留學生dissertation網(wǎng)_一個有效的拉丁美洲游客分割營銷研究-Benefits Segmentation of Visitors t

留學生dissertation網(wǎng)提供一個有效的拉丁美洲游客分割營銷研究，這項研究提供了一個有效的拉丁美洲游客分割，從而提供了關于該地區(qū)的旅游目的地營銷戰(zhàn)略計劃的寶貴意見。四個不同的分割基礎上確定的相關利益要求。Emine Sarigöllü and Rong Huang
Benefits Segmentation of Visitors to Latin America

On behalf of:Travel and Tourism Research Association
Additional services and information for Journal of Travel Research can be found at:
Email Alerts: http://jtr.sagepub.com/cgi/alerts
Subscriptions: http://jtr.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007
J1FO0E.UB11R7NU7A/0LR04 YO7 F28 0T70R50A4V27E2L0 R32ESEARCH
Benefits Segmentation ofVisitors to Latin America
EMINE SARIGÖLLÜ and RONG HUANG
This research presents an effective segmentation of LatinAmerican tourists and thereby provides invaluable input andguidance for destination marketers in regard to strategicplanning for the region’s tourist provision. Four distinct segmentsare identified based on the benefits sought, and theseare profiled with respect to demographics, travel behavior,expectations about the infrastructure, local environment,services, and costs, as well as the visitors’ personalities andinterests. Globally, the most important decision drivers arefound to be safety, flight and accommodation availability,and affordability. This research also presents an overview ofbenefits segmentation literature on tourism.

Keywords: benefits segmentation; destination marketing;Latin America
With abundant natural endowments of sand, sea, and sun;diverse cultures; and a rich historical heritage, Latin Americancountries have a great potential in the inbound travelmarket (Strizzi and Meis 2001). The Latin American regionattracted 18.9 million international travelers (3.6% of theworld total) in 2001 (World Tourism Organization 2002).oreover, the growth rate of international tourist arrivals inLatin America was 6.8% in 2000. Nonetheless, the growthrate for visitors from North America dropped from 8.9% in2000 to 1.8% in 2001, which is a cause for anxiety becausealmost three-quarters of all international tourist arrivals areeither from North America or intraregional in origin (Strizziand Meis 2001). According to Augusto Huéscar, the World
Tourism Organization’s (WTO) chief of market intelligenceand promotion, “Although North American travellers composeda great part of the international travellers for Latin
America, increased competition from other regions in theworld and the safety of tourists are the two greatest challengesfor the Americans to overcome” (Bambrad 2001, p. 1).Furthermore, Strizzi and Meis (2001) noted that the Latin#p#分頁標題#e#
American tourism industry is being adversely affected byvarious factors including economic and financial instability,ongoing urbanization, safety and security risks, healththreats, and aviation infrastructure limitations. In fact, Latin
American tourism is expected to face further challenges in
the new millennium, specifically due to problems in regionaleconomic growth and political, social, and environmentalfactors. In short, macrolevel evidence points toward difficulttimes for Latin American tourism. At the microlevel, however,there is a lack of systematic research. Microlevel analysisentails investigation of demand from the perspective ofthe visitor and thus presents an opportunity not only forobtaining an in-depth understanding of the visitor but alsofor developing the key strategic and managerial implications,
particularly in regard to product offerings, targeting, andpromotion. Hence, research with a microfocus is needed tocomplement the extant macroresearch for a complete assessmentof the tourist demand for Latin America. This researchrepresents a step in that direction.
This research provides a thorough assessment of NorthAmerican visitors to Latin America. It specifically (1) identifies
http://m.elviscollections.com/dissertation_writing/Tourism/benefits that are important for the tourists’ destinationchoice; (2) segments tourists based on the benefits sought;
(3) delineates different segments with respect to demographics,travel behavior, services, infrastructure and cost expectations,personality, and interests; and (4) proposes the managerialimplications for destination marketers for designing,targeting, and promoting the Latin American tourist productfor the segments identified. Hence, this research has importantimplications for the stakeholders in the Latin Americantourism industry because it provides valuable input andguidance toward strategic planning for the region’s tourist
offering.
The contribution of this research is twofold. First, it providesan overview of the benefits segmentation literature intourism. Specifically, prior studies are categorized in twodimensions, depending on whether (1) benefits wereobtained via direct questioning as opposed to some form ofindirect/inferential analysis, and (2) they are destinationspecificor general. Second, through focusing on tourists toLatin America (a region characterized by a lack of systematicstudy to date), this research incorporates a more comprehensiveset of attributes than those used in typical tourism segmentationstudies. Specifically, in addition to the standardattributes used in tourism segmentation studies (e.g., age,gender, and place of residency), we also consider variousaspects of the infrastructure, local environment, services,
Emine Sarigöllü (Ph.D., Wharton School, University of Pennsylvania;M.A., University of Pennsylvania; MBA, Bogazici University)
is associate professor of marketing and director of theMcGill Institute of Marketing (MIM) at the Faculty of Management,#p#分頁標題#e#
McGill University in Montreal, Canada. Her primary research domainis consumer choice. Rong Huang is a doctoral student at the
Faculty of Management, McGill University in Montreal, Canada.Her research interests focus on branding. This research was supportedby funds from the Organization of American States (OAS)and MIM. The authors thank Daniel Perna, George Vincent, andthe reviewers for their helpful comments, and Dr. Anthony Miyazakifor his assistance in data collection.Journal of Travel Research, Vol. 43, February 2005, 277-293
DOI: 10.1177/0047287504272032
© 2005 Sage Publications
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007
and costs, as well as the visitors’ personalities, interests, andtravel behavior. Hence, this study provides an in-depthunderstanding of Latin American visitor segments throughan unusually rich profiling. Consequently, this research produces
practical operational information on each segment thatis translatable into strategy, specifically in terms of thedesign, promotion, and targeting of the tourist product.
LITERATURE REVIEW
Market segmentation is the process of classifying customersinto groups based on different needs, characteristics,
or behavior, and it has strategic implications for customertargeting and product positioning. Market segmentation is
widely implemented in the tourism industry, using visitordemographics (Morrison et al. 1996; Mudambi and Baum1997), psychographics (Bieger and Laesser 2002; Cha,McCleary, and Uysal 1995; Dodd and Bigotte 1997; Kau and
Lee 1999; May et al. 2001; Mo, Havitz, and Howard 1994;Sirakaya, Uysal, and Yoshioka 2003), behavior (Court andLupton 1997; Fodness and Murray 1998; Formica and Uysal
1998; Goldsmith and Litvin 1999; Meric and Hunt 1998;
Pritchard and Howard 1997), and benefits (Kanstenholz,Davis, and Paul 1999; Tian, Crompton, and Witt 1996).
Benefit segmentation is a powerful method for groupingconsumers (Kotler and Turner 1993). It has been proposedthat because different segments of consumers may desire differentbenefits from using a product, and the travel relatedbenefits sought also vary (Woodside and Jacobs 1985). Furthermore,
Haley (1968) claimed that the benefits people seekare the basic reason for the existence of true market segmentsand thus provide better determinants of behavior than otherapproaches. Indeed, benefit segmentation is shown to predictbehavior better than demographics and geographic segmentation(Haley 1968; Kastenholz, Davis, and Paul 1999;
Young, Ott, and Feigin 1978). Moreover, based on predictivefactors and combined with key descriptive variables,
benefit segmentation provides a clear insight into marketingand communication strategy formulation (Loker and Perdue1992). Accordingly, benefit segmentation is becomingincreasingly popular in tourism markets, particularly as an#p#分頁標題#e#
effective foundation for marketing strategy.
In the early tourism literature, benefits were defined asvisitor ratings of desired amenities and activities (Tian,Crompton, and Witt 1996). This approach was used in imagestudies to describe and evaluate potential visitors’ perceptionof destinations (Crompton 1979; Fakeye and Crompton1991; Gartner 1986; Hunt 1975; Keown, Jacobs, andWorthley 1984; Mayo 1973; Um and Crompton 1992).
Other researchers, however, focused on tourists’ motivationto travel and conceptualized various attributes as conduits tofacilitate the desired psychological benefit outcomes. Earlyresearch on tourists’ motivation includes Lundberg (1971)and Crompton (1979), in which, respectively, 18 and 9 motivationswere identified as influencers of the travelers’ decisions.
Other researchers followed, including Bieger andLaesser (2002); Cha, McCleary, and Uysal (1995); Formicaand Uysal (1998); Iso-Ahola (1982); Loker and Perdue(1992); Pearce and Caltabiano (1983); Sirakaya, Uysal, and
Yoshioka (2003); Tian, Crompton, and Witt (1996); andWoodside and Jacobs (1985). Adopting a broad perspectiveon benefit segmentation, the present study considers priorresearch that took either of the following approaches: desiredamenities/activities or motivation for travel. Prior studies inthis domain can be categorized in terms of whether benefitswere obtained via direct questioning or through some formof indirect/inferential analysis.1 Furthermore, they can alsobe categorized depending on whether they were destinationspecificor general. Table 1 provides an overview of extantresearch on benefit segmentation in tourism.
Typically, direct questioning was used to obtain the benefitssought from travel. For example, Kastenholz, Davis, andPaul (1999), in their segmentation study of north and centralPortugal, identified 27 dimensions of rural experiences, suchas entertainment/nightlife, opportunities for socializing,opportunities for families with children, and so on, by direct
questioning. The direct approach is based on the premise thatrespondents can reliably sum up their vacation experiences.
As Dann (1981) has pointed out, however, tourists may beunable or unwilling to reflect on and/or express real travelmotives to themselves or to professional interviewers. Inaddition, managerial implications may be hampered whesome questions are not directly related to vacation activities;managers may not be able to determine what specific activitieswould bring the desired benefits to tourists and, hence,would not be able to design and improve the activitiesaccordingly.
Various forms of indirect/inferential analysis were alsoused to derive the benefits sought (Bryant and Morrison
1980; Davis and Sternquist 1987; Dybka 1987; Johns andGyimothy 2002; Moscardo et al. 2000; Pearce and
Caltabiano 1983; Shoemaker 1994). Typically, researchersderive benefits from what tourists did or plan to do duringtheir vacations. For example, Bryant and Morrison (1980)derived certain benefits from reports of participation in recreationaland sightseeing activities during the vacation.#p#分頁標題#e#
Alternatively, Pearce and Caltabiano (1983) inferred travelmotivations indirectly from travel experiences using
Maslow’s hierarchy of needs. The indirect approach typicallyinvolves a factor analysis and needs to be implemented
very carefully to ensure that the set of activities identified iscomplete and comprehensive; otherwise, the range of benefits
might be unduly constrained. Furthermore, it is likelythat the derived benefits are rather obvious and simple innature. It was demonstrated that the “push-pull” relationship(i.e., the relationship between the functional benefits and themore abstract psychological and social benefits) might influenceand drive travel behavior (Klenosky 2002). In addition,different activities might offer the same or similar psychologicalor social benefits to tourists. Finally, because mostsuch work uses factor analysis to infer the benefits soughtand a given activity can only load on a single factor, the possibilitythat the activity in question might be associated with
several different psychological or social benefits cannot beexplored.2
As stated earlier, benefit segmentation research can also
be classified according to whether the study is destinationspecific.In destination-specific studies, tourists to the samedestination are classified into segments based on the benefitssought. For example, Formica and Uysal (1998) studied visitors
of the Spoleto Festival in Italy and identified six benefit
factors (from the original 23 variables): socialization anentertainment, event attraction and excitement, grouptogetherness, cultural/historical significance, family
278 FEBRUARY 2005
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007togetherness, and site novelty. Using these factors, two furthersegments were identified: moderates and enthusiasts. Innondestination research, researchers study what tourists seek
in general for their vacation. Such studies are typically conductedon tourists from a specific country or region. For
example, Cha, McCleary, and Uysal (1995) studied thetravel motivation of Japanese overseas travelers and identifiedsix motivational factors: relaxation, knowledge, adventure,travel bragging, family, and sports, based on which
three segments were found: sports seekers, novelty seekers,and family and relaxation seekers.
The current study is destination-specific. Visitors toLatin America were segmented using benefits sought as thebasis for study. Specific benefits were derived through aninferential approach from a comprehensive set of vacationactivities including sightseeing, entertainment, and sports.Benefit segmentation was then implemented using the standardstatistical methodology. What differentiates this
research from the literature is what follows the segmentationprocess, specifically, the offering of rich practical implicationsto destination marketers. This was achieved through theuse of a rich set of variables that provided the opportunity to#p#分頁標題#e#
JOURNAL OF TRAVEL RESEARCH 279
TABLE 1
BENEFITS SEGMENTATION IN TOURISM: PREVIOUS LITERATURE
Destination Specific Not Destination SpecificBenefits obtained by Kastenholz, Davis, and Paul (1999) Bieger and Laesser (2002)direct questioning 27 benefits including culture and traditions and rural Travel motives including nightlife,environment, for example peaceful/quiet comfort, partner, family, nature,atmosphere, unpolluted environment, and healthy culture, liberty, sports, and sunlifestyleCha, McCleary, and Uysal (1995)
Formica and Uysal (1998) 30 motives including relaxation,
23 motives including socialization and entertainment, knowledge, adventure, travelevent attraction, excitement, group togetherness, bragging, family, and sportscultural, and historical
Tian, Crompton, and Witt (1996)18 benefits including relaxation, entertainment,socializing, bonding, social recognition, selfesteem,
and education
May et al. (2001)
26 benefits including enjoying nature, achievement,stimulation, escape, personal/social pressure,and being with family and friends
Anderson, Prentice, and Watanabe (2000)
21 motives including novelty, independence, prestige,relaxation, understanding, and utility
Loker and Perdue (1992)
12 benefits including escape, relaxation, naturalsurroundings, excitement variety, family, and friends
Benefits derived by Bryant and Morrison (1980) Shoemaker (1994)some form of indirect/ Participation in vacation outdoor and sightseeing 39 benefits including educationalinferential analysis activities including hiking, fishing, tennis, historic sites, possibilities, environmentalprofessional sports, and cultural activities aspects, resort set, sun sports,popularity of destination, value,
留學生dissertation網(wǎng)Davis and Sternquist (1987) scenery, friend, relatives,10 benefits including sports, sightseeing, rest, shopping, entertainment, and conveniencefood, entertainment, and cultureDybka (1987)
Johns and Gyimothy (2002) Content of the questionnaire12 benefits including social, cultural aspects, nature and is not providedscenery, relaxation and slower pace of life, andenthusiast activities
Moscardo et al. (2000)
20 benefits including relaxation, resort, warm sunnyweather, beach activities, and environmental activities
Woodside and Jacobs (1985)
26 benefits including value of money, relaxation,
educational, family, and friends
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007validate and delineate the segments in relation to aspectsother than the benefits sought. Specifically, this study consideredvisitors’ expectations regarding a comprehensive setof infrastructure, local environment, service, and cost factors,as well as travel behavior and typical demographics,then profiled the segments accordingly. In addition, thisresearch also delineated the segments in terms of the visitors’personalities and interests. The profiling of segments by differentvalue and personality types complements the benefitssought and offers a deeper understanding of each segment,thereby helping travel suppliers to create packages that aremore compatible with the motivations, attitudes, and opinionsof the travelers, and also provides better prediction oftravel behavior (Abbey 1979; Keng and Cheng 1999; Woodsideand Pitts 1976). In short, this research incorporates awider variety of vacation activity, infrastructure, environment,service, and cost aspects, as well as taking into accountthe personalities, interests, and demographics of the visitors,than were considered by the prior research. Consequently,this research offers rich managerial implications in relationto destination marketers in terms of the design, targeting, andpromotion of the tourism product—more so than theprevious research, which lacks the breadth and scope ofvariables needed to draw such strategy implications.#p#分頁標題#e#
METHODOLOGY
This research performs a market segmentation ofNorth American tourists to Latin America based on thebenefits sought. Requisite information was obtained by aquestionnaire.
Questionnaire Development
The questionnaire contains two parts. The first part
inquired about visitors’ general vacation preferences, such as
the frequency of travel, the traveling party, informationsources, and package types. In addition, respondents wereasked to rate the importance of vacation activities that werepresented in three categories: sightseeing, sporting activities,and entertainment. Under each category, specific activitieswere listed and respondents were asked to rate their relativeimportance from 1 (not at all important) to 5 (very important).Finally, respondents rated the importance of variousother decision drivers, including 15 items relating to infrastructure,local environment, services, and costs. The secondpart contained questions about the respondents’ personalities,interests, and demographics.
Three categories of vacation activity—sightseeing, sportingactivities, and entertainment—and specific personalityand interests items were selected jointly with the sponsors ofthe research, namely, the Inter-Sectoral Unit for Tourism ofthe Organization of American States (OAS), following aconsultation process with the representatives of memberstates and a review of the extant literature in academic journals.
Although the activities (25 in total) are quite comprehensive,they are not necessarily exhaustive. The questionnairewas pretested first on a sample of 15 MBAstudents andthen 20 respondents recruited by travel agencies in Montreal,then revised accordingly. Although the questionnaire wasshortened following comments from the respondents, thefinal version was still quite demanding (10 pages) due to theextensive information needs of the OAS members. Toencourage an enthusiastic response, a prize draw was introducedwith a first prize of $300 and second and third prizes of$100 each.
Data Collection
The survey was developed by McGill Institute of Marketing
(MIM) of McGill University through a research projectsponsored by the Inter-Sectoral Unit for Tourism of theOAS. The survey was initially administered with the help oftravel agencies in the United States and Canada. Concerns,however, about the quality of the data due to sampling error(e.g., responses from cruise visitors who did not stay overnight,or visitors who have not yet visited Latin America),and substantial missing information led to the rejection ofmany responses. Consequently, 94.3% of the data (250responses) were collected at Miami airport from respondents
who were awaiting their connections. Data collection wasplanned around the scheduled arrival times of airlines fromvarious Latin American countries, and respondents were personallyapproached. Effective control of the data-collectionprocess at the airport resulted in a good response rate andverifiable data. In all cases, attempts were made to excludebusiness travelers. Included were visitors who havevacationed in Latin America during the past year and stayeda minimum of 3 nights. The minimum stay criterion wasimplemented to ensure that the respondents had sufficientopportunity to experience various aspects of the destination,as assessed by the questionnaire. In total, 265 North Americans(Americans or Canadians) responded, with 99%originating from the United States.#p#分頁標題#e#
Analysis
Data were analyzed in three stages. First, the variousdimensions underlying the benefits sought were uncoveredby a factor analysis using the principal component methodwith varimax rotation. These methods “were supported in theliterature and yielded the most interpretable results” (Lokerand Perdue 1992, p. 31). In fact, factor analysis has beenwidely used in visitor segmentation research (Cha,
McCleary, and Uysal 1995; Formica and Uysal 1998; Johnsand Gyimothy 2002; Kastenholz, Davis, and Paul 1999;Loker and Perdue 1992; Madrigal and Kahle 1994; Shoemaker1994). Typically, factor analysis is implementedbecause it allows data reduction and substantive interpretation
(Churchill and Iacobucci 2002). In this research, datareduction was specifically useful in the next stage of analysis,involving cluster analysis, because it eliminated correlationamong the variables (which would have been problematicin cluster analysis). Furthermore, factor analysis helpedidentify the constructs that underlie the variables, providinga global view of the most substantive benefits sought usingsuch constructs.
The factor analysis produced orthogonal factors that
summarized the 25 vacation activities. Then, the factor
scores for each respondent were saved and consequently
used in stage 2 for clustering them into market segments.
Individuals were clustered such that those within each cluster
280 FEBRUARY 2005
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007
were more similar to each other than to those in other clusters,
thereby creating a situation of homogeneity within clusters
and heterogeneity between clusters (Hair et al. 1998).
Specifically, the K-means cluster method, which is quite
common in visitor segmentation research, was implemented
(e.g., Cha, McCleary, and Uysal 1995; Formica and Uysal
1998; Kau and Lee 1999; Madrigal and Kahle 1994). The Kmeans
clustering method produces results that are less susceptible
to outliers in the data, the distance measure used, and
the inclusion of irrelevant or inappropriate variables (Hair
et al. 1998).
During the third stage, segment characteristics were
delineated by various univariate and multivariate statistical
procedures. Specifically, the differences among clusters in
demographics, travel behavior, the ratings of various decision
drivers, and the various personality and interest dimensions
were assessed by suitable analyses, including
ANOVA, discriminant, and chi-square. These analyses typically
entail cluster analysis for the purpose of validation and
segment profiling (Cha, McCleary, and Uysal 1995; May et
al. 2001; Formica and Uysal 1998; Madrigal and Kahle
1994). Different statistical tests are conducted according to#p#分頁標題#e#
the characteristics (metric, categorical) of the variables.
ANOVA is used to identify if there are any differences
among the clusters, as measured by a comparison of mean
ratings (for metric variables). Discriminant analysis is specifically
used to understand how the members in one segment
differ from those in another and/or to correctly classify
individuals into segments. In the current study, decision
drivers, personality, and interests items (all metric) are used
to find the best discriminators among the identified clusters
(i.e., segments). Finally, chi-square analysis is used to
explore the differences between clusters in terms of
categorical variables, such as demographics and travel
behavior.
FINDINGS
The findings will be presented in three sections. First, the
dimensions underlying the benefits sought will be revealed.
This will be followed by results from the benefits segmentation.
Finally, the segments will be profiled in terms of demographics,
travel behavior, infrastructure, service and cost
expectations, and personality and interests.
Dimensions Underlying Benefits Sought
The assessment of the benefits sought was obtained by
asking respondents how important it was whether a vacation
destination should provide specific sightseeing, sporting,
and entertainment activities. In total, 25 activities were
included in the survey for the sake of completeness. It is
unlikely, however, that visitors consider each of these
dimensions independently when selecting a destination;
rather, they are likely to evaluate some activities similarly.
Thus, it is conceivable that a relatively fewer number of
dimensions represent the information found in the original
25 items. To explore the dimensions underlying visitors’
benefits sought, a factor analysis (using the principal components
method with varimax rotation) was performed on the
importance ratings of the 25 vacation activities. Because the
major objective of the factor analysis was to reduce a large
number of variables to a smaller set of uncorrelated variables
for subsequent use in cluster analysis, Hair et al. (1998) indicated
that the orthogonal rotation methods, such as varimax,
are appropriate. Specifically, varimax rotation ensures a
clearer separation of the factors, and it has proved very
successful as an analytic approach to obtain an orthogonal
rotation of factors (Hair et al. 1998).
In keeping with the conventions for factor analysis, we
used the following criteria: (1) factor loadings equal to or
above 0.50, (2) eigenvalues equal to or above 1.0, and (3)
results of the factor analysis explaining at least 50% of the
total variance (Hair et al. 1998). Table 2 displays factor loadings,
eigenvalues, and the explained variance. In addition,#p#分頁標題#e#
alpha coefficients for items in each factor are provided.
The factor analysis grouped together items that received
similar ratings and thus revealed five factors accounting for
65% of the total variance. The factor solution is acceptable
“since it is acceptable to consider a solution that accounts for
60% of the total variance as satisfactory (and in some
instances even less), in the social science research” (Hair et al.
1998, p. 104).
The resultant five factors represent specific dimensions
of the benefits that respondents seek when they go on vacation.
The results confirm that visitors are indeed seeking a
variety of vacation activities. The first factor includes a variety
of typical tourist activities and is named fun & sun benefits.
Specifically, 8 variables were loaded on this factor,
explaining about 27% of the total variance. The second factor
summarizes ecotourism and culture-related benefits and
is labeled ecotourism in keeping with the highest loaded variables.
Labeled as performing arts and events, the third factor
includes 4 variables, including concerts and theaters. The
fourth factor consists of 3 variables, representing outdoor
adventure benefits, including hiking, camping, and extreme
sports. Finally, the fifth factor includes 3 variables that relate
to general sightseeing, such as the exploration of nature and
scenery, and small towns and villages. Only one variable,
golf/tennis, did not load on any factor because its loading
was below the threshold of 0.5. This indicates that the level
of importance attributed to golf/tennis is not related in any
way to the importance attributed to the other variables.
Moreover, the respondents’ ratings for this variable do not
account for a significant portion of the total variance in the
importance ratings data. Hence, golf/tennis will be excluded
from further analyses.
Benefit Segmentation
Having recovered the major dimensions underlying the
benefits sought in selecting a vacation destination, the focus
next turns to visitors’ preferences in relation to these benefits.
By analyzing the diversity in benefits sought, visitors
who valued similar benefits were grouped together. A
nonhierarchical cluster analysis (SPSS 10.0 Classify Kmeans
Cluster) was implemented on the five benefits sought,
and the dimensions identified above and the solutions with
two, three, four, and five clusters were explored. The typical
criteria for effective segmentation were considered in the
analysis as follows. Effective segments
1. consist of consumers with homogeneous needs, attitudes,
and responses to marketing variables (McCarthy
1982);
2. are distinctive from one another (Weinstein 1987);
JOURNAL OF TRAVEL RESEARCH 281
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.#p#分頁標題#e#
Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007
3. are substantial, that is, large enough to be profitable
(McCarthy 1982); and
4. provide operational data that are practical, usable, and
readily translatable into strategy (Weinstein 1987).
A four-cluster solution (i.e., four visitor segments) was
the most readily interpreted, most favorably met the above
criteria, and provided the best statistical results (when variables,
other than the ones used to form clusters, were used to
test for cluster differences). In addition, the validity of the
cluster solution was confirmed by splitting the sample into
two groups and then comparing the results.
The four segments are named: adventurer, multifarious,
fun&relaxation seeker, and urbane. The mean benefit factor
scores and statistics regarding tests of the differences
between the four segments’ scores are displayed in Table 3.
Univariate ANOVA tests confirm that the segments statistically
differ in their mean benefit scores. Follow-up ad hoc
tests indicate that almost all the segments, with a few exceptions,
differ from each other in pairwise tests. Specifically, in
regard to two benefit dimensions, namely ecotourism and
outdoor adventure, all the segments differ pairwise without
exception.
In segment 1, the adventurer mainly seeks outdoor
adventure and general sightseeing activities. This segment’s
lack of interest in fun and sun activities differentiates it from
the other segments. This is the smallest segment; nonetheless,
it constitutes 10% of the sample. The multifarious
282 FEBRUARY 2005
TABLE 2
BENEFITS SOUGHT DIMENSIONS: FACTOR LOADINGS
Factor
Performing Factor 4 Factor 5
Factor 1 Factor 2 Arts and Outdoor General
Item Fun and Sun Ecotourism Events Adventure Sightseeing
Dance/bar 0.781
Beach, sunbathing 0.754
Casino/gambling 0.742
Sailing/boating 0.698
General entertainment 0.687
General sport 0.671
Snorkeling/scuba diving 0.622
Amusement park 0.555
Botanic/zoologic garden 0.739
Environmental/ecology excursion 0.721
Local boat tours 0.704
Guided city tour 0.691
Art gallery/museum 0.686
History/archaeology 0.662
Musical concert 0.819
Theater play 0.796
Local event 0.634
Dining/restaurant 0.513
Hiking/camping 0.820
Extreme sports 0.796
Fishing 0.586
Nature/scenery 0.769
General sightseeing 0.705
Small towns/villages 0.682
Eigenvalue 6.716 4.219 2.717 1.548 1.008
% of variance 26.864 16.877 10.867 6.193 4.033
Alpha coefficient 0.869 0.834 0.779 0.756 0.738
TABLE 3
VISITOR SEGMENTS: MEAN BENEFIT SCORES
Segment 1: Segment 2: Segment 3: Segment 4:
Adventurer Multifarious Fun and Relaxation Urbane#p#分頁標題#e#
Benefit Dimension (n = 20) (n = 72) Seeker (n = 50) (n = 62) F a Post Hoc
Fun & sun –1.40843 0.18714 0.45112 –0.12680 23.79 All but 2-3
Ecotourism –0.67843 0.35345 0.01243 –0.20163 7.58 All
Performing arts & events –1.02172 0.67945 –0.81398 0.19698 52.73 All but 1-3
Outdoor adventure 0.96037 0.37067 –0.79832 –0.09645 28.48 All
General sightseeing 0.58638 0.52088 0.40250 –1.11864 82.21 All but 1-2 and 1-3
a. All reported F values are significant at .000.
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007
makes up the largest segment, with 35% of the sample. They
like to explore the destination fully and, hence, seek diverse
benefits, including performing arts and events, general sightseeing,
outdoor adventure, ecotourism, and fun and sun
activities. In short, this segment represents tourists that seek
a bit of everything. Segment 3 predominantly seeks fun and
relaxation and is thus labeled accordingly. They also seek
general sightseeing activities and constitute 24.5% of the
sample. Finally, the urbane section comprises the visitors
who exclusively seek performing arts and local events. They
do not seem much interested in other benefits, including general
sightseeing, fun and sun, ecotourism, and outdoor
adventure. Interestingly, this is a large segment, consisting of
30% of the sample.
So far, it has been established that the four segments differ
in terms of the benefits sought. The segmentation analysis,
however, would be meaningful only if the segments can
also be differentiated in terms of characteristics other than
the benefits sought. As indicated previously, in addition to
having homogeneous product needs, consumers in a good
market segment should also possess homogeneous attitudes
and responses to marketing variables (McCarthy 1982). Furthermore,
the various segments must be distinct from one
another with respect to other consumer characteristics
(Weinstein 1987). Finally, operational data are needed to
provide practical, usable, and readily translatable information
for each segment. Hence, to further examine the differences
among segments and provide practical information to
formulate marketing strategy, we next turn to exploring
whether the four segments in fact differ in terms of demographics,
travel behavior, expectations about various
infrastructures, service and cost factors, and personality
attributes and interests.
Delineating Segments: Demographics
Segment demographics are presented in Table 4. Female
respondents were concentrated in the urbane and multifarious
segments, whereas men were prominent in the adventurer#p#分頁標題#e#
segment. Fun and relaxation seekers were split evenly
between males and females. Significant differences were
found in the mean age of the respondents across each of the
four segments. The post hoc analysis indicated that the multifarious
segment is the oldest among the four. The segments
also differed significantly in terms of their members’ mean
household income. Post hoc analysis indicated significant
differences between three segments, namely, the urbane, the
adventurer, and the multifarious. The multifarious has the
highest household income level, whereas the adventurer has
the lowest. Most respondents are employed full-time. Those
who work part-time, students, or the unemployed/retired are
more likely to belong to the multifarious rather than the other
segments. The segments did not differ significantly in terms
of educational level.
Delineating Segments: Travel Behavior
Significant differences were found in travel behavior
among the four segments as displayed in Table 5. Whereas
more than half of the adventurer section travel more than
once a year, only 13% of the urbane do so. Most members of
the multifarious section travel once a year. A substantial
majority of visitors have not visited Latin America previously,
yet most of the adventurers were repeat visitors. The
JOURNAL OF TRAVEL RESEARCH 283
TABLE 4
SAMPLE DELINEATION: DEMOGRAPHIC CHARACTERISTICS
Sample Adventurer Multifarious Fun & Relaxation Urbane
(100%) (10%) (35%) (25%) (30%)
Sex
Female 57 40 56 51 63
Male 43 60 44 49 37
Age (F =10.44, p = .000)
18-34 46 45 41 52 51
35-54 40 45 41 34 39
55+ 14 10 18 14 10
Education
Elementary 1 0 0 2 0
High school 18 25 14 17 22
Technical/vocational 9 0 7 8 10
University 53 50 62 63 52
Postgraduate 19 25 17 10 16
Occupation
Employed full-time 64 79 57 71 61
Employed part-time 10 5 13 8 10
Self employed 7 5 3 4 13
Full-time student 7 5 13 6 6
Unemployed/retired 12 6 14 11 10
Household income per year
(in dollars; F = 3.321, p =.02)
Less than 20,000 9 10 10 6 10
20,000 to 49,999 32 53 31 40 23
50,000 to 69,999 25 11 18 27 41
70,000+ 34 26 41 27 26
© 2005 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.
Downloaded from http://jtr.sagepub.com at University of Westminster on February 19, 2007
different segments differed considerably in regard to their
interest in various types of vacations. Although most visitors
indicated an interest in self-organized vacations, the multifarious
were the most interested and the fun and relaxation
seekers were the least interested. Major contrasts were
observed in the segments’ interest in all-inclusive vacations;#p#分頁標題#e#
whereas the adventurers indicated almost no interest in this
concept, the multifarious and the fun and relaxation seekers
were very interested. Similar, yet less extreme, results were
observed in relation to the interest shown in cruise vacations.
Most respondents travel with their spouses. Those who travel
with family and friends are, however, more likely to be multifarious,
and those who travel alone are more likely to be in
the adventurer or urbane segments. Interestingly, those who
indicated a relatively higher interest in self-organized tours,
namely, the multifarious and adventurers, were less likely to
consult travel agents than the other segments. The Internet
seems to be a major source of travel information for adventurers,
whereas the multifarious seem to rely mostly on
friends and family, along with a variety of other sources. In
contrast, the urbane and the fun and relaxation seekers tend
to consult travel agents and brochures.
Delineating Segments: Decision Drivers
Decision drivers refer to various local services, infrastructure,
and cost items that visitors may consider in choosing
a vacation destination abroad. These aspects complement
the typical vacation benefits sought such as fun and sun,
sightseeing, and outdoor adventure. In total, 14 items were
included. The respondents were asked to rate the importance
of each item by assigning a number from 1 = not at all important
to 5 = very important. The four segments statistically
differ in their importance ratings of decision driver items, as
indicated by univariate ANOVA tests in Table 6.
In general, the adventurers’ mean ratings for decision
drivers were the lowest among the four segments, indicating
that they are less concerned than the others about
284 FEBRUARY 2005
TABLE 5
SAMPLE DELINEATION: TRAVEL BEHAVIOR
Sample Adventurer Multifarious Fun and Relaxation Urbane
Frequency of travel
(%; chi-square = 19, sig. = .004)
Less than 1 per year 38 25 32 42 48
1 per year 39 20 48 31 39
More than 1 per year 23 55 20 27 13
Previous visit to Latin America
(%; chi-square = 12 , sig. =.06)
No 71 40 65 80 82
Yes 29 60 35 20 18
Interest in self-organized vacations
(%; chi-square = 16, sig. =.02)
None 6 5 1 12 7
Somewhat 22 15 13 26 32
Very 72 80 86 62 61
Interest in all-inclusive vacations
(%; chi-square = 73, sig. = .000)
None 13 75 9 6 7
Somewhat 23 25 16 18 33
Very 64 0 75 76 60
Interest in cruises
(%; chi-square = 51, sig. = .000)
None 11 55 4 12 3
Somewhat 30 30 29 25 37
Very 59 15 6