hypotheses. Regarding Japan, seven of eight

hypotheses were supported, as shown in Figure 7. Information source usage had

signiﬁcant direct inﬂuences on cognitive image of Japan and perceived risks of

vacationing in Japan (support for H1 and H3). This ﬁnding indicates that respondents

who obtained more information on Japan had a more positive cognitive image of Japan

and lower perceived risks of vacationing in Japan. The direct inﬂuence of information

source usage on cognitive image (B = .3 8, p < .01) was larger than this factor’s inﬂuence

on perceived risks (B = -.29, p < .01). Only the hypothesis regarding the relationship

82

between information source usage and affective image (H2) was unsupported. No
signiﬁcant direct inﬂuence of information source usage on affective image was found (B
= .06, p > .01).

Cognitive image of Japan and perceived risks of vacationing in Japan had a
signiﬁcant direct inﬂuence on affective image of Japan (support for H4 and H5). This
ﬁnding indicates that respondents who had a positive cognitive image of Japan and lower
perceived risks of vacationing in the country appeared to have a more positive affective
image of Japan. The direct inﬂuence of cognitive image on affective image was very
strong (B = .57, p < .01), which supported previous studies showing the strong association
between cognitive and affective image (Baloglu and McCleary, 1999). Perceived risks of
vacationing in Japan had a signiﬁcant and negative direct inﬂuence on affective image (B
= - 22, p < .01), but this direct inﬂuence of perceived risks was smaller than that of
cognitive image on affective image.

Cognitive and affective image of Japan and perceived risks of vacationing in
Japan had signiﬁcant direct inﬂuences on respondents’ intentions to visit Japan (support
for H6, H7, and H8). This result indicates that respondents who had more positive
cognitive and affective images and lower perceived risks regarding Japan appeared to
have greater intention to visit Japan. Among the three factors inﬂuencing Visit intention to
Japan, cognitive image had the strongest direct inﬂuence on visit intention (B = .34, p
< .01); affective image had the second strongest (B = .28, p < .01). The direct inﬂuence of
perceived risks of vacationing in Japan on intention to Visit (B = -.20, p < .01) was smaller

than that of the two destinations’ image factors.

83

Multiple squared correlations. As shown in Figure 7, 49 percent of variance in
intention to visit Japan was explained by the direct inﬂuences of cognitive image,
affective image, and perceived risks. Information source usage explained 14 percent of
variance in cognitive image of Japan, but only 8 percent of variance in Americans’
perceived risks of vacationing in Japan. The large portion of unexplained variance in
cognitive image and perceived risks suggests the possibility of including other factors
which may inﬂuence cognitive image and perceived risks in future research. Fifty-two
percent of variance in affective image was explained by cognitive image and perceived

risks.

Indirect and total inﬂuences of the factors on visit intention. The standardized

indirect inﬂuences of the factors on intention to visit Japan are presented in Table 28.

Table 28. Standardized estimates for direct, indirect, and total inﬂuences for Japan.

 

Standardized Standardized Standardized

 

 

 

 

 

 

 

Exogenous Endogenous Direct Indirect Total
variables variables . . .
Inﬂuence inﬂuence . Inﬂuence
Affective Image .06 .28 .34
Information Cognitive Image .38 --- .38
Source Perceived risks -.29 --- -.29
Usage

Intention to visit --- .28 .27
Affective Image Intention to visit "28 _...,._ --- .28
Cognitive Image Affective Image .57 --- .57
Intention to visit 34 .16 .50
Perceived risks Affective Image -.22 --- -.22
Intention to visit -.20 -.06 -.26

 

Information source usage on Japan had a positive indirect inﬂuence on intention to
visit Japan through perceived risks and cognitive image (B = .27). In other words,

information source usage on Japan had a positive direct inﬂuence on cognitive image,

84

which, in turn, had a positive direct inﬂuence on Visit intention. Cognitive image of Japan
and perceived risks also had indirect inﬂuences on visit intention through affective image.
The indirect inﬂuence of cognitive image on intention to visit (B = .16) was larger than
that of perceived risks (B = - .06).

Examination of the standardized total inﬂuence combining both direct and indirect
standardized inﬂuences, as presented in Table 28, showed that cognitive image had the
largest standardized total inﬂuence (B = .50). Affective image had the second largest, but
the standardized total inﬂuence of affective image (B = .28) was much smaller than that of
cognitive image. Perceived risks of vacationing in Japan had the smallest standardized

total inﬂuence on intention to visit (B = -.26).

85

The South Korea Model

The results of hypotheses testing are presented in Table 29, including standardized
path coefﬁcients, critical ratio, and p values. A path diagram is also presented in Figure 8.
The path diagram shows standardized path coefﬁcients, representing the direction and
strength of the direct inﬂuence of one factor on another, and squared multiple correlations

indicating the total variance explained by direct inﬂuences between factors.

Table 29. Results of hypotheses testing for South Korea.

 

 

Paths Coggfiem Critical Ratio P-value Supporta
5.5555555552553555 5 '33 7'31 0'00 X
5555555555555555 5 '0‘ '23 0'81 Y
55255550515555153555 5 “22 '4'96 0'00 X
3m?) X
5555555 15115:: (:15) "28 "6'15 0'00 X
gm?) X
35:31::56515553H7) '18 2'96 0'00 X
11555115515515558318) "8 2'96 0'00 X

 

Note: 1’X = Supported hypothesis, Y = Did not support hypothesis

86

 

R-square=.l 1

Cognitive
Image

.54 .28

 

R-square=.45

‘
NEE . 1 8 Intention
Image v To
/ Visit
R-square=.51

 

 

-.28 -.33
Perceived
Risks
— signiﬁcant
R-square=.05 -------- non signiﬁcant

 

 

 

Figure 8. Results of testing the proposed structural model with standardized path
coefficients and squared multiple correlations (R2) for South Korea.

Note: Squared multiple correlations indicate the total variance explained by the direct
inﬂuence of one factor on another.

Direct inﬂuences related study hypotheses. Regarding South Korea, seven of

eight hypotheses were supported, as shown in Figure 8. Information source usage had

 

signiﬁcant direct inﬂuences on cognitive image of South Korea and on perceived risks of
vacationing in South Korea (support for H1 and H3). This ﬁnding indicates that
respondents who used more information appeared to have more positive cognitive image
of the country and lower perceived risks of vacationing there. The direct inﬂuence of

information source usage on cognitive image (B = .33, p < .01) was larger than this

87

factor’s inﬂuence on perceived risks (B = -.22, p < .01). Only the hypothesis regarding the
relationship between information source usage and affective image (H2) was rejected (B
= .01, p > .01).

Cognitive image of South Korea and perceived risks of vacationing in South
Korea had a signiﬁcant and positive direct inﬂuence on affective image of South Korea
(support for H4 and H5). This ﬁnding indicates that respondents who had a more positive
cognitive image of the country and lower perceived risks of vacationing there appeared to
have a more positive affective image of South Korea. The direct inﬂuence of cognitive
image on affective image was very large (B = .54, p < .01), which supported previous
studies showing the strong association between cognitive and affective image. Perceived
risks of vacationing in South Korea had a signiﬁcant and negative direct inﬂuence on
affective image of the country (B = - .28, p < .01) but this direct inﬂuence of perceived
risks was smaller than that of cognitive image.

Cognitive and affective image of South Korea and perceived risks of vacationing
in South Korea had signiﬁcant direct inﬂuences on respondents’ intentions to visit South
Korea (support for H6, H7, and H8). This indicated that respondents who had a more
positive cognitive image of the country and lower perceived risks of vacationing there
appeared to have greater intention to visit South Korea. Among the three factors
inﬂuencing visit intention to South Korea, perceived risks had the strongest direct
inﬂuence on visit intention (B = -.33, p < .01); cognitive image had the second strongest
direct inﬂuence (B = .28, p < .01). The direct inﬂuence of affective image on intention to

visit (B = .18, p < .01) was smaller than that of cognitive image and perceived risks.

88

 

 

Multiple squared correlations. As shown in Figure 8, 45 percent of the variance
in intention to visit South Korea was explained by the direct inﬂuences of cognitive
image, affective image, and perceived risks. Information source use explained 11 percent
of variance in cognitive image of South Korea, but only 5 percent in variance for
perceived risks. The large portion of the unexplained variance in cognitive image and
perceived risks suggests the possibility of including other factors which may inﬂuence
cognitive image and perceived risks in future studies. Fifty-one percent of variance in

affective image was explained by cognitive image and perceived risks.

Indirect and total inﬂuences of the factors on visit intention. The standardized
indirect inﬂuences of the factors on Americans’ intention to visit South Korea are

presented in Table 30.

Table 30. Standardized estimates for direct, indirect, and total inﬂuences for South Korea.

 

Standardized Standardized Standardized

 

 

 

 

 

 

 

 

 

Exogenous Endogenous Direct Indirect Total
vanables varlables . . .
Inﬂuence Inﬂuence Inﬂuence
Affective Image .01 .24 .25
Information Cognitive Image .33 --- .33
Source Perceived risks -.22 --- -.22
Usage

Intention to visit --- .21 .21
Affective Image Intentiqp_!.9_Xi.§i__t.__~_.._-_-_wW ,V -13 .18
Cognitive Image Affective Image .54 --- .54
Intentlon to VISIt 28W. .09 .37
Perceived risks Affective Image -.28 --- -.28
Intention to visit -.33 -.O6 -.39

 

Information source usage had a positive indirect inﬂuence on intention to visit
South Korea through perceived risks and cognitive image (B = 0.21). In other words,

information source usage on South Korea had a positive direct inﬂuence on cognitive

89

image, which, in turn, had a positive direct inﬂuence on visit intention. Cognitive image
of South Korea and perceived risks of vacationing there also had indirect inﬂuences on
Visit intention through affective image. The indirect inﬂuence of cognitive image and
perceived risks on intention to visit South Korea was small (B = .09 for cognitive image
and B = - .06 for perceived risks).

Examination of the standardized total inﬂuence combining both direct and indirect
standardized inﬂuences, as presented in Table 30, showed that perceived risks had the
largest standardized total inﬂuence on intention to Visit South Korea (B = -.39); cognitive
image had the second largest inﬂuence (B = .3 7). The total inﬂuence of affective image (B

= .18) was much smaller than that of cognitive image and perceived risks.
Comparison of the results of hypotheses testing for the three East Asian

countries. Overall, seven of eight hypotheses were supported in the three models. These

results are summarized in Table 31.

90

Table 31. Results of hypotheses testing across three East Asian countries.

 

Std. Path Coefﬁcients

i
Paths China Japan South Support

 

 

Korea
Information Source Usage 9 33,” 38** 33" X
Cognitive Image (H1) ' ° '
Information Source Usage 9
Affective Image (H2) '04 '06 '01 Y
Information Source Usage 9 _ H _ H _ H
Perceived Risks (H3) '26 '29 '22 X
Cognitive Image 9 Affective Image (H4) .56” .57“ .54" X
Perceived Risks 9 Affective Image (H5) -.28** -.22"‘* -.28** X
Cognitive Image 9 Intention to Visit (H6) .19" .34" .28" X
Affective Image 9 Intention to Visit (H7) .41" .28” .18” X
Perceived Risks 9 Intention to Visit (H8) -.26** -.20** -.33** X

 

Note: signiﬁcant at ** p < .01, a X = Supported hypothesis, Y = Did not support
hypothesis

As shown in Table 31, the ﬁnding that information usage had an insigniﬁcant
direct inﬂuence on affective image was the same across the three countries. All other
hypotheses were supported for China, Japan, and South Korea. This similarity indicates
that the same model can be applied to all three countries in order to explain Americans’
intentions to visit them. Similarly, across China, Japan, and South Korea, cognitive image
had a very strong direct inﬂuence on affective image, with a standardized direct inﬂuence
greater than .50. This ﬁnding indicates that cognitive image was the strongest factor
inﬂuencing affective image. A third ﬁnding across all three countries is that the direct
inﬂuence of information sources on cognitive image was greater than this factor’s
inﬂuence on perceived risks, a ﬁnding which indicates that increasing information usage
played a more important role in improving respondents’ cognitive image of these
countries than did decreasing perceived risks of vacationing there.

As well as similarities, there were differences in the results of hypotheses testing

across the three models. One such difference is the factor exhibiting the greatest direct

91

inﬂuence on intention to visit. For China, this factor was affective image; for Japan,
cognitive image; for South Korea, perceived risks. Interestingly, perceived risks had the
least direct inﬂuence on intention to Visit for China and Japan, while affective image had
the least direct inﬂuence on intention to visit for South Korea, ﬁndings which suggest
South Korea’s present inability to instill prospective American tourists with knowledge
and good feeling about South Korea, in comparison to China and Japan. It remains to be
explained why cognitive image of Japan exerted more inﬂuence on Americans’ intentions
to visit this country, while affective image did so for China.

Another difference in the results of the hypotheses testing was the strength of the
total inﬂuence combining both the direct and indirect inﬂuences on intention to visit, as
displayed in Table 32. In both the China and Japan models, cognitive image had the
greatest total inﬂuence on intention to visit. In the South Korea model, perceived risks
had the greatest total inﬂuence on intention to visit South Korea. These ﬁndings suggest
that the information prospective American tourists obtained on China and Japan produced
in them more knowledge and good feeling about these countries than did the information
obtained on South Korea perhaps because prospective American tourists obtained

information on China and Japan more frequently than they did on South Korea.

Table 32. Comparison of total inﬂuence.

 

 

 

Exogenous variables Endogenous Standardized total inﬂuence
variables China Japan South Korea
Informatlon Source Intention to visit .25 .28 .21
Usage
Affective Image Intention to visit .41 .28 .18
Cognitive Image Intention to visit .42 .50 .37
Perceived Risks Intention to visit -.37 -.26 -.39

 

92

Additional Analysis:
The Comparative Inﬂuence of Different Types of Information Source on
Destination Image and Perceived Risks

As shown in the results of the above SEM analyses, cognitive image, affective
image, and perceived risks were all important factors in promoting American vacationers’
intention to visit China, Japan, and South Korea, even though the strength of their
inﬂuence varied across the three countries. Given the importance of these three factors in
inﬂuencing Visit intention, the next task was to determine which type of information
source most directly inﬂuenced cognitive image and perceived risks and, through these
two factors, intention to visit. Affective image was excluded from this assessment, given

that information source usage had been shown to have no direct inﬂuence on affective

image. These correlations are reported in Table 33.

Table 33. Correlation between types of information source used and cognitive image and
perceived risk.

 

 

 

Country Infol Inf02 Info3 Info4 Inf05 lnfo6 Info7
Cogn- China .12M .15** .2011tb .19" .28W .18" .01
itive Japan .14" .21M .23“b .18" 29*“ .20" .02
Image 3. Kor. .15" .21M b .19" .16" .27W 21"" .11*
Perce- China -.04 -.11** -.13** -.15**" -.17**“ -.10** -12"
ived Japan -.12** -.17** -.19="=tb -.20**° -.18** -.l6" -.04
Risks s. Kor. -.1 l” _21" a -.12** -.16** -.20Mb -.11** -.12**

 

Signiﬁcant at * p< 0.05, ** < 0.01
Note: a Most correlated; bSecond most correlated. .
Infol=TV/radio, Inf02=Word of mouth, Info3=Books and movies, Info4=lntemet,
Inf05=Cu1turaI experiences (in art museums, restaurants, concert halls, courses),
Info6=Newspapers/magazines, Info7=Professional advice

As shown in Table 33, cultural experiences in, for example, art museums,

restaurants, concert halls, and courses had the highest correlation with respondents’

cognitive image of all three countries. The type of information source having the second

93

highest correlation with cognitive image was books and movies for China and Japan, and
word of mouth and newspapers/magazines for South Korea.

However, the type of information source having the highest correlation with
perceived risks varied across the three countries. Cultural experience had the highest
correlation with perceived risks for China, followed by the Internet. The Internet had the
highest correlation with perceived risks for Japan, followed by cultural experiences and
books and movies. Finally, word of mouth had the highest correlation with perceived risks
for South Korea, followed by the cultural experiences.

These ﬁndings suggest that, in general, Americans’ perceptions of the risks of

vacationing in China, Japan, or South Korea depend most heavily on the type of

 

information source from which they have obtained the most information on each of these
three countries. If this primary information source produces knowledge and good feeling
about China, Japan, and South Korea, it would seem that this information source lowers
prospective American tourists’ sense of potential travel-related risk in these East Asian
countries. The results of these correlation analyses will be further discussed in terms of

their implications for both tourism researchers and tourism marketers in Chapter Five.

 

"Sun .1”

94

CHAPTER 5

DISCUSSION, IMPLICATIONS, AND CONCLUSIONS

The present dissertation study examined the inﬂuences of destination image,
perceived risks, and information source usage on Americans’ intentions to visit China,
Japan, and South Korea in order to identify which factors most inﬂuenced Americans’
intentions to visit these East Asian countries. Chapter Five has four sections. In the ﬁrst
section, a summary and discussion of the ﬁndings are provided. In the second section,
both theoretical and practical implications of the ﬁndings are discussed. In the third
section, limitations of this research and suggestions for future research are presented. The

ﬁnal section provides conclusions.

Summary and Discussion of the Findings

The purpose of this study was to develop and empirically test a model to explain
Americans’ intentions to visit three East Asian countries (China, Japan, and South Korea).
It was proposed that Americans’ visit intentions could be explained by the interrelations
among cognitive and affective destination image of these countries in general, perceived
risks of vacationing in these countries, and information source usage regarding these
countries. Before testing the proposed model against the data collected, it was necessary
to determine if statistically signiﬁcant differences existed among the data on China, Japan,
and South Korea. If so, then the proposed model would have to be tested separately
against data for each country. If not, data could be pooled for testing. Statistical analyses

of the mean differences across the variables for each of the three countries indicated

95

signiﬁcant differences. Thus, data on China, Japan, and South Korea were separated in
order to test the proposed conceptual model.

This section consists of three subsections. In the ﬁrst subsection, ﬁndings about
mean differences across the variables for China, Japan, and South Korea are summarized.
In the second subsection, results of hypotheses testing are summarized and discussed. In
the ﬁnal subsection, ﬁndings about relationships between information source usage and

destination image and perceived risks are summarized and discussed.

Mean Differences across the Variables for Three East Asian Countries
Signiﬁcant mean differences were found across the variables (destination image,
perceived risks, information source usage, and visit intention) for China, Japan, and
South Korea. First, signiﬁcant mean differences in cognitive image were found across
the three countries. A comparison revealed that study participants’ had a more positive
cognitive image of China and Japan than they did of South Korea. South Korea was
ranked last for all cognitive image variables. Japan was ranked ﬁrst as offering many
things to see and do, easy to navigate, quality accommodation, friendly local people, and
unique cultural resources, while China was regarded as the best East Asian country in
terms of natural scenic beauty, appealing local food and drink, and interesting historical
attractions. While mean differences in cognitive image were observed across the three
countries, the strengths of each were similar. Japan and China had the highest mean
scores for many things to do and see, and interesting historical attractions, and South
Korea had the highest mean scores for interesting historical attractions and unique
cultural resources, ﬁndings which suggest that these East Asian countries cultural and

historical resources might distinguish them from other destinations.

96

Second, signiﬁcant differences in affective image were found across China, Japan,
and South Korea. The mean scores for affective image items for Japan were the highest,
while the mean scores for affective image items for South Korea were the lowest.
However, when compared to the mean scores for cognitive image items, the mean scores
for affective image items were found to be smaller, suggesting that not all knowledge
about a place (cognitive destination image) produces feeling about a place (affective
destination image).

Third, signiﬁcant differences were observed in six of the seven perceived risk
items. The mean differences for language barrier across the three countries were
insigniﬁcant; however, regarding all three countries, this perceived risk item received the
highest mean score. This indicates that respondents perceived that they would have
problems in communicating with people in the three East Asian countries. The mean
scores for terrorism and violence were highest for South Korea, while the mean scores for
likelihood of illness, political instability, and potential health problems were highest for
China. The mean score for natural disaster was highest for Japan, while the mean scores
of other perceived risk items for Japan were lower, or equal to, those for China and South
Korea. Overall, South Korea and China were perceived as being riskier countries than
Japan.

Fourth, signiﬁcant differences across China, Japan, and South Korea were found
regarding information source usage. The mean scores for all types of information source
usage were lowest for South Korea. The mean scores for TV/radio, word of mouth, the
Internet, and professional advice were highest for Japan, while the mean scores for
books/movies, and cultural experiences were highest for China. This indicates that

respondents were more unfamiliar with South Korea than with China and Japan. While

97

there were mean differences in information source usage across the three countries, the
mean scores for all three countries were not high. This ﬁnding indicates a lack of
information on all three countries, possibly due to a limited search for such information.
Finally, statistically signiﬁcant differences in visit intention across China, Japan,
and South Korea were found. Japan had the highest scores for intention to visit, followed
by China and South Korea. This result makes sense when considering the mean scores
for cognitive and affective destination image, perceived risks, and information source
usage. It can be inferred that Japan was the ﬁrst-choice destination of study participants,
since Japan had the highest mean scores for most destination image variables, including
all affective image variables, and the lowest mean scores for most perceived risk
variables. Similarly, it can be assumed that China was their second-choice destination,
since its positive destination image variables were somewhat offset by its perceived risk
variables. South Korea, it seems, was the last-choice destination, since South Korea
received the lowest mean scores for all destination image variables and the highest mean

scores for most perceived risk variables.

98

Results of Hypotheses Testing
As presented in Figure 8, the conceptual model this study proposed to explain

Americans’ intentions to visit China, Japan, and South Korea included eight hypotheses.

 

Cognitive

Image

     

 

 

 
   

H4 H6
Information Intention
Source - Affective H7 W To.
Usage Image VISIt
H3 H 5 H8

Perceived

Risks

 

 

 

 

 

Figure 9. Proposed conceptual model

The results of Structural Equation Modeling analyses showed similarities and
differences across participants’ responses regarding China, Japan, and South Korea. For
all three countries, seven of the eight hypotheses were supported. For these seven
hypotheses, the sign/direction of the relationships was found to be consistent across all
three countries; however, the magnitude of these relationships differed. Results of

hypotheses testing are summarized below.

99

 

Hypotheses 1 & 2: Increased usage of information sources increases
Americans ’ (emotionally neutral) cognitive image of the destination in general, and
increased usage of information sources increases Americans ’ (emotionally positive)
aﬂective image of the destination in general. Across all three countries, the hypothesis
regarding the direct inﬂuence of information source usage on cognitive image was
supported (H1); however, the hypothesis regarding the direct inﬂuence of information
source usage on affective image was not supported (H2). These ﬁndings provide 5
empirical evidence in support of previous studies that noted the inﬂuence of information

source usage on cognitive destination image (Baloglu, 1999; Gartner, 1993). These

 

ﬁndings also indicate that respondents having greater access to information on each of the Ii:
three countries had a higher cognitive image of them. The strength of the direct inﬂuence

of information source usage on cognitive image was similar across the three countries (B

= .33 for China, B = .38 for Japan, and B = .33 for South Korea), a ﬁnding consistent with

Baloglu’s (2001) study, in which he found that the more varied the information sources

consulted on a destination, the higher the cognitive image of that destination.

Hypothesis 3: Increased usage of information sources decreases Americans ’
perceived risks of vacationing in the destination. Hypothesis 3 was supported for all
three countries. Information source usage had a signiﬁcant direct inﬂuence on perceived
risks. This ﬁnding indicates that respondents who frequently accessed various types of
information sources had lower perceived risks of vacationing in China, Japan, and South
Korea. The strength of the direct inﬂuence of information source usage on perceived risks
was similar across the three countries (B = -.26 for China, B =.-.29 for Japan, and B = -.22

for South Korea). The direct inﬂuence of information source usage on perceived risks was

100

slightly weaker when compared to that of cognitive image. This ﬁnding also provides
empirical evidence in support of research showing that the more information travelers
have on a destination, the lower their perceived risks of vacationing in that destination
(Fodness & Murray, 1999; Gemunden, 1985; Maser & Weiermair, 1998; Sonmez &

Graefe, 1998b).

Hypotheses 4 & 5: Increased (emotionally neutral) cognitive destination image

increases Americans ’ (emotionally positive) affective destination image, and increased
perceived risks of vacationing in the destination decreases Americans’ (emotionally
positive) affective image of the destination in general. Hypotheses 4 and 5 were
supported across China, Japan, and South Korea. While cognitive image had a signiﬁcant
and positive direct inﬂuence on affective image for each country, perceived risks had a
signiﬁcant and negative direct inﬂuence on affective image for each country. This
ﬁnding indicates that respondents who had a higher cognitive image of China, Japan, and
South Korea and lower perceived risks of vacationing in these countries exhibited more
positive affective images. The strength of the direct inﬂuences of cognitive image on
affective image was similar across the three countries. Cognitive image had a very strong
direct inﬂuence on affective image, with a standardized direct inﬂuence of greater than .5
across China, Japan, and South Korea (B = .56 for China, B = .57 for Japan, and B = .54
for South Korea). The strength of the direct inﬂuences of perceived risks on affective
image was also similar across the three countries. Perceived risks moderately inﬂuenced
affective image across the three countries (B = -.28 for China, B =.-.22 for Japan, and B = -
.28 for South Korea). This ﬁnding regarding the signiﬁcant inﬂuence of cognitive image

on affective image is consistent with Baloglu and McCleary’s (1999a) study, which

101

 

 

asserted that travelers’ intention to visit a destination was indirectly inﬂuenced by
cognitive image of the destination. However, the present study is perhaps the ﬁrst to test
and conﬁrm the widely held assumption that Americans’ perceived risks of vacationing in

a destination inﬂuences their affective image of that destination.

Hypotheses 6 & 7: Increased (emotionally neutral) cognitive destination image
increases Americans ’ intention to visit the destination, and increased (emotionally
positive) affective destination image increases Americans ’ intention to visit the
destination. Hypotheses 6 and 7 were supported across all three countries. Both cognitive
and affective image had signiﬁcant and positive direct inﬂuences on Americans’
intentions to visit each country. The higher Americans’ cognitive and affective images of
a destination, the greater their intention to visit the destination. This ﬁnding suggests that
both destination image components contributed to travelers’ intentions to visit China,
Japan, and South Korea, and thus conﬁrms the ﬁndings of previous tourism studies
(Baloglu & McCleary, 1999a; Baloglu, 2000; Pizam, 1991; Vogt & Andereck, 2003).

While the sign/direction of the inﬂuence between the two destination image
components and Americans’ intentions to visit was consistent across the three countries,
the strength of the relationship differed across China, Japan, and South Korea. Affective
image had the strongest direct inﬂuence on Americans’ intention to visit China (B = .41
for affective image and B = .19 for cognitive image), while cognitive image had the
strongest direct inﬂuence on Americans’ intention to visit Japan (B = .28 for affective
image and B = .34 for cognitive image). The direct inﬂuence of destination image on
Americans’ visit intentions was less strong for South Korea (B = .18 for affective image,

and B = .28 for cognitive image) than it was for China and Japan.

102

Only one study has compared the inﬂuences of both affective and cognitive image
on travelers’ intentions to visit a destination (four Mediterranean countries) (Baloglu,
I999). Baloglu (1999) found that, before the ﬁrst visit, cognitive image was more
inﬂuential on visit intention than was affective image. Findings for the Japan and South
Korea models support Baloglu’s study; however, the ﬁndings from the China model do
not, suggesting that the relative inﬂuences of affective and cognitive image may differ

depending on the destination.

Hypothesis 8: Increased perceived risks of vacationing in a destination
decreases American ’3 intention to visit the destination. Hypothesis 8 was supported for
all three countries. Perceived risks were found to have signiﬁcant and negative direct
inﬂuences on Americans’ intentions to Visit each country. This ﬁnding indicates that
respondents who perceived higher risks of vacationing in the three countries had
diminished intentions to visit them. This result supports the ﬁndings of previous tourism
studies (Bramwell & Rawding, 1996; Coshall, 2003; Dann, 1996; Sonmez & Graefe,
1998a, 1998b). However, the strength of perceived risks’ direct inﬂuence on intention to
visit differed across the three countries. It was stronger for South Korea (B = -.33) than it
was for China and Japan (B = -.26 for China, and B = -.20 for Japan). In comparison to the
direct inﬂuence exerted by destination image on visit intention, perceived risks had a
stronger direct inﬂuence for South Korea, but destination image (either cognitive or

affective) had a stronger direct inﬂuence for China and Japan.

103

Overall Pattern of Relationships among the Factors across the Three Countries

The ﬁndings derived from the structural equation model reveal that information
source usage, cognitive and affective destination image, and perceived risks signiﬁcantly
inﬂuenced American tourists’ intentions to visit China, Japan, and South Korea in both
direct and indirect ways: (1) information source usage inﬂuenced visit intention indirectly
through cognitive image and perceived risks; (2) cognitive image and perceived risks
inﬂuenced visit intention both directly and indirectly through affective image; and (3)
affective image inﬂuenced visit intention directly. These results were found to be
consistent for all three East Asian countries.

However, the total inﬂuence, which combines both the direct and the indirect
inﬂuences of each factor on visit intention, varied across the three countries. The factors
which had the strongest total inﬂuence were affective and cognitive destination image for
China, cognitive destination image for Japan, and perceived risks for South Korea. This
result makes sense in terms of the social and political instability in South Korea at the
time when respondents were surveyed. (Between May 20 and June 20, 2006, the US.
regarded North Korea’s attainment of nuclear capability as a threat to national security:
President George W. Bush named North Korea one of three “axes of evil.” Confusion of
North Korea with South Korea possibly had a negative effect on study participants’
responses.) American respondents had a less developed cognitive image of South Korea,
as shown in the previous section, which seems to have allowed perceived risks to
dominate their affective image of the country, and, in turn, their visit intentions. This was
not the case for China and Japan, countries about which Americans had more varied

information and thus more positive feelings.

104

 

Relationships between Information Source Usage and Destination Image and
Perceived Risks

As shown above, cognitive image, affective image, and perceived risks were all
important factors in promoting Americans’ intentions to Visit China, Japan, and South
Korea, though the strength of inﬂuence for each of these factors varied across the three
countries. Thus, it was important to examine which types of information source most
inﬂuenced these three factors. In this study, information sources were found to have
signiﬁcant direct inﬂuences on cognitive image and perceived risks, and, through them,
an indirect inﬂuence on affective image. In other words, information sources had direct
inﬂuences on both cognitive image and perceived risks, while they had indirect inﬂuence
on affective image through cognitive image and perceived risks. Thus, it was important
to identify what types of information source used were more inﬂuential in improving
Americans’ cognitive images and in reducing their perceived risks.

It was found that cultural experiences (in art museums, restaurants, concert halls,
and courses) had the highest correlation with cognitive image across the three countries.
Books and movies were the most inﬂuential information source on cognitive image for
China and Japan, and word of mouth and newspapers/magazines were the most
inﬂuential information sources on cognitive image for South Korea. These same
information sources inﬂuenced perceived risks for the three countries: for China, cultural
experiences, followed by the Internet, was the most inﬂuential; for Japan, the Internet,
followed by cultural experiences and books and movies, was the most inﬂuential; and,
for South Korea, word of mouth, followed by cultural experiences, was the most

inﬂuential.

105

Implications
The ﬁndings of this study have both theoretical and practical implications. This
section presents, ﬁrst, the theoretical contributions of this study to existing tourism

literature, then its practical implications for East-Asia-based tourism marketers.

Theoretical Implications

From a theoretical perspective, this study contributes to current understanding of
American tourists’ decision-making processes regarding China, Japan, and South Korea.
One important implication of this study is the explanatory model it provides. Speciﬁcally,
this model clariﬁes the mutual inﬂuence of general cognitive image of a destination and
perceived risks of vacationing there on affective image. Both of these factors are directly
inﬂuenced by information source usage, particularly cultural experiences, and all three of
these factors directly inﬂuence American tourists’ intentions to Visit China, Japan, and
South Korea for the ﬁrst time. Thus, this study provides a model for tourism researchers
that may be tested in relation to other tourist destinations.

A second theoretical implication of this study emerges from its operationalization
of the destination image construct. Recent studies have suggested that destination image
includes two distinct but interrelated components: cognitive destination image and
affective destination image (Baloglu and McCleary, 1999b; Gartner, 1993; Reilly, 1990).
However, most studies of destination image have focused on cognitive destination image.
The present study operationalized destination image as a construct comprising both
cognitive and affective destination image, and empirically proved the distinctiveness of
these two image components, using conﬁrmatory factor analysis. This study also

demonstrated the interrelatedness of these two image components, speciﬁcally showing

106

 

 

that cognitive image can either create affective image (as its antecedent) or modify
existing affective image. Thus, the ﬁndings of this study suggest that, instead of using
either cognitive or affective image as measures of destination image, tourism researchers
should use both image components. This two-fold analysis reveals the complexity of
destination image to tourism researchers.

A third theoretical implication of this study is its contribution to existing literature
on destination image formation. Several destination image studies have emphasized the
importance of information sources in a prospective tourist’s development of a destination
image (Baloglu, 1999; Beerli & Martin, 2004a; Fakeye & Crompton, 1991; Gunn, 1972;
Hanlan & Kelly, 2005). For example, Baloglu (1999, 2001) found that the variety of the
information sources consulted contributed positively to travelers’ cognitive image of a
destination. Baloglu also demonstrated that different types of information sources have
varying degrees of effect on tourists’ cognitive images of four Mediterranean countries.
The present study conﬁrms Baloglu’s ﬁndings, demonstrating that a variety of
information source usage signiﬁcantly inﬂuences both prospective American tourists’
cognitive destination image of China, Japan, and South Korea and their perceived risks of
vacationing in these countries. This study also extends Baloglu’s ﬁndings in
demonstrating that different types of information sources are more inﬂuential depending
on the destination. For example, books and movies were more highly related to perceived
risks of vacationing in Japan than they were for China and South Korea.

A fourth theoretical implication of this study is that the importance of factors
directly inﬂuencing visit intention (i.e., cognitive image, affective image, and perceived
risks) varied from destination to destination. For China, affective image was the most

inﬂuential; for Japan, cognitive image; and, for South Korea, perceived risks. Future

107

 

‘II

 

research regarding other tourism destinations might also conclude that limited
information on a destination allows perceived risks to dominate affective image and thus

to inﬂuence American tourists’ intentions to visit the destination most heavily.

Practical Implications

Additionally, the present study provides important information for East-Asia-
based tourism marketers. One important practical implication of this study emerges from
comparing prospective American tourists’ responses to South Korea to their responses to
China and Japan. Because study participants lacked information on South Korea, their
perceived risks of vacationing there dominated their affective image of the country in
general and ultimately weakened their intention to visit South Korea. The opposite was
true for China and Japan. Based on this ﬁnding, South-Korea-based tourism marketers
should focus on increasing Americans’ cognitive image of their country, particularly by
promoting cultural experiences, an information source which was found to have a positive
inﬂuence on cognitive image for all three countries, but more so for China and Japan.
South Korean tourism marketers should invest in museum exhibits, concerts, study
abroad programs, and South Korean cuisine festivals in the U. S. Chinese and Japanese
tourism marketers should increase their efforts in this domain as well to attract even more
American tourists.

A second practical implication of this study emerges from comparison of the three
countries. It would seem that increased investment in U.S.-based cultural experiences
promoting China, Japan, and South Korea might improve prospective American tourists’
cognitive images of these three countries. Not only should East Asian tourism marketers

increase their promotional efforts in the US, but they should also place special emphasis

108

 

 

on distinguishing their country’s attractions from those of other East Asian countries.
(Regarding South Korea, respondents perceived South Korea as having interesting
historical attractions and unique cultural resources: two strengths of this country.
However, China and Japan were also perceived as having these strengths.) Thus, it would
seem that East-Asia-based tourism marketers must also focus on presenting their
destinations as being unique.

One promotional strategy would be to provide prospective American tourists with
more detailed information about China, Japan, and South Korea, helping them to
recognize both that each country has cultural and historical resources different from the
other two, and that each country’s culture and history are closely related to those of the
other two, thereby warranting joint trips to all three destinations. For example, American
tourists could visit different Buddhist temples in all three countries, admiring their
variation.

Cultural experiences orchestrated in the US. might also serve this purpose. For
example, a traveling museum exhibit on the tea ceremony in East Asia might be jointly
funded by tourism marketers in China, Japan, and South Korea. Americans could, then,
sample black tea from China, green tea from Japan, and ginseng tea from South Korea.
For the sports-oriented crowd, a similar traveling exhibit, this time targeting sports arenas,
could showcase similarities and differences in East Asian martial arts traditions.
Americans could watch shows of kungfu from China, karate from Japan, and tae kwon
do from South Korea. Although cultural experiences would be created by these traveling
exhibits, so would word of mouth, an information source on which South Korea relies.

To promote books and movies, East-Asia-based tourism marketers could sponsor

public readings and book signings from Chinese, Japanese, and South Korean authors

109

who treat similar themes in their work; or a ﬁlm festival in which movies from China,
Japan, and South Korea would compete for a prestigious award. Competitive, “Iron
Chef”-style food tastings at which dishes would be prepared by celebrity chefs from
China, Japan, and South Korea might also be arranged to promote cookbooks or cooking-
related videos.

Increasing prospective American tourists’ cognitive images of these three East
Asian countries is crucial in that, as this study demonstrates, lack of information results in
the overshadowing of positive affective image by perceived risks, an effect which
weakens visit intentions. Tourism marketers in China, Japan, and South Korea have to

work particularly hard to counterbalance market-damaging phenomena, such as SARS,

 

bird ﬂu, natural disasters, and terrorism. Sonmez and Graefe (1998a, 1998b) found that
perceived risks was a critical predictor of American tourists’ avoiding a particular
destination. This ﬁnding, combined with the results of the present dissertation study,
suggest that East-Asia-based tourism marketers should design promotional tools that
address each of the perceived risks identiﬁed in this study, as well as positive cultural
experiences.

Finally, this study offers East-Asia-based tourism marketers information on how

to invest more effectively in marketing programs that target prospective American

 

tourists. For example, tourism marketers based in China, Japan, and South Korea should
invest in programs that increase cognitive image and decrease perceived risks; however,
such investments would especially beneﬁt the South Korean tourism industry. As the
Internet was a key source of information concerning perceived risks for the respondents
in this study, East-Asia-based tourism marketers should also consider designing Web sites

that highlight the safety and the ease of navigation offered by these three countries. All

110

marketing strategies should emphasize the “English-friendliness” of China, Japan, and
South Korea by presenting cultural events in English and by providing bilingual resources
on promotional Web sites, including on-site translation services. Although tourism
marketers in China, Japan, and South Korea face similar challenges, they should not

forget to highlight for prospective American tourists their countries’ unique attractions.

Limitations of this Study and Suggestions for Future Research

One limitation of this study is associated with the sampling frame; only residents
of Michigan and Illinois were selected for inclusion. If residents of other states had
participated in the study, responses might have differed. Thus, the generalizablity of the
ﬁndings is limited to similar populations. Future research would need to sample a wider
diversity of Americans in order to conﬁrm or deny claims made in this study about
Americans’ perceptions of East Asian countries.

A second limitation of this study relates to the information source usage items in
the survey. Although these items measured the frequency with which each type of
information source was consulted, it did not assess the exact content of each information
source. (The scale used in the present dissertation was relative, not absolute.) Thus, future
research would need to capture the content of each information source in order to advance
understanding of how information sources inﬂuence destination image and perceived
risks regarding China, Japan, and South Korea.

A third limitation concerns the scope of this study, which sampled only
Americans who had not previously visited China, Japan, and South Korea. As discussed
in the literature review, visitors might have responded differently regarding destination

image, perceived risks, and intention to visit. If a sample of Visitors were surveyed, an

111

 

 

altered model would likely be needed to explain the relationships assessed in this study.
However, this possibility remains to be investigated in future research.

A fourth limitation of this study also concerns scope: Other factors remain to be
incorporated into the model. As indicated in previous research, socio-demographics, such
as age, educational background, and gender, affect prospective tourists’ intentions to Visit
a destination, due to their potential interactions with destination image (Baloglu and
McCleary, I999b; Fakeye and Crompton, 1991; Goodrich, 1978; Gunn, 1989; Javalgi,
Thomas, and Rao, 1992). Additionally, previous research has shown that travelers may
have different motives for international tourism, and these motives inﬂuence their visit
intention; for example, travelers who seek either familiarity or novelty, associate different
levels of risk with intentional tourism (Lepp & Gibson, 2003). Given the as yet
unexplained variances in Americans’ intentions to visit the three East Asian countries,
especially regarding South Korea, future research would need to consider adding other
potential factors to the conceptual model developed in this study, such as perceived
behavioral control, or the perceived expenditure of time and money a vacation to East

Asia would require.

Conclusions
The ﬁndings from this study demonstrate that American tourists’ intentions to
visit China, Japan, and South Korea are inﬂuenced by the interrelationships among
information source usage, cognitive and affective destination image, and perceived risks.
They also show that, while all of these factors have signiﬁcant inﬂuences on Americans’

intentions to vacation in all three countries, different factors have a more direct inﬂuence

112

 

on visit intention for each country (i.e., affective destination image for China, cognitive
destination image for Japan, and perceived risks for South Korea).

The results of this study have both theoretical and practical value in that they ﬁll
gaps in previous tourism research on visit intention, destination image, perceived risks,
and information source usage, and they address an understudied American tourist region
(East Asia). This study also provides important information for East-Asia-based tourism
marketers, information suggesting how these practitioners might invest more effectively
in marketing programs to attract ﬁrst-time American tourists to their countries.

Future research, based on this study, should (I) extend this model by
incorporating other possible factors that may inﬂuence prospective American tourists’
intentions to Visit China, Japan, and South Korea, (2) replicate this study with a
probabilistic survey sample, and (3) test this model on other potential destinations for

American tourists.

113

 

APPENDIX

 

 

114

Appendix A: Survey Questions

Americans’ Reasons for Vacationing in East Asia

Dear Sir/ Madam:

I am a doctoral student in Michigan State University’s tourism program. The purpose
of my study is to develop and test a model explaining Americans’ reasons for vacationing
in East Asia. My related survey will provide the empirical basis of my dissertation.

Your participation in this survey will take approximately 15 minutes. This study is for
research purposes only. Your responses will not be associated with you in any way and
will remain strictly conﬁdential. Your identity will not be linked to the data you provide.
There are no anticipated risks associated with participation. You consent to voluntarily
participate in this study by completing this survey and you may choose not to participate

at all or you may refuse to answer certain questions.

Please direct any questions about this study to J eonghee Noh at nohieong@msu.edu.,
phone (517) 333-3662, fax (517) 432-2296.

Jeonghee Noh

PhD. Candidate

Department of Community,

Agriculture, Recreation and Resource

Studies

172 Natural Resources Building

Michigan State University
East Lansing, MI 48824-1222

MICHIGAN STATE

 

UNIVERSITY

1.Have you ever visited the countries listed below? If yes, how many times?
Please check the appropriate box(es) for each of the countries listed below.

 

 

 

 

 

 

 

o o 6)
Country Havilyéou vrsrted ........... Y. 6.8 ..... Number of visits
CHINA D ’? timet's)
. JAPAN . 1:1 [I ___—time(s)
SOUTH KOREA 1:1

 

 

 

115

 

., .A_,ti.r_u9(§)

 

 

 

2. Please rate your image of China, Japan, and South Korea according to the following
features. Please circle only one number for each country.

1 2 3 4 5
Strongly Somewhat Neither Agree Nor Somewhat Strongly

 

3. Please express your feelings about each country as a vacation destination.
Please circle only one number for each set of opposite adjectives to indicate how you feel.

I) CHINA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Extremely Svtrﬂet- Neither 83%? Extremely
Unpleasant 1 2 3 4 5 Pleasant
Depressing l 2 3 4 5 Em
Boring 1 2 3 4 5 Fascinating
.msgessing 1 2 3 4 5 ML...
2 JAPAN S S
Extremely walling- Neither gﬂet- Extremely
Unpleasant 1 2 3 4 5 Pleasant _ .
Depressing 1 2 3 4 5 Energizing
Boring 1 2 3 4 5 Fascinating
.DigLressing 1 2 3 4 5 Relaxmg' .
3)South Korea
Extremely 5‘3]??? Neither S3113;- Extremely
__Uppleasant 1 2 3 4 5 Pleasant
Depressing 1 2 3 4 5 Energizing.
”Boring 1 2 3 4 5 Fascinating
Distressing l 2 3 4 5 Relaxipg

 

116

4. Please read each statement and indicate your level of agreement or disagreement, for
each country using the scale below. Please circle only one number for each country.

1 2 3 4 5
Strongly Somewhat Neither Agree Nor Somewhat Strongly

CHINA JAPAN

1234 234512345
34 2345

234512345

3 4
5 W3”?
3 4

withothers 1 2 3 4 5

 

5. How oﬁen do you obtain information on the following countries from the sources listed
below? Please circle only one response for each statement.

 

 

 

 

 

 

 

 

 

1 2 3 4 5
Never Rarer Sometimes Often Frguently
China Japan South Korea

TV/ Radio 1 2 3 4 5 l 2 3 4 5 1 2 3 4 5
Word of mouth Farnil , Friends,
assassin; , ,4 5 .1 3,. .3__._:__3.3,:_3
Books and Movies 1 2 34 ’5 _l 2 3 4 5 l 2 3 4 5
Professional advice
Altamlsgnalltentssr Broshltres) 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Internet 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5
Cultural experiences (Art Museums, j
Theaters Restaurants, Concert Halls, 1 2 3 4 5 1 2 3 4 5 l 2 3 4 5
WCl;as_s_es, etc.) .
I c r r ’” u ' l 2 3 4 5 1 2 3 4 5 l 2 3 4 5

 

 

 

 

 

6. How interested are you in visiting the following vacation destinations in the next 5
years? Please circle only one number for each country.

 

 

Completely Somewhat Neither Somewha Strongly

Unintereste Unintereste Interested t Interested
CHINA 1 2 3 4 5
SOUTH 1 2 3 4

 

117

 

7. How likely is it that you will visit the following vacation destinations in the next 5
years? Please circle only one number for each country.

 

 

 

 

 

Very Somewhat No Somewhat Ve
UnlikelL Unlikely Opinion Likely Lik'el'y
CHINA 1 2 -3 4 5
1 JAPAN '1 2 ,-. . W. 3 _ .._._.4 5
SOUTH l 2 3 4 5

 

ABOUT YOU:
(These data will be kept in the strictest conﬁdence and used for statistical purpose only).
1. Are you: [3 Female [3 Male

2. What is your age:

 

3. What was the FINAL level of school completed? (please check one)

D Middle/Junior High School
[:1 High School

El College (Undergraduate)

D Master’s Studies

E] Doctoral Studies

a Post-doctoral Studies

4. What was your approximate annual household income before taxes in 2005? (please
check one)

[3 Less than $ 20,000 C] $ 60,000 - $ 79,999 D $ 120,000 - $ 139,999
a $ 20,000 - $ 39,999 E] $ 80,000 - $ 99,999 D $ 140,000 or more
D $ 40,000 - $ 59,999 D $ 100,000 - $ 1 19,999

5. What is the zip code of your primary residence?

 

6. Which one of the following best describes you? (please check one)

:1 African American/ Black [:1 European American/ White

CI Asian American C] Latino/ Latina

1:] Native American! Native Alaskan [:1 Native Hawaiian or Other Paciﬁc Islander
:1 Other (Please Specify)

7. Marital Status (Please check one)

a Married/Committed relationship 13 Divorced
E] Widowed/Widower/Never remarried [:1 Single, never married
CI Separated ,

8. Please write the number of children (aged 18 & under) living in your home:

 

Thank you so much for your gracious effort in completing this questionnaire

118

 

REFERENCES I:

 

"_IL-w-' -' .-

 

119

REFERENCES

Ahmed, Z. U. (1991). The inﬂuence of the components of a state's tourist image on
product positioning strategy. Tourism Management, 12(4), 331-340.

Baloglu, S. (1999). A path analytic model of visitation intention involving information
sources, socio-psychological motivations, and destination image. Journal of
Travel & Tourism Marketing, 8(3), 81-90.

Baloglu, S. (2001). Image variations of Turkey by familiarity index: informational and
experiential dimensions. Tourism Management, 22(2), 127-133.

Baloglu, S., & Brinberg, D. (1997). Affective images of tourism destinations. Journal of
Travel Research, 35(4), 11-15.

Baloglu, S., & McCleary, K. W. (1999a). A model of destination image formation.
Annals of Tourism Research, 26(4), 868-897.

Baloglu, S., & McCleary, K. W. (1999b). U.S. international pleasure travelers' images of
four mediterranean destinations: A comparison of visitors and nonvisitors. Journal
of Travel Research, 38(2), 144-152.

Basala, S. L., & Klenosky, D. B. (2001). Travel-style preferences for visiting a novel
destination: A conjoint investigation across the novelty-familiarity continuum.
Journal of Travel Research, 40(2), 172-182.

Beerli, A., & Martin, J. D. (2004a). Factors inﬂuencing destination image. Annals of
Tourism Research, 31(3), 657-681.

Beerli, A., & Martin, J. D. (2004b). Tourists' characteristics and the perceived image of
tourist destinations: a quantitative analysis - a case study of Lanzarote, Spain.
Tourism Management, 25(5), 623-636.

Bramwell, B., & Rawding, L. (1996). Tourism marketing images of industrial cities.
Annals of Tourism Research, 23(1), 201-221.

Brown, T. J ., Mowen, J. C., Donavan, D. T., & Licata, J. W. (2002). The customer
orientation of service workers: Personality trait effects on self and supervisor
performance ratings. Journal of Marketing Research, 39(1), 110-119.

Burnkrant, R., & Thomas, P. (1982). An examination of the convergent, discriminant, and

predictive validity of Fishbein's behavioral intention model. Journal of Marketing
Research, 19(4), 550-561.

120

 

 

Byme, B. M. (2001). Structural Equation Modeling with AMOS: Basic Concepts,
Applications, and Programming. Mahwah, NJ: Lawrence Erlbaum Associates.

Chen, J. S. (2000). A comparison of information usage between business and leisure
travelers. Journal of Hospitality & Leisure Marketing, 7(2), 65-76.

Choi, W. M., Chan, A., & Wu, J. (1999). A qualitative and quantitative assessment of
Hong Kong's image as a tourist destination. Tourism Management, 20(3), 361-365.

Chon, K.-S. (1991). Tourism destination image modiﬁcation process: Marketing
implications. Tourism Management, 12(1), 68-72.

Chon, K.-S., & Olsen, M. D. (1991). Functional and symbolic congruity approaches to '-
consumer satisfaction/dissatisfaction in tourism. The Journal of the International r

Academy of Hospitality Research, 3, 2-23.

Chon, K. S., Weaver, P. A., & Kim, C. Y. (1991). Marketing your community: Image .-
analysis in Norfolk. Cornell Hotel and Restaurant Administration Quarterly, '. s
31(4), 31.

 

Coshall, J. T. (2003). The threat of terrorism as an intervention in international travel
ﬂows. Journal of Travel Research, 42(1), 4-12.

Court, B., & Lupton, R. A. (1997). Customer portfolio development: Modeling

destination adopters, inactives, and rejecters. Journal of Travel Research, 36(1),
35-43.

Crompton, J. L. (1979). An assessment of the image of Mexico as a vacation destination
and the inﬂuence of geographical location upon that image. Journal of Travel
Research, 1 7(Spring), 18-23.

Crompton, J. L., Fakeye, P. C., & Lue, C.-C. (1992). Positioning: The example of the we
lower Rio Grande Valley in the winter long stay destination market. Journal of
Travel Research, 31(2), 20-26.

 

Dann, G. M. S. (1996). Tourists images of a destination: An alternative analysis. Journal ' ___
of Travel & Tourism Marketing, 5(1/2), 41-55.

Dolnicar, S. (2005). Understanding barriers to leisure travel: Tourist fears as a marketing
basis. Journal of Vacation Marketing, 11(3), 197-208.

Echtner, C. M., & Ritchie, J. R. B. (1993). The Measurement of Destination Image: An
Empirical Assessment. Journal of Travel Research, 3 1 (Spring), 3-13.

F akeye, P. C., & Crompton, J. L. (1991). Image differences between prospective, ﬁrst-

time, and repeat visitors to the lower Rio Grande Valley. Journal of Travel
Research, 30(2), 10-16.

121

Faulkner, B. (2001). Towards a framework for tourism disaster management. Tourism
Management, 22(2), 13 5-147.

Floyd, M. F., Gibson, H., Pennington-Gray, L., & Thapa, B. (2003). The effect of risk
perceptions on intentions to travel in the aftermath of September 11, 2001.
Journal of Travel & Tourism Marketing, 15(2/3), 19-38.

Fodness, D., & Murray, B. (1997). Tourist information search. Annals of Tourism
Research, 24(3), 503-523.

Fodness, D., & Murray, B. (1998). A typology of tourist information search strategies.
Journal of Travel Research, 37(2), 108-119.

Fodness, D., & Murray, B. (1999). A model of tourist information search behavior.
Journal of Travel Research, 3 7(3), 220-230.

Fridgen, J. D. (1987). Use of cognitive maps to determine perceived tourism regions.
Leisure Sciences, 9, 101-117.

Garson, G. D. (2006). Structural Equation Modeling Example Using WinA MOS,
2006.08.01. Retrieved from :
http://www2.chass.ncsu.edu/garson/PA765/semAMOSl .htm

Garter, W. C. (1993). Image formation process. Journal of Travel & Tourism Marketing,
2(2/3), 191-215.

Gartner, W. C., & Hunt, J. D. (1987). An Analysis of State Image Change Over 3 Twelve-
Year Period (1971-1983). Journal of Travel Research, 26(2), 15-19.

Gartner, W. C., & Shen, H. (1992). The impact of Tiananmen Square on China's tourism
image. Journal of Travel Research, 30(4), 47-52.

Gemuden, H. G. (1985). Perceived risk and information search: A systematic meta-

analysis of empirical evidence. International Journal of Research in Marketing, 2,
79 -100.

Goodrich, J. N. (1978). A new approach to image analysis through multidimensional
scaling. Journal of Travel Research, 16(3), 3-7.

Gunn, C. A. (1972). Vacationscape; designing tourist regions. Austin: Bureau of
Business Research, University of Texas at Austin.

Gursoy, D. (2003). Prior product knowledge and its inﬂuence on the traveler's

information search behavior. Journal of Hospitality and Leisure Marketing,
10(3/4), 113-131.

122

Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data
analysis (Fiﬂh ed.). New Jersey: Prentice Hall.

Hanlan, J ., & Kelly, S. (2005). Image formation, information sources and an iconic
Australian tourist destination. Journal of Vacation Marketing, 11(2), 163-177.

Hsu, C. H. C., Wolfe, K., & Kang, S. K. (2004). Image assessment for a destination with
limited comparative advantages. Tourism Management, 25(1), 121-126.

Hunt, J. D. (1975). Image as a factor in tourism development. Journal of Travel Research,
13, 1-7.

Javalgi, R. G., Thomas, E. G., & Rao, S. R. (1992). US pleasure travelers' perceptions of
selected European destinations. European Journal of Marketing, 26(7), 45-64.

Joppe, M., Martin, D. W., & Waalen, J. (2001). Toronto's image as a destination: A
comparative importance-satisfaction analysis by origin of visitor. Journal of
Travel Research, 39(3), 252-260.

 

Joreskog, K. G., & Sorbom, D. (1993). LIRSEL 8: Structural equation modeling with the
SIMPLIS command language. Chicago: Scientiﬁc Software International.

Kim, S. S., & Morrsion, A. M. (2005). Change of images of South Korea among foreign
tourists aﬁer the 2002 F IFA World Cup. Tourism Management, 26(2), 233-247.

Kline, R. B. (1998). Principles and Practice of Structural Equation Modeling. New York:
Guilford.

Korea Tourism Organization (KTO) (2004). Annual Report 2003, Seoul, KTO.

Lam, T., & Hsu, C. H. C. (2006). Predicting behavioral intention of choosing a travel
destination. Tourism Management, 27(4), 589-599. h

Lang, T. C. (2000). The effect of the Internet on travel consumer purchasing behavior
implications for travel agents. Journal of Vacation Marketing, 6(4), 368-385.

 

Lepp, A., & Gibson, H. (2003). Tourist roles, perceived risk and international tourism. I
Annals of Tourism Research, 30(3), 606-624.

Litvin, S. W., & Ling, S. N. S. (2001). The destination attribute management model: an
empirical application to Bintan, Indonesia. Tourism Management, 22(5), 481-492.

Lo, A., Cheung, C., & Law, R. (2004). Information search behavior of Mainland Chinese
air travelers to Hong Kong. Journal of Travel & Tourism Marketing, 16(1), 43-51 .

123

Maser, B., & Weiermair, K. (1998). Travel decision-making: From the vantage point of
perceived risk and information preferences. Journal of Travel & Tourism
Marketing, 7(4), 107-121.

Mazzocchi, M., & Montini, A. (2001). Earthquake effects on tourism in central Italy.
Annals of Tourism Research, 28(4), 1031-1046.

Mihnan, A., & Pizam, A. (1995). The role of awareness and familiarity with destination:
The Central Florida case. Journal of Travel Research, 33(3), 21-27.

Money, R. B., & Crotts, J. C. (2003). The effect of uncertainty avoidance on information
search, planning, and purchases of international travel vacations. Tourism
Management, 24(2), 191-202.

Phelps, A. (1986). Holiday destination image-the problem of assessment: An example
developed in Menorca. Tourism Management, 7(3), 168-180.

Pike, S. (2002). Destination image analysis - A review of 142 papers from 1973 to 2000.
Tourism Management, 23(5), 541-549.

Pike, 8., & Ryan, C. (2004). Destination positioning analysis through a comparison of
cognitive, affective, and conative perceptions. Journal of Travel Research, 42(4),
333-342.

Pizam, A., Jafari, J ., & Milman, A. (1991). Inﬂuence of tourism on attitudes: US students
visiting USSR. Tourism Management, 12(1), 47-54.

Reilly, M. D. (1990). Free elicitation of descriptive adjectives for tourism image
assessment. Journal of Travel Research, 28(4), 21-26.

Richter, L. K. (2003). International tourism and its global public health consequences.
Journal of Travel Research, 41(4), 340-347.

Rittichainuwat, B. N., Qu, H., & Brown, T. J. (2001). Thailand's international travel
image. Cornell Hotel and Restaurant Administration Quarterly, 42(22), 82-95.

Roehl, W. S., & Fesenmaier, D. R. (1992). Risk perception and pleasure travel: An
exploratory study. Journal of Travel Research, 30(4), 17-26.

Ross, G. F. (1993). Destination evaluation and vacation preferences. Annals of Tourism
Research, 20, 477-489.

Russell, J.A., Ward, L.M., & Pratt, G. (1981). Affective quality attributed to
environments: A factor analytic study. Environment and Behavior, 13(3), 259-288.

124

 

 

Snepenger, D. J ., & Snepenger, M. (1993). Information search by pleasure travelers. In
Mahmood A. Khan, M. D. Olsen & T. Var (Eds), VNR’s encyclopedia of
Hospitality and Tourism (pp. 830-835). New York: Van Nostrand Reinhold.

Sonmez, S. F ., & Graefe, A. R. (1998a). Determining future travel behavior from past
travel experience and perceptions of risk and safety. Journal of Travel Research,
37(2), 171-177.

Sonmez, S. F., & Graefe, A. R. (1998b). Inﬂuence of terrorism risk on foreign tourism
decisions. Annals of Tourism Research, 25(1), 112-144.

Stern, E., & Krakover, S. (1993). The formation of a composite urban image.
Geographical Analysis, 25(2), 130- l 46.

Suh, Y. K., & Gartner, W. C. (2004). Perceptions in international urban tourism: An
analysis of travelers to Seoul, Korea. Journal of Travel Research, 43(1), 39-45.

Tabachnick, B. G., & Fidell., L. S. (2001). Using multivariate statistics (4th ed.). Boston:
Allyan and Bacon.

 

Travel Industry Association of America (TIA) (2005). 2005 International Outlook for
Travel and Tourism. Washington, DC: TIA.

Um, S., & Crompton, J. L. (1992). The roles of perceived inhibitors and facilitators in
pleasure travel destination decisions. Journal of Travel Research, 30(3), 18-25.

Vogt, C. A., & Fesenmaier, D. R. (1998). Expanding the functional information search
model. Annals of Tourism Research. 25(3). 551-578.

Vogt, C. A., Andereck, KL. (2003). Destination perceptions across a vacation. Journal of
Travel Research, 41, 348-354.

Woodside, A. G., & Lysonski, S. (1989). A general model of traveler destination choice.
Journal of Travel Research, 27(4), 8-14.

World Tourism Organization (WTO) (2006). Tourism 2020 Vision. Retrieved from:
http: //www. world- tourism. org/facts/menu. html.

 

Yamamoto, D., & Gill, A. M. (1999). Emerging trends in Japanese package tourism.
Journal of Travel Research, 38(2), 134-143.

125

      
 

    

       
   

  

G

11111111

93

8

1111111