ARTICLE IN PRESS

0261-5177/$ - se

doi:10.1016/j.to

�CorrespondE-mail addr

h.qu@okstate.e

Tourism Management 29 (2008) 624–636

www.elsevier.com/locate/tourman

Examining the structural relationships of destination image, touristsatisfaction and destination loyalty: An integrated approach

Christina Geng-Qing Chia,�, Hailin Qub

aSchool of Hospitality Business Management, Washington State University, Pullman, WA 99164-4742, USAbSchool of Hotel and Restaurant Administration, Oklahoma State University, 210 HESW, Stillwater, OK 74078, USA

Received 25 September 2006; accepted 25 June 2007

Abstract

The objective of this study was to offer an integrated approach to understanding destination loyalty by examining the theoretical and

empirical evidence on the causal relationships among destination image, tourist attribute and overall satisfaction, and destination loyalty.

A research model was proposed in which seven hypotheses were developed. The empirical data was collected in a major tourism

destination in the state of Arkansas—Eureka Springs. A total of 345 questionnaires were returned and the data were analyzed using

Structural Equation Modeling (SEM). The results supported the proposed destination loyalty model: (1) destination image directly

influenced attribute satisfaction; (2) destination image and attribute satisfaction were both direct antecedents of overall satisfaction; and

(3) overall satisfaction and attribute satisfaction in turn had direct and positive impact on destination loyalty. The theoretical and

managerial implications were drawn based on the study findings, and recommendations for future researchers were made.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Destination image; Destination loyalty; Tourist satisfaction; Attribute satisfaction; Mediation; Structural equation modeling (SEM)

1. Introduction

The link between customer satisfaction and companysuccess has historically been a matter of faith, andnumerous satisfaction studies have also supported thecase. Customer satisfaction has always been considered anessential business goal because it was assumed that satisfiedcustomers would buy more. However, many companieshave started to notice a high customer defection despitehigh satisfaction ratings (Oliver, 1999; Taylor, 1998). Thisphenomenon has prompted a number of scholars (e.g.,Jones & Sasser, 1995; Oliver, 1999; Reichheld, 1996) tocriticize the mere satisfaction studies and call for aparadigm shift to the quest of loyalty as a strategicbusiness goal.

As a result, satisfaction measurement has recently beendisplaced by the concept of customer loyalty, primarily

e front matter r 2007 Elsevier Ltd. All rights reserved.

urman.2007.06.007

ing author. Tel.: +1405 744 6711; fax: +1 405 744 6299.

esses: cgengqi@yahoo.com (C.G.-Q. Chi),

du (H. Qu).

because loyalty is seen as a better predictor of actualbehavior. Two of the three measures making up mostCustomer Loyalty Indices (CLI) are behavior based, suchas ‘‘likelihood to repurchase the product or service’’ and‘‘likelihood to recommend a product or service to others.’’The third element of a CLI is usually ‘‘overall satisfaction’’itself (Taylor, 1998).The move to measure loyalty is based on a desire to better

understand retention, which has a direct link to a company’sbottom line. Studies have documented that a 5% increase incustomer retention can generate a profit growth of 25–95%across a range of industries (Reichheld, 1996; Reichheld &Sasser, 1990). In addition, retaining existing customersusually has a much lower associated costs than winning newones (Fornell & Wernerfelt, 1987), so a larger proportion ofthe gross profit counts towards the bottom line. Further-more, loyal customers are more likely to act as free word-of-mouth (WOM) advertising agents that informally bringnetworks of friends, relatives and other potential consumersto a product/service (Shoemaker & Lewis, 1999). In fact,WOM referrals account for up to 60% of sales to new

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636 625

customers (Reichheld & Sasser, 1990). With such excep-tional returns, loyalty becomes a fundamental strategiccomponent for organizations.

However, in the context of travel and tourism, a reviewof literature reveals an abundance of studies on touristsatisfaction; and destination loyalty has not been thor-oughly investigated (Oppermann, 2000). Therefore, it istime for practitioners and academics to conduct morestudies of loyalty in order to have greater knowledge of thisconcept, to understand the role of customer satisfaction indeveloping loyalty, the impact of other non-satisfactiondeterminants on customer loyalty, and their interrelation-ships.

Understanding the determinants of customer loyalty willallow management to concentrate on the major influencingfactors that lead to customer retention. A number ofstudies have examined the antecedents or causes of repeatpurchase intentions (Backman & Crompton, 1991; Cronin,Brady, & Hult, 2000; Petrick, Morais, & Norman, 2001).Results of this body of research have shown thatsatisfaction, quality/performance, and different other vari-ables are good predictors of customer intended loyalty. Themore satisfied the customers are, the more likely they are torepurchase the product/service and to encourage others tobecome customers. In order to retain customers, organiza-tions must seek to satisfy them, but a further objective mustbe to establish customer loyalty.

In the tourism context, satisfaction with travel experiencescontributes to destination loyalty (Alexandris, Kouthouris,& Meligdis, 2006; Bramwell, 1998; Oppermann, 2000;Pritchard & Howard, 1997). The degree of tourists’ loyaltyto a destination is reflected in their intentions to revisit thedestination and in their willingness to recommend it(Oppermann, 2000). Tourists’ positive experiences of service,products, and other resources provided by tourism destina-tions could produce repeat visits as well as positive WOMeffects to friends and/or relatives. Recommendations byprevious visits can be taken as the most reliable informationsources for potential tourists. Recommendations to otherpeople (WOM) are also one of the most often sought typesof information for people interested in traveling.

Given the vital role of customer satisfaction, one shouldnot be surprised that a great deal of research has beendevoted to investigating the antecedents of satisfaction(Churchill & Surprenant, 1982; Oliver, 1980; Oliver &DeSarbo, 1988; Tse & Wilton, 1988). Most early researchwork focused on satisfaction at the global level (e.g.,Oliver, 1980). Recently, there emerged an attribute-levelconceptualization of the antecedents of satisfaction (e.g.,Oliver, 1993). Under an attribute-level approach, overallsatisfaction is a function of attribute-level evaluations.These evaluations typically capture a significant amount ofvariation in overall satisfaction (e.g., Bolton & Drew, 1991;Oliver, 1993). Overall satisfaction and attribute satisfactionare distinct, though related, constructs (Oliver, 1993). Thisstudy focused on overall evaluation, attribute satisfaction,and the relationship between the two.

Furthermore, previous studies (e.g. Baloglu &McCleary,1999; Chon, 1990, 1992) showed that destination image willinfluence tourists in the process of choosing a destination,the subsequent evaluation of the trip and in their futureintentions. Destination image exercises a positive influenceon perceived quality and satisfaction. A positive imagederiving from positive travel experiences would result in apositive evaluation of a destination. Tourist satisfactionwould improve if the destination has a positive image.Destination image also affects tourists’ behavioral inten-tions. More favorable image will lead to higher likelihoodto return to the same destination.To sum up, the following sequence could be established:

destination image - tourist satisfaction - destinationloyalty. Destination image is an antecedent of satisfaction.Satisfaction in turn has a positive influence on destinationloyalty. In an increasingly saturated marketplace, thesuccess of marketing destinations should be guided by athorough analysis of destination loyalty and its interplaywith tourist satisfaction and destination image. Never-theless, the tourism studies to date have addressed andexamined the constructs of image, satisfaction, and loyaltyindependently (Bigne, Sanchez, & Sanchez, 2001); studiesdiscussing the causal relationships among destinationimage, tourist satisfaction, and destination loyalty arelacking.To bridge the gap in the destination loyalty literature,

the main purpose of this study was to offer an integratedapproach for understanding destination loyalty and toexamine the theoretical and empirical evidence on thecausal relationships among destination image, touristsatisfaction, and destination loyalty. A research modelwas proposed and tested. The model investigated therelevant relationships among the constructs by using astructural equation modeling (SEM) approach. Theempirical data for the study was collected in a majortourism destination in the state of Arkansas—EurekaSprings, which is known as ‘‘the little Switzerland of theOzarks’’ and was named one of the 12 DistinctiveDestinations by the National Trust for Historic Preserva-tion. The city has been attracting people of all ages from allaround the country for over 100 years.

2. Literature review

The main purpose of this study was to develop and test atheoretical model, which represented the elements con-tributing to the building of destination loyalty: destinationimage, attribute satisfaction, and overall satisfaction.Below is a brief overview of the interrelationships of theconstructs in the model.It has been widely acknowledged that destination image

affects tourists’ subjective perception, consequent behavior,and destination choice (e.g., Baloglu & McCleary, 1999;Castro, Armario, & Ruiz, 2007; Chon, 1990, 1992; Echtner& Ritchie, 1991; Milman & Pizam, 1995; Woodside &Lysonski, 1989). Tourists’ behavior is expected to be partly

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636626

conditioned by the image that they have of destinations.Image will influence tourists in the process of choosing adestination, the subsequent evaluation of the trip, and intheir future intentions.

The influence of image on destination choice processhas been studied by various authors (e.g., Crompton &Ankomah, 1993; Gartner, 1989; Goodall, 1988). It isbelieved that destinations with more positive images willmore likely be included in the process of decision making.In addition, destination image exercises a positive influenceon perceived quality and satisfaction. More favorableimage will lead to higher tourist satisfaction. In turn, theevaluation of the destination experience will influence theimage and modify it (Chon, 1991; Echtner & Ritchie, 1991;Fakeye & Crompton, 1991; Ross, 1993). Lastly, destinationimage also affects the behavioral intentions of tourists. Forexample, Court and Lupton (1997) found that the image ofthe destination under study positively affects visitors’intention to revisit in the future.

Kotler, Bowen, and Makens (1996) established thefollowing sequence: image - quality - satisfaction. Inthis model, image would affect how customers perceivequality—a more positive image corresponds to a higherperceived quality. Perceived quality will in turn determinethe satisfaction of consumers (Fornell, Johnson, Anderson,Cha, & Bryant, 1996; Kozak & Rimmington, 2000),because satisfaction is the result of customers’ assessmentof the perceived quality. To test the relationship betweendestination image and tourist satisfaction, the followinghypotheses were proposed:

H1. Destination image positively influences tourists’ over-all satisfaction.

Most early research work concentrated on satisfaction atthe global level (e.g. Oliver, 1980). Until recently,researchers started to pay attention to attribute-levelconceptualization of the antecedents of satisfaction (e.g.Oliver, 1993). According to Oliver (1993), overall satisfac-tion and attribute satisfaction are distinct but relatedconstructs. Attribute satisfaction has significant, positive,and direct effects on overall satisfaction; and it capture asignificant amount of variation in overall satisfaction(Bolton & Drew, 1991; Oliver, 1993; Spreng, Mankenzie,& Olshavsky, 1996).

Satisfaction research in tourism and recreation hasindicated that tourists’ satisfaction with individual compo-nent of the destination leads to their satisfaction with theoverall destination (e.g. Danaher & Arweiler, 1996;Hsu, 2003; Mayer, Johnson, Hu, & Chen, 1998; Ross &Iso-Ahola, 1991). It is important in tourism to distinguishoverall satisfaction from satisfaction with individualattributes; because the particular characteristics of tourismhave a notable effect on tourist satisfaction (Seaton &Benett, 1996). Beyond the generic characteristics thatdistinguish services from goods, such as intangibility,inseparability, heterogeneity, and perishability (Zeithaml,Parasuraman, & Berry, 1985), there are some further

differences between tourism and other services. Forexample, Middleton and Clarke (2001) highlighted inter-dependence-sub-sector interlinkage of tourism products.Tourists experience a medley of services such as hotels,restaurants, shops, attractions, etc.; and they may evaluateeach service element separately. Satisfaction with variouscomponents of the destination leads to overall satisfaction(Kozak & Rimmington, 2000). Overall satisfaction with ahospitality experience is a function of satisfactions with theindividual elements/attributes of all the products/servicesthat make up the experience, such as accommodation,weather, natural environment, social environment, etc.(Lounsbury & Hoopes, 1985; Pizam & Ellis, 1999).Therefore, it was postulated that:

H2. Attribute satisfaction positively influences overallsatisfaction.

Two more hypotheses were drawn to test the relationshipbetween destination image, attribute satisfaction, andoverall satisfaction:

H3. Destination image positively influences tourists’ attri-bute satisfaction.

H4. Attribute satisfaction partially mediated the relation-ship between destination image and overall satisfaction.

The link between satisfaction and post-purchase behaviorhas been well established by prior literature (Hallowell,1996; LaBarbera & Mazursky, 1983; Rust & Zahorik, 1993).It is generally believed that satisfaction leads to repeatpurchase and positive WOM recommendation, which aremain indicators of loyalty. Marketing literature has paidmuch attention to the relationship between customersatisfaction and loyalty, and a number of studies haveconfirmed a significant positive relationship between custo-mer satisfaction and loyalty/retention (Anderson & Sullivan,1993; Cronin et al., 2000; Taylor & Baker, 1994). Ifconsumers are satisfied with the product/service, they aremore likely to continue to purchase, and are more willing tospread positive WOM.In tourism industry, there are empirical evidences that

tourists’ satisfaction is a strong indicator of their intentionsto revisit and recommend the destination to other people(Beeho & Prentice, 1997; Bramwell, 1998; Juaneda, 1996;Kozak, 2001; Kozak & Rimmington, 2000; Ross, 1993;Yau & Chan, 1990; Yoon & Uysal, 2005). Satisfied touristsare more likely to return to the same destination, and aremore willing to share their positive traveling experiencewith their friends and relatives. WOM recommendationsare especially critical in tourism marketing because they areconsidered to be the most reliable, and thus are one of themost sought-after information sources for potential tour-ists (Yoon & Uysal, 2005). Therefore, the followinghypotheses were drawn:

H5. Overall satisfaction positively influenced destinationloyalty.

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636 627

H6. Overall satisfaction fully mediated the relationshipbetween destination image and destination loyalty.

H7. Overall satisfaction fully mediated the relationshipbetween attribute satisfaction and destination loyalty.

To sum up the seven hypotheses, the following pathscould be established: (1) destination image - attributesatisfaction - overall satisfaction; (2) destination image- overall satisfaction - destination loyalty; (3) attributesatisfaction - overall satisfaction - destination loyalty.

3. Methodology

3.1. Survey instrument

This study employed a causal research design using across-sectional sample survey. The survey questionnaireconsisted of the following major sections: questions thatmeasured the following constructs—destination image,tourists’ attribute satisfaction, overall satisfaction, destina-tion loyalty, and questions designed to gather tourists’demographic information and travel behavior.

(1)

Destination image. A combination of structured andunstructured techniques was used in order to capturevarious aspects of the respondents’ perceptions ofEureka Springs as a travel destination, including acomprehensive review of previous destination litera-ture, content analysis of tourism literature, promotionbrochures, and websites for Eureka Springs, and theemployment of qualitative research techniques such asfocus group sessions, unstructured personal interviews,and managerial judgment. The selected 53 destinationitems were rated on a 7-point Likert scale where1 ¼ Strongly disagree (SD) and 7 ¼ Strongly agree

(SA).

(2) Attribute satisfaction. Drawing upon the most relevant

tourism literature and destination attributes applicableto the Eureka Springs situation, an attribute listconsisting of 33 items was established. The destinationattributes encompassed seven domains of tourismactivities: accommodation, dining, shopping, attrac-tions, activities and events, environment, and accessi-bility. The choice of attributes within each domainvaried with the chosen mix of the seven tourismactivities. Along seven-point Likert-scales, tourists wereasked to evaluate their satisfaction with each tourist-attracting attribute (1 ¼ Very dissatisfied and 7 ¼ Very

satisfied ).

(3) Overall satisfaction. A number of studies have used a

summative overall measure of satisfaction (e.g. Bloemer& Ruyter, 1998; Bolton & Lemon, 1999; Fornell et al.,1996). A single overall measure of satisfaction was used inthis study for its ease of use and empirical support. Therespondents were asked to rate their satisfaction with theoverall traveling experience on a 7-point Likert scale with1 being Very dissatisfied and 7 being Very satisfied.

(4)

Destination loyalty. Attitudinal measurement, includingrepeat purchase intentions and WOM recommenda-tions were most usually used to infer consumer loyalty,and were found to be the pertinent measure (Hawkins,Best, & Coney, 1989; Jones & Sasser, 1995). Priorresearch has shown that loyal customers are more likelyto repurchase a product/service in the future (Hughes,1991; Petrick et al., 2001; Sonmez & Graefe, 1998). Ithas also been suggested that loyal visitors are morewilling to recommend the product/service to others(Shoemaker & Lewis, 1999). In addition, good correla-tion has been found between consumers’ repurchaseintentions and positive WOM referrals (Oh, 2000; Oh &Parks, 1997). Therefore, repurchase and referral inten-tions make up the most CLI (Taylor, 1998). In thisstudy, two single-item measures were used for assessingtourist destination loyalty as the ultimate dependentconstruct: tourists’ intention to revisit Eureka Springsand their willingness to recommend Eureka Springs asa favorable destination to others, with 7-point Likertscale (1 ¼ most unlikely; 7 ¼ most likely).

3.2. Reliability

A pilot test was conducted to test the internalconsistency of the questionnaire items. The first draft ofthe survey instrument was distributed to 50 randomlyselected visitors who stayed at Eureka Springs’ hotels andmotels. A total of 32 completed surveys were returned.A reliability analysis (Cronbach’s alpha) was performed for‘destination image’ and ‘attribute satisfaction,’ resulting ina robust a of 0.96 and 0.93, respectively. An alpha of 0.7 orabove is considered acceptable as a good indication ofreliability (Nunnally & Bernstein, 1994). Based on theresults of the pilot test and feedbacks from Eureka SpringsChamber of Commerce, the final version of the surveyinstrument was developed.

3.3. Sampling plan

The target population was all the visitors who stoppedby the Eureka Springs Welcome Center, stayed at hotels,motels, and B&B, and they also visited souvenir shops/artgalleries during a 2-month survey period. Confidenceinterval approach was used to determine the sample size(Burns & Bush, 1995). The formula for obtaining 95%accuracy at the 95% confidence level is

n ¼z2ðpqÞ

e2¼

1:962ð0:5� 0:5Þ

0:052¼ 385,

where z is the standard error associated with chosen level ofconfidence (95%); p the estimated variability in thepopulation (50%); q ¼ 1�p; and e the acceptable error75% (desired accuracy 95%). The amount of variability inthe population is estimated to be 50%, which is widely usedin social research (e.g. National opinion polls in the USA).Assuming a response rate of 50% and an unusable rate of

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636628

10%, a total of 963 (385/0.4) people were approached toparticipate in the survey.

A two-stage sampling approach was used: proportionatestratified sampling was applied for deciding on the stratasample size, and systematic random sampling (SRS) wasused to select the survey participant within each stratum,which involved choosing every kth element after a randomstart.

3.4. Data analysis

Exploratory factor analysis (EFA) was employed toderive the underlying dimensions of destination image andvisitors’ attribute satisfaction. Confirmatory factor analysis(CFA) and SEM were used to test the conceptual modelthat examined the antecedents of destination loyalty.

4. Results

A total of 345 questionnaires were returned, about 90%of the targeted sample size. The vast majority of therespondents (99%) were domestic visitors from 21 differentUS states. The majority (94%) of the respondents weretraveling with partners (family and friends), and vacation/leisure was quoted as the major purpose of the trip (79%).One-third (33%) of the respondents were first-time visitors.Previous visits (37%) and WOM (33%) emerged as the twokey information sources for respondents to learn about thetravel destination.

4.1. Underlying dimensions of ‘destination image’ and

‘attribute satisfaction’

The EFA was performed to determine the underlyingdimensionality of ‘destination image’ by analyzing patternsof correlations among the 53 image attributes. Principleaxis factoring extraction method with oblimin rotation wasadopted because (1) oblique rotation is best when the goalof the factor analysis is to obtain several theoreticallymeaningful factors; (2) oblique rotation assumes thatfactors are correlated to each other, which is morejustifiable and more realistic in social sciences (Hair,Anderson, Tatham, & Black, 1998).

A range of cutoff criteria were used to determine thenumber of factors derived, such as eigenvalues, scree plot,percentage of variance, item communalities, and factorloadings (Hair et al., 1998). Items with loadings lower than0.4 and with loadings higher than 0.4 on more than onefactor were eliminated. A nine-factor solution, with 37variables being retained, was chosen representing approxi-mately 75.9% of the total variance (see Table 1). Althoughonly seven factors had eigenvalues greater than 1.0,the scree plot showed that the cutoff point for maximumfactors to extract might be nine. Both seven-factorand nine-factor solutions were analyzed, and the loadingsof the nine-factor model presented a cleaner and moreinterpretable solution. The last two factors included in the

nine-factor model also represented important aspects of‘destination image.’The communalities of the 37 variables ranged from 0.45

to 0.90, suggesting that the variances of each originalvariable (from 45% to 90%) were reasonably explained bythe nine-factor solution. Factor loadings of the variablesranged from 0.41 to 0.96, above the suggested thresholdvalue of 0.30 for practical and statistical significance (Hairet al., 1998). The Cronbach’s alpha for the nine factorsvaried from 0.81 to 0.93, suggesting high internalconsistency. The nine factors were labeled based on thecore variables that constituted them: travel environment,natural attractions, entertainment and events, historicattractions, travel infrastructure, accessibility, relaxation,outdoor activities, and price and value. Nine compositevariables were created and used as indicators for the latentconstruct ‘destination image’ in the subsequent SEM.The same EFA procedure was used to verify the pre-

specified dimensions of tourist satisfaction. Seven factorswith eigenvalues above 1.0 were generated, which explainedabout 71% of the total variance (see Table 2). Thecommunalities varied from 0.50 to 0.92, suggesting thatthe variance in each original variable was reasonablyexplained by the seven common factors taken together. Thefactor loadings for the 33 variables ranged from 0.38 to0.89, within the threshold value suggested by Hair et al.(1998). The loadings also presented a clean and highlyinterpretable solution: the 33 variables loaded significantlyon seven factors as the researchers conceptualized—lodging, dining, shopping, attractions, activities and events,environment, and accessibility; no variables loaded sig-nificantly on more than one factor. The Cronbach’s alphasfor the seven factors were robust, ranging from 0.85 to0.91, well above the generally agreed upon lower limit of0.60 for research at exploratory stage (Nunnally &Bernstein, 1994), indicating high internal consistencyamong the variables within each factor. Seven summatedscales were created and used as manifest variables for thelatent variable ‘attribute satisfaction’ in the subsequentSEM analysis.

4.2. Testing the destination loyalty model

SEM was applied for testing the destination loyaltymodel in which seven hypotheses were developed based ona comprehensive review of literature. Various measures ofoverall model goodness-of-fit and measurement model fitwere assessed to determine if the proposed conceptualmodel was acceptable.

4.2.1. Overall model fit

Overall model fit depicts the degree to which thespecified indicators represent the hypothesized constructs.The w2 value (690.67 with 149 degrees of freedom) has astatistical significance level of 0.0. This statistic failed tosupport that the differences of the predicted and actualmodels were non-significant. However, it is generally

ARTICLE IN PRESS

Table 1

Underlying dimensions of ‘destination image’

Eigenvalue Variance explained (%) Cronbach’s a Factor loadings Communalities

F1 Travel environment 15.44 41.74 0.86

Safe and secure environment 0.73 0.71

Clean and tidy environment 0.64 0.71

Friendly and helpful local people 0.56 0.67

Tranquil and restful atmosphere 0.55 0.70

Pleasant weather 0.41 0.49

F2 Natural attractions 3.48 9.40 0.93

Scenic mountain and valleys �0.82 0.74

Breathtaking scenery and natural attractions �0.80 0.80

Gorgeous gardens and springs �0.79 0.79

Fabulous scenic drive �0.68 0.69

Picturesque parks/lakes/rivers �0.67 0.78

Unspoiled wilderness and fascinating wildlife �0.58 0.66

Spectacular caves and underground formations �0.47 0.56

F3 Entertainment and events 2.40 6.50 0.90

Wide arrays of shows/exhibitions �0.75 0.76

Tempting cultural events and festivals �0.75 0.74

Excellent quality and fun country/western music �0.74 0.69

Colorful nightlife �0.59 0.61

Wide variety of entertainment �0.58 0.65

F4 Historic attractions 1.43 3.87 0.83

Distinctive history and heritage 0.80 0.79

Vintage buildings 0.69 0.62

F5 Infrastructure 1.39 3.74 0.84

Wide selection of restaurants/cuisine �0.77 0.71

Wide variety of shop facilities �0.68 0.71

Wide choice of accommodations �0.53 0.52

F6 Accessibility 1.16 3.14 0.81

Well communicated traffic flow and parking information 0.73 0.73

Available parking downtown 0.62 0.59

Easy access to the area 0.56 0.59

Easy-to-use and affordable trolley system 0.41 0.45

F7 Relaxation 1.04 2.80 0.84

Relaxing day spa and healing getaway 0.70 0.68

Great place for soothing the mind and refreshing the body 0.67 0.68

Spiritual rejuvenation 0.66 0.68

F8 Outdoor activities 0.90 2.43 0.88

Exciting water sports/activities (boating, fishing, etc) 0.79 0.80

Terrific place for hiking/picnicking/camping/hunting 0.64 0.69

Enormous opportunities for outdoor recreation 0.45 0.66

Good facilities for golfing 0.43 0.65

F9 Price and value 0.84 2.27 0.89

Reasonable price for food and accommodation �0.96 0.90

Good value for money �0.70 0.76

Reasonable price for attractions and activities �0.67 0.73

Good bargain shopping �0.46 0.66

C.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636 629

agreed that the w2 value should be used as a guide ratherthan an absolute index of fit due to its sensitivity to samplesize and model complexity (Anderson & Gerbing, 1982).Thus, other indices should also be assessed. Incremental FitMeasures assess the incremental fit of the model comparedto a null model that usually specifies no relation among theconstructs and variables. These were the Comparative FitIndex (CFI), the Tucker–Lewis Index (TLI), and the

Normed Fit Index (NFI), which were 0.95, 0.95, and0.94, respectively. These measures were above the recom-mended level of 0.90, indicating support for the proposedmodel. Another measure to assess the model fit is the RootMean Square Error of Approximation (RMSEA), whichprovides a measure of fit that adjusts for parsimony byassessing the discrepancy per degree of freedom in themodel. The RMSEA value was a marginal 0.11.

ARTICLE IN PRESS

Table 2

Underlying dimensions of ‘attribute satisfaction’

Eigenvalue Variance explained Cronbach’s a Factor loadings Communalities

F1 Shopping 14.87 45.06 0.85

Quality of merchandise 0.82 0.74

Reasonable price of merchandise 0.63 0.59

Variety of shops 0.58 0.61

Friendliness of service 0.61 0.54

F2 Activities and events 2.63 8.00 0.86

Variety of special events/festivals 0.85 0.76

Variety of spa/massage/healing options 0.72 0.55

Variety of evening entertainment 0.62 0.53

Variety of outdoor recreation 0.58 0.61

Reasonable price for activities and events 0.38 0.63

F3 Lodging 1.79 5.42 0.90

Uniqueness of lodging 0.89 0.82

Variety of lodging options 0.80 0.72

Historic interests of lodging 0.61 0.50

Service in lodging facilities 0.53 0.71

Reasonable price of meals 0.45 0.50

Quality and cleanliness of lodging facilities 0.45 0.64

F4 Accessibility 1.59 4.82 0.91

Availability of local parking 0.73 0.56

Convenience of local transportation 0.71 0.69

Availability of travel information 0.67 0.73

Helpfulness of welcome center 0.62 0.62

Ease of access 0.60 0.56

F5 Attractions 1.14 3.45 0.85

Variety of historic/cultural sites �0.73 0.92

Variety of natural attractions �0.64 0.79

Variety of cultural options �0.55 0.72

Reasonable price for sightseeing �0.38 0.64

F6 Environment 1.11 3.37 0.89

Peaceful and restful atmosphere 0.69 0.75

Cleanliness 0.65 0.77

Friendliness of local people 0.62 0.70

Safety and security 0.48 0.58

F7 Dining 1.08 3.26 0.87

Quality of food 0.83 0.77

Variety of cuisine 0.75 0.63

Service in restaurants 0.74 0.69

Convenience of meals 0.71 0.58

Reasonable price of meals 0.62 0.62

C.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636630

4.2.2. Measurement model fit

The measurement model provides meaning to theconstructs (latent variables) in the model. Proper evalua-tion of the measurement model is a pre-requisite to theevaluation of the structural model (Anderson & Gerbing,1982). The convergent validity of the measurement scalewas examined via the following tests. First, for eachvariable the t value associated with each of the loadingswas significant at the 0.01 level (see Table 3). The resultsindicated that all variables were significantly related totheir specified constructs, verifying the posited relation-ships among indictors and constructs. Second, squaredmultiple correlation coefficients (SMC) for the y- andx-variables were assessed. SMCs lie between 0 and 1 (the

closer to 1, the better the variable acts as an indicator of thelatent construct). Table 3 revealed that the SMCs fory-variables ranged from 0.52 to 0.92 and for x- variablesfrom 0.30 to 0.65, indicating fairly high reliability (con-vergent validities) of the measurement model.The construct reliability (CR) and the average variance

extracted (AVE) were also computed for the latentconstructs. For both CR and AVE, all three constructssurpassed the threshold value of .70 and .50, respectively.Therefore, it can be concluded that the indicators for allthree constructs were sufficient in terms of how themeasurement model was specified. To examine thediscriminant validity of the measurement model, the AVEvalues for the latent constructs were compared to the

ARTICLE IN PRESS

Table 3

LISREL results for measurement model

Std. loadings SMC CR AVE

Exogenous: destination image 0.91 0.52

Travel environment 0.80 0.63

Natural attractions 0.70 0.49

Entertainment and events 0.76 0.58

Historic attractions 0.55 0.30

Infrastructure 0.73 0.54

Accessibility 0.72 0.51

Price and value 0.81 0.65

Outdoor activities 0.72 0.52

Relaxation 0.68 0.47

Endogenous: attribute satisfaction 0.91 0.60

Lodging 0.75 0.56

Attractions 0.84 0.71

Shopping 0.83 0.69

Dining 0.75 0.56

Activities and events 0.72 0.52

Accessibility 0.76 0.57

Environment 0.76 0.58

Endogenous: destination loyalty 0.90 0.62

Revisit intention 0.84 0.70

Recommend intention 0.96 0.92

OverallSatisfaction

AttributeSatisfaction

DestinationImage

DestinationLoyalty

H1

H2

H3

H5

Fig. 1. Theoretical ‘destination loyalty’ model (MT).

Competing Model M1

Competing Model M2

OverallSatisfaction

AttributeSatisfaction

DestinationImage

DestinationLoyalty

OverallSatisfaction

AttributeSatisfaction

DestinationImage

DestinationLoyalty

Fig. 2. Competing ‘destination loyalty’ models.

C.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636 631

squared correlations between the corresponding constructs(Fornell & Larcker, 1981), and none of the squaredcorrelations surpassed the AVE. The above tests indicatedthat the discriminant validity was upheld for the measure-ment model.

4.2.3. Structural model parameters

The most obvious examination of the structural modelinvolves the significance tests for the estimated coefficients(paths), which provide the basis for accepting or rejectingthe proposed relationships between latent constructs. TheLISREL results showed that all the paths proposed in the‘destination loyalty’ model were statistically significant andof the appropriate direction (positive): (1) destinationimage positively influenced overall satisfaction (g2,1 ¼ 0.29;t ¼ 4.15); (2) attribute satisfaction positively affectedoverall satisfaction (b2,1 ¼ 0.20; t ¼ 2.94); (3) destinationimage positively influenced attribute satisfaction (g1,1 ¼0.71; t ¼ 11.66); (4) overall satisfaction positively affecteddestination loyalty (b3,2 ¼ 0.74; t ¼ 12.34); and from theresults of (1)–(3), it can be concluded that (5) attributesatisfaction partially mediated the relationship betweendestination image and overall satisfaction (Baron &Kenny, 1986). The hypotheses 1–5 could not be rejected,which proposed causal relationships among destinationimage, attribute satisfaction, overall satisfaction, anddestination loyalty.

The fit of the structural model was also assessed by theSMCs for structural equations, which indicate the amountof variance in each endogenous latent variable accountedfor by the antecedent variables in the relevant structuralequation. The SMC for ‘attribute satisfaction’ was 0.51,

indicating that 51% of the variance in attribute satisfactionwas explained by ‘destination image.’ About 25% of theuncertainties in ‘overall satisfaction’ were accounted for by‘destination image’ and ‘attribute satisfaction’ (SMC ¼0.25). ‘Destination image,’ ‘overall satisfaction,’ and‘attribute satisfaction’ explained approximately 44% ofthe variance in ‘Destination loyalty’ (SMC ¼ 0.44).

4.2.4. Competing models

The final approach to model assessment was to comparethe proposed theoretical model (MT) (see Fig. 1) with aseries of competing models, which acted as alterativeexplanations to the proposed model. The objective was todetermine the best fitting model from a set of models.In this study, two alternative models were proposed (seeFig. 2): M1 and M2. M1 added the path between ‘attributesatisfaction’ and ‘destination loyalty.’ M2 further addedanother path between ‘destination image’ and ‘destinationloyalty.’The sequential Chi-square (w2) difference tests (SCDTs)

were performed to assess whether there were significantdifferences in estimated construct covariances explained bythe three structural models (Joreskog & Sorbom, 1995).The w2 difference test examined the null hypotheses of no

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636632

significant difference between two nested structural models(denoted as M1–MT ¼ 0 and M1–M2 ¼ 0). The differencebetween w2 statistic values (Dw2) for nested models wasitself asymptotically distributed as w2, with degrees offreedom equal to the difference in degrees of freedom forthe two models (Ddf). If the null hypothesis was upheld, themore constrained model of the two would be tentativelyaccepted. The w2 difference test between MT and M1

(Dw2 ¼ 5.43; Ddf ¼ 1) suggested that M1 was performingsignificantly better than the theoretical model MT; and thew2 difference test between M1 and M2 (Dw2 ¼ 0.28;Ddf ¼ 1) suggested that M2 was not performing signifi-cantly better than M1.

The results of the w2 difference tests favored the competingmodel M1 to the proposed theoretical model MT and thealternative model M2 (saturated model). To further detect theeffect of adding more causal relationships (paths), it wasnecessary to examine the statistical significance of theparameter coefficients for the additional paths for M1 andM2. The causal relationship between ‘attribute satisfaction’and ‘destination loyalty’ was significant (b ¼ 0.12; t ¼ 2.32);whereas the causal path from ‘destination image’ to‘destination loyalty’ was not deemed significant (g ¼ 0.04;t ¼ 0.54). This suggested that there should be a direct pathbetween ‘attribute satisfaction’ and ‘destination loyalty’ as thecompeting model M1 proposed. This relationship could betheoretically justified because tourists’ satisfaction withvarious components of a destination could directly lead totheir loyalty with the destination. The findings supported thefull mediation role of overall satisfaction on the relationshipbetween destination image and destination loyalty (H6 couldnot be rejected), but failed to support the full mediation roleof overall satisfaction on the relationship between attributesatisfaction and destination loyalty (H7 could not besupported). Therefore, overall satisfaction partially mediatedthe relationship between attribute satisfaction and destinationloyalty (Baron & Kenny, 1986).

As another means of comparison, a set of goodness-of-fitmeasures were also compared to determine which of thethree models had the best model fit (see Table 4). The fitindices such as RMSEA, CFI, and PNFI for the threecompeting models were almost identical, indicating that thethree competing models achieved approximately the samelevel of model fit. Thus it was concluded that the competing

Table 4

Fit indices for competing models

Theoretical (MT) M1 M2

Chi-square 690.67 685.24 684.96

Degrees of freedom 149 148 147

RMSEA 0.11 0.11 0.11

RMR 0.062 0.056 0.056

GFI 0.81 0.81 0.81

CFI 0.95 0.95 0.95

NNFI 0.95 0.95 0.94

PNFI 0.82 0.81 0.81

model M1 could be retained as a viable alternative foracceptance. Considering the attenuation in the fit measuresfor large models and large sample sizes, the final model M1

(see Fig. 3), though not achieving the most desirable levelsof fit (especially the overall model fit indices), mayrepresent the best available model until further researchidentifies improvements in theoretical relationships ormeasurement of the constructs.

5. Implications

5.1. Theoretical implications

The SEM analysis offered support for the statisticallysignificant relationships between destination image andoverall satisfaction (H1), attribute satisfaction and overallsatisfaction (H2), destination image and attribute satisfac-tion (H3), and overall satisfaction and destination loyalty(H5). The SEM analysis also confirmed the partialmediation role attribute satisfaction played betweendestination image and overall satisfaction (H4), and thefull mediation role overall satisfaction played betweendestination image and destination loyalty (H6). The onlyhypothesis (H7) that was not supported pointed to overallsatisfaction as a partial mediator, rather than a fullmediator as originally proposed, between attribute satis-faction and destination loyalty. The destination loyaltymodel outlined in the conceptual framework was corrobo-rated. Therefore it can be said that tourist overallsatisfaction was determined by destination image andattribute satisfaction, tourist attribute satisfaction was alsodirectly influenced by destination image, and destinationloyalty was in turn influenced by overall satisfaction. Inaddition, the newly proposed direct path from attributesatisfaction to destination loyalty was shown to besignificant; thus, attribute satisfaction was also a directantecedent of destination loyalty. The findings confirmedthat tourists’ loyalty was enhanced by positive destinationimage and high satisfaction, consistent with the image -satisfaction- loyalty scheme that conceptually guided thisstudy.The empirical results of this study provided tenable

evidence that the proposed structural equation modeldesigned to consider simultaneously destination image,overall and attribute satisfaction, and destination loyaltywas acceptable. Tourism destination loyalty had causalrelationships with image and satisfaction. Additionally, theattribute satisfaction separately from the overall satisfac-tion influenced the destination loyalty. This study makes itclear that destination image plays an essential role inachieving the loyalty of an individual, and tourists’satisfaction must be handled proactively in order todevelop it into a lasting relationship beneficial to bothparties. Destination image had a positive effect on touristsatisfaction as well as destination loyalty. An improvementin the overall image of a place held by an individualincreased his/her propensity to make a positive assessment

ARTICLE IN PRESS

Destination image

X1

X2

X3

X5

X4

X6

X7

X9

X8

Attribute satisfaction

Overall satisfaction

Destination loyalty

Y9

Y8

Y10

0.71 (11.56)

0.29 (4.85)

0.58 (17.58)

0.83 (19.00)

0.54 (12.69)

0.78 (15.63)

0.42 (10.11)

0.66 (15.65)

0.81 (15.8)

0.68 (12.81)

0.63 (12.93)

0.19 (10.76)

0.42 (12.51)

0.50 (11.60)

0.46 (12.61)

0.35 (11.50)

0.53 (11.52)

0.67 (12.19)

0.56 (12.63)

0.28 (9.92)

Y1 Y2 Y3 Y4 Y5 Y6 Y7

0.20 (2.40)

0.12 (2.20)

0.67 (10.76)

0.62 0.74

(13.95)0.81

(15.49) 0.84 (15.37)

0.71 (12.82)

0.58 (16.26) 0.78

(13.98)

0.00

0.60 (7.90)

0.09 (1.74)

1.00

1.10 (17.20)

1.16

0.32 (11.46)

0.41 (11.38)

0.28 (10.06)

0.31 (10.12)

0.51 (11.70)

0.27 (11.52)

0.46 (11.42)

Where:

* X1…..X9: travel environment, natural attractions, entertainment and events, historic attractions, travel

infrastructure, accessibility, relaxation, outdoor activities, and price and value

* Y1…..Y10: lodging, dining, shopping, attractions, activities and events, environment, accessibility, overall

satisfaction, revisit intention, referral intention

* Values in parenthesis are t-statistics (t critical value at 0.05 level = 1.96)

Fig. 3. Results of destination loyalty model (M1).

C.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636 633

of the stay. It also enhances his or her intention to returnand to recommend it in the future. Consequently, withregard to the sequence image - satisfaction - loyaltysuggested by the review of the literature, the analysis of theinterrelationships as a whole confirmed the proposedmodel.

In the literature, although it has been acknowledged thatdestination loyalty is important, not much has been doneto investigate its measurement, or its structural relation-ships with image and satisfaction. This study revealed andconfirmed the existence of the critical relationships amongdestination image, attribute/overall satisfaction, and desti-nation loyalty. The findings suggested that it would beworthwhile for destination managers to make greaterinvestments in their tourism destination resources, in orderto continue to enhance tourists’ experiences. It is believedthat this study has a substantial capability for generatingmore precise applications related to destination behavior,especially concerning tourists’ loyalty.

5.2. Managerial implications

Destinations today are facing steep competitions and thechallenges are getting greater in the years to come.

Therefore, it is essential to gain a better understanding ofwhy travelers are loyal to a destination and what drives theloyalty. The major findings of this study have significantmanagerial implications for tourism and hospitality man-agers and marketers.First of all, the exploratory and confirmatory factor

analyses revealed that destination image was consisting ofnine latent dimensions, and attribute satisfaction had sevenunderlying factors. These results could help destinationmarketers better understand the factors contributing totourist satisfaction and loyalty so that they are able tocarefully deliver appropriate products and services thataccommodate tourists’ needs and wants. Thus, it issuggested that destination suppliers and managers considerthe practical implications of these latent variables, whichmay be fundamental elements in increasing tourists’ overallsatisfaction and loyalty.Furthermore, the SEM findings provided guidance for

the success of marketing destinations. First of all, image isshown in this study to be a key factor in the hands ofdestination managers. It is a direct antecedent of attributesatisfaction and overall satisfaction as well as a majorfactor in influencing destination loyalty. Therefore, desti-nation managers must strive to improve the image tourists

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636634

hold of a destination if they are to compete successfully inthe competitive holiday market. Adding to the fact thatonce an image is formed, it is difficult to change; it becomesmore important for destinations to present the right imageand then maintain it.

Because the image that tourists hold of a destination willaffect tourists’ satisfaction with the travel experiences, theWOM communication that takes place after the trips aswell as the intention to return in the future, destinationmarketers should take a serious approach to manage theimage. Although it is not possible to control all theelements contributing to the shaping of the image of adestination, it is possible to manipulate some of them suchas advertising and promoting tourist attractions, organiz-ing cultural events that appeal to tourists, administeringservice quality provided by tourism infrastructure such ashotels, restaurants, tourist centers, retail establishments,etc.

Since image is modified by each new piece of informationor stimulus received by an individual, one’s own experienceor that of friends, acquaintances, or family will helpestablish more diversified, detailed and realistic image of adestination. Because tourists tend to rely more on thisimage for satisfaction evaluation and destination choicedecisions, all efforts should be aimed at improving thatexperience. To conclude, tourism destinations must takespecial care of the image that they attempt to convey andthe quality of the services and products that they offer, asall these will affect visitors’ satisfaction and their intentionsfor future behavior.

Secondly, destination managers should consider the roletourist satisfaction played in developing destinationloyalty. It is intuitively assumed that if tourists are satisfiedwith their travel experiences, they are more willing to revisita destination as well as spread positive WOM. This studyprovided empirical evidence supporting this assumption:satisfaction was found to directly affect destination loyaltyin a positive direction. Higher tourist satisfaction will leadto higher destination loyalty, which prompts tourists tovisit a destination again and/or recommend the destinationto others. Therefore, destination managers should focus onestablishing a high tourists’ satisfaction level so as to createpositive post-purchase tourist behavior and improve/sustain destination competitiveness.

Since attribute satisfaction affected destination loyaltyboth directly as an immediate antecedent and indirectlythrough overall satisfaction, its measurement and improve-ment are critical to destination managers. The specialcharacteristics of tourism determine that many elementsare involved in the formation of tourists’ satisfaction, fromthe providers of specific services of accommodation,transport, leisure, among others, to the tourism informa-tion offices, the local residents, natural and artificialresources, etc. The situations become even more compli-cated when a single unpleasant incident leads to a negativeoverall evaluation, depending on how important theincident is to the tourist. Therefore, in order to achieve a

high overall level of satisfaction, it is essential for all partiesinvolved to have smooth coordination and co-operationand be fully aware of the critical importance of providingquality services/products as well as diagnosing the servicequality. In addition, appropriate destination products andservices should be delivered to tourists in order to enhancedestination competitiveness.

6. Limitations and future research recommendations

The results presented in this study need to be qualified inlight of several limitations. First, the study was conductedin the summer, thus findings were limited to summertravelers. Tourists who travel in different seasons may formdifferent opinions of a destination. Seasonality restricts thegeneralizability of tourism research findings, and shouldalways be taken into consideration in the interpretationstage. To overcome this limitation, future researchers couldconduct similar surveys in different seasons. The surveyresults can then be compared to identify similarities anddifferences in them. Further, the population of this studywas limited to visitors of a tourist destination in thesouthern US. Therefore, the results from the study may notbe generalized beyond this population. Replicating similarstudies in other tourist destinations would be imperativefor increasing the generalizability of these findings.Secondly, overall satisfaction, repurchase, and referral

intention (used to infer destination loyalty) were allmeasured by a single question. The use of a multiple-itemmeasurement scale in future studies may enhance theinterpretation and prediction of overall satisfaction anddestination loyalty. For example, for destination loyalty,questions can be asked regarding tourists repurchase andreferral intentions in short term (1–2 years), medium term(3–5 years), and long term (5 and more years). Thedevelopment of more complete and psychometricallysound measures would strengthen the reliability of findingsand assist future tourism research projects with scale andtheory development about tourist satisfaction and loyalty.Thirdly, ‘destination image,’ ‘attribute satisfaction,’ and

‘overall satisfaction’ were studied as antecedents todestination loyalty. There might be additional factorsinfluencing and interacting with tourists’ loyalty. Futureresearchers are advised to investigate additional antece-dents of tourist loyalty. This may lead to the uncovering ofomissions and misrepresentation of the relationships testedin the current study and to further conceptual refinementand extension.In addition, since the survey was conducted by the staffs

working at different local destinations such as the welcomecenter, hotels/motels, and shops/galleries, it was not surethat all respondents would have completed their travelingexperiences with Eureka Springs while replying to thesurvey. Tourists’ perceptions of satisfaction or destinationimage may be under risk of being colored with theiradditional experiences in Eureka Springs. It is advisable forfuture studies to add a question in the survey instrument to

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636 635

determine at which traveling stage data were collected (e.g.during the trip or end of the trip).

Lastly, data collected from the present study wereneither experimental nor longitudinal. As such, the cause-and-effect relationships reported herein should be inter-preted with caution. Although SEM allows one topostulate causal relationships, the present study’s modelspecification was based on previous research and theory,not on the actual data. As a consequence, the cause-and-effect relationships suggested by the model in this studymay not represent the true causal nature of the relation-ships among the constructs. Future research will benefitfrom the collection of longitudinal data to more preciselymeasure change across time and the direction of causalityamong relationships. Ideally, this research would begintracking tourists from one trip until the next trip. Inaddition, it may be useful to manipulate factors of interestexperimentally, thereby enabling more definite conclusionsabout causal relationships to be drawn.

References

Alexandris, K., Kouthouris, C., & Meligdis, A. (2006). Increasing

customers’ loyalty in a skiing resort: The contribution of place

attachment and service quality. International Journal of Contemporary

Hospitality Management, 18(5), 414–425.

Anderson, E. W., & Sullivan, M. W. (1993). The antecedents and

consequences of customer satisfaction for firms. Marketing Science, 12,

125–143.

Anderson, J., & Gerbing, D. (1982). Some methods of respecifying

measurement models to obtain unidimensional construct measure-

ment. Journal of Marketing Research, 19, 453–460.

Backman, S. J., & Crompton, J. L. (1991). The usefulness of selected

variables for predicting activity loyalty. Leisure Science, 13, 205–220.

Baloglu, S., & McCleary, K. W. (1999). A model of destination image

formation. Annals of Tourism Research, 26(4), 868–897.

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable

distinction in social psychological research: Conceptual, strategic, and

statistical considerations. Journal of Personality and Social Psychology,

51, 1173–1182.

Beeho, A. J., & Prentice, R. C. (1997). Conceptualizing the experiences of

heritage tourists. Tourism Management, 18(2), 75–87.

Bigne, J. E., Sanchez, M. I., & Sanchez, J. (2001). Tourism image,

evaluation variables and after-purchase behavior: Inter-relationship.

Tourism Management, 22(6), 607–616.

Bloemer, J., & Ruyter, K. (1998). On the relationship between store image,

store satisfaction and store loyalty. European Journal of Marketing,

32(5/6), 10.

Bolton, R. N., & Drew, J. H. (1991). A multistage model of Customers’

assessments of service quality and value. Journal of Consumer

Research, 17(March), 375–384.

Bolton, R. N., & Lemon, K. N. (1999). A dynamic model of customers’

usage of services: Usage as an antecedent and consequence of

satisfaction. Journal of Marketing Research, 36(2), 171–186.

Bramwell, B. (1998). User satisfaction and product development in urban

tourism. Tourism Management, 19(1), 35–47.

Burns, A. C., & Bush, R. F. (1995). Marketing research. New Jersey:

Prentice-Hall.

Castro, C. B., Armario, E. M., & Ruiz, D. M. (2007). The influence of

market heterogeneity on the relationship between a destination’s image

and tourists’ future behavior. Tourism Management, 28, 175–187.

Chon, K. S. (1990). The role of destination image in tourism: A review and

discussion. Tourist Review, 45(2), 2–9.

Chon, K. S. (1991). Tourism destination image modification process.

Tourism Management, 12, 68–72.

Chon, K. S. (1992). The role of destination image in tourism: An

extension. Revue du Tourisme, 1, 2–8.

Churchill, G. A., Jr., & Surprenant, C. (1982). An investigation into the

determinants of customer satisfaction. Journal of Marketing Research,

19(November), 491–504.

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

Modeling destination adopters, inactives, and rejecters. Journal of

Travel Research, 36(1), 35–43.

Crompton, J. L., & Ankomah, P. K. (1993). Choice set propositions in

destination decisions. Annals of Tourism Research, 20, 461–476.

Cronin, J. J., Jr., Brady, M. K., & Hult, G. T. M. (2000). Assessing the effects

of quality, value and customer satisfaction on consumer behavioral

intentions in service environments. Journal of Retailing, 76(2), 193–218.

Danaher, P. J., & Arweiler, N. (1996). Customer satisfaction in the tourist

industry: A case study of visitors to New Zealand. Journal of Travel

Research, 35(1), 89–93.

Echtner, C. M., & Ritchie, J. R. B. (1991). The meaning and measurement

of destination image. The Journal of Tourism Studies, 2(2), 2–12.

Fakeye, 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), 10–15.

Fornell, C., Johnson, M. D., Anderson, E. W., Cha, J., & Bryant, B. E.

(1996). The American customer satisfaction index: Nature, purpose,

and findings. Journal of Marketing, 60(Oct.), 7–18.

Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation

models with unobservable variables and measurement error. Journal of

Marketing Research, 18(February), 39–50.

Fornell, C., & Wernerfelt, B. (1987). Defensive marketing strategy by

customer complaint management. Journal of Marketing Research,

24(November), 337–346.

Gartner, W. C. (1989). Tourism image: Attribute measurement of state

tourism products using multidimensional scaling techniques. Journal of

Travel Research, 28(2), 16–20.

Goodall, B. (1988). How tourists choose their holidays: An analytical

framework. In B. Goodall, & G. Ashworth (Eds.), Marketing in the

tourism industry: The promotion of destination regions (pp. 1–17).

London: Routledge.

Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998).

Multivariate data analysis with readings (5th ed). Englewood Cliffs,

NJ: Prentice-Hall International.

Hallowell, R. (1996). The relationship of customer satisfaction, customer

loyalty, profitability: An empirical study. International Journal of

Service Industry Management, 7(4), 27–42.

Hawkins, D. I., Best, R. J., & Coney, K. A. (1989). Consumer behavior:

Implications for marketing strategy (4th ed). Boston, MA: Richard D.

Irwin.

Hsu, C. H. (2003). Mature motor-coach travelers’ satisfaction:

A preliminary step toward measurement development. Journal of

Hospitality and Tourism Research, 20(10), 1–19.

Hughes, K. (1991). Tourist satisfaction: A guided cultural tour in north

Queensland. Australian Psychologist, 26(3), 166–171.

Jones, T. O., & Sasser, W. E., Jr. (1995). Why satisfied customers defect.

Harvard Business Review, 73(November/December), 88–99.

Joreskog, K. G., & Sorbom, D. (1995). LISREL 8: User’s reference guide

and LISREL 8: The SIMPLIS command language and PRELIS 2:

User’s reference guide. Chicago, IL: Scientific Software International.

Juaneda, C. (1996). Estimating the probability of return visits using a

survey of tourist expenditure in the Balearic Islands. Tourism

Economics, 2(4), 339–352.

Kotler, P., Bowen, J., & Makens, J. (1996). Marketing for hospitality and

tourism. New Jersey: Prentice-Hall.

Kozak, M. (2001). Repeaters’ behavior at two distinct destinations. Annals

of Tourism Research, 28(3), 784–807.

Kozak, M., & Rimmington, M. (2000). Tourist satisfaction with Mallorca,

Spain, as an off-season holiday destination. Journal of Travel Research,

38(1), 260–269.

ARTICLE IN PRESSC.G.-Q. Chi, H. Qu / Tourism Management 29 (2008) 624–636636

LaBarbera, P. A., & Mazursky, D. (1983). A longitudinal assessment of

consumer satisfaction/dissatisfaction: The dynamic aspect of the

cognitive process. Journal of Marketing Research, 20(November),

393–404.

Lounsbury, J. W., & Hoopes, L. L. (1985). An investigation of factors

associated with vacation satisfaction. Journal of Leisure Research,

17(1), 1–13.

Mayer, K. J., Johnson, L., Hu, C., & Chen, S. (1998). Gaming customer

satisfaction: An exploratory study. Journal of Travel Research, 37(2),

178–183.

Middleton, V. T. C., & Clarke, J. (2001). Marketing in travel and tourism

(3rd ed). Oxford: Butterworth-Heinemann.

Milman, A., & Pizam, A. (1995). The role of awareness and familiarity

with a destination: The central Florida case. Journal of Travel

Research, 33(3), 21–27.

Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed).

New York: McGraw-Hill.

Oh, H. (2000). The effects brand class, brand awareness, and price on

customer value and behavioral intentions. In Journal of Hospitality

and Tourism Research: Vol. 24(2). Advances in consumer behavior and

research in hospitality and tourism (pp. 136–162) (special issue).

Oh, H., & Parks, S. C. (1997). Customer satisfaction and service quality:

A critical review of the literature and research implications for the

hospitality industry. Hospitality Research Journal, 20(3), 35–64.

Oliver, R. L. (1980). A cognitive model of the antecedents and

consequences of satisfaction decisions. Journal of Marketing Research,

17(4), 46–49.

Oliver, R. L. (1993). Cognitive, affective, and attribute bases of the

satisfaction response. Journal of Consumer Research, 20(December),

418–430.

Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63,

33–44.

Oliver, R. L., & DeSarbo, W. S. (1988). Response determinants in satisfaction

judgments. Journal of Consumer Research, 14(March), 495–507.

Oppermann, M. (2000). Tourism destination loyalty. Journal of Travel

Research, 39(1), 78–84.

Petrick, J. F., Morais, D. B., & Norman, W. (2001). An examination of

the determinants of entertainment vacationers’ intentions to visit.

Journal of Travel Research, 40(1), 41–48.

Pizam, A., & Ellis, T. (1999). Customer satisfaction and its measurement

in hospitality enterprises. International Journal of Contemporary

Hospitality Management, 11(7), 326–339.

Pritchard, M., & Howard, D. R. (1997). The loyal traveler: Examining a

typology of service patronage. Journal of Travel Research, 35(4), 2–10.

Reichheld, F. F. (1996). The loyalty effect. Boston, MA: Harvard Business

School Press.

Reichheld, F. F., & Sasser, W. E. (1990). Zero defections: Quality comes

to services. Harvard Business Review, 68(September/October),

105–111.

Ross, G. F. (1993). Destination evaluation and vacation preferences.

Annals of Tourism Research, 20, 477–489.

Ross, L. D., & Iso-Ahola, S. E. (1991). Sightseeing tourists’ motivation

and satisfaction. Annals of Tourism Research, 18(2), 226–237.

Rust, R. T., & Zahorik, A. J. (1993). Customer loyalty, customer retention

and market share. Journal of Retailing, 69(2), 193–215.

Seaton, A. V., & Benett, M. M. (1996). Marketing tourism products:

Concepts, issues, cases. London: International Thomson Business

Press.

Shoemaker, S., & Lewis, R. C. (1999). Customer loyalty: The future of

hospitality marketing. International Journal of Hospitality Manage-

ment, 18, 345–370.

Sonmez, S. F., & Graefe, A. R. (1998). Determining future travel behavior

from past travel experience and perceptions of risk and safety. Journal

of Travel Research, 37(2), 171–178.

Spreng, R. A., Mankenzie, S. B., & Olshavsky, R. W. (1996). A re-

examination of the determinants of consumer satisfaction. Journal of

Marketing, 60(3), 15–32.

Taylor, S. A., & Baker, T. L. (1994). An assessment of the relationship

between service quality and customer satisfaction in formation of

consumers’ purchase intentions. Journal of Retailing, 70(2), 163–178.

Taylor, T. B. (1998). Better loyalty measurement leads to business

solutions. Marketing News, 32(22), 41–42.

Tse, D. K., & Wilton, P. C. (1988). Models of consumer satisfaction

formation: An extension. Journal of Marketing Research, 25, 204–212.

Woodside, A. G., & Lysonski, S. (1989). A general model of traveler

destination choice. Journal of Travel Research, 27(4), 8–14.

Yau, O. H. M., & Chan, C. F. (1990). Hong Kong as a travel destination

in Southeast Asia: A multidimensional approach. Tourism Manage-

ment, 11(2), 123–132.

Yoon, Y., & Uysal, M. (2005). An examination of the effects of

motivation and satisfaction on destination loyalty: A structural model.

Tourism Management, 26(1), 45–56.

Zeithaml, V. A., Parasuraman, A., & Berry, L. L. (1985). Problems and

strategies in services marketing. Journal of Marketing, 49(2), 33–47.