The preference analysis for tourist choice of destination: A case study of Taiwan

  • Published on

  • View

  • Download


  • ead,ad, Hng 4


    n a ra1. Staat intein the2005,

    economy from one that is industry oriented to one that is leisureand entertainment oriented, Taiwans government has committedto making tourism development a major policy. While Taiwanalready has so much to offer to international travelers - its natural

    their choices. The decision-making process leading to the nalchoice of a travel destination is a very complex process, andunderstanding what inuences a traveler to choose a destination isimportant in developing appropriate marketing strategies.Recently, the Tourism Bureau, in its effort to promote Taiwansinternational tourism brand imagewith Taiwan, Touch your heartas the international marketing slogan, has named 8 tourist spots asFocus on Taiwan Tourism-Flagship Tourist Attraction. We selected

    Contents lists availab

    Tourism Ma


    Tourism Management 30 (2009) 288297* Corresponding author. Tel.: 886 960007707.all-time record of 806 million people traveling internationally.Growth is expected to continue in 2007 at a pace of around 4%worldwide. As the worlds second fastest-growing region in therst 8 months of 2006 (8.3%), the Asia and Pacic region remainsstrong. Notwithstanding this fact, research into Asian travel moti-vations has not attracted as much attention as European and UStravel motivations (Kim & Prideaux, 2005). Empirical ndingsrelating to Asia regions remain insufcient. Taiwan, as part of theemergingmarket in tourism,may be a small island, but its economyhas a big impact internationally. Taiwan is the 7th largest tradepartner of the US, and is the worlds 14th largest trading entity in2005. In tandem with the tremendous diversication of Taiwans

    tourist destination. Much more collaboration between the publicand private sectors is expected with the implementation of thepolicy. According to the Bureau of Tourism, the year 2005 yieldeda bumper harvest for Taiwans tourism industry. Total visitorarrivals for 2005 reached more than 3.3 million, exceeding thetarget of 3.25 million that had been set for the year. Total annualamount in visitor expenditures in tourism is 4035 million USdollars, whereas the annual growth rate is 3.78%. The TourismBureau has launched a project entitled Doubling of Tourist ArrivalPlan (20022007) aimed at doubling the number of touristsvisiting Taiwan by year 2008. To achieve this goal, it is important tounderstand why people travel to Taiwan and what factors inuence1. Introduction

    The Asia Pacic region has beedestination in the world since 200TourismOrganization (2006) show thregion grew to a total of 578 millionup 4.5% from 553 million in those ofE-mail address: (Y.-F.

    0261-5177/$ see front matter 2008 Elsevier Ltd.doi:10.1016/j.tourman.2008.07.011pidly growing tourismtistics from the Worldrnational tourists in therst 8 months of 2006,a year which shows an

    beauty; a rich culture; amazing technological advancement; greatcuisine; and friendly citizens; in order to compete as a prime touristdestination in the world market, it needs to draw up a blueprint.The governments great commitment is reected in the Ministry ofTransportation and Communications decision to focus its newadministration on tourism. The policy will serve as a guideline indeveloping Taiwans new international image as a leisure andTOPSISTaiwan 2008 Elsevier Ltd. All rights reserved.DestinationAHP

    inbound tourists to Taiwan, price is the least important and Taipei 101 is the rst priority for travelers.Case Study

    The preference analysis for tourist choicof Taiwan

    Tzu-Kuang Hsu a, Yi-Fan Tsai b,*, Herg-Huey Wu c

    aDepartment of International Business, Chung Hwa University, 707, Section 2, WuFu Rob Institute of Technology Management, Chung Hwa University, 707, Section 2, WuFu RocDepartment of Accounting, Feng Chia University, 100 Wen Hwa Road, Seatwen Taichu

    a r t i c l e i n f o

    Article history:Received 6 July 2007Accepted 28 July 2008

    Keywords:Tourist motivation

    a b s t r a c t

    This study identies thepreferences of tourists forwas proposed and testedimportance of pre-selectedgiven destinations correspindicate that visiting friend

    journal homepage: www.eTsai).

    All rights reserved.of destination: A case study

    HsinChu, 300 TaiwansinChu, 300 Taiwan07, Taiwan

    ors that inuence the tourists choice of destination and evaluates theinations. A 4-level AHP model, consisting of 22 attributes on the 4th level,ng data collected from tourists visiting Taiwan to establish the relativectors (criteria). By using fuzzy set theory and TOPSIS, the preference of 8ing to each criterion can be evaluated and given nal ranking. Resultselatives and personal safety appear to be the 2 most important factors for

    le at ScienceDirect


    evier .com/locate/ tourmanthese 8 tourist spots to be the alternatives in our case study.

  • anagA 4-level Analytic Hierarchy Process (AHP) model, consisting of 22attributes on the 4th level, was proposed and tested using datacollected from tourists visiting Taiwan to establish the relativeimportance of pre-selected factors (criteria). By using fuzzy settheory and the technique for order preference by similarity to idealsolution (TOPSIS), the preference of 8 given destinations corre-sponding to each criterion are evaluated and given a nal ranking.

    Prioritizing factors and attributes affecting destination choice isa complex multi-criteria decision-making process. The AHP,a prevalent multi-criteria, multi-objective decision-makingprocess, is particularly suitable for situations where most of theessential data is subjective. It can be consistently introduced intothe priority setting and deals with decision problems involvingmultiple-criteria dimensions. AHP is unique in the sense that itrecognizes bias and inconsistencies in subjective judgments. Theseinconsistencies can be tested and improved, resulting in a moreconsistent nal ranking. For years AHP has been used in tourismplanning (Moutinho & Curry, 1994), and convention site selection(Chen, 2006); however, there have been no empirical studies usingAHP in destination choice. This study presents a decision-makingmodel based on AHP for destination choice; it provides not onlya general understanding of decision factors but also evaluates therelative weight of critical attributes affecting destination choice.The AHP converts individual preferences into ratio-scale weights;these resultant weights are used in ranking the alternatives andassisting the decision maker in making choices or forecasting anoutcome. The drawback of AHP is the necessity of pairwisecomparison that could result in a tedious comparison process ifthere are many alternatives to evaluate. Therefore, this studyutilizes TOPSIS to evaluate the alternatives. TOPSIS, developed byHwang and Yoon (1981), was based on the concept that theselected best alternative should have the shortest distance fromthe ideal solution and the farthest distance from the negative-idealsolution in a geometrical (Euclidean) sense. In other words, theideal alternative has the best level for all attributes considered,whereas the negative ideal is the one with all the worst attributesvalue. A TOPSIS solution is dened as the alternative that issimultaneously farthest from the negative-ideal and closest to theideal alternative. The TOPSIS has two main advantages: its math-ematical simplicity and very large exibility in the denition of thechoice set. When solving real-life problems, or representing realworld phenomena, linguistic variable usually appears to be animportant output of the process. The descriptions and judgmentson destination are in linguistic terms represented by fuzzynumbers in our study. The fuzzy set theory has been applied to theeld of management science; however, it is scarcely used in theeld of destination choice. Thus, this study that includes a fuzzymultiple-criteria decision-making process provides a coherentprocess for incorporating subjective views into an explicit decisionprocess.

    In Section 2 of this paper, we review the past studies on desti-nation choice and state the objective of the study. In Section 3, wediscuss the research methodology of evaluation. Section 4 presentsa case application, and the nal section presents the conclusion.

    2. Literature review and objective of this study

    2.1. Destination choice

    According to Dellaert, Etterma, and Lindh (1998), tourists deci-sions are complex multi-faceted decisions in which the choices fordifferent elements are interrelated and evolve in a decision processover time, and most studies of tourists travel choice address touristdestination choice as the key element in the travel decision-making

    T.-K. Hsu et al. / Tourism Mprocess. The decision-making process is inuenced by a number ofpsychological (internal) and non-psychological (external) variables,and consists of a number of different stages that are marked byspecic actions. Sirakaya andWoodside (2005) provided a compre-hensive qualitative review of the tourist decision-making literature,and integrated the main conceptual and empirical work that hasbeen reported in the tourism literature. According to their analysis,the destination choice set model developed by Um and Crompton(1990) is simpler and more theoretically and methodologicallysound than the others in tourism decision research. In this model,a tourists destination choice is made through a 3-stage sequentialand funnel-like process: a composition of awareness set (an initialset of destinations that a tourist is aware of at any given time), anevoked set (late consideration set), and nal destination choice. Theevoked set is developed fromthe awareness set. It consists of variousdestinations that people actively seek information about for alter-natives to best meet their needs. According to the choice set model,the destination should be included in each choice set stage in orderto be selected as a nal destination. The criteria that affect thisprocess include personal (push) factors, destination attributes (pullfactors), and constraints. Crompton and Ankomah (1993) suggestedthat one might use 2 or 3 criteria to reduce the number of alterna-tives from the awareness set to the evoked set; otherwise, therecould be too many attributes to compare. In addition, Lam and Hsu(2006) mentioned that the complex decision-making processleading to the choice of a travel destination had not been wellresearched. Past studies related to destination choice mainly focuson identifying important attributes affecting destination choice;professional judgment and factor analysis are the main methods(Goossens, 2000; Heung, Qu, & Chu, 2001; Kim & Prideaux, 2005;Kozak, 2002). These studies have contributed to identifying manyfactors; the 5-point (or 7-point) Likert scale was used for rating theimportance of each factor, and the factors extracted are arranged inorder of decreasing variance, but little has been learned about therelative importance of each one by pairwise comparison. Nicolauand Mas (2006) pointed out the choice of tourist destination thatdistinguishes between various approaches to the denition oftourist destination; they showed an overview of the empiricalevidence of destination choice with revealed and stated preferenceprobabilistic models respectively, as demonstrated in Tables 1 and 2of their paper. Eithermultinomial logitmodel or nestedmultinomiallogitmodel is used to investigate the choice of destination inmost ofthose papers. In addition to the above, the literature of destinationchoice is centered on the direct impact of destination attributes suchas prices and distance (Nicolau & Mas, 2006), climate (Hamilton &Lau, 2004), quality and pricing (Goossens, 2000). Furthermore,a number of studies were concerned with identifying pleasuremotivations which inuence the destination choice; however,empirical choice literature has devoted little attention to the impactof tourist motivations on the selection of destinations (Nicolau &Mas, 2006). In this study, destination choice can be conceptualizedas a tourists selection of a destination froma set of alternatives; thatselection is determined by various motivational factors. The studyincludes fuzzy multiple-criteria decision-making theory tostrengthen the rationality and comprehensiveness of the decision-making process. The AHP is a non-compensatory model becausea decision may be determined by an objects score on a singleattribute, irrespective of its score onother attributes. Fuzzy TOPSIS isa compensatory model because it entails that bad attribute valuescan be compensated by good values on their attributes, and thealternative chosen is superior to the other alternatives in the sum ofthe weighted utilities of all the attributes considered and leads tomaximization of utilities. Using AHP combining with fuzzy TOPSIS,which consider both psychological compensatory and non-compensatoryeffects in theprocess of decision, canbe considered asamore rational andmore efcientmodel for destination choice. The

    ement 30 (2009) 288297 289method applied in our study is another approach to destinationdecision-making, with a view to illustrating the same process in

  • anaga different way and achieving results that are useful and comple-ment, rather than contradict, one another.

    2.2. The concept of destination

    Since many studies on the choice of tourist destination distin-guished between various approaches to the denition of touristdestination, the destination should be dened. Traditionally,destinations are regarded as well-dened geographical areas, suchas country, an island or a town (Hall, 2000). Some researchersviewed destination as a product or a brand (Kozak, 2002; McIntosh& Goeldner, 1990; Yoon & Uysal, 2005); for example, Van Raaij(1986) viewed the travel destination as a product, which is partlygiven and partly man-made. The given part refers to naturalfeatures of tourist destinations such as climate, scenery, beaches,mountains, historic-culture buildings, and so forth. Whereas theman-made part refers to features such as hotels, package tours,transportation facilities, and facilities for sports and recreationwhich can all be adapted to customer preferences, subject to budgetrestrictions. Kim (1998: 340) stated that, A destination can beviewed as a uniquely complex product of the tourism industrycomprising, among other factors, an areas climate, infrastructureand superstructure, services, and natural and cultural attributes.Despite this complexity, it is nevertheless a product. Beerli andMortin (2004) pointed out tourist destinations must be conceivedas brands that have to be managed from a strategic point of view.Buhalis (2000) regarded destination as a dened geographicalregion which is understood by its visitors as a unique entity, witha political and legislative framework for tourism marketing andplanning; destinations offer an amalgam of tourism products andservices, which are subsumed under the brand name of the desti-nation. For the purpose of this study, we follow the denition ofBuhalis. Furthermore, the government is trying to provide Taiwanwith a recognizable international brand; therefore, in our study,Taiwan can be viewed as a corporate brand (or, more generally, anorganization brand) wherein the 8 domestic destinations aredifferent products of an organization. Thus, Taipei 101 can beviewed as a product brand. When an individual decides to traveloverseas, he/she has to choose which country to visit, while thechoice to visit cities or scenic spots may be made during the trip.This case study includes 2 stages; rst, we evaluate the preferenceof tourist who travels to Taiwan for motivational factors by usingAHP; second, we measure the preference of tourists for domesticdestinations with respect to each factor by using TOPSIS. Ourstrategy is to link the corporate brand with the product brand, andpromote the two brands together to potential tourists. In the study,the use of the AHP is more like a branding strategy for a corporatebrand, while the use of TOPSIS is more like a branding strategy fora product brand.

    2.3. Travel motivation

    Determining the factors that inuence peoples choice ofdestination is essential in developing appropriate marketingstrategies. Age, income, gender, personality, education, cost,distance, nationality, risk, and motivation, etc., are factors thataffect ones choice of destination. Of these factors, travel motivationhas been an important area of study in the tourism literature fordecades. As motivation is a dynamic concept, it may vary from oneperson to another, from one market segment to another, from onedestination to another, as well as from one decision-makingprocess to the next (Uysal & Hagan, 1993). It is therefore notsurprising that the concept of motivation is considered as anelement of market segmentation in tourism in many empirical

    T.-K. Hsu et al. / Tourism M290investigations (Kozak, 2002; Yavuz, Baloglu, & Uysal, 1998). Onepopular typology for understanding travel motivation is the pushand pull model by Crompton (1979). The push motivations havebeen thought useful for explaining the desire for travel while thepull motivations have been thought useful for explaining the actualchoice of destination. Crompton drew seven socio-psychological(push) motivations (escape, self-exploration, relaxation, prestige,regression, kinship-enhancement, and social interaction) and twocultural (pull) motivations (novelty and education). Uysal andJurowski (1994) summarized internal (push) and external (pull)motivators to travel. Internal motivators include desire for escape,rest, relaxation, prestige, health and tness, adventure, and socialinteraction. External motivators were based on attractiveness of thedestination, including tangible resources (beaches, recreationalactivities, and cultural attractions), and travelers perceptions andexpectations (novelty, benet expectations, and marketing image).In more recent studies, researchers have added shopping asa motivational characteristic of the destination (Hanqin & Lam,1999; Sirakaya, Uysal, & Yoshioka, 2003). Oh, Uysal, and Weaver(1995) noted good shopping was considered as a pull item, anattribute of the destination. There are still other important factorssuch as destination image, food, and safety. Milman and Pizam(1995) pointed out that destination image is the visual or mentalimpression of a place held by the general public. Goossens (2000)discussed in depth the role mental imagery plays in the content ofthe pull force. Eating is one of the most enjoyable activities thattourists undertake during their holidays (Ryan, 1997). Quan andWang (2004) found that food can act as either a primary orsecondary trip motivation and adds value to the image of a desti-nation. Safety is a major concern for tourists (Middleton, 1994).Heung et al. (2001) found that safety appeared to be the top priorityfor both Hong Kong and Taiwan travelers. A recent trend has shownthat people from developed countries are seeking treatment fromthe health professionals in developing countries. This preferentialshift is broadly termed as Medical Tourism which has becomeone of the star industries that attract attention. Therefore, weconsider medical treatment as a motivational factor in our case.The study of tourist motivations in this case is based on the conceptof Cromptons pushpull model. The main factors effecting desti-nation choice are grouped into categories. The push motivationrelating to internal forces consists of 4 factors (psychological,physical, social interaction, and seeking/exploration) and undereach a total of 11 items are included, while the pull motivationrelating to external forces consists of 2 factors (tangible andintangible). Tangible factors include 9 items, and intangible factorsinclude 2 items (see Fig. 1).

    2.4. Preference of tourist

    According to Goodall (1991), motivations initiate actions andguide satisfactory behavior but more precise lters of choices areexercised by decision makers preferences. Personal preferences,like motivations, may be both intrinsic, reecting individual likesand dislikes, and extrinsic, or socially conditioned. Pearce (1988)stated that preferences are more specic than motivations, and arerevealed by where travelers go and what travelers do. There havebeen several studies on the preference of travelers; these studiesused conjoint analysis (stated preference method) that has beensuccessfully applied in tourism as a technique to describe andforecast tourist choice behavior (Suh & McAvoy, 2005). However,preferences were not specically addressed in the modelsregarding the decision-making process of destination choice. FromTables 1 and 2 of Nicolau and Mas (2006), research has mostfrequently used destination attributes and personal characteristicsas explicative dimensions, and dened choice alternatives (desti-nations) based on destination type (nature parks) or administrative

    ement 30 (2009) 288297unit (countries) as being amongst the most important preferences.Decrop (2000) noted that preference is a special case of attitudes

  • hoic

    of d

    anagwhere product alternatives are compared and the one is chosenover the other. Tourists have to choose which of the destinationsthey wish to visit and which to skip. The result is the creation oftypical consumption patterns of the tourism product based onpreferences. Different statistical techniques are available todecompose the overall preferences or choices as provided by therespondents into utility weights associated with the factors. Forexample, MONANOVA, Logit model, Probit model, and non-metricregression are some techniques that can be used to analyse rankeddata. Which technique to use is typically determined by themeasurement scale of the response variable, the form of the utilityfunction, the availability of the software and the purpose of theresearch (Kroes & Sheldon, 1988). The preference analysis in thestudy includes tourists preference for destination choice factors(criteria) and tourism destinations (alternatives). This analysis isable to relate motivational factors to tourists preference ratings ofdestination and to understandwhat factors are driving preferences.It is important for potential tourists to make a decision on desti-nation choice so as to enable destination marketers to betterunderstand the competitive positioning of destination in the

    The c




    Rest &RelaxationMedicaltreatmentHealth &Fitness


    Internal force

    Visitingfriends and/or relativesMeetingnew people

    Level 1

    Level 2

    Level 3

    Level 4

    Fig. 1. The hierarchy

    T.-K. Hsu et al. / Tourism Mmarket place and combine this information with touristpreferences.

    3. Methodology of evaluation

    The evaluation procedure of this study consists of several steps.The detailed descriptions of each step are given in the followingsubsections.

    3.1. The Analytic Hierarchy Process (AHP)

    The AHP was developed by Saaty (1980). It rst breaks downa complex multi-criteria decision-making problem into a hierarchy,in which each level is composed of specic elements. The overallobjective is at the top, and the criteria, sub-criteria, and alternativesare at each descending level of the hierarchy. Saaty (1986) andHarker and Vargas (1987) provided a detailed statement of AHPaxioms that should be stresses in any AHP application. Once thehierarchy has been constructed, the decision-maker begins theprioritization procedure to determine the relative importance ofthe elements in each level. It is assumed that the decision makerhas thorough knowledge and understanding of these elements.Elements in each level are compared pairwise in terms of theirimportance to an element in the next higher level. Starting at thetop and working down, a number of square matrices, preferencematrices, are created in the process of comparing elements ata given level. Saaty (1980) describes the development 9-point ratioresponse scale that is integral to the AHP. The decision maker canexpress his/her preference between every 2 elements verbally asequally important, moderately more important, strongly moreimportant, very strongly more important, and extremely moreimportant. These descriptive preferences would then be translatedinto numerical ratings 1, 3, 5, 7, and 9, respectively, with 2, 4, 6, and8 as intermediate values for compromising two successive quali-tative judgments. This scale is insensitive to small change ina decision makers preferences, thereby minimizing the effect ofuncertainty in evaluation. After forming the preference matrices,the relative weights of the elements of each level with respect to anelement in the next higher level are computed as the componentsof the normalized eigenvector associated with the largest eigen-value of their comparison matrix; this value may be estimated bythe geometric mean of each row in the preference matrix (Saaty,1980). The composite weights of the alternatives are then deter-mined by aggregating the weights throughout the hierarchy. One

    e of the destination



    Transportation facilitiesFriendliness of peopleQuality & variety of foodAccommodation facilitiesPersonal safetyPriceCulture & historicalresources


    NoveltyseekingCultureexplorationAdventureseekingEnjoying nightlife &shopping


    External force

    Good shoppingEnvironmental safety& quality

    estination selection.

    ement 30 (2009) 288297 291important advantage of using AHP is that it canmeasure the degreeto which the pairwise comparisons are consistent. This measure,consistency ratio (CR), allows managers to detect inadvertentmisjudgments in comparisons. Not only does this reduce carelesserrors, but also it can reveal to the manager his/her own unsus-pected bias or exaggeration concerning one or more of thecomparisons. If the CR is larger than 0.1, it is recommended thatthe decision maker reevaluate the comparisons, since some of thejudgments are contradictory. Forman and Gass (2001) pointed outthe AHP is more than just a methodology for choice situation. It isnot just another analysis tool, it is theoretically sound, readilyunderstood, easily implemented, and capable of producing resultsthat agree with expectations. The 3 primary AHP functions ofstructuring complexity, measurement on a ratio scale, andsynthesis make the AHP suitable for a wide range of applications.Saaty sought a simpleway to deal with complexity: the hierarchicalstructuring of complexity into homogeneous clusters of factors. Heused paired comparisons of the hierarchical factors to derive(rather than assign) ratio-scale measure that can be interpreted asnal ranking priorities (weights). This is necessary because thepriorities (weights) of the elements at any level are determined bymultiplying the priorities of the element in that level by thepriorities of the parent element. This is particularly important if thepriorities are to be used not only in choice applications, but also for

  • where4 is the add operation of fuzzy numbers.

    tion selection, including internal and external factors. Various sets of

    anag3.2.3. Defuzzication of fuzzy averageThe aggregation dened by a triangular average number very

    often has to be expressed by a crisp value which represents best thecorresponding average. This operation is called defuzzication. Theoperation defuzzication cannot be dened unique (Bojadziev &other types of applications such as forecasting. An equally impor-tant function is the AHPs ability to measure and synthesize themultitude of factors in a hierarchy; there is no other methodologythat facilitates synthesis as does the AHP.

    3.2. Fuzzy set theory

    The concept of a fuzzy set was introduced by Zadeh (1965). Heintroduced the notion of a fuzzy set and some of the ideas thatplayed a fundamental role in the evolution of the fuzzy settheory. Zadeh (1976) presented further ideas concerning fuzzydecision-making and fuzzy optimization. The applications of thefuzzy set theory in this study are elaborated in the followingsubsections.

    3.2.1. Linguistic variableA linguistic variable is a variable with linguistic expression as its

    values. For example, in our case, the respondents were asked toevaluate the given destinations with respect to each criterion byusing a linguistic variable. The possible value for this variable couldbe any of the following: very good, good, fair, poor, verypoor. Each linguistic term can be characterized by a triangularfuzzy number of representing its approximate value range between0 and 10. These numbers can be further assigned as lower value,modal value and upper value, denoted as (l,m,u), where0 lm u10, m is the most likely value of the linguistic term,and l, u are lower and upper bounds, respectively, to reect thefuzziness of the term.

    3.2.2. Fuzzy numbersA fuzzy number is a convex fuzzy set, characterized by a given

    interval of real numbers, eachwith a grade of membership between0 and 1. In this study, triangular fuzzy number will be used in ourmodel. The membership function is dened as:



    x1m1 l x mxumu m x u0 otherwise


    For example, in the context of this study, the respondent believesthat Taipei 101 is a very good place to escape, the respondent cansubjectively assume his/her personal range of the linguistic variable(very good) (8,9,10), and the membership functions can becalculated according to the Eq. (1).

    In order to form the decision matrix, assume that we have pdecision makers as responsible for evaluating m alternatives withrespect to n criteria. Consider p triangular numbers Dkij l

    kij ;

    mkij ; ukij , where k 1, 2,.,p, i 1, 2,.,m, j 1,2,.,n.

    Using addition of triangular numbers and division by a realnumber gives the triangular averageDij (Bojadziev & Bojadziev,1997: River Edge Decrop, 2000: New York; Hall, 2000: Harlow)

    Dij 1p

    D1ij 4,,,4D




    lkij ;1p


    mkij ;1p




    T.-K. Hsu et al. / Tourism M292Bojadziev, 1997: 69), in our study the defuzzied value of a fuzzynumber can be obtained from the formula:AHP. The highest level of the hierarchy is the overall goal. Under theoverall goal, the 2nd level represents the factors affecting destina-Xij lij 2mij uij



    3.3. The Technique for Order Preference by Similarity to IdealSolution (TOPSIS)

    Once the weights of criteria are obtained, the TOPSIS approachcanbeproposed for conducting the rankingprocess. TOPSISassumesthat each attribute has a tendency towardmonotonically increasingor decreasing utility. Therefore, it is easy to locate the ideal andnegative-ideal solutions. The TOPSIS method evaluates the decisionmatrix, Xwhich refers tomalternatives and is evaluated in termsofncriteria, an element xij of X is the performance measure of the ithalternative in terms of the jth criterion from equation (2). Then theTOPSIS method consists of the following steps.

    Step 1: Construct the normalized decision matrix R: an elementrij of R is calculated as rij xij=

    Pmi1 x2ij

    qwhere i 1,2,.,m;

    j 1,2,.,n.Step 2: Construct the weighted normalized decision matrix V:

    a set of weights W w1;w2; ,,,;wn,Pn

    j1wj 1, from thedecisionmaker is accommodated to the decisionmatrix in this step.Anelement vijofV is calculatedas vij wjrij, i 1, 2.,m; j 1, 2.,n.

    Step 3: Determine the ideal A*and the negative-idealA solutions:

    A* maxi

    vijjJ; min

    ivijjJ0 for i 1;2;/;m

    A mini

    vijjJ; max

    ivijjJ0 for i 1;2;/;m

    whereJ {j 1,2,.,nj j associated with the benet criteria}J0 {j 1,2,.,nj j associated with the cost criteria}Step 4: Calculate the separation measure: the separation of

    each alternative from the ideal one is then given by

    S*i Pn

    j1vij A*2q

    i 1;2;/;m:Similarly, the separation from the negative-ideal one is given by

    Si Pn

    j1vij A2q

    i 1;2;/;m:Step 5: Calculate the relative closeness to the ideal solution: the

    relative closeness of alternativeAiwith respect to the ideal solutionA*is dened as C*i Si =S*i Si 0 C*i 1 i 1;2;/; m. It isclear thatC*i 1 if and only if Ai A*, and C*i 0 if and only ifAi A.

    Step 6: Rank the preference order: a set of alternatives can nowbe preference ranked according to the descending order of C*i .

    4. Case application

    Using Taiwan as the application case, the procedures of researchare shown as follows.

    4.1. Developing the evaluation hierarchy

    The key factors consist of 22 items that were used to constructa 4-level hierarchy, building on the tourist decision-making andrelated literature, consulting with experts (both professional andacademic sections of the tourism sector), and based on 4 axioms of

    ement 30 (2009) 288297criteria associated with each factor in the 2nd level are linked to the3rd level. Various sets of sub-criteria (attributes) associated with

  • each criterion in the 3rd level are linked to the 4th level. As seen inFig. 1, there are a total of 22 attributes on the 4th level.

    4.2. The alternatives

    A brief description of 8 tourist spots is set forth below, thelocation of which is shown in Fig. 2.

    Taipei 101, located in Xinyi District of Taipei, has 101 storiesabove ground (hence the name) and 5 basements. The structuralshowpiece built in segments inspired by bamboo, holds thefollowing records: ground to structural top, ground to the highestoccupied oor, fastest ascending elevator speed, setting a record of37 seconds for a ride from the 5th oor to the 89th oor and thelargest count-down clock on New Years Eve. Shopping at the Taipei101 is a satisfying experience especially for well-heeled fashion-istas, who enjoy checking out the upscale boutiques of all the majorinternational fashion brands and labels. The Page One Bookstoreboasts of having the largest number of English-language titles inTaiwan. Jasons Market Place offers a selection of premium beef,international grocery items, and wines as no other place does. The

    Kenting National Park

    T.-K. Hsu et al. / Tourism Managfood court in the basement provides a good variety of eatingoptions for the daily crowd in Xinyi District. More elegant diningoutlets open on the 85th to 88th oors to serve Chinese and Italiancuisines. Alain Robert, dubbed the French Spiderman, with hisdaredevil solo climbing act puts Taipei 101 in the world newsspotlight. In addition to Taipei 101, the National Palace Museum, setin the hills north of Taipei, houses more than 650 000 pieces ofChinese bronze, jade, calligraphy, painting, and porcelain, is rankedas one of the worlds top four museums. The collection, estimatedto be one-tenth of Chinas cultural treasures, is so extensive thatonly 1% is on display at any time.

    Sun Moon Lake is Taiwans largest natural lake, the pearl ofCentral Taiwan, with its eastern part shaped like the sun, and itswestern part shaped like a crescent moon, hence the name. Thebeauty of the lake is accented by the surrounding mountains whichappear dark with vegetation and rise from 600 to 2000 m. Its mainattractions include the Formosan Aboriginal Culture Village, whichoffers a variety of educational and recreational activities; Sun MoonVillage, where the Shao People live, they are the smallest pop-ulation (about 400) among the 13 aboriginal tribes in Taiwan; thePuli Winery, known for its pleasant wine; the Chung-Tai Temple,which is the largest Buddhist temple in Southeast Asia; and the

    National Palace Museum

    Sun Moon Lake

    Taroko National Park

    Yushan National Park


    Love River

    Taipei 101

    Pacific Ocean


    an S

    traitBashi ChannelFig. 2. Location of 8 local destinations.Taiwan Endemic Species Research Institute, which is an experi-mental and research organization dedicated to rehabilitation ofTaiwans endemic, rare and endangered species.

    Taroko National Park, in the eastern part of Taiwan, covers anarea of more than 92000 ha. It has many high mountains and steepgorges, with many peaks towering above 3000 m in height. Thespectacular Taroko Gorge is the largest in the world. The manywaterfalls, diverse forms of plant and animal life, and the indige-nous Atayal people, together create the rich texture of this uniquenatural ecosystem.

    Yushan National Park is located in the center of subtropical Tai-wan. However, due to the mountainous terrain, the elevation variesfrom 300 to 3952 m. Thus, it has 3 climatic zones: subtropical,temperate, and frigid. It is the largest national park in Taiwan; itsrange covers one-third of the noted mountains all over the islandand extends across 4 counties. The area encompasses Mt. Jade, thehighest mountain in North-East Asia, and is an important water-shed for most of the main river systems of central, southern, andeastern Taiwan. There is also a hot spring here.

    Alishan is an internationally well-known tourist scenic spot.AlishanMountain Railway is one of only 3 alpine railways left in theworld. It carries people through 4 climatic zones: tropical,subtropical, temperate, and frigid; and each section of the linepasses by truly unique scenery. One can experience a corkscrewride as it winds its way up one of the mountains along its line. Itends up at the Alishan National Scenic Areawhere one can view thespectacular sunrise, sea of clouds, sacred trees, owers, andmountain pathways. The Bunun are one of the most energetic,widespread, and best adapted for the mountain climate among thenative tribes of Taiwan. Half of them dwell in mountainous areasaround an altitude of 1500 m.

    Love River is located in Taiwans 2nd biggest city, Kaohsiung thatis preparing to host the 2009 World Games; the Love River isconsidered the soul of Kaohsiung City. The scent of coffee, thesound of music, the sight of the Love Boat, and the museums onboth sides of the riverbank accompany the spectacular river view.At night, the stunning view of Love River with its colorful lights isjust another attraction. Different water sport competitions,appealing to sailing, windsurng, and regatta enthusiasts havelikewise transformed the river into a popular city arena for excitingspectator events in southern Taiwan.

    Kenting National Park, Taiwans rst national park, is located atthe southern tip of Taiwan. Of the 6 national parks in Taiwan, it isthe only one that encompasses coastal waters. Under the waves areclear, sediment-free waters and one will nd a magnicentunderwater coral palace in which coral reef sh of every color andall sorts of shellsh and algae thrive. Its colossal waves are perfectfor surng and a suitable area for snorkeling and diving. The richora of the park support a variety of wildlife, and the park is alsorich in precious cultural heritage sites. The park includes moun-tains, forests, pasture, lake, sand dunes, limestone caves, beaches,and fringing reefs and is famous for its coral reef cliffs.

    4.3. Method

    4.3.1. Questionnaire designThe questionnaire was composed of two parts. Part one con-

    sisting of 22 attributes was designed to consider a tourists moti-vations for traveling to a destination. The respondents were askedto assign corresponding numerical values based on the relativeimportance of the attribute with respect to the overall goal ofselecting the destination. For example, when asked With respectto internal force and external force, which is more important to youin selecting a travel destination? the verbal judgment from equally

    ement 30 (2009) 288297 293important to extremely more important was then rated with a scalerange of 19. After doing all pairwise comparisons at level 2, the

  • anagpairwise comparison matrix is constructed. Similarly, the pairwisecomparison procedure is then applied to all factors with respect tothe level 3 and then level 4. A brief description of the 22 items in thequestionnaire is also given to provide a better understanding of theanalysis. In part two, the respondents were asked to evaluate the 8given local destinations with respect to each criterion by usinglinguistic variables very good, good, fair, poor, or verypoor. In order to establish the membership function associatedwith each linguistic variable, the respondents were asked to indi-cate with numbers within the scale range of 010 correspondent toeach linguistic variable subjectively and further assigned thesenumbers as lower value, modal value and upper value. For example,the respondents were asked, How would you evaluate Taipei 101with respect to escape? Would you say very good, good, fair,poor, or very poor?, then the respondents were asked to assigna value for very good, good fair, poor and very poorindicated by the numbers from 0 to 10 subjectively and assigned aslower value, modal value, and upper value.

    4.3.2. Data collectionAsurveywas conductedbyusinga convenience samplingapproach

    during summer vacation from June to September 2006. Data weregathered in the lounges of the international tourist hotels and the localhotels proposed by the tour guides working for the travel agenciesspecializing in local tours for overseas travelers at Taipei, Kaohsiung,andTaichungcity. These3citieswerechosenrstlybecause theyare, indescending order, the 3 biggest cities of Taiwan, located in the north,south and central metropolitan areas of Taiwan, respectively, wheretravelnetworksarerelativelywelldeveloped inaddition tobeinghometo important events, festivals, exhibitions, conferences, and concerts.Second, of the two International Airports in Taiwan, one is located inKaohsiung and the other is in Toaynan (the latter is very close toTaipeicity).Bothserveasa touristgatewaytoTaiwan.Lastly,3outof8selecteddestinations are located in central Taiwan, 2 out of 8 are inTaipei, and 1is in Kaohsiung. The other 2 use Taipei or Kaohsiung as a gateway. Dueto the sampling cost, only individuals who are 20 years or up, withoverseas travel experience and knowledge about 8 selected localdestinations andhad traveled to at least 3 (including1national park) ofthe 8 destinations are selected. People from overseas who had spentmore than a year in Taiwan were not considered tourists. OverseasTaiwanese and Mainland Chinese (including people from Hong Kong/Macao) were excluded. Only one person can be chosen from a travelparty. Therefore, inbound foreign visitors on vacation and/or on busi-ness who had spent their spare time sightseeing during their visit toTaiwan are included as our population. The reason for restricting thequalications of respondents was that we wanted the opinion ofexperienced travelers who were knowledgeable about the entirecriteria and destinations to be evaluated, as required by AHP method.Among those who knew the 8 destinations well and were able toremember and compare them appropriately, a total of 36 (a responserate of 70%) replied to the questionnaire by face-to-face interview. Inorder to create a representative and accurate sample of the individualforeign tourists who visited Taiwan, 4 out of the 36 who had a consis-tency ratio greater than 0.1 when applying the AHP approach werediscarded. Taiwan is not so well known to foreigners. A number oftourists come to attend some events or festivals; however, only a fewhad thorough knowledge and understanding about all 8 destinations.Although the sample might not necessarily represent all foreign tour-ists in Taiwan, it did not exhibit any systematic bias that might createdifculties in drawing accurate comparisons of the 8 destinations. Byusingan expert survey,wewereable to followthebasicconceptionofthe AHP methods.

    4.3.3. The weights of evaluation criteria

    T.-K. Hsu et al. / Tourism M294Once the hierarchy structure has been constructed, the priori-tization procedure begins to determine the relative importance ofthe elements in each level. Computations were done by computersoftware; Expert Choice was used in our case. Table 1 shows therelative weights for each criterion.

    4.3.4. Overall preference measure of destinationsThe preference of alternatives corresponding to each criterion

    evaluatedby respondents ismeasuredas fuzzynumberwith triangularmembership function. We use Eq. (2) to defuzzify the fuzzy numbers(see Table 2) into crisp numbers so as to conduct TOPSIS rankingprocedure. The nal ranking of destinations is shown in Table 3.

    There are 3 reasons for choosing TOPSIS as the ranking tech-nique instead of using a full AHP method in the study. First, theconcept is straightforward and easy to use. Second, it avoids anunreasonably large number of pairwise comparisons. Lastly, deci-sion makers ranking n alternatives in n (n 1)/2 pairwisecomparisons using the 9-point semantic scale are unsuitable foruse with the eigenvector method when more than 4 alternativesare compared, unless the adjacent alternatives are closer in valuethan slightly more important (Kuenz Murphy, 1993).

    4.4. Empirical results and discussions

    Table 1 shows the relative weights for each criterion. Of the 22motives, visiting friends/relatives, personal safety, escape, rest andrelaxation, destination image, and environment safety and qualityappear to be the 6 most important factors for inbound tourists toTaiwan. Among the internal forces, the most important factors areescape, rest and relaxation, visiting friends and/or relatives. That is,inbound visitors were seeking rest (both physically and mentally)offered by the destination. This indicated that even though Taiwanis a small island; it is a good place for relaxation and rest. Somestudies showed that visiting friends/relatives did not seem to bea priority during a vacation (Heung et al., 2001; Mok & Armstrong,1995). Their studies focused on Hong Kong outbound travelers andJapanese leisure travelers to Hong Kong, respectively. On thecontrary, visiting friends and/or relatives appears to be an importantfactor in our study. That is, due to the political situation, Taiwan isnot so well known to foreign tourists, especially European, Aus-tralia, and New Zealand travelers. Individuals came to Taiwan forthe purposes of visiting friends whom they have known or met inthe past in other countries or through the Internet in recent years.Among the external forces, the most important factors are personalsafety, destination image, and environment safety and quality; theresults support previous tourism studies, which suggest that bothpersonal and environment safety are common motives for travel. Ofthe 22 motives, price is the least important factor. This is probablybecause beautiful scenery, quality of food, accommodation facili-ties, different cultures, and discovering new places are moreimportant factors than price when tourists are at the initial stage ofselecting a destination for taking a vacation. This is consistent withthe studies of Stevens (1992) and Nicolau and Mas (2006).

    The preference for 8 destinations and the nal ranking are asshown in Table 2 and Table 3. The nding shows that Taipei 101 isthe rst priority for travelers, followed by Kenting National Park, SunMoon Lake, National Palace Museum, Love River, then come TarokoNational Park, Alishan, and Yushan National Park. It is no doubt thatTaipei 101 is ranked the number one favorite local destination. Thatis, Taipei 101 is worthy of visiting because it is the highest buildingin the world, is located in the business section of Taipei city, andalso a Taipei landmark. As shown in Table 2, Taipei 101 is consideredthe most desirable destination with respect to self-actualization,meeting new friends, medical treatment, novelty seeking,enjoying night life, transportation facilities, quality and variety offood, accommodation facilities, good shopping, personal safety

    ement 30 (2009) 288297and benet expectations. One matter which merits mentioning isthat a number of tourists visit Taiwan for medical treatment since

  • Table 1Weights of the factors and criteria

    Goal Factors Criteria Global Sub-criteria (Local) Global Rank

    Choice of Destination Internal (0.521) Psychological 0.141 Escape (0.716) 0.101 3(0.263) Self-actualization (0.284) 0.040 10Physical (0.195) 0.104 Rest and relaxation (0.613) 0.064 5

    Medical treatment (0.101) 0.011 21Health and tness (0.286) 0.030 16

    Social interaction (0.294) 0.157 Visiting friends/relatives (0.692) 0.109 1Meeting new people (0.308) 0.048 8

    Seeking/exploration (0.248) 0.132 Novelty seeking (0.248) 0.132 11Culture exploration (0.226) 0.030 15Adventure seeking (0.114) 0.015 19Enjoying night life (0.382) 0.051 7

    External (0.488) Tangible (0.764) 0.355 Transportation facilities (0.118) 0.042 9Friendliness of people (0.043) 0.015 20Quality and variety of food (0.100) 0.036 12Accommodation facilities (0.097) 0.035 13Personal safety (0.295) 0.105 2Price (0.031) 0.011 22Culture and historical resources (0.075) 0.027 18Good shopping (0.087) 0.031 14Environmental safety and quality (0.154) 0.055 6

    Intangible (0.236) 0.110 Destination image (0.740) 0.081 4Benets expectations (0.260) 0.029 17

    T.-K. Hsu et al. / Tourism Management 30 (2009) 288297 295this country is equippedwith world-class medical technologies andmedical charges are relatively low as compared to other countries.For example, the charges for assisted reproduction, the surgery feefor liver transplant (Taiwan was the rst Asian country to performsuch surgery successfully), for hip joint replacement, laser eyesurgery, dental treatment, cosmetic surgery, health examinations,and cancer screening are considerably competitive. Moreover, thehepatic medicine in Taiwan is highly acclaimed worldwide. Asa number of hospitals and clinics in Taiwan which provide specialmedical services for international visitors, such as Chang GungMemorial Hospital, Cathay General Hospital, Shin Kong MemorialHospital, and National Taiwan University Hospital are located inTaipei city, not far from Taipei 101. This should probably explainwhyTaipei 101 is considered the desirable destination with respect tomedical treatment. Kenting National Park is a very good place withrespect to friendliness of people (got 6.95 in preference measure);

    it is quite obvious that people in southern Taiwan are very

    Table 2Overall preference measures of destinations

    Criteria Taipei 101 National Museum Sun moon Lake

    Escape 6.23 5.42 7.16a

    Self-Actualization 7.16a 6.89 6.30Rest/relaxation 5.38 5.05 7.59a

    Medical treatment 6.40a 4.41 4.30Health and tness 5.05 4.10 6.57Visiting friend/relative 7.96 5.54 8.54a

    Meeting new people 7.45a 5.55 5.98Novelty seeking 8.04a 6.08 5.35Culture exploration 6.27 7.87a 6.54Adventure seeking 3.97 3.43 5.96Enjoying night life 8.56a 5.53 4.49Transportation facilities 8.60a 8.03 6.20Friendliness of people 6.48 6.52 6.68Quality/variety of food 8.85a 6.41 5.70Accommodation 7.81a 7.26 6.67Environment safety 6.10 6.38 6.98Personal safety 7.36a 7.29 5.80Price 4.39 5.12 5.55Culture and historical 6.11 8.47a 6.78Good shopping 8.04a 5.99 4.89Destination image 7.78 7.27 7.94a

    Benets expectations 7.67a 6.80 5.41

    a Is the best preference out of the 8 destinations.enthusiastic, and give tourists a very warm welcome. The samehospitality awaits tourists who visit Taroko National Park (got 6.93),located in eastern Taiwan, facing the Pacic Ocean. Eastern Taiwanhas a small population, is inhabited by many ethnic groups; theyare polite and follow their custom of sharing to show theirfriendliness. Sun Moon Lake is a very good place with respect toescape, relaxation, visiting friends, and destination image. Inour study, Yushan National Park, a mountainous national park, rankslast (or is the least preferred destination) despite being loved bymountain climbers who come here one after another.

    The push and pull items in the study do not include all possibledimensions of motivation and here theoretical considerations and theavailability of variables are decisive. Future analysis for other domesticdestinations within other countries (e.g. countries in the Asian Pacicregion) can improve the reliability of the ndings by adding severalmore factors or using a different set of factors contributing to desti-

    nation choice. Since the government in Taiwan is promoting these 8

    Alishan Yushan Park Taroko Park Love River Kenting Park

    6.80 5.94 7.08 5.99 6.926.09 5.22 6.41 5.66 6.457.06 6.13 7.05 6.36 7.304.34 3.02 4.09 5.38 4.736.98 6.34 7.15a 5.17 6.577.01 5.88 6.84 6.86 7.425.55 4.37 5.61 5.76 6.805.06 4.16 5.22 5.61 6.296.41 5.42 6.28 5.66 6.456.56 6.30 6.67 4.94 6.73a

    3.95 3.43 4.41 6.45 5.805.28 4.04 5.45 6.78 6.706.84 5.95 6.93 6.33 6.95a

    5.58 4.28 5.80 7.16 6.885.72 4.31 6.63 7.36 7.526.81 6.07 7.20a 6.24 6.735.02 4.55 5.38 5.80 6.595.64 5.88 6.05a 5.30 5.306.72 6.22 6.63 5.60 6.494.79 3.63 5.40 6.13 6.117.06 6.53 6.92 6.38 7.145.41 4.60 5.32 6.48 6.67

  • anagdomestic destinations,wechoose themas thealternatives in the study.According to Swait (2001) the true choice set of travelers is normallyunknown to the analysts, as only the chosen alternative can beobserved. Biases in the choice set can occur, if an alternative is presentthat in reality is impossible for the traveler to choose. The AHP proce-dures are applied to individual decision setting in this study; it can alsobe applied to group decision setting.

    5. Conclusion

    From a methodological point of view, the results of this studythat adopted the AHP approach reveal that travel motivations havean inuence on choice of destinations. The ndings demonstratethat the AHP approach is a useful tool to help support a decision indestination choice. It integrates the opinion and evaluation ofexperts and devises the complex decision-making system intoa simple element hierarchy system. As visitors become bettereducated and informed about tourist destinations, the AHPmethodshould be a more accurate measurement of the differences inattribute preference for tourists and the results of this approachwill be better than that of other approaches. The ranking of the 8destinations in terms of their overall performance in our study issomewhat similar to the 2005 annual survey report of the TourismBureau on the ranking of favorite scenic spots of inbound visitors.This nding demonstrates that TOPSIS is an adequate tool forselecting the best alternative. Nevertheless, this study also has itslimitation since it covers only the selected samples from touristsvisiting Taiwan, and the data were collected over a short period oftime. To verify the validity of this studys nding, results should becompared with those from future studies that use differentsamples. Second, the use of AHP in selecting a destination requiresthat respondents had the traveling experience, knowledge andunderstanding about all the destinations to be evaluated. Therefore,the sample size is limited.

    In conclusion, developing a model that ts all decision-makersand every decision situation may not be realistic (Sirakaya &Woodside, 2005). Each destination has different roles and level ofinvolvement with the tourism industry. The study does not attemptto recommend an ideal destination position, but rather providesdestinationmarketers with information that can lead to destination

    Table 3Final ranking of destinations

    Destination Rank Similarity to ideal solution

    Taipei 101 1 0.75Kenting National Park 2 0.63Sun Moon Lake 3 0.55National Palace Museum 4 0.48Love River 5 0.47Taroko National Park 6 0.43Alisan 7 0.37Yushan National Park 8 0.16

    T.-K. Hsu et al. / Tourism M296repositioning. One of the targets of Doubling of Tourist ArrivalPlan is to raise the total number of visitor arrivals to at least 5million by 2008. The governments institution of a 5-day visa-freeentry program to 15 countries also helped to increase arrivals toTaiwan. In addition, the government continued offering free half-day tours to transit passengers and free full-day tours to interna-tional participants in conferences, along with their accompanyingpersons. However, destination marketing should not only aim atincreasing the number of tourists traveling to a region, but also aimto facilitate sustainable tourism development. The study can helpindividuals make decisions during their visit to Taiwan, and allowdestination marketers to better understand how tourists andpotential tourists view and choose their destinations. The empiricalexamination of tourist motivation will help to identify the

    27, 982996.Oh, H., Uysal, M., & Weaver, P. (1995). Product bundles and market segments basedon travel motivations: a canonical correlation approach. International Journal ofHospitality Management, 14(2), 123137.

    Pearce, D. (1988). Yourism time budgets. Annuals of Tourism Research, 15(1), 106121.Quan, S., & Wang, N. (2004). Towards a structural model of the tourist experience:

    an illustration from food experience in tourism. Tourism Management, 25(3),297305.attributes that are to be promoted so as to match tourist motiva-tions. In addition, this study can provide important information tothe international tourist who still thinks that Taiwan is merely anindustrial island without any status in the realm of tourism.


    Beerli, A., & Mortin, J. D. (2004). Tourists characteristics and the perceived image oftourist destinations: a quantitative analysis a case study of Lanzarote, Spain.Tourism Management, 25, 623636.

    Bojadziev, G., & Bojadziev, M. (1997). Fuzzy logic for business, nance, and manage-ment. NJ: World Scientic.

    Buhalis, D. (2000). Marketing the competitive destination of the future. TourismManagement, 21, 97116.

    Chen, C. F. (2006). Applying the analytical hierarchy process approach to conventionsite selection. Journal of Tourism Research, 45, 167174.

    Crompton, J. (1979). Motivations for pleasure travel. Annual of Tourism Research, 6,408424.

    Crompton, J., & Ankomah, P. (1993). Choice set propositions in destination decision.Annuals of Tourism Research, 20, 461476.

    Decrop, A. (2000). Tourists decision-making and behavior processes. In A. Pizam, &Y. Mansfeld (Eds.), Consumer behavior in travel and tourism. NY: The HaworthHospitality Press.

    Dellaert, B. G. C., Etterma, F., & Lindh, C. (1998). Multi-faceted tourist travel deci-sions: a constraint-based conceptual framework to describe tourists sequentialchoice of travel components. Tourism Management, 19(4), 313320.

    Forman, E. H., & Gass, S. I. (2001). The analytic hierarchy process an exposition.Operations Research, 49(4), 469486.

    Goodall, B. (1991). Understanding holiday choice. In C. Cooper (Ed.), Progress intourism, recreation and hospitality management (pp. 103133). London:Belhaven.

    Goossens, C. (2000). Tourism information and pleasure motivation. Annals ofTourism Research, 27, 301321.

    Hall, C. M. (2000). Tourism planning: Policies, processes, relationships. U.K.: PrenticeHall.

    Hamilton, J. M., & Lau, M. A. (2004). The role of clime information in tourist desti-nation choice decision-making. Working Paper FNU-56. Center for Marine andClimate Research. Hamburg, Germany: Hamburg University.

    Hanqin, Z., & Lam, T. (1999). An analysis of Mainland Chinese visitors motivationsto visit Hong Kong. Tourism Management, 20, 587594.

    Harker, P. T., & Vargas, L. G. (1987). The theory of ratio scale estimation: Saatysanalytic hierarchy process. Management Science., 36(3), 13831403.

    Heung, V. C. S., Qu, H., & Chu, R. (2001). The relationship between vacation factorsand socio-demographic and traveling characteristics: the case of Japaneseleisure travelers. Tourism Management, 22(3), 259269.

    Hwang, C. L., & Yoon, K. (1981). Multiple attribute decision making. Berlin/Heidel-berg/New York: Springer-Verlag.

    Kim, H. B. (1998). Perceived attractiveness of Korean destinations. Annual of TourismResearch, 25(2), 340361.

    Kim, S. S., & Prideaux, B. (2005). Marketing implications arising from a comparativestudy of international pleasure tourist motivations and other travel-relatedcharacteristics of visitors to Korea. Tourism Management, 26, 347357.

    Kozak, M. (2002). Comparative analysis of tourist motivations by nationality anddestinations. Tourism Management, 23, 221232.

    Kroes, E. P., & Sheldon, R. J. (1988). Stated preference methods: an introduction.Journal of Transport Economics and Policy., 22, 1125.

    Kuenz Murphy, C. (1993). Limits on the analytic hierarchy process from itsconsistency index. European Journal of Operational Research, 65(1), 138139.

    Lam, T., & Hsu, H. C. (2006). Predicting behavioral intention of choosing a traveldestination. Tourism Management, 27(4), 589599.

    McIntosh, R. W., & Goeldner, C. R. (1990). Tourism: Principles, practices, philosophies.New York: Wiley.

    Middleton, V. M. (1994). Marketing in travel and tourism (2nd ed.). Oxford: Butter-worth-Heinemann.

    Milman, A., & Pizam, A. (1995). The role of an awareness and familiarity witha destination: the central Florida case. Journal of Tourism Research, 33(3), 2127.

    Mok, C., & Armstrong, R. W. (1995). Leisure travel destination choice criteria ofHong Kong residents. Journal of Travel & Tourism Marketing, 4(1), 99104.

    Moutinho, L., & Curry, B. (1994). Modeling site location decisions in tourism. Journalof Travel & Tourism Marketing, 3(2), 3556.

    Nicolau, J. L., & Mas, F. J. (2006). The inuence of distance and prices on choice oftourist destinations: the moderating role of motivations. Tourism Management,

    ement 30 (2009) 288297Ryan, C. (1997). The tourist experience: A new introduction. London: Cassell.Saaty, T. L. (1980). The analytical hierarchy process: Planning, priority setting, resource

    allocation. New York: McGraw-Hill Book Co.

  • Saaty, T. L. (1986). Axiomatic foundation of the analytic hierarchy process.Management Science., 36(3), 259268.

    Sirakaya, E., Uysal, M., & Yoshioka, C. (2003). Segmenting the Japanese tour marketin Turkey. Journal of Tourism Research, 41, 293304.

    Sirakaya, E., & Woodside, A. G. (2005). Building and testing theories of decisionmaking by travelers. Tourism Management, 26, 815832.

    Stevens, B. F. (1992). Price value perceptions of travelers. Journal of Travel Research,31, 4448.

    Suh, Y. K., & McAvoy, L. (2005). Preferences and trip expenditures-a conjointanalysis of visitors to Seoul, Korea. Tourism Management, 26, 325333.

    Swait, J. (2001). Choice set generation within the generalized extreme value familyof discrete choice models. Transportation Research B, 35(7), 643667.

    Um, S., & Crompton, J. L. (1990). Attitude determinants in tourism destinationchoice. Annuals of Tourism Research, 17(3), 432448.

    Uysal, M., & Hagan, L. R. (1993). Motivation of pleasure to travel and tourism. InM. Khan, M. Olsen, & T. Var (Eds.), Encyclopedia of hospitality and tourism(pp. 798810). New York: Van Nostrand Reinhold.

    Uysal, M., & Jurowski, C. (1994). Testing the push and pull factors. Annual of TourismResearch, 21(4), 844846.

    Van Raaij, W. F. (1986). Consumer research on tourism: mental and behaviouralconstructs. Annual of Tourism Research, 13, 19.

    World Tourism Organization. (2006). Available at: January 2007.

    Yavuz, N., Baloglu, S., & Uysal, M. (1998). Market segmentation of European andTurkish travelers to North Cyprus. An International Journal of Tourism andHospitality, 9(1), 418.

    Yoon, Y., & Uysal, M. (2005). An examination of the effects of motivation andsatisfaction on destination loyalty: a structural model. Tourism Management, 26,4556.

    Zadeh, L. A. (1965). Fuzzy sets. In J. Klir, & B. Yuan (Eds.), Fuzzy sets, fuzzy logic, andfuzzy systems (pp. 1934). Singapore: World Scientic.

    Zadeh, L. A. (1976). The linguistic approach and its application to decision analysis.In J. Klir, & B. Yuan (Eds.), Fuzzy sets, fuzzy logic, and fuzzy systems (pp. 260281).Singapore: World Scientic.

    T.-K. Hsu et al. / Tourism Management 30 (2009) 288297 297

    The preference analysis for tourist choice of destination: A case study of TaiwanIntroductionLiterature review and objective of this studyDestination choiceThe concept of destinationTravel motivationPreference of tourist

    Methodology of evaluationThe Analytic Hierarchy Process (AHP)Fuzzy set theoryLinguistic variableFuzzy numbersDefuzzification of fuzzy average

    The Technique for Order Preference by Similarity to Ideal Solution (TOPSIS)

    Case applicationDeveloping the evaluation hierarchyThe alternativesMethodQuestionnaire designData collectionThe weights of evaluation criteriaOverall preference measure of destinations

    Empirical results and discussions



View more >