Development of a Measurement Scale For Analysing Destination Choice Criteria in the Tourist Decision Process

Citation:

Jane M. Summers and Janet R. McColl-Kennedy (1995) ,"Development of a Measurement Scale For Analysing Destination Choice Criteria in the Tourist Decision Process", in E - European Advances in Consumer Research Volume 2, eds. Flemming Hansen, Provo, UT : Association for Consumer Research, Pages: 406-411.

European Advances in Consumer Research Volume 2, 1995      Pages 406-411

DEVELOPMENT OF A MEASUREMENT SCALE FOR ANALYSING DESTINATION CHOICE CRITERIA IN THE TOURIST DECISION PROCESS

Jane M. Summers, University of Southern Queensland

Janet R. McColl-Kennedy, University of Queensland

Examination of the tourist's decision process is of particular importance to tourism marketers and managers. Despite this, very few studies have examined tourist destination choice criteria. No study to date has attempted to develop a destination choice scale. This paper outlines the methodology used to develop a measurement scale for analysing choice criteria used by tourists in selecting a travel destination. Seven steps are discussed including: specifying the domain of the construct from the literature; generation of sample items from literature review, experience surveys and focus groups; collection of data; purifying the measure through coefficient alpha and factor analysis; collection of data; assessing reliability through coefficient alpha and split-half measures and assessing validity through item-to-total correlation. The paper concludes by providing direction for subsequent application in destination choice research.

Tourism is a major international business today. Increased real income, longer holiday periods with pay, improved opportunities for mobility, better education and wider dissemination of information have all contributed toward changing attitudes about travelling away from home (Goodall, 1988b).

Examination of the traveller's decision process is of particular importance to tourism marketers and managers. Understanding how tourists acquire, organise and use information in the process of choosing a destination, including the criteria they use to evaluate alternatives, enables tourism marketers to better design, position and promote products.

This paper will detail the methodology used to develop a measurement scale for analysing choice criteria used by tourists in selecting a travel destination. For the purposes of this paper, the term tourist will refer to an international pleasure or holiday traveller.

DESTINATION CHOICE

Destination choice, transportation methods, accommodation styles, activities, budget factors and reservation requirements, are referred to as 'subdecisions' (Moutinho, 1987) that tourists need to make during their holiday planning.

Subdecisions

The tourist's subdecisions are affected by a range of influencing factors and considerations. Things like, perceived image, reasons for travel, previous experience, travel destination promotion and the impact of advice from others can all contribute to the final outcome.

Moutinho (1987) suggests that the 'subdecision' of destination choice is generally made initially, at the information gathering stage and finally, at the choice stage of the decision process. He further posits that perceived image of the destination, tourism destination marketing, previous experience and travel intermediaries advice, all impact on ultimate choice of destination for the international tourist.

Although image represents a very personal, composite view of a destination's tourism potential, the images held by any person are not static. At any given time, a person possesses a certain accumulation of images about a great number of holiday experiences, some personal, but many second hand (Goodall, 1988a). These images, for each individual, will be modified with each additional experience and by further exposure to a variety of information sources.

Destination choice could be considered to take place within the broader context of the tourist decision making process. This process would consist of a series of steps, whose sequence would not differ greatly from the general model proposed by Engel, Blackwell and Miniard (1993). The process should also incorporate such concepts as motivation, image and perceptions as proposed by Moutinho (1987) and Goodall (1988b).

Essentially, the decision process begins with a recognition of travel needs and desires, which results in a motivation to travel. This then stimulates a search phase and finally a choice of holiday destination and type is made.

The decision process is also influenced by a number of external and internal factors. Moutinho (1987) identified self-concept, motivation, culture, personality, perception and cognition, perceived risk, attitude and intention and family influences, as critical factors that effect the travel decision at various stages

The tourist buying decision is unique in that it is usually an investment with no tangible rate of return, and the purchase is often prepared and planned through savings made over a considerable period of time. That is, the vacation tourist will invest, with no expectation of material and economic return on their purchase, in an intangible satisfaction (Moutinho, 1987).

The implications for tourist destinations are clear. Unless a destination is counted among the tourist's current set of mental images, it has no chance of being selected as the holiday destination. Secondly, the image of the destination must be very positive in the tourist's mind for it to be selected from other alternatives. Finally, the actual experience of the tourist with that destination, once chosen, must match the expectations created by the image, otherwise the tourist generally will not wish to return, and will not recommend that destination to others.

Destinations compete, even in growing markets for tourists, since their products are often close substitutes. A positive image and high profile are essential to keep tourists coming, and this is something that needs constant maintenance from tourism marketers.

The 'image' of a destination is therefore critical in terms of its successful marketing. Images form part of the consumer's decision process, significantly influencing the search and evaluative criteria stages. Basic psychological, physical, cultural, social and economic motivators govern behaviour and these are conditioned by experiences, information and individual preferences to create images of reality (Stabler, 1988).

When attempting to measure or analyse destination images, Meinung (1989), suggests the attractiveness of a tourism region can be grouped into three major categories. These are:

(1) Primary or static factors - the landscape, forms of landscape, the climate, the means of travel to and within the region and the culture of the region;

(2) Secondary or dynamic factors - tourism supply (accommodation, catering, entertainment), administrative and political settings, and the market position of the region (growth, decline etc.); and

(3) Tertiary or current decision factors - marketing of the region, prices in the target region and country of origin and the organisation of tourism within the region in relation to administration and economic organisation

Meinung concludes his observations by stating that interrelationships among these factors are greatest where variability is greatest (between the secondary and tertiary factors).

DEVELOPING A SCALE FOR ANALYSING DESTINATION CHOICE

Very few studies have examined tourist destination choice criteria (Lancaster, 1966; Goodall, 1988; Meinung, 1989; and Stabler, 1988). All of these studies examine choice criteria for European destinations.

The classic Languedoc-Roussillon study (Goodall 1988b) concentrates on a very specific area of France and fails to consider the general destination choice criteria used by potential tourists.

A recent study by Yau and Chan (1990), attempts to examine destination choice criteria for South-East Asia with respect to Hong Kong and thus adopts a more generalist approach. Despite the interest in destination choice, no studies to date have developed a destination choice scale.

The current study attempts to develop a valid and reliable scale for measuring and testing destination choice criteria for international holiday travellers to Australia.

METHODOLOGY

The fundamental objective in measurement is to produce X₀ scores which approximate X_T scores as closely as possible. Unfortunately, the researcher never knows for sure what the X_T scores are. Rather, measures are always inferences. The quality of these inferences depends directly on the procedures that are used to develop the measures and on the evidence supporting their "goodness". This evidence typically takes the form of some reliability or validity index (Churchill, 1979).

There appears to be general agreement amongst empirical researchers that for considerations of item selection, reliability and validity are of primary importance for the construction of measurement scales and their ultimate applicability in further research (Lundstrom, 1974).

In operationalising this, the procedure suggested for developing multi-item marketing measures by Churchill (1979) was used. He lists seven critical steps or stages and for most of these a number of recommended coefficients or techniques that can and should be used.

The first step is to specify the domain of the construct, through a comprehensive literature review. The second and third steps are to generate a sample of items for the scale and to collect some data to analyse. The recommended techniques for these stages are literature reviews; experience surveys; insight stimulating examples, critical incidents and focus groups.

The fourth step is to purify the measure using a combination of coefficient alpha and factor analysis. Step five is to collect additional data which is then assessed for reliability and validity in steps six and seven. Once again coefficient alpha is recommended for this along with split-half reliability, multitrait-multimethod matrix and criterion validity.

The final step is to develop norms of the measure and for this the average and summary distribution of scores is the recommended technique.

Specify the Domain of the Construct

The first step in the procedure was to specify the domain of the construct being measured. In this case the scale would be used to measure dimensions of vacation destination criteria. From the literature (Goodall, 1988a; Meinung, 1989; Moutinho, 1987; and Stabler, 1988) this construct has been specified as follows.

Holiday destinations have two dimensions: a physical dimension and an intangible dimension. The physical dimension comprises the actual geographic location, its physical features, climate, infrastructure and landscape. The intangible dimension comprises the tourist's perceived image of the physical features, travel motives, previous experience, destination marketing and reference group influence.

Thus in order to determine the specific criteria used by a tourist in their choice of Australia as a destination it is necessary to consider a large range of potential variables.

Generate a Sample of Items

The second step in the procedure was to generate items which captures this construct. Stellitz et al. ( 1976) suggests that for generating possible items in exploratory research, literature searches, experience surveys and focus groups are generally the most productive. In this study all three methods were used.

The literature review revealed a number of possible items that were either used in similar studies or were suggested as areas for further research. These items were then tested against and compared to those items that emerged from four (4) focus groups each with 10 overseas students who were studying in Australia. Three of the groups consisted of Malaysian students and one of a mixture of other nationalities (Taiwanese, Korean, Japanese and American). For the majority of students within these groups (98%), this was their first visit to Australia, though most (67%) had travelled to other overseas locations.

Lundstrom (1974), suggests a number of generally acceptable methods in item selection: (1) non-response, (2) judgement, (3) t-test, (4) biserial correlation and (5) point biserial correlation (the latter two deal only with dichotomous responses and thus were discarded).

The non-response method was employed next during a series of four (4) in-depth interviews with a selection of overseas students studying in Australia. Two of the students used in the interviews had participated in the focus groups and two had not. Analysis suggested that there appeared to be no major differences between their input.

During these interviews the items generated from the literature and from the focus groups were discussed as to their potential for consideration in choice of Australia as a holiday destination. Those items that had a high level of non-responses during the in-depth interviews were considered to be unreliable and were subsequently discarded.

Meinung's (1989) three categories of destination criteria: (1) primary factors; (2) secondary factors; and (3) tertiary factors, were then used as hypothetical groupings for the items selected. As a further test of the appropriateness of these groupings and the items themselves, 25 students (both overseas and Australian) were asked to assign all items from the total item pool of 30 items to one of these three categories as they saw fit. A "does not fit any" category was included as well in an attempt to decrease the possibility of bias.

All responses were then collated and the frequencies of item allocations for each of the four possible groupings considered. Any items that were allocated more than 50% of the time to the "does not fit any" category (that is a frequency of 12.5 or more) were discarded. Items needed to receive a count of more than 12.5 (50%) to be included in a grouping.

Next, a different group of 25 students was given these selected groupings of items and asked to rank on a 5 point rating scale, how appropriately they felt each item matched its recommended grouping. This data was then collated and an item by item summary statistic for each was computed, giving the mean and standard deviation.

TABLE 1

T-TEST RESULTS FOR ITEM SELECTION

Edwards (1957) suggests sorting respondent's scores of an item. Then selecting the first quartile and the fourth quartile of the respondents for comparison. This is based on the underlying assumption that these two groups provide the criterion for discriminating between individual items. If there is a significant difference between the mean scores on an item the item will be accepted for inclusion in the scale. If not, the item will be discarded accordingly. A t-test for item selection was performed and all items were significant at the level 0.01. They were thus included in the scale for further testing. Table 1 lists these items and their respective scores.

Collect Data

A sample of data was then collected from 124 overseas students asking them to rank the importance of the items selected if they were considering a vacation to Australia. The data was gathered by self-administered questionnaires and all respondents were university business major students. The profile of the sample is shown in table 2.

Purify the Measure

In an attempt to confirm the hypothesised groupings of items, the data was explored by factor analysis with varimax rotation using the Statistical Package for the Social Sciences (SPSS/PC+). The eigenvalue criterion was used to determine the number of factors. The factor analysis extracted 5 factor groupings (not the three groupings as originally hypothesised), which in total accounted for 55 percent of the variance in the data.

These factors and their loadings are shown in Table 3. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy (an index for comparing the magnitudes of the observed correlation coefficients to the magnitudes of the partial correlation coefficients) was 0.79. This is considered to be more than acceptable by Kaiser's (1974) standards for the factor analysis to continue.

Coefficient alpha was calculated for each of the proposed groupings and the results are shown in Table 3. For factors one, three and four the reliability coefficients could be regarded as high (Nunnally, 1967; Peterson, 1994), indicating that they could be considered a highly reliable grouping of destination criteria. The coefficient for factor two was considered to be good, whilst that calculated for factor 5 was considered to be marginal (0.5 being the generally accepted cut-off point) (Nunnally, 1967).

This indicates that perhaps this measurement dimension needs to be improved before further research use, possibly through further exploration of factor loadings. As shown in Table 3, no item was found to have a correlation coefficient which was not significant at the 0.01 level.

Reliability

Two methods were used to test the overall reliability of the scale: the alpha coefficient and split-half method. Table 4 shows that the reliability coefficient was 0.8096. This high score (Nunnally, 1967; Peterson, 1994), indicates that the scale could be considered a highly reliable measure of these destination choice criteria. The alpha coefficients for the two halves of the scale are above the minimally acceptable level set by Nunnally (1967), though the difference in the two values, does indicate some instability (Table 4).

TABLE 2

SUMMARY OF SAMPLE CHARACTERISTICS

TABLE 3

FACTOR LOADING FOR DESTINATION CHOICE VARIABLES

Validity

The internal validity of the destination choice criteria scale was assessed by the item-to-total correlation method on the 24 items in the scale. None of the item-to-total correlations for all items of the scale (Table 5) were found to have a correlation coefficient that was not significant at the 0.01 level. This result suggests that the items exhibit internal consistency.

TABLE 4

RELIABILITY COEFFICIENT FOR THE DESTINATION CHOICE CRITERIA SCALE

TABLE 5

ITEM-TO-TOTAL CORRELATION: DESTINATION CHIOCE CRITERIA

SUMMARY

In order to establish meaningful measures of destination choice criteria, it is necessary to have reliable measures. The primary method for measuring the reliability of the multi-item measures in this study was Cronback alpha coefficient. In general the reliability coefficient for the sample of data collected (0.8096) compared very favourably with those achieved in other studies (Peterson, 1994) and was of a high enough level to warrant confidence in the measures (Nunnally, 1967). Furthermore, results of the item-to-total correlation analysis also showed that internal consistency of the research instrument was prevalent. As such, this study has contributed significantly to the field by providing a scale which could be considered to be reliable and valid for further research use.

LIMITATIONS

One of the major limitations of the study is the question of external validity. As there was no other known scale measuring the same constructs, testing of external validity using a correlation of this scale with one of parallel form was not possible.

Similarly, there could be some concern over the sample size and the sampling frame chosen for this study. The data was collected using a convenience sampling method, which was deemed appropriate for this form of exploratory research.

IMPLICATIONS FOR MARKETERS AND DIRECTIONS FOR FUTURE RESEARCH

There is no doubt that the image of a destination or region is a significant influencer in a tourist's demand for commodities and services, and thus the level and pattern of their expenditures (Stabler, 1988). The images that tourists have of a destination or region are, in turn, significantly dependant on their experiences, on marketing information and on information provided by others (friends or travel intermediaries).

Destination areas are aggressively competing to attract holiday-makers and being able to better understand the criteria that are important to them in making their decisions will greatly benefit both destination marketers and consumer researchers alike.

REFERENCES

Churchill JR., G.A. (1979), 'A Paradigm for Developing Better Measures of Marketing Constructs', Journal of Marketing Research, Vol. XVI ( February), pp. 64-73.

Edwards, A.L., ( 1957), Techniques of Attitude Scale Construction, New York:Appelton-Century-Crofts.

Engel, J.F., Blackwell R.D.and Miniard, P.W.,(1993), Consumer Behaviour, 7th edn., Fort Worth: Harcourt Brace College Publishers.

Goodall, B. (1988a), 'How Tourists Chose their Holidays: An Analytical Framework', in B. Goodall and G Ashworth (Eds.), Marketing in the Tourism Industry; Promotional destination regions, London, Croom Helm, pp 1- 10.

Goodall, B. (1988b), 'Changing Patterns and Structure of European Tourism', in B. Goodall and G Ashworth (Eds.), Marketing in the Tourism Industry; Promotion of destination regions, London, Croom Helm, pp 18 - 36.

Lancaster, K.J. (1966), 'A new Approach to Consumer Theory', Journal of Political Economy, Vol. 84, pp 132-157.

Lundstrom, W.J., (1974), The Development of a scale to Measure Consumer Discontent, Unpublished Ph.D. Dissertation, The University of Colorado.

Meinung, A., (1989), 'Determinants of the attractiveness of a tourism region', in S.F Witt and L Moutinho,(Eds.), Tourism Marketing and Management Handbook, Englewood Cliffs: Prentice Hall Inc., pp 99 - 101.

Moutinho, L. (1987), 'Consumer Behaviour in Tourism', European Journal of Marketing, Vol 21, No. 10, pp 3 - 44.

Nunnally, J.C. ( 1967), Psychometric Theory, 2nd Ed. New York:McGraw-Hill.

Peterson, R.A. ( 1994), 'A Meta-analysis of Cronbach's Coefficient Alpha', Journal of Consumer Research, Vol21, September, pp 381-391.

Stabler, M.J., ( 1988), ' The Image of Destination Regions: Theoretical and Empirical Aspects', in B. Goodall and G Ashworth (Eds.), Marketing in the Tourism Industry; Promotional destination regions, London, Croom Helm, pp133-161.

Stellitz, C. Wrightsman L.S. and Cook S.W. (1976), Research Methods in Social Relations, 3rd edn., New York: Holt, Rinehart and Winston.

Yau, O.H.M and Chan, C.F.,(1990), 'Hong Kong as a Travel Destination in South-East Asia: a multidimensional approach', Tourism Management, June, pp. 123-132.

Yau, O.H.M.,(1994), Consumer Behaviour in China: Customer Satisfaction and Cultural Values, London:Routledge.

----------------------------------------