rezende morrisonismailjvm brazil

17

Click here to load reader

Upload: a-morrison

Post on 18-May-2015

423 views

Category:

Education


2 download

TRANSCRIPT

Page 1: Rezende morrisonismailjvm brazil

Dazed and confused? An exploratory study ofthe image of Brazil as a travel destination

Aline M. Rezende-Parker, Alastair M. Morrison� and Joseph A. IsmailReceived (in revised form): 5th August, 2002Anonymously refereed paper

�Department of Hospitality and Tourism Management, Purdue University, West Lafayette,

IN 47907-1266, USA

Tel: +1 765 494 7905; Fax: +1 765 494 0327; E-mail: [email protected]

Aline M. Rezende-Parker is a former masters

degree student in the Department of Hospitality

and Tourism Management, Purdue University,

USA.

Alastair M. Morrison is Professor of Marketing/

Tourism in the Department of Hospitality and

Tourism Management, Purdue University, USA.

Joseph A. Ismail is an assistant professor in

tourism economics in the Department of Hospital-

ity and Tourism Management, Purdue University,

USA.

ABSTRACT

KEYWORDS: destination image, destinationselection, factor and cluster analysis, Brazil

The tourism industry in Brazil has been growingsharply in the past few years. Among its visitors,US citizens are one of the most important andattractive markets for the Brazilian tourism indus-try. It is believed, however, that most Americansstill know little about Brazil and what it has tooffer. This study had three objectives: determine theimages that American citizens interested in traveland tourism have of Brazil as a travel destination,determine if they differentiate Brazil from neigh-bouring Hispanic countries and group them accord-ing to similar images of Brazil. Online discussiongroups on travel and tourism were used to obtain asample of this population. The results indicatedthat ‘natural attractions/interest’ and ‘vacation

atmosphere/exoticness’ were the two most impor-tant images of Brazil among the respondents. Theresults also indicated that respondents were unableto differentiate Brazil from Hispanic countries.

BRAZIL AS A TRAVEL DESTINATIONDespite its imposing territorial proportions,growing economic importance and countlessattractions, Brazil still is not well known byUS residents. The proximity to Hispaniccountries in South America and the increas-ing popularity of ‘Latin music and culture’that reflects very little about Brazil help toblur further US residents’ images of Brazil.For Brazil’s tourism industry, it is very

important to understand and improve theimage that US residents have of the countryas a travel destination. American touristsspend the most while visiting Brazil amongall international visitors. They are also amongthose who stay the longest (11.8 days onaverage) and thus represent an important andgrowing market for Brazil.

1Also, as Reilly

2

emphasises, ‘An accurate assessment of pro-duct image is a prerequisite to designing aneffective marketing strategy’, and this helps acountry to offer what its visitors are reallyexpecting or create more realistic expecta-tions if necessary. But little research has beenconducted to find out more about Americanvisitors and what they think of Brazil as aplace to visit. Partially this is because of thedisinterest, until a few years ago, of the

Page 243

Journal of Vacation Marketing Volume 9 Number 3

Journal of Vacation MarketingVol. 9 No. 3, 2003, pp. 243–259,& Henry Stewart Publications,1356-7667

Page 2: Rezende morrisonismailjvm brazil

Brazilian government and private sector inpromoting Brazilian tourism abroad, andpartially it is due to the lack of Brazilianacademic and research institutions related totourism and travel as an economic activity.Therefore, the three main objectives of thisstudy were defined as:

— determine the image that American citi-zens interested in travel and tourism haveof Brazil as a travel destination;

— determine if Americans can differentiateBrazil from neighbouring Hispanic coun-tries;

— identify groups of Americans who holdsimilar images of Brazil.

DESTINATION IMAGE AND

SELECTION

Destination image and its componentsIt is accepted among tourism scholars thatimage is a crucial component in the processof destination selection.

3As Baloglu and

Bringberg4stated, ‘image differentiates tour-

ist destinations from each other and is anintegral and influential part of the traveler’sdecision process’. The definition of imageand its components, however, varies amongscholars and there is still no consensus. Afteran extended review of the literature on thetourism image concept, Ko and Park

5found

that there are usually three major emphasesamong researchers on the definition of desti-nation image. Some researchers, such asCrompton,6 emphasise the composite struc-ture of the image construct, defining it as thesum of beliefs, ideas and impressions that aperson has of a destination. Others use attitu-dinal components such as cognition andaffect to access the concept of image.

7A

third group sees it as an overall impression ofan object, as, for example, Milman andPizam,8 who defined destination image asthe visual or mental impression of a place, aproduct or an experience held by the generalpublic.Another aspect is what the components of

tourism image are, and again there are differ-ent views. Some authors have suggested thatimage structure is made up of one single

component, named cognitive. The cognitivecomponent is defined as an evaluation ofknown attributes of the product or the un-derstanding of the product in an intellectualway.

9Studies using the cognitive dimension

to measure image were conducted bySchneider and Sonmez,

10who measured

tourist images of Jordan; Bignon et al.,11who

measured the French image of the USA as adestination country; and Dimanche andMoody,

12who studied the image that Amer-

ican travel intermediaries had of New Or-leans as a tourism destination.Baloglu and Bringberg,

13however, in

agreement with other scholars, recognised anaffective dimension as a second componentof tourist image structure. The affectivecomponent is conceptualised as the feelingsthat an individual has about an object orplace.There is also a third stream who believe

the structure of destination image consists ofthree components. Gartner

14is part of this

group and has suggested that the three com-ponents of image are affective, cognitive andconative. According to him, the conativeimage component ‘is analogous to behaviorbecause it is the action component. After allinternal and external information is pro-cessed a decision is reached’. Baloglu andMcCleary

15suggested that the three compo-

nents were cognitive, affective and an overallimage formed by an interaction between thecognitive and affective components.Echtner and Ritchie

16presented a unique

three-dimensional model which separatedimage into two components (attribute-basedand holistic) (Figure 1). Each componenthad items that could have functional orpsychological characteristics, and each char-acteristic could reflect common or uniquefeatures of the destination. In this sense, athree-dimensional diagram would be thebest way to illustrate the components ofdestination image. As Figure 1 shows, thediagram has three scales. The first conti-nuum, attribute-holistic, is based on the ideathat image is composed of specific attributes(such as good restaurants, expensiveness, re-ceptiveness of the people) and mental pic-tures (holistic) such as people dancing on the

Page 244

An exploratory study of the image of Brazil as a travel destination

Page 3: Rezende morrisonismailjvm brazil

beach or couples having dinner in a romanticsetting.The functional-psychological continuum

divides perception of image into measurablecharacteristics (such as beaches, shoppingfacilities, national parks) and more intangibleor psychological characteristics (such as fame,mood, friendliness, etc), while the third con-tinuum identifies unique attractions of thedestination (such as the pyramids of Egypt orthe Eiffel Tower in France) and commonattractions (mountains, beaches and naturallife).

Measuring destination imageMost image studies dealing with tourismhave used attribute lists to measure the cog-nitive component of destination image.

17

Gartner18

used this structured approach tothe measurement of the image of four desti-nations in terms of recreation activities andattractions. Several items (perceptions of his-

torical sites, fishing activities, etc) were ratedusing a Likert scale that ranged from one tofive, and mean scores were used as inputs formulti-dimensional scaling. Chaudhary

19

studied the image of India as a travel destina-tion using a list of India’s attributes devel-oped from reports and articles about India inthe media and also from a small survey oftourists, who were asked to describe India intheir own words.Depending on the researcher’s conceptual

definitions and understanding of the struc-ture of image, however, measurementmethodologies and techniques vary. Reilly,

20

for example, studying the affective compo-nent of image, employed an unstructuredapproach. This method is an alternativetechnique that uses open-ended questionsand allows respondents to describe impres-sions of the product more freely. As hefurther explains:

‘Here, the purpose is not so much to

FUNCTIONAL CHARACTERISTICS

PSYCHOLOGICAL CHARACTERISTICS

ATTRIBUTES HOLISTIC

UNIQUE

COMMON

FUNCTIONAL CHARACTERISTICS

PSYCHOLOGICAL CHARACTERISTICS

ATTRIBUTES HOLISTIC

UNIQUE

COMMON

Figure 1The components of destination image

Source: Echtner and Ritchie, 1993

Page 245

Rezende-Parker, Morrison and Ismail

Page 4: Rezende morrisonismailjvm brazil

reveal hidden or unconscious motivationalstates but rather to allow the respondentsto describe the target stimulus in terms thatare salient to the respondents, rather thanresponding to the researcher’s pre-deter-mined image dimensions.’

Echtner and Ritchie,21

however,suggested that in order to capture all thecomponents of destination image (attribute-based/holistic, functional/psychological, com-mon/unique), a combination of structuredand unstructured methodologies should beused. In their study they demonstrated thatthe response to open-ended questions bettercaptured the holistic (affective), psychologi-cal and unique characteristics of a destina-tion; while scale items developed overseveral stages (content analysis, focus groups,literature review, etc) more effectively meas-ured the common, attribute-based compo-nents. This model was used by Murphy

22to

determine backpackers’ images of Australiaas a holiday destination. Some of the uniqueelements of Australia, for example, werefound to be kangaroos, the Great BarrierReef, the outback, Aboriginal culture andAyers Rock/Uluru. This Echtner andRitchie model was also adopted in this study.The authors acknowledge that several otherresearch studies have been completed ondestination image and its measurement, andthat the preceding review has encompassedonly a selected number of these previousworks.

METHODOLOGY

Population and sampleThe population of this study was Americancitizens interested in travel and tourism. Tosample this population, online discussiongroups on travel and tourism were used. Atotal of 37 online discussion groups werecontacted and messages were posted to theirsubscribers. The total number of subscribersto the lists was 9,183. Not all subscribersreceived the message in their mailboxes,however, since there is an option availablefor subscribers to read just posted messages

according to their convenience. It was notpossible to estimate exactly how many sub-scribers received and read the message. Atotal of 246 responses were received. As ageneral rule, for factor analysis the sampleshould be four or five times as many obser-vations as there are variables to be analysed.

23

Thus, for this study a sample of at least 190respondents was expected and surpassed.Following the procedures suggested by

Echtner and Ritchie,24

a combination ofstructured and unstructured methodologieswas applied to develop the questionnaire.The questionnaire was divided into threeparts. The first part presented three open-ended questions borrowed from Echtner andRitchie’s study on destination image andintended to capture the more affective orholistic components of image. The secondpart comprised a set of 39 statements orimage attributes of Brazil. Three of thosestatements were about Brazilian culture, cos-tumes, music and dance, and were includedin order to address the second objective ofthis study (the perceived differences betweenBrazil and neighbouring Hispanic countries).Respondents were asked to rate their per-

ceptions of the 39 Brazil attributes on a five-point Likert scale from one, for stronglydisagree, to five, for strongly agree. Threemethods were used to generate the list ofattributes: a content analysis of package toursto Brazil, a review of the literature on desti-nation image measurement and a focusgroup. The content analysis of package toursto Brazil was used to find out more aboutthe Brazilian tourism products that werebeing sold in the USA by tour operators andtravel agencies. The information identifiedincluded the most popular destinations, mainactivities and attractions featured, tripcharacteristics (trip length, transportationmodes, types of accommodation and type oftour) and the photographic images used todepict Brazil as a destination. The focusgroup was held at a large Midwestern uni-versity. The participants were asked toarticulate their images, feelings and percep-tions of Brazil, and their comments wereused to produce a more accurate list ofattributes.

Page 246

An exploratory study of the image of Brazil as a travel destination

Page 5: Rezende morrisonismailjvm brazil

The third part of the survey sought demo-graphic information, and included questionsregarding respondent interest in visiting Bra-zil in the future.

Data collectionThe survey was posted on the internet dur-ing the period from November 2001 toFebruary 2002. To contact the subjects,LISTSERVs (online discussion groups) wereused. Online discussion groups related totravel and tourism were identified throughsearch engines and directories (includingYahoo! and Google). Subscription e-mailswere sent to the owners of the lists. Whenthe subscription was accepted, a message waseither directly posted or sent to the modera-tors/owners, which they forwarded to sub-scribers once approved. Responses weresubmitted online and automatically sent toan electronic file. One follow-up postingwas made to the more active discussiongroups.

Data analysisThe data were entered in the SPSS 10.1software package and the analysis startedwith a grouping and counting of the wordsused by the respondents to answer the threeopen-ended questions. This process gener-ated tables with frequencies and percentagesbased on the number of people who an-swered the questions. More than one de-scription/word per person was allowed.Later, the most frequent answers were com-bined with the answers to the attribute-basedquestions to compose the diagram of thecomponents of destination image elaboratedby Echtner and Ritchie

25(Figure 1). Next

was the analysis of the 39 attributes to con-dense the information contained in the ori-ginal variables into a smaller set of newdimensions or constructs. Before proceedingwith the analysis, tests of correlations, in-cluding a Kaiser’s measurement of samplingadequacy (MSA ¼ 0.83) and a Bartlett’s testof sphericity (chi-square ¼ 2,834, df ¼ 74,p , 0.01), were used to verify the appropri-ateness of factor analysis. A principal compo-

nents analysis with varimax rotation wasperformed, and factor loadings smaller than0.3 were removed from the analysis.To achieve the third objective of this

study, cluster analysis was prepared on theresulting factors to determine groups ofAmericans with similar images of Brazil.A combination of hierarchical and non-hierarchical procedures was used. Ward’shierarchical clustering procedure was used apriori and the results suggested a four-factorsolution. Next, a K-mean analysis was runon the major clusters to confirm and definecluster membership. In addition, analysis ofvariance (ANOVA) was used to determine ifthe clusters were significantly different fromeach other. Chi-square tests were run todetermine if there were significant differ-ences among the four clusters in demo-graphic characteristics and the proportion ofvisitors and non-visitors to Brazil.

RESULTS OF THE EXPLORATORY

STUDY

Demographic characteristics of

respondentsThe majority of the respondents to the sur-vey were aged 18 to 34 years (54.7 per cent)and male (54.4 per cent). Some 45.1 per centof the respondents were married and 37.6per cent were single (15.1 per cent weredivorced/separated and 2.2 per cent werewidowed). Most of the respondents (81.4per cent) had some college experience; 73.9per cent had earned college degrees and 13.7per cent had postgraduate degrees. Similarproportions for low and median incomes(44.4 per cent) were found.

Interest in visiting BrazilOverall, most respondents were interested invisiting Brazil: 43.3 per cent said they were‘very interested,’ 37.8 per cent said that theywere ‘interested’ and 10.0 per cent were ‘notinterested’. This does not mean, however,that a high percentage of all Americans areinterested in visiting Brazil. Possibly it wasbecause the sample was composed of people

Page 247

Rezende-Parker, Morrison and Ismail

Page 6: Rezende morrisonismailjvm brazil

interested in travel and tourism and thusmore interested in knowing about foreigncultures and destinations than other people.The proportion of visitors to Brazil in thesurvey was 35 per cent, while the non-visitors comprised 65 per cent.A significant difference at p , 0.01 was

found among the answers of visitors andnon-visitors (chi-square ¼ 17.27, df ¼ 3) re-garding respondent interest in visiting Brazil.Previous visitors were more interested inreturning to Brazil than non-visitors were ingoing to Brazil for a first visit.

Images of BrazilOverall, positive images (means above 4)were found for nine statements: many placesof interest to visit (4.3); several natural parksand much wildlife (4.3); good place forgoing to the beach (4.3); variety of musicand dance (4.4); natural and scenic beauty(4.5); friendliness of the people (4.1); goodnightlife (4.3); adventure (4.0); and severalopportunities to increase knowledge aboutanother culture (4.3).Negative images of Brazil (means of less

than 3) were found for five statements: se-curity of public transportation (2.9); crowd-edness of the cities (2.3); local standards ofcleanliness and hygiene (2.8); beggars (2.5);and language barriers (2.9) (‘many peoplespeak English in Brazil’).Significant differences between visitors

and non-visitors were found for 24 items ofthe 39 attribute-based scales. Compared tonon-visitors, visitors had a more positiveimage of Brazil in all aspects. For the state-ment ‘most people visit Brazil because of theCarnival in Rio’, visitors showed a tendencyto disagree (2.9) while non-visitors weremore likely to agree (3.3).Regarding the comparative statements

(Table 1), the overall results showed thatAmericans were not able to differentiateBrazil from Hispanic countries. The meansfor the three comparative questions were 3.2for the statement on music and dance; 3.3for culture, art and costumes; and 3.5 for thestatement on Brazilian food. The number 3on the scale stood for ‘neither agree or T

able

1:C

om

para

tive

state

ments

:vis

itors

,non-v

isitors

and

all

resp

ondents

Sta

tem

ents

Visitor

mea

nN

o.N

on-v

isitor

mea

nN

o.T

otal

mea

nN

o.

Musicanddance

inBrazilarenotsimilarto

thoseofother

South

andCentralAmerican

countries.�

3.71

83

3.00

161

3.24

244

Theculture,artandcustomsofBraziliansarenotverysimilarto

thoseofother

South

and

CentralAmericacountries.�

3.81

84

3.16

159

3.38

243

Brazilian

foodisnotspicyandisdifferentfrom

someMexican

food.�

4.07

82

3.25

159

3.53

241

Note:Theseitem

swerereversecoded

foranalysis.Item

sweremeasuredonafive-pointLikertscalerangingfrom

1(stronglydisagree)to

5(strongly

agree).

�Significantlydifferentat

p,

0.01

Page 248

An exploratory study of the image of Brazil as a travel destination

Page 7: Rezende morrisonismailjvm brazil

disagree.’ This showed that respondentsseemed not fully to understand Brazil’s dif-ferences. A closer look to the results, how-ever, revealed that visitors were more certainof the differences between Brazil and Hispa-nic countries. Visitors’ mean responses to thethree items were significantly higher thannon-visitors, showing that they seemed todifferentiate Brazil better than non-visitors.

Subjective statements about BrazilTable 2 presents a summary of the answersprovided by the respondents to the threeopen-ended questions. For the first question,respondents were asked to provide theimages or characteristics that came to theirminds when they thought of Brazil as a traveldestination. The most frequent answers tothis question were beaches/beautiful beaches(42.1 per cent), Amazon jungle/Amazon

River/rainforest (40.8 per cent), Carnival(20.8 per cent) and Rio de Janeiro (19.1 percent). Among the answers, one negativecharacteristic, insecurity, was mentioned by10.6 per cent of the respondents.The second question asked the respon-

dents to describe the mood or atmospherethat they would expect to experience inBrazil. Fun/partying/festive/celebrating wasthe most popular answer (27.0 per cent),followed by friendly and relaxed/relaxing/laid back/loose (both 22.3 per cent). Againthere was a concern for safety, reflected by9.8 per cent of the respondents who ex-pected a frightening/dangerous/scary atmo-sphere in Brazil.The third question was ‘Please list any

distinctive or unique attractions that you canthink are in Brazil.’ Of those who answeredthis question, 57.5 per cent mentioned Ama-zon jungle/Amazon River/rainforest as a

Table 2: Most frequent answers to open-ended questions

Images evoked when thinking of Brazil as a vacation destination (n ¼ 235)• Beaches/beautiful beaches (42.1 per cent)• Amazon jungle/Amazon River/rainforest (40.8 per cent)• Carnival (20.8 per cent)• Rio de Janeiro (19.1 per cent)• Climate (hot/warm/humid/tropical) (18.3 per cent)• Natural features (mountains/open spaces/beautiful scenery) (14.0 per cent)• Wonderful music/dancing people/samba (14.0 per cent)• Nature/biodiversity/flora/fauna (13.0 per cent)• Great food/good food/examples of Brazilian food (13.2 per cent)• Insecurity (crime/violence/high crime rates) (10.6 per cent)

Descriptions of the atmosphere or mood expected while visiting Brazil (n ¼ 233)• Fun/partying/festive/celebrating (27.0 per cent)• Friendly (22.3 per cent)• Relaxed/relaxing/laid back/loose (22.3 per cent)• Hospitable/accommodating/welcoming/warm (12.8 per cent)• Frightening/dangerous/scary (9.8 per cent)

Distinct or unique attractions in Brazil (n ¼ 214)• Amazon jungle/Amazon River/rainforest (57.5 per cent)• Rio de Janeiro (31.3 per cent)• Christ the Redeemer/Corcovado (26.2 per cent)• Waterfalls/Iguacu Falls (22.4 per cent)• Beaches (21.0 per cent)• Carnival (14.9 per cent)• Sugar Loaf Mountain (14.0 per cent)• Sao Paulo (7.0 per cent)

Page 249

Rezende-Parker, Morrison and Ismail

Page 8: Rezende morrisonismailjvm brazil

unique attraction in Brazil. The second mostcited answer was Rio de Janeiro (31.3 percent), followed by Christ the Redeemer/Corcovado (26.2 per cent), which is anattraction located in Rio de Janeiro. Threeother answers associated with Rio de Janeirowere provided but presented separately sincethe same respondent could have mentionedthem simultaneously: Sugar Loaf Mountain(14.0 per cent), Copacabana Beach (6.5 percent) and Ipanema Beach (6.0 per cent).Iguacu Falls was mentioned by 22.4 per centof the respondents.The most frequent answers to the three

open-ended questions, as well as the attrib-ute-based statements with a mean of 4.0 orhigher, were placed on Echtner andRitchie’s

26three-dimensional model of desti-

nation image (Figures 2, 3 and 4). Figure 2presents the attribute-holistic and functional-psychological components of Brazil’s image;Figure 3 provides the functional-psychologi-cal and common-unique components; andFigure 4 illustrates the common-unique andattribute-holistic component. The placementof the items within the diagram was subjec-tive, but still it gives a clearer idea of how

the respondents pictured Brazil in theirminds in terms of its unique attractions andholistic images as, for instance, people relax-ing on beautiful beaches, having fun andcelebrating the Carnival or contemplatingthe beauty of the rainforest.

Results of factor analysisFactor analysis was applied to determine theunderlying dimensions of Brazil’s image thatlater were used to group respondents withsimilar images of Brazil. Eleven factors witheigenvalues greater than one emerged fromthe first analysis and the variance explainedwas 62.1 per cent. The Cronbach’s alphacoefficients of the last five factors were low(0.55, 0.55, 0.5, 0.4 and 0.2), however, andthe overall alpha was 0.75. To improve thisresult, an analysis of the correlations betweeneach attribute statement and the total scorefrom the scale was conducted to find attri-butes that did not correlate well with theoverall score from the scale. One attribute,‘Beaches are Brazil’s main attraction’, had alow correlation (-0.0927) and was removedfrom further analysis. Then the changes in

Interesting places to visit (4.36)Natural parks and wildlife (4.33)Beaches (4.33)Natural scenic beauty (4.55)Nightlife (4.33)Variety of dance and music (4.42)

People on beautiful beaches (42.1%)Jungle, rivers and rainforest environment (40.8%)Celebration of Carnival (20.8%)

Friendly people (4.13)Adventure (4.02)Opportunity to increase knowledge (4.37)

Festive and partying environment (27.0%)with friendly people (22.3%) andalso relaxing and laid back (22.3%)

FUNCTIONAL CHARACTERISTICS

PSYCHOLOGICAL CHARACTERISTICS

ATTRIBUTES HOLISTIC

Interesting places to visit (4.36)Natural parks and wildlife (4.33)Beaches (4.33)Natural scenic beauty (4.55)Nightlife (4.33)Variety of dance and music (4.42)

People on beautiful beaches (42.1%)Jungle, rivers and rainforest environment (40.8%)

Friendly people (4.13)Adventure (4.02)Opportunity to increase knowledge (4.37)

Festive and partying environment (27.0%)with friendly people (22.3%) andalso relaxing and laid back (22.3%)

FUNCTIONAL CHARACTERISTICS

PSYCHOLOGICAL CHARACTERISTICS

ATTRIBUTES HOLISTIC

Figure 2 The attribute/holistic and functional/psychological components of Brazil’s image

Page 250

An exploratory study of the image of Brazil as a travel destination

Page 9: Rezende morrisonismailjvm brazil

COMMON

Variety of dance and music (4.42)

FUNCTIONAL CHARACTERISTICS

PSYCHOLOGICAL CHARACTERISTICS

Interesting places to visit(4.36)Natural parks and wildlife (4.33)Beaches (4.33)Natural scenic beauty (4.55)Nightlife (4.33)

Amazon jungle and Amazon River (57.5%)Rio de Janeiro (31.3%)Christ Redeemer/Corcovado (26.2%)Iguaçu Falls/waterfalls (22.4%)

Friendly people (4.13)Adventure (4.02)Opportunity to increase knowledge (4.37)

FUNCTIONAL CHARACTERISTICS

PSYCHOLOGICAL CHARACTERISTICS

UNIQUE

Festive and partying environment (27.0%)with friendly people (22.3%) andalso relaxing and laid back (22.3%)

Figure 3The common-unique and functional/psychological components of Brazil’s image

Interesting places to visit (4.36)Natural parks and wildlife (4.33)Beaches (4.33)Natural scenic beauty (4.55)Nightlife (4.33)Friendly people (4.13)Adventure (4.02)Opportunity to increase knowledge (4.37)

Beautiful beaches (42.2%) withfriendly and festive people (22.3%)

Variety of music and dance (4.42)Rio de Janeiro (31.3%)Christ Redeemer/Corcovado (26.2%)Carnival (20.8%)

COMMON

UNIQUE

ATTRIBUTES HOLISTIC

COMMON

UNIQUE

Adventure (4.02) and contemplation of the rainforestand its flora and fauna (40.8%)

Figure 4The attribute/holistic and common/unique component of Brazil’s image

Page 251

Rezende-Parker, Morrison and Ismail

Page 10: Rezende morrisonismailjvm brazil

Cronbach’s alpha after removing particularattributes were evaluated and one morestatement, ‘Most people visit Brazil becauseof the Carnival in Rio’, was also removedfrom the analysis due to its great influence inlowering the reliability coefficient. Afterthese procedures, the factor analysis wasrepeated and ten factors emerged in a newsolution. To retain an attribute statement, acut-off point of 0.3 for factor loadings wasestablished, and all the attributes satisfied thiscriterion.The Cronbach’s alpha coefficients were

again measured and two factors were re-moved due to poor reliability (0.2 and 0.4).The factor analysis was repeated for a thirdtime and the final solution had eight factorsand 33 items, which explained 57.7 per centof the variance. The overall Cronbach’s al-pha coefficient showed a high level of relia-bility at 0.89 (Table 3). Table 4 shows thefactor solutions with their means, percen-tages of variance explained, eigenvalues andloadings. Factor scores were computed bytaking the averages for the attribute state-ments within each factor.The first factor extracted was composed of

six attribute statements and was named ‘nat-ural attractions/interest’, based on the twostatements concerning natural parks, wildlifeand natural scenic beauty, and one statementregarding historic attractions. The otherthree statements were related since theyemphasised the idea that there are many

places to see and things to do (shopping) inBrazil, and that the country has enoughinteresting attractions to be visited by itselfand not just while touring neighbouringcountries. This factor had the highest eigen-value (7.76) and explained 10.36 per cent ofthe variance.The second factor, ‘safety and comfort’,

included statements on relaxation, friendli-ness and safety. This factor had an eigenvalueof 2.67 and explained 8.64 per cent of thevariance. Factor three grouped the threestatements determining whether or not re-spondents differentiated Brazil from otherSouth and Central American countries.Named ‘cultural comparisons’, the factor in-cluded statements involving art, costumes,food, music etc. It had an eigenvalue of 2.12and explained 7.44 per cent of the variance.‘Facilities and information’ was the fourth

factor and included six attribute statementsrelated to travel and tourist facilities, such asrestaurants, hotels, guides and availability oftourist information. Its eigenvalue was 1.57and the variance explained was 7.02 percent.Factor five was named ‘vacation atmo-

sphere/exoticness’, had an eigenvalue of1.39 and explained 6.58 per cent of the vari-ance. ‘Economic and social conditions’ wasthe label attached to the sixth factor, whichincluded five attribute statements regardingpolitical stability, beggars, transportationconditions, crowdedness of the cities and

Table 3: Final results of factor and reliability analysis

FactorsNumber ofitems

Reliability coefficients(alpha)

1 Natural attractions/interest 6 0.742 Safety and comfort 5 0.723 Cultural comparisons 3 0.784 Facilities and information 5 0.625 Vacation atmosphere/exoticness 4 0.626 Economic and social conditions 5 0.637 Transportation 2 0.508 Adventure/learning 3 0.52

Total 33 0.89% of variance explained 57.72

Page 252

An exploratory study of the image of Brazil as a travel destination

Page 11: Rezende morrisonismailjvm brazil

Table

4:Perc

enta

gesofvari

ance

expla

ined,eig

envalu

es,

and

loadin

gsoffa

cto

rs

Item

Fac

tor

Var

iation

Eig

enva

lue

Loa

ding

Factor1:Naturalattractions/interest(4.11)

10.36

7.76

2Thereareseveralnaturalparksto

visitandmuch

wildlifeto

see(4.33)

0.78

1Therearemanyplacesofinterestto

visitin

Brazil(4.36)

0.70

38

Brazilisnotonlyworthvisitingwhiletouringneighbouringcountries�

(3.97)

0.64

9Braziloffersmuch

interm

sofnaturalscenicbeauty

(4.55)

0.61

3Brazilhasmanyhistoricsitesandmuseums�

(3.83)

0.53

11

Brazilisagoodplace

forshopping(3.65)

0.41

Factor2:Safetyandcomfort(3.48)

8.64

2.67

35

Brazilisagoodplace

totakechildren(3.48)

0.58

33

Brazilisarestfulandrelaxingplace

tovisit(3.68)

0.65

32

Brazilian

peoplearefriendly(4.13)

0.61

27

Ingeneral,Brazilisasafeplace

tovisit(3.35)

0.59

6TheAmazonisnotadangerousplace

tovisit(3.27)

0.65

Factor3:Culturalcomparisons(3.38)

7.44

2.12

8Musicanddance

inBrazilaresimilarto

thoseofother

South

andCentralAmerica

countries�

(3.24)

0.87

10

Theculture,artandcustomsofBraziliansareverysimilarto

thoseofother

South

and

CentralAmericacountries�

(3.38)

0.88

31

Brazilian

foodisspicyandsimilarto

someMexican

food�(3.53)

0.65

Factor4:Facilitiesandinform

ation(3.50)

12

Goodtouristinform

ationisreadilyavailableforBrazil(3.24)

7.02

1.57

0.72

13

Tourswithguides

areavailablein

Brazil�

(3.85)

0.61

18

Good-qualityrestaurantsareeasy

tofindin

Brazil(3.80)

0.32

26

Travellingto

Brazilisconvenientandeasy

(3.07)

0.34

36

Good-qualityhotelsareeasy

tofindin

Brazil(3.56)

0.38

(con

tinued

over

leaf)

Page 253

Rezende-Parker, Morrison and Ismail

Page 12: Rezende morrisonismailjvm brazil

Table

4:(continued

)

Item

Fac

tor

Var

iation

Eig

enva

lue

Loa

ding

Factor5:Vacationatmosphere/exoticness(4.12)

6.58

1.39

4Brazilisagoodplace

forgoingto

thebeach

(4.33)

0.67

10

Brazilhasgoodnightlife(4.33)

0.55

22

Brazilisan

exoticdestination(3.99)

0.62

14

Therearemanysportsandrecreationalopportunitiesin

Brazil(3.84)

0.43

Factor6:Economicandsocialconditions(2.74)

6.54

1.35

23

Localstandardsofcleanlinessandhygienearehighin

Brazil(2.87)

0.40

25

Thereisnotalotofpoliticalinstabilityin

Brazil�

(3.02)

0.70

24

Therearenotmanybeggarsin

Brazil�(2.52)

0.73

21

Brazilian

citiesarenotcrowded�(2.34)

0.58

17

Publictransportationissafein

Brazil�

(2.95)

0.37

Factor7:Transportation(3.12)

5.99

1.27

16

Transportationin

Brazilisinexpensive�

(3.16)

0.74

15

Transportationwithin

Brazilisconvenient(3.09)

0.60

Factor8:Adventure/learning(3.94)

5.11

1.04

34

Thereareseveralopportunitieswithin

Brazilto

increase

myknowledgeaboutanother

culture

(4.36)

0.45

37

Brazilisadestinationthatpeopledream

ofvisiting(3.46)

0.55

28

Aholiday

inBrazilisarealadventure

(4.01)

0.70

� Theseitem

swerereversecoded

foranalysis

Page 254

An exploratory study of the image of Brazil as a travel destination

Page 13: Rezende morrisonismailjvm brazil

standards of cleanliness. Its eigenvalue andpercentage of variance explained were 1.35and 6.54 per cent respectively. Factor sevenwas called ‘transportation’, and had two asso-ciated attribute statements and an eigenvalueof 1.27. Factor eight was called ‘adventure/learning’ and had an eigenvalue of 1.04. Thelast two factors explained together 11.1 percent of the variance.The factors with the highest average

attribute mean scores were ‘naturalattractions/interest’ (4.11) and ‘vacationatmosphere/exoticness’ (4.12). The factorwith the lowest mean score was ‘economicand social conditions’ at 2.74.

Identification of clustersCluster analysis was applied to identify andclassify respondents according to similarimages of Brazil based on the eight dimen-sions found in the factor analysis. Table 5shows the four clusters found through thisanalysis. The respondents in Cluster I werecharacterised as possessing generally positiveimages of Brazil as a vacation destination.This cluster had the highest mean score forthe ‘safety and comfort’ factor at 3.97, whichwas significantly different statistically fromthe other three clusters. Cluster I also hadthe highest scores for ‘cultural comparisons’(4.18), ‘facilities and information’ (3.88) and‘vacation atmosphere/exoticness’ (4.53).This cluster, however, had negative imagesfor two of the factors, ‘transportation’ (2.48,significantly lower than the others) and ‘eco-nomic and social conditions’ (2.78).The respondents in Cluster I clearly differ-

entiated Brazil from other South and CentralAmerican countries (4.18). They were malein the majority (57.89 per cent), 35 to 49years old (50 per cent) and married (58.8 percent). This cluster had the smallest portion ofsingle people (17.6 per cent) and the largestportion of divorced people (23.5 per cent)(Table 6). Some 60 per cent of Cluster I’srespondents had median incomes ($40,000–79,000), and 80 per cent had college degrees.Most importantly, it had more respondentsthat had been to Brazil than any other cluster(65 per cent). T

able

5:M

eansofim

age

facto

rsam

ong

the

fourgro

upsofre

spondents

Dim

ension

s(f

acto

rs)

Clu

ster

In¼

20

Clu

ster

IIn¼

93

Clu

ster

III

48

Clu

ster

IVn¼

53

F-r

atio

Sig

.le

vel

Naturalattractions/interest

4.31a

4.4

a4.3

a3.38b

55.88

,0.0001

Safetyandcomfort

3.97b

3.56a

3.4

a3.15a

7.11

,0.0001

Culturalcomparisons

4.18b

3.1

a4.03b

2.94a

38.55

,0.0001

Facilitiesandinform

ation

3.88a

3.67a

3.28b

3.2

a14.94

,0.0001

Vacationatmosphere/exoticness

4.53a

4.3

a4.16a

3.6

b10.50

,0.0001

Economicandsocialconditions

2.78

2.91a

2.52b

2.6

10.07

,0.0001

Transportation

2.48a

3.20b

3.31c

3.0

b23.97

,0.0001

Adventure/learning

4.03a

4.22b

3.81a

3.5

a8.00

,0.0001

Note:Valuesaremeansbased

onafive-pointLikertscale.Meanswithadifferentsuperscript(a,bandc)aresignificantlydifferentfrom

each

other,based

onScheffeTestsat

p,

0.05.Forexam

ple,fortransportation,Cluster

IIIissignificantlydifferentfrom

ClustersI,IIandIV

.ClustersIIandIV

arenot

significantlydifferent.Cluster

Iissignificantlydifferentfrom

ClustersII,IIIandIV

.

Page 255

Rezende-Parker, Morrison and Ismail

Page 14: Rezende morrisonismailjvm brazil

Clusters II and III were intermediate clus-ters in terms of positive to negative images.Cluster II had the highest means for ‘naturalattractions/interest’ (4.4) and ‘adventure/learning’ (4.22). These respondents had neu-tral images of ‘safety and comfort’ (3.56),‘cultural comparisons’ (3.1), ‘facilities andinformation’ (3.67) and ‘transportation’(3.2). The majority of the members of Clus-

ter II (72.8 per cent) had never been toBrazil, 59 per cent were male and 68 percent had college degrees.Cluster III had negative images of Brazil

for ‘economic and social conditions’ (2.52),and neutral images for ‘safety and comfort’(3.4), ‘facilities and information’ (3.28),‘transportation’ (3.31) and ‘adventure/learn-ing’ (3.81). This cluster demonstrated posi-

Table 6: Characteristics of clusters

Characteristics

Cluster 1(n ¼ 20)%

Cluster 2(n ¼ 93)%

Cluster 3(n ¼ 48)%

Cluster 4(n ¼ 53)%

Gender�Male 57.89 59.14 69.57 34.62Female 42.11 40.86 30.43 65.38

Chi-square ¼ 13.390, df ¼ 3, p ¼ 0.004Age

18–19 5.0 9.7 22.9 19.620–24 25.0 19.57 20.83 17.6525–34 15.0 27.2 8.3 15.735–44 25.0 14.13 20.83 5.845–49 25.0 10.8 14.5 19.650–54 5.0 13.0 8.3 15.755–64 – 5.43 4.1 5.865 and over

Marital statusMarried 58.8 43.0 41.8 48.0Widowed – 1.1 2.3 2.0Divorced 23.5 13.9 11.6 18.0Separated – 1.1 2.3 –Single 17.6 40.7 41.8 32.0

EducationLeft high school – 5.4 2.1 3.7High school 15.0 16.3 16.6 13.2Some college 5.0 9.7 8.3 5.6Associates degree 40.0 34.7 27.1 35.8Bachelors degree 30.0 22.8 25.0 26.4Masters degree 10.0 10.8 20.8 11.3Doctoral degree – – – 3.7

IncomeLow (less than $39,000) 25.0 48.9 42.2 46.9Median ($40,000–79,000) 60.0 41.1 51.1 38.8High ($80,000 and more) 15.0 6.7 6.7 8.2Unwilling to answer – 3.3 – 6.1

Visit Brazil�Yes 65.0 27.2 54.2 13.2No 35.0 72.8 45.8 86.8

Chi-square ¼ 29.631, df ¼ 3 , p ¼ 0.000

�Significant at alpha ¼ 0.01 by chi-square analysis

Page 256

An exploratory study of the image of Brazil as a travel destination

Page 15: Rezende morrisonismailjvm brazil

tive images for ‘natural attractions/interest’(4.3), ‘cultural comparisons’ (4.03) and ‘va-cation atmosphere/exoticness’ (4.16). Some69 per cent were male, 51 per cent were inthe median income bracket ($40,000–79,000) and 54.2 per cent had visited Brazil.Additionally, this was the cluster with thehighest portion of respondents in the 18–19age bracket (22.9 per cent).Cluster IV was characterised as having

more neutral to negative images of Brazil.This cluster had statistically significant lowermeans for ‘natural attractions/interest’ (3.38)and ‘vacation atmosphere’ (3.6). It also hadthe lowest means for ‘safety and comfort’(3.15) ‘cultural comparisons’ (2.94), ‘facilitiesand information’ (3.2), and ‘adventure/learning’ (3.5). This group had the highestproportion of non-visitors to Brazil (86.8 percent), and was composed of females in themajority (65.3 per cent). Some 77.2 per centhad college degrees, and 46.9 per cent hadincomes of $39,000 or less.

CONCLUSIONSThe first objective of this study was todetermine the image that Americans inter-ested in travel and tourism have of Brazil as atravel destination. To achieve this, a combi-nation of structured (attribute-based scale)and unstructured (open-ended questions)methods of destination image measurementwas applied, as suggested by Echtner andRitchie.

27The attribute-based scale was used

to measure the common, attribute-basedcomponents of image, while the open-endedquestions were used to measure the uniqueand holistic components.The variables contained in the attribute-

based scale (structured approach) were factoranalysed and summarised into a smaller set ofnew composite dimensions. Five of the eightfactors that emerged from the analysis wereequivalent to the factors found by Echtnerand Ritchie

28in their study on the measure-

ment of destination image. These resultssupport their suggestions that this methodprovides a generalised framework that can beused to compare and contrast the images ofmost tourist destinations.

The results indicated that ‘natural attrac-tions/interest’ and ‘vacation atmosphere/exoticness’ were the two most prevalentimages of Brazil held by respondents. Theanalysis of the open-ended questions furtherclarified this result by revealing that theimages that came to respondents’ mindswhen they thought of Brazil as a place tovisit were mainly of beautiful beaches andthe rainforest/jungle. The two most popularunique attractions in Brazil were found to bethe Amazon jungle and Rio de Janeiro. Upto now, these two attractions (along withIguacu Falls) have been intensely promotedby the majority of the most important touroperators in the USA, as was indicated bythe content analysis of package tours toBrazil. These are, however, two contrastingimages that attract people with very differentinterests and profiles.In terms of negative images, the factor

analysis and the answers to the open-endedquestions indicated that respondents do nothave good perceptions of the economic andsocial conditions in Brazil, and they have aconcern for their safety while visiting thecountry. Words such as crime, violence,poverty, dangerous and frightening wereused by approximately 10 per cent of re-spondents to express the images and atmo-sphere that they would see and experience inBrazil. Therefore, it is very important forBrazil to address these concerns if it is toimprove the images held by prospective andactual visitors and become more competitivein the international market.Regarding the second objective, to deter-

mine if respondents differentiated Brazil fromneighbouring Hispanic countries, the resultswere close to the midpoint of the scale. Theanswers to the three comparative statementswere neutral, indicating that respondentswere not able to pinpoint clearly the differ-ences between Brazil and Hispanic countries.A closer look at the answers of visitors andnon-visitors, however, showed that visitorswere more certain of the differences thannon-visitors, particularly with regard to Bra-zilian food. Furthermore, the results of thecluster analysis revealed that there was onegroup of respondents (Cluster I) who clearly

Page 257

Rezende-Parker, Morrison and Ismail

Page 16: Rezende morrisonismailjvm brazil

differentiated Brazil from Hispanic countries.Past visitors to Brazil composed the majorityof this group. The inconclusive results for theoverall analysis might be the lack of sufficientknowledge of the Brazilian culture amongthe majority of respondents. Another possibleexplanation could be the unclear and mislead-ing images gained from the media or otherinformation sources.The third objective was to group respon-

dents based upon similar images of Brazil.The cluster analysis did not reveal an empha-sis on one or more specific images amongdifferent groups of respondents; rather, itindicated that images varied from positive tonegative among the clusters. Cluster I had anoverall positive image, Clusters II and III hadintermediate positive to neutral images,while Cluster IV had neutral to negativeimages of Brazil for the eight image factors.The ‘natural attractions/interest’ and ‘vaca-tion atmosphere/exotic’ factors had veryhigh mean attribute statement scores forthree of the four clusters.

Marketing implications and future

research recommendationsWhile this research has definite marketingimplications for Brazilian tourism, there arebroader implications for destinations in gen-eral. The first of these is that past visitors to adestination may not only hold the most posi-tive images of that destination, but also havethe most accurate understanding of the desti-nation’s limitations, such as transportationwithin Brazil. Testimonials and other formsof endorsements by these past visitors couldprove useful in clarifying potential visitors’images of the destination, and in alleviatingany misapprehensions among these people.The second implication arises from the find-ing that Americans were generally unable todifferentiate Brazil clearly from neighbouringSpanish-speaking countries. This may be in-dicative of a general tendency for people toassume that neighbouring destinations arealike, unless they have specific informationto refute this belief. Therefore, it is incum-bent on marketing managers to develop im-

age and communications strategies thatclearly separate their destinations from near-by areas, as well as from the principal com-petitors.A third area of implications is the dilemma

raised by the findings about Brazil’s image as avacation destination. Should, for example,the marketing approaches be changed forpeople with different images of the country,especially for those with negative images?Alternatively, should a ‘one-size-fits-all’ ap-proach be employed in which the same imageis communicated to all potential visitors? Thecurrent movement toward greater destinationbranding appears to suggest the second strat-egy. Moreover, trying to promote two ormore different images to an overall audience,such as US travellers, could be confusing andineffective. If the destination is difficult topicture and categorise in travellers’ minds, asappears to be the case for Brazil, then it maynot even make it into travellers’ destinationselection processes. Destinations might alsohave a strong appeal to specialised nichemarkets, however, including, for example,the attractiveness of Brazil’s rainforest, riversand waterfalls to nature-based tourists. In thiscase, marketers need to develop further andexpand upon the broader destination imageto match each niche market’s interests andperceptual expectations.In Brazil’s case, the broader destination

image has not yet been effectively establishedin the US market and additional investmentin image development seems warranted.Specifically, more effort is needed to changeand improve Americans’ perceptions of Bra-zil. For example, special attention will haveto be given to the first image cluster, whichhad negative images of transportation andeconomic and social conditions. Since thisgroup was composed of past visitors to Brazilin the majority, these respondents’ percep-tions were based upon real experiences andindicate a weakness in the Brazilian tourismproduct that must be addressed by Braziliantourism authorities.It should be acknowledged that this study

was exploratory in nature, and as such theresults need to be verified through furtherempirical research. It is also recognised that

Page 258

An exploratory study of the image of Brazil as a travel destination

Page 17: Rezende morrisonismailjvm brazil

the sample had some limitations. First, theresponse rate to this online survey was low,while being sufficient to meet the conditionsfor using the selected statistical analysis pro-cedures. Second, the population under studywas defined as being US citizens interestedin travel and tourism. Therefore, the findingscannot be generalised to the entire US popu-lation, since the participants in the sampleare expected to have a better understandingand knowledge of differences among desti-nations than other Americans.Finally, participation in the study was vo-

luntary. This might suggest that some peoplealready had an interest in Brazil and weremore willing to participate in the survey thanothers. This can be verified by the fact thatthe proportion of visitors to Brazil was higherin the sample than for the US population as awhole. Even given these limitations, thisstudy makes a contribution by further validat-ing the Echtner and Ritchie approach to themeasurement of destination image.

REFERENCES

(1) Embratur (2001) ‘Estudo da Demanda Tur-istica Internacional 2000, Brasilia’.

(2) Reilly, M. D. (1990) ‘Free elicitation ofdescriptive adjectives for tourism image as-sessment’, Journal of Travel Research, Vol. 28,No. 4, pp. 21–26.

(3) Baloglu, S. and McCleary, K. W. (1999) ‘Amodel of destination image formation’, An-nals of Tourism Research, Vol. 26, No. 1, pp.868–897; Crompton, J. L. (1979) ‘An assess-ment of the image of Mexico as a vacationdestination and the influence of geographicallocation upon that image’, Journal of TravelResearch, Vol. 17, No. 4, pp. 18–23; Gart-ner, W. C. (1989) ‘Tourism image: Attrib-ute measurement of state tourism productsusing multidimensional techniques’, Journalof Travel Research, Vol. 28, No. 2, pp. 16–20;Ko, D. W. and Park, S. H. (2000) ‘Fiveaspects of tourism image: A review’, Interna-tional Journal of Tourism Science, Vol. 1, No. 1,pp. 79–92; Reilly, ref. 2 above.

(4) Baloglu, S. and Bringberg D. (1997) ‘Affec-tive image of tourism destinations’, Journal ofTravel Research, Vol. 35, No. 4, pp. 11–15.

(5) Ko and Park, ref. 3 above.

(6) Crompton, ref. 3 above.(7) Dobni and Zinkhan, 1990, and Mazursky

and Jacoby, 1986, both cited by Ko andPark, ref. 3 above.

(8) Milman, A. and Pizam, A. (1995) ‘The roleof awareness and familiarity with a destina-tion: The central Florida case’, Journal ofTravel Research, Vol. 33, No. 3, pp. 21–27.

(9) Gartner, W. C. (1993) ‘Image formationprocess’, Journal of Travel and Tourism Mar-keting, Vol. 2, Nos 2/3, pp. 191–215.

(10) Schneider, I. and Sonmez, S. (1999), ‘Ex-ploring the touristic image of Jordan’, Tour-ism Management, Vol. 20, No. 4, pp. 539–542.

(11) Bignon, V. N., Hammitt, W. E. and Nor-man, W. C. (1998) ‘French perceptions andimages of the United States as a destinationcountry’, Tourism Analysis, Vol. 3, No. 3/4,pp. 159–171.

(12) Dimanche, F., and Moody, M. (1998) ‘Per-ceptions of destination image: A study ofLatin American intermediary travel buyers’,Tourism Analysis, Vol. 3, No. 3/4, pp. 173–180.

(13) Baloglu and Bringberg, ref. 4 above.(14) Gartner, ref. 9 above.(15) Baloglu and McCleary, ref. 3 above.(16) Echtner, C. M., and J. R. B. Ritchie (1993)

‘The measurement of destination image: Anempirical assessment’, Journal of Travel Re-search, Vol. 31, No. 4, pp. 3–13.

(17) Evidence for the assertion made in thissentence comes from Gartner, ref. 9 above,and Ko and Park, ref. 3 above.

(18) Gartner, ref. 9 above.(19) Chaudhary, M. (2000), ‘India’s image as a

tourist destination — A perspective of for-eign tourists’, Tourism Management, Vol. 21,No. 3, pp. 293–297.

(20) Reilly, ref. 2 above.(21) Echtner and Ritchie, ref. 16 above.(22) Murphy, L. (1999) ‘Australia’s image as a

holiday destination: Perceptions of back-packer visitors’, Journal of Travel and TourismMarketing, Vol. 8, No. 3, pp. 21–45.

(23) Hair, J. F., Anderson, R. and Tatham, R.(1987) ‘Multivariate Data Analysis withReadings’, Macmillan, New York, pp.233–348.

(24) Echtner and Ritchie, ref 16 above.(25) Ibid.(26) Ibid.(27) Ibid.(28) Ibid.

Page 259

Rezende-Parker, Morrison and Ismail