The construct lifestyle in market segmentation Essay
The construct lifestyle in market segmentation
Keywords Lifestyles, Market segmentation, Consumer behaviour, Tourism Abstract The swift and wide-ranging changes that present-day society is undergoing are leading to an increasing personalization in consumer behaviour patterns. These are increasingly less well explained by socio-demographic and economic criteria. This effect seems to be particularly well reflected in tourism. As one of the chief characteristics of this market is its heterogeneity, there is a need to include other variables, such as lifestyle, in order to segment it more adequately.
This would permit a greater depth of knowledge of the variables influencing tourist behaviours, rendering them accessible to businesses, which could thus better satisfy tourists’ needs and wants by matching the services they offer more efficiently to them. Defines the construct “lifestyle” based on the activities, interests and opinions approach. Offers in addition certain innovations of scholarly interest, but also of practical use for business. Although this variable is not a brand-new concept, it is still being actively researched.
The explanatory value of traditional criteria is in steady decline. This is because the individuals who together make up the market every day provide examples of very similar purchase behaviour patterns on the part of people differing considerably in socio-economic and demographic terms and vice versa, with growing personalization of consumer habits being observed.
Hence, such criteria are tending to be used only as variables describing the market as a whole. The deep and wide-ranging changes which present-day society is undergoing must also be taken into account, among them being the increasing numbers of single-person households, a decrease in the average number of members making up families, the entry of women (better and better qualified) into the labour market, and the growth in purchasing power of households. This leads to the conclusion that the evolution of society is the principal cause of the need to include further criteria (lifestyles) as essential variables in detailed knowledge of consumers. This paper considers the place that these variables hold in overall consumer behaviour.
Integrating lifestyles into overall consumer behaviour models Generalisations on global models There are various models that endeavour to explain consumer behaviour. Their origin lies in the considerable differences that separate consumers, the various kinds of purchase decision and the contexts in which they are made, in response to unique and differing purchase circumstances, which nonetheless usually share some similarities. The need for such models arises from the help that they can offer in commercial decision-making, since they provide a simplification of the process by which a purchase decision is reached, especially those of them which are global models. Among others, two are of particular interest; those developed by Howard and Sheth (1969) and by Engel et al. (1978). The first includes the notion of culture and cultural norms, while the second accepts the direct influence exercised over a purchase decision by a larger range of variables, taking into account, on the one hand, the surroundings, including cultural norms and values, and, on the other, personal variables, such as personality and lifestyles. This model was updated in Engel et al. (1990), with a view to responding to behavioural processes in situations of both high and low risk and implications.
The model put forward by Alonso Rivas (1997), like earlier models, includes variables such as socially established values, as it sees them as very significant in understanding the culture of a society. They bring out tendencies to act or to respond, personality and psychographic facets related to motivation, and can be used in defining markets.
Individuals will adopt given behaviour patterns representative of their lifestyles, and as a consequence will purchase different types of products or services (Howard, 1993; Howard and Sheth, 1969; Engel et al., 1978).
These scholars thus consider that behavioural variations in purchases, even if there is no question of a mix of socio-demographic variables coming into play, lead to a need for research into lifestyle as a potentially influential factor. Specifications of a given service: tourism
These global models of consumption become specific when purchase behaviours relating to particular products or services are analysed. In this case it is necessary to take into account the characteristics peculiar to each market. Tourism is a strategic sector world-wide, with Spain, moreover, holding a prominent place in the rankings. Hence, it is of interest to determine what place lifestyle variables have in tourist behaviour, particularly in the light of the special idiosyncrasy and characteristics of both tourism and tourist demand. Initially, with the aim of establishing general guidelines to support later justifications based on lifestyles, it is necessary to delimit the market that is to be analysed, since many definitions have been put forward for tourism.
Thus, there follow details of those generally accepted, as adopted in this study. First, reference may be made to the definition of tourism by the Statistics Committee of the United Nations as the set of activities carried out by people in their trips and stays in places other than their usual place of residence for less than one continuous year for leisure, business or other purposes. More recently, at the International Conference on Travel and Tourism Statistics organised by the Canadian Government and the World Tourism Organization (WTO), held in Ottawa in June 1991, new definitions were laid down, so as to standardize the measurement of Tourism (WTO, 1999):
Tourism: The activities of persons travelling to, and staying in, places outside their usual environment for not more than one continuous year for leisure, business, and other purposes.
Traveller: Any person on a trip between two or more locations (WTO, 1995).
Visitor: Any person travelling to a place other than that of his/her usual environment for fewer than 12 consecutive months and whose main purpose of travel is not to work for pay in the place visited. Tourist (overnight visitor): Visitor staying at least one night in collective or private accommodation in the place visited.
Same-day visitor (excursionist): Visitor who does not spend the night in collective or private accommodation in the place visited.
Furthermore, tourism is a service with its own unique nature, thanks to its chief characteristics. Among these are the way it is perceived when being purchased and later consumed, the immobility of factors going to form it, the limits on capacity, its seasonal nature and the impossibility of trying out holiday products before purchasing them (Bull, 1996).
All of this, together with personal and social factors characterizing demand for tourism, as essentially human, leads to many forms of consuming tourism products. This stimulates diversity in supply, whether current or potential. The most striking characteristics of the demand for tourism are: .
Constant positive development, although not always steady growth. .
Both temporal and spatial concentration. There is concentration in time, because, despite changes currently under way, most tourism occurs on a limited range of dates of the year, mainly in summer. There is spatial concentration because the countries of origin of tourists and those hosting them (including Spain) do not vary greatly.
Heterogeneity. Demand consists of many types of tourists with different behaviours arising from their wide range of motivations and the varying
factors influencing them.
The varying combinations that travellers can choose at present, in terms of destinations, accommodation, means of transport, services and activities, have led to the creation of a wide range of differentiated products. These are adapted to the new forms of demand constantly arising, as a result of the very varied needs which are satisfied by tourism.
A good theory of tourist motivation must consider the total needs of travellers and not limit its focus to one need, must be able to manage dynamic changes within individuals and society, and must balance needs influenced by other people with those determined by individuals themselves. An important theory of tourist motivation is that of the travel career ladder (Pearce, 1988, 1991).
This theory argues that travel behaviour reflects a hierarchy of travel motives (see Figure 1). Based on Maslow’s hierarchy of needs, this approach retains the premise that lower levels on the ladder usually have to be satisfied before the person moves to higher levels of the ladder. Nevertheless, it is the total pattern of travellers’ motives which describes them rather than a focus on any one single motive. By expanding and extending the range of specific needs at each ladder level, a comprehensive
and rich catalogue of the many different psychological needs and motives can be realized.
In this model, destinations are considered as different travel experiences, where travellers select destinations, transport, accommodation and activities according to their lifestyles, personality and motivational profile.
Furthermore, the travel career ladder is a multi-motive model with a flexibility and variability that recognize that motivation may change over time and across situations.
The satisfaction of this range of needs has led, as previously noted, to different forms of tourism arising. Among these, the following are of particular note: .
Domestic versus international travel, according to whether the destination chosen is inside or outside the frontiers of the tourist’s own country. .
Leisure versus business travel, depending on the purposes of the trip. Business travel can be defined as including meetings, conferences and conventions, training and sales missions, and general promotional and professional contact work. Pleasure travel, by way of contrast, is seen as either independently organized or organized through a travel agent and is likely to have a greater regional spread with more diverse destinations. In general, pleasure travel tends to be dispersed across destinations, while business travel is concentrated in key economic centres.
Sun-and-sea holiday, with the chosen destination on the coast in an area where the climate is usually good. This is also known as mass tourism. .
Cultural tourism. This is focused on arts events, monuments, museums, exhibitions, visits to historical and archaeological sites, festivals or folklore events, and the like. It responds to cultural or artistic motivations on the part of those involved.
Sports tourism. The motivation here is either to participate in some sport (water, mountain, golf, hunting, fishing and so forth) or to attend any sports event.
Health tourism. The aim of such trips is to undergo some form of health treatment, normally involving specially qualified staff. In this category would fall visits to health farms, spas and thermal baths.
Rural tourism. This is the planned use of resources in a rural zone with the aim of increasing the general welfare of both the community and the visitor, while preserving the environment.
Apart from those mentioned above, there are other types of tourism of some interest such as religious pilgrimages, travel with the aim of attending a foreign language course, adventure holidays, trips awarded as incentives, alternative tourism, multi-site or touring holidays, and so on. These and the forms noted above are not mutually exclusive, because in many cases tourists can enjoy several forms simultaneously, responding to a combination of motivations. If the previously mentioned diversity of behaviours on the part of tourists does occur, particularly as a consequence of differentiated lifestyles, then this would provide ratification and support for segmenting the market on the basis of such variables. Thus, the overall theme that will be investigated is that of how to improve lifestyle assessment techniques in order to use this criterion for segmenting the tourist market.
Application of attitudes, interests, opinions (AIO) to studying lifestyle The construct AIO Once the potential explanatory power that lifestyle variables appear to have within overall consumer behaviour models has been demonstrated, along with the relevance and specific characteristics of tourism, this study will move on to develop a methodological line based on activities, interests and opinions, rather than other approaches to lifestyles, for various reasons.
The first reason is that beliefs, opinions and preferences, being highly specific, allow forecasts of behaviour to the extent that they reveal predispositions (Hustad and Pessemier, 1974; Triandis, 1979). That is, these variables form a part of the modern concept of an attitude as an affective state of an individual linked to a disposition to act in a certain way when faced with some stimulus, as it was put by Vazquez Casielles et al. (1994, p. 135). This definition has its origins in the views Â
expressed by Allport (1935), who stated that it is a mental and neurological state of ability to react, organized through experience, which exerts a direct and dynamic influence on individuals’ responses to those circumstances in which they find themselves. Another source for the definition is the ideas put forward by Triandis (1979), who considered that it is an emotion-charged idea predisposing to a particular course of action in a given situation. These definitions indicated that attitudes have an impact on individuals’ behaviours, differing as a function of the circumstances arising.
As for activities, these are a concept relating to the use made of time available by any individual. They may be part of a job, obligatory or necessary actions in the individual’s day-to-day life, work in the home, or leisure (Feldman and Hornik, 1981). However, since the first three of these are related to statements including interests and opinions, this section includes only those activities which are undertaken in spare time.
This lifestyle methodology, based on research into motivation, is shaped by variables from rational, concrete, behavioural psychology and yields an
overall view of the consumer. Individuals normally adopt a lifestyle in accordance with the dominant features of the social groups to which they belong or wish to belong (Baudrillard, 1970).
The methodology establishes typical profiles of lifestyles as a function of work and leisure habits and the world-view individuals themselves hold. It casts light on the behaviours and predispositions immediately prior to an action, and for this reason is classified within the behaviourist current. The construct “lifestyle” is defined by including variables like activities, referring to the way in which individuals spend their time and money; interests, which are those things in their immediate surroundings they consider more or less important; and opinions, the view they have of themselves and of the world around them (Pessemier and Tigert, 1966; Wells and Tigert, 1971; Tigert, 1971; Wind and Green, 1974; Unger, 1982).
Practical applications of this methodology have revealed those behaviours and predispositions closest to action (core interests and buying intentions),
this mostly having been done in a sectorial context, as can be seen from Table I. Apart from these applications relating to a range of areas, there have been others specifically focused on tourism. Although by no means an exhaustive list, Table II picks out those of greatest interest for the present study. As may be seen from the list given in Table II, use of lifestyles is not a new idea. However, businesses look for swift results, and so there is a lack of theoretical studies that have been empirically validated.
Moreover, consideration of research relating to tourism points to the presence of two methodologies, depending on the degree to which the information is more general or more specific:
(1) Segmentation of the population in accordance with general lifestyles. This permits the definition of broad categories of consumer in response to the lifestyle criterion chosen, yielding information on their way of life and outlook. From this an overall definition of their structure of
necessities and consumption can be obtained.
(2) Market segmentation according to features of lifestyle specific to the product or service concerned. The segmentation study is based on questions linked to a given type of consumption. A more appropriate name might be “consumption style”.
The second methodology has been more extensively used, owing to demands for information from businesses. If the items included in these researches are considered, this methodology would really be measuring the style of tourism in which travellers are engaged. These results do not extend to other products or services consumed by individuals, because their lifestyles are not truly being measured. However, market segmentation by general lifestyle would allow more in-depth awareness of variables influencing consumers’ behaviour, regardless of the product or service consumed. This would bring with it the possibility of more wide-ranging applications of the methodology. As a generalist perspective of this sort has not been widely employed, the fact of choosing this method can in itself be considered an important contribution to this research.
A critical evaluation of the AIO line
The methodological problems associated with this approach relate to a range of features. Surveys are over-long not just because of the large amount of data but also, perhaps even more, because of the scales used, mostly based on scoring, for each and every item in the questionnaire. This means that the respondents have to stop and think hard so as to rate each single item, leading to great fatigue, often not really necessary. Questionnaires are customarily selfadministered, which is an added handicap, as the questions used are sometimes not easy to understand, and this form of survey does not allow for clarifications. Both these factors, length and difficulty in understanding questions, lead to the spoiling of a great many questionnaires, especially those filled in by elderly folk, and to considerable biasing of data by flagging
The construct attention during the long time-span required for completion (Valette-Florence, 1994). The bulkiness of the ensuing data also triggers a need for multiple statistical analyses, including factorial analysis of principal components, used with the aim of reducing the number of variables. These retain only a part of the information, to which further techniques are applied, thus leading in the end to a considerable amount of unused residual information, as the raw data are not handled directly, but undergo successive impoverishment.
In respect of the items used, it should be stated which constitute independent variables and which dependent, so as to allow comparison of studies and to be able to interpret the results obtained from them correctly. However, a good few pieces of research have left this unclear, which means that the interpretations put on them are doubtful and difficult to check.
The impact of lifestyles on segmentation of the tourism market Aims and hypotheses of the research
The comments made above concerning the criticisms brought forward against the variables traditionally used in segmenting the market for tourism, particularly their limited explanatory value in the context of developed countries (Ritchie and Goeldner, 1987; Fisher, 1990; Mitchman, 1991; Witt and Moutinho, 1994; Lambin, 1995), justify the aims of the present study. These are an improvement in the techniques for measuring lifestyle, followed by use of the construct (as defined in this paper) to divide up tourists, provided that a relationship can be demonstrated between lifestyle and consumer behaviour. This would yield differentiated character-product-destination groups and permit deeper awareness of the characteristics of markets. The intention would be to draw up a more human picture of purchasers and to explain their behaviour patterns. In order to attain these overall aims, a set of more limited objectives specifically linked to them was established:
An investigation of the possible existence of a direct relationship between the variables under consideration, lifestyle and tourist behaviour. .
Grouping of individuals into sets as a function of the lifestyle construct. .
A segment-by-segment review of the variables under consideration to check for the existence of a direct relationship.
A characterisation of groups as a function of lifestyle variables and tourists’ behaviour.
These were carried out taking into account the difference between short trips (fewer than four nights) and long trips (four or more nights), in both cases for leisure purposes. This was done because major changes are taking place in the consumption of tourism, running counter to its traditional seasonality. There is, for instance, an increasing trend towards leisure trips on weekends and public holidays and the spacing-out of leave entitlements over the whole of the year. The double effect of this is to reduce the summer holiday peak period and to extend tourism beyond the high season.
The lifestyle variable would offer in-depth knowledge of individuals, allowing their characteristics to be related to their behaviours. Analysis of this relationship would permit the definition of strategies in accord with this type of variable, adapted to the segments delimited by the study, and the provision for them of those products or services that suit their needs and wants. In order to confirm these assumptions, it is necessary first to check whether there is a significant correlation between the variables, lifestyle and travel. This check is to be performed by testing two hypotheses:
H1. There is a direct relationship between the lifestyles of individuals, as measured by activities, interests and opinions, and the behaviour of consumers of short-stay tourism.
H2. There is a direct relationship between the lifestyles of individuals, as measured by activities, interests and opinions, and the behaviour of consumers of long-stay tourism.
In order to attain the aims and objectives laid down for this research project, an empirical check on the proposed methodology was performed. This was done by carrying out a personal survey and interview at the homes of the subjects selected as a sample.
The universe of the study is made up of those people aged over 15 years who are resident in provincial capitals with a population greater than 100,000 in one autonomous community in Spain. The sample size was 400 valid surveys, with a sample error of 5 percent, and a confidence level of 95.5 percent in the worst-case scenario. The design of the sample was based on a multi-stage stratified random selection, with the method of choosing subjects making up the set being by random routes.
In view of the kind of study involved, a structured questionnaire was drawn up, including, among others, the sections shown in Figure 2. In this research, as a method for gaining knowledge of the consumer through the variables affecting the decision-making process, once these were defined, various multivariate statistical techniques were applied (see Figure 3). They suited the purpose, as there are a number of factors with an impact on, and giving shape to, behaviour in regard to tourism.
In order to overcome one of the most serious problems previously noted, the excessive length of lifestyle questionnaires, the variables under study were measured with two different types of scale: ordinal and nominal. The ordinal scales used were five-point Likert scales. These indicate the weighting assigned by individuals to a set of statements about interests and opinions, from strong agreement to strong disagreement. As an example, the following can be cited: Say how far you agree or disagree with the following comments and whether each of them affects you personally: (Interviewer: show Card 1: 1. Strongly disagree; 2. Disagree slightly; 3. Neither agree nor disagree; 4. Agree slightly; 5. Strongly agree; X Do not know/no answer): .
Why did you choose that destination?
Similar information was requested in respect of longer trips (four nights or more), again using open questions.
The use of variables measured on nominal scales allowed attainment of one of the main aims of this study, which was to obtain the same amount of information in considerably less time than usual with other research into lifestyle. By cutting down on the time spent answering the questionnaire,
the quality of the information gathered was improved, since it was possible to avoid the bias arising from the excessive time needed. Moreover, it was possible to achieve a substantial reduction in the cost of fieldwork and thus of the whole research project, saving resources and making it more affordable for companies.
A set of techniques easily put into operation by businesses was used for the statistical treatment of data.
The first of these techniques was factor analysis of principal components with varimax rotation, applied to interests and opinions, those lifestyle-shaping variables that were measured by ordinal Likert scales (43 items), the statistical package used for this being SPSS (Statistical Package for the Social Sciences). The second technique was factor analysis of multiple correspondences, used on the activities variables, nominal variables that round out the construct lifestyle (110 items).
Finally, two independent multiple correspondence analyses were used on variables covering behaviour during trips. One related to trips of fewer than four nights and the other to leisure travel of longer duration (308 items in each case). This technique was carried out using the statistical package SPAD (Systeme Portable pour l’Analyse des Donnees).
The main intention behind using the two factor analyses was to homogenise the information, and thus obtain new, continuous variables (factor scores), to constitute the inputs for later multivariate analyses, without loss of statistical information. This is because all of the factors are extracted, so that the entire variability of each variable is explained. Furthermore, these factors are linearly independent and so any possible collinearity that might be present was eliminated, revealing the structure underlying the data.
As the aims previously noted were not merely to reduce the number of variables, it is of no real interest to explain the factors. Since the hypothesis adopted was to look for the existence of some direct relation
and behaviour during tourist or leisure journeys, in order to support segmentation of the market using this criterion (as measured by the variables and scales explained in this paper, and organized so as to be easily applied by business, as noted above), the choice fell on an application of multivariate technique. This was the analysis of canonical correlations, with the statistical package used for it being BMDP (Biomedical Package).
This approach allows the detection of any relationship between a set of independent variables, or “predictors”, and another set of dependent variables, or “criteria”. It can also establish whether this relationship is significant, and measure its intensity, with its most important purpose being the analysis of dependency (Burtschy, 1994, p. 216). In order to check the hypotheses put forward, Bartlett’s chi-squared statistical test is used, acting as an approximation to the null hypothesis. In this case the canonical correlations are nil, which means that the groups of variables are independent of one another but, when this is not so, it shows up those relationships which are significant and quantifies them.
This technique was applied to the factor results, rather than directly to the original variables, so as to eliminate any potential risk of multi-collinearity, because the factors are linearly independent. In this way also any possible instability in the canonical weights was avoided and matrices of identical structure were yielded.
The number of factors included in canonical correlation analysis is not the totality of them, as would be the case with other techniques, but rather criteria are adopted to select fewer factors. If this were not done, the number of variables would be larger than the number of individuals, so that the results would not be interpretable. For each case the criteria adopted are indicated below.
For factor analysis of principal components, the number of factors which should be included in later canonical analysis was determined using Kaiser’s (1961) empirical criterion. This requires retention of those factors with an eigenvalue greater than the average of these values, that is greater than one. It should be noted in addition that a sufficiently large proportion of the variance is explained, more than 60 percent of the accumulated variance, this being an desirable percentage for social sciences (Hair et al., 1998).
As for the analyses of multiple correspondences, first, the empirical rule of analysing the histogram of decreasing eigenvalues was followed. Note was taken of where discontinuities occurred because of the assumption that a nonrandom circumstance is showing up in these cases. Second, a fresh calculation of the inertia rates was made on the basis of pseudo-eigenvalues, in accordance with the guidelines put forward by Benzecri (1979). This scholar points out that Â
the original value calculated through the technique is very pessimistic in respect of the information represented, and considers it advisable to perform a new calculation of variance using the formula quoted below. The aim is to get a truer perspective of the variance explained through multiple correspondences analysis (Lebart et al., 1997). Thus, use was made of the following formula:
In this formula, s indicates the number of active questions and the eigenvalue used in analysing correspondences arising from Burt’s Table. Both allowed calculation of the cut, taking into account not only this discontinuity but also the explanation obtained from this number of factors (almost 100 percent of the variance).
The number of factors included in statistical techniques used thereafter is specified for each specific case.
At this point in the research, the canonical correlation technique was applied to the whole of the sample, covering the set of factor results on lifestyle, as explanatory, and the new variables on consumption of tourism, as explained, in order to test the hypothesis (see Figure 4).
In the analysis of principal components used on items relating to interests and opinions, the first 23 factors, explaining 75.3 percent of the variance, were chosen. In respect of the activities variable that completes the definition of lifestyle, 39 factors emerging from factor analysis of multiple correspondences and explaining 72.16 percent of the variance were used. All these factors taken together go to form the group of independent variables. The other input of information related to behaviour in respect of journeys, which was supposed to be dependent upon the lifestyle, and
consisted of 39 factors that accounted for 62.81 percent of the variance in short trips and 39 factors explaining 70.37 percent of the variance in long trips (see Figure 5).
In order to test the first hypothesis, which required checking whether there was a dependency relationship between individuals’ lifestyles and their behaviour in respect of short leisure journeys, Bartlett’s chi-squared test was used. The result yielded was a test value of 2,685.14 with 2,418 degrees of freedom. As a probability below 0.05 was obtained, it was inferred that at least one canonical correlation was non-zero. The test value for the second canonical correlation was 2,461.64 with 2,318 degrees of freedom, and the probability here also was less than 0.05. It was thus concluded that both canonical correlations were significantly different from zero. This means that there is a relationship between individuals’ lifestyles and their behaviour as consumers of short-stay
leisure journeys, for the whole of the sample, and that the first hypothesis put forward is positive when tested.
The value of the maximum canonical correlation between the first and second sets of variables, as defined above, is 0.68 (see Table III). The same analytic procedure was then applied, once more using as independent variables the factors that had been picked out as relating to lifestyle. The dependent variables in this case were the first 39 factors yielded by factor analysis of multiple correspondences applied to the variables covering behaviours on long-stay tourist journeys, these accounting for 70.37 percent of the variance (see Figure 6).
The result obtained is a value for Bartlett’s chi-squared statistic of 2,840.43 with 2,418 degrees of freedom. The probability for this is under 0.05, so that the first canonical correlation is non-zero, implying that a relationship exists. The second value for this statistic is 2,638.97 with 2,318 degrees of freedom; thus the second canonical correlation is likewise non-zero. The next two values for the statistic are also high, with a probability less than 0.05. Therefore, a significant canonical correlation does exist between the two groups of variables. The greatest correlation obtained showed a value of 0.66 (see Table IV). The general conclusion emerging from both analyses is that there is a significant relationship between individuals’ lifestyles, as measured in this research, and their behaviours as consumers of tourism on leisure journeys, whether long-stay or short-stay, this applying to the sample as a whole.
Although the value for this is not extremely high, the hypotheses put forward do nonetheless test positively. The question arises whether this relationship would be stronger or weaker if market segments defined by the lifestyle criterion, that is with differentiated lifestyles, were identified. If the relationship were stronger, this would bear out the need to segment the market on the basis of the variables under consideration in this paper, deducing from the different lifestyles of each segment the corresponding behaviour pattern in respect of tourism. It would then be necessary to develop differentiated strategies; drawing on the information available from the variables determining lifestyles.
If, on the contrary, it were to be weaker, this would mean that overall there does exist a significant link between the two sets of variables, as was indeed observed when testing the hypotheses. It would not, however, demonstrate that differing lifestyles cause different behaviours in respect of tourism, so that it would be unnecessary to apply differentiated strategies, and more appropriate to adopt a non-differentiated strategy.
Thus, the next step was to put consumers into homogeneous groups, segmenting the market on the basis of the criterion under investigation. Each of these might be chosen as a target segment to which to apply a differentiated marketing mix, should the assumptions of the study prove correct.
To achieve segmentation of the reference market, the cluster analysis technique was employed, using all of the lifestyle factors as segmentation criteria to differentiate between consumers of tourism. The objective was to reveal the latent structure of the market in relation to the variable of individuals’ lifestyles and thus to discover the specific features of each group and the relation of this variable to tourist behaviour.
The method adopted was the K-means algorithm, owing to the sample size, and the measure of closeness chosen was Euclidean distance, as the factors derived from the analysis of principal components and multiple correspondences were found in systems of orthonormal reference. Assignment of individuals to clusters was done by following the criterion of closest centroid. As the optimum number of clusters was not initially known, and with the intention of choosing the most appropriate figure objectively, several successive iterations were performed, in which the number of segments was varied from two up to eight.
The reason for the choice of five clusters was the search for a high degree of association among the elements of any given group and the lowest possible association across different clusters. This degree of association was measured by analysis of the variance of the clusters formed. In order to attain this objective, the nil hypothesis (that the averages of the variables are equal in the different clusters formed) was checked using the F-test with (g±1, n±g) degrees of freedom:
where n = sample size and g = number of groups.
Once this process had been performed, the appropriate number of clusters proved to be five, since the null hypothesis was rejected the greatest number of times on this basis.
This classification was later validated by discriminant analysis. This technique showed that the overall percentage of correct classification carried out by cluster analysis in this research was 94 percent and this permitted reclassification of those individuals who should be in another group in accordance with the discriminant functions (Table V). The system for selecting most discriminant variables that was used was stepwise, using the Mahalanobis distance. The factors selected during the process were 22 in total, mostly related to interviewees’ interests and opinions.
These 22 factors constitute the key predictive variables that can be used to segment the holiday market by firms operating in this sector. They are connected with the following features: self-realization in work, enterprising attitude, fashion, independence, concern for the environment,
conformism, value for money, responsibility at work, emancipation, novelty, liberalism, hedonism, safety on the streets, development of society, pragmatism, solidarity, caution, attachment to home, familiarity, materialism, ambition and conservatism. The intention was to demonstrate a need to segment the market and gain a more in-depth acquaintance with individuals’ lifestyles as systems, not only predicting their behaviour as tourists, but setting up differentiated marketing strategies, directly related to these variables. The next step had as its aim to test the same hypotheses for each of the segments.
The method used in hypothesis testing for each market segment is identical to that used for the sample as a whole. Therefore, the statistical technique is canonical correlation analysis, used independently on each cluster, and searching for a dependency relationship between the groups of variables established.
For each of the segments to which this methodology was applied, it should be noted that owing to budget limitations the sample size, although representative of the population, did not run to a very large number of individuals for any single group. Hence, hypothesis testing was restricted to segments four and five only, because their volume of components was larger than in the others. Indeed, in the rest of the segments, the number of variables was larger than the number of individuals. Nevertheless, the procedure to be used would have been the same if this limitation had not been present and in that case the results would have been fully comparable.
The fact that it was used on two of the segments did not have the effect of increasing the correlations calculated, for two reasons. First, because the application for each segment was totally independent and referred to the dependency relationship between two groups of variables analysed for each case individually. Second, because once the sample size is large enough, and in this instance such a minimum was very easily exceeded, the test for significance of the coefficient of correlation does not change. Thus, sample size does not have an influence over whether a correlation is significant or not (Sachs, 1978).
The data inputs included retain exactly the same structure as in hypothesis testing for the sample as a whole. This is because the aim is to compare the two results, so that only their outcomes will be commented upon. Analysis of the relationship between individuals’ lifestyles, as measured with activities, interests and opinions variables, and their behaviours as consumers of tourism during weekend and public holiday short breaks yields for segments four and five the results presented below.
For group four the null hypothesis is rejected for Bartlett’s chi-squared statistical test, which means that the canonical correlations are non-zero and there is thus a dependency relationship. The first value given by the test is 2,673.20 with a probability lower than 0.05, from which it may be inferred that
the relationship between the two sets of variables is significant. The second value is 2,449.80, likewise with a probability less than 0.05. At least two canonical correlations are therefore significantly different from zero. The greatest canonical correlation reckoned among the variables under consideration is 0.95. The hypothesis put forward in the research thus tests positive: there is a direct relation between them, and moreover it is of high intensity.
The results obtained for the fifth segment show four canonical correlations differing significantly from zero, with probabilities under 0.05 and high values for Bartlett’s chi-squared statistic. It can be stated that there is a direct relationship between the groups of variables under study, with the maximum value being 0.97 (see Table VI).
A comparison of these results with those obtained for short-stay trips by the same procedure from the sample as a whole shows that in both cases the canonical correlation is significant. However, the maximum value for the correlation calculated between the dimensions covering individuals’ lifestyles and those indicating behaviour relating to short breaks, of under four nights’ duration, increases notably here. This implies a much stronger dependency between these variables, when the group of individuals has
homogeneous characteristics in respect of lifestyle.
In order to extend these results to cover the totality of behaviours relating to tourism, it is necessary to take a look in addition at the outcomes of an analysis of canonical correlations applying to the set of new lifestyle variables and those carrying information about long-stay tourism.
As can be seen from the results yielded by this analysis, there is a significant relation between these variables for segment four, since the probability of getting a high value for Bartlett’s chi-squared statistic, such as 2,689.40 or 2,471.00, is in both cases lower than 0.05. The maximum canonical correlation found is 0.95.
In the case of segment five this relationship is likewise significant. The value for the chi-squared statistic is 3,017.05 with 2,418 degrees of freedom. The probability of this is 0.0000, from which it is clear that at least one canonical correlation is non-zero. The next three canonical correlations are also significantly different from zero. This implies that there is a direct relationship between the lifestyle variables and long-stay trips for this cluster, with the maximum value reaching 0.98 (see Table VII).
The set of results presented above demonstrates the fact that the hypotheses put forward test positive by segments, showing a really significant increase in the value of the relation. This points to the suitability of the method of assessing lifestyle put forward in this paper for segmentation of the market, in that the existence of differentiated lifestyles is made plain by segments related to similarly differentiated behaviours in respect of tourism. Once such a segmentation has been put in place, knowledge of the lifestyle of given individuals will permit prediction of their behaviours as consumers of tourism, whether shorter or longer leisure trips (see Table VIII).
The suitability of the methodology proposed in this paper, based on two specific types of measurement for the construct lifestyle (ordinal and nominal scales) relating to activities, interests and opinions, was thus demonstrated. This was also true of the statistical techniques involving factor analysis of principal components and multiple correspondences, generating new, continuous variables for lifestyle, usable as inputs for market segmentation. In addition a dependency relationship between this construct and tourist behaviour was shown as proposed. At this point a brief description of each segment will be provided as a function of the variables considered in this research, so as to round out the information presented in this study. In respect of how the descriptions supplied emerge from the analyses undertaken, it should be noted that the DEMOD module of the statistical package SPAD was applied directly to the original data.
This tool permits description of individuals and variables present in the data matrix, in this Bartlett’s test instance the original variables, in such a way as to give a picture of voluminous data sets.
The key concept in organizing characteristic elements is that of the test value. This evaluates the weight of an element in characterizing a category of individuals on the basis of the chi-squared statistic calculated for the sample. Tables are calculated by cross-referencing the variable to be characterized with the other nominal variables, putting them in decreasing order of test value. In this way a full characterization of the nominal variable is achieved.
A further characterization is produced for each class of individuals by use of the nominal variables, comparing by means of the chi-squared statistic the profile of distribution of the group of individuals within the variable with the overall profiles for that variable, and putting the set of nominal variables in decreasing order of test value. In addition the nominal variable is characterized by the modalities of other variables, which are distributed over the modalities of the variable to be characterized. By means of the chi-squared statistic the profile of the individuals is distributed with the overall profile of the variable to be characterized. The variables best characterizing the nominal variable are likewise put in decreasing order of test value.
Description of the segments arising from this study
The optimum number of segments emerging from this analysis is five, as stated above. Once their characteristics had been considered, a label was applied to each segment so as to describe the general style of life associated with it. However, the information available for each is much more extensive, so it is difficult to encapsulate in a single term. Hence, a short characterization for each cluster is presented below.
Home-loving (first segment)
Lifestyle. This segment is made up of individuals fundamentally focused on family life. They channel their efforts into enjoying a quiet and happy private life, laying special emphasis on having children and taking responsibility for their upbringing. In fact, this is one of their chief concerns. For this reason they work at home rather than elsewhere, and get satisfaction from carrying out the tasks involved.
These are people of an inflexible nature who are conservative in their views on life. They set store by religion. Although they are not materialists, they are cautious in their attitudes towards the future, as they consider it uncertain and take precautions to protect themselves.
From a consumer perspective they are the most demanding segment, particularly in relation to items for the family. They consider that quality is always more important than price, and this is reflected in their decisions about purchases. By way of pastimes, they enjoy cultural activities, such as regular visits to exhibitions, monuments or places of outstanding natural beauty. They have a wide range of preferences in reading material: books; magazines on home, fashion, gossip and health; and newspapers, usually local or regional.
As for the other media, it is striking that on television they prefer news, documentaries, current affairs, debates and travel programmes. On the radio they also prefer news programmes. They have a strong aversion to sport in any mass media.
Short trips (under four nights). This is the segment which makes the smallest number of short trips. When such trips are taken, the principal type of destination chosen would be inland and urban, either large cities or provincial capitals near their home town. The typical accommodation would be hotels, especially those of middling to modest ratings. .
Long journeys (four or more nights). This group makes the greatest number of long holiday journeys, domestic destinations being the most frequent. In fact, this is the segment that took the lowest percentage of trips outside Spain during the one-year period considered.
The places chosen for preference are on the coast, mainly in the North of Spain. It must be stressed that this is the group with the lowest percentage of long journeys inside their own autonomous region. Such people stay with family or friends or in a second home, usually accompanied by their families, for a holiday stay of two weeks’ duration. Marketing implications. People in this segment would be interested in enjoying holiday arrangements corresponding to the needs of the family as a whole. This would allow them to channel their leisure time towards improving relationships between the parents and children. Hence, they choose destinations permitting parents to relax, but at the same time with some activities intended for children.
They often return to the same destinations year after year. This ensures that children continue to take holidays with their parents without family arguments, as they will have established friendships there with people of the same age and interests.
Packages proposed by travel firms for this group ought to involve a fortnight’s stay, in summer, so that all members of the family can travel together. Destinations should be mostly seaside, near the place of habitual residence, and with a mild and pleasant climate. The type of accommodation best suited would give some degree of comfort and quality, but within the customers’ expectations on prices. Tourist apartments might be the first choice, as these would allow the family unit to remain together for fairly long periods in conditions quite like those normal at home. The activities in which they would participate at their destinations would be organized by the customers themselves. They would generally be trips in the area, trying out local foods, and enjoying any cultural offerings available.
It should be kept in mind that this is a rather conservative group, whose members in principle would prefer to organize their own trips, as they are very familiar with the destinations they visit. This implies that there might be a need
to increase communication. Thus, travel operators could use direct marketing campaigns, through mailings. These could be backed up in other mass media, principally radio and television advertising around the times of news programmes, it being of particular interest to draw this group’s attention to the existence of other alternatives. With regard to print media, regional newspapers and home and fashion magazines would seem the most appropriate. As the message would be connected to comfort for the family, enjoyed in a tranquil spot, it could be adapted to youngsters as well as to adults.
Idealistic (second segment)
Lifestyle. Those in the idealistic segment believe that the attainment of personal success is rooted in a determination to achieve a better world and
to fight against injustice. They have a genuine commitment to this themselves. They set special store by matters relating to the workplace. They take an active collaborative part in the enterprises employing them. This is a result of their desire to have a job about which they can be enthusiastic. However, they are not interested in taking on managerial responsibilities. Their character is flexible, responsible and tolerant in all matters, whether of politics, religion or law and order. As consumers, besides giving quality priority over price, they are the most innovative group.
With regard to the activities in which they participate, they enjoy sport, music, especially classical, going to concerts or the theatre, and dancing. They read magazines on politics as well as local and national papers. On television, they dislike programmes related to society gossip, their favourites being sports broadcasts. Their preference in radio would be music channels, while they never listen to game shows and virtually never to comedies.
Short trips (under four nights). For such travel this group mainly chooses inland destinations, making the second greatest number of visits to rural zones.
Generally speaking, members of this group are unwilling to spend much money on weekend trips, and so stay with relatives or friends, or in guesthouses. They are usually accompanied either by family or by friends, with trips involved at the same time being a noteworthy feature.
Long journeys (four or more nights). This segment is especially fond of country villages, particularly those near their place of residence. Members of it mostly go to destinations within Spain.
They do not spend much money on accommodation, mainly staying in the homes of relatives or friends, or in houses which they themselves own or rent. The typical journey is with family and lasting for a week.
Marketing implications. Tour operators who would have the greatest likelihood of targeting this market segment would be those offering a wide range of countryside holidays. They could offer this group a combination of destinations, including interesting villages where consumers like this could get in touch with the rural economy, through use of rooms in country homes shared with the owners, together with towns where they could enjoy a wider selection of cultural activities and accommodation would be on the basis of renting an entire house. A combination like this would address segment members’ interest in rural society, somewhat different from what exists in their places of residence. Moreover, the types of accommodation proposed would meet their requirements for a high level of quality at very affordable prices and would be rather unlike the categories of accommodation usually used, and so would appeal to this grouping’s innovative tendencies.
These holidays should include options allowing a chance to take an active part in rural activities, on the one hand, and, on the other, giving access to various types of sport that can be practised in contact with nature. The message to be employed should refer to getting to know, participating in and giving a hand with rural economies. The proposed packages could be aimed at small groups, made up of a family and a few friends, and with a duration of eight to ten days. Any media could be used, but on radio and television,
adverts during sports and music programmes would be best, while provincial and national newspapers might be the most effective channel. Autonomous (third segment)
Lifestyle. Members of this segment see success as linked fundamentally with individual freedom and independence. Indeed, a large percentage of them consider themselves independent. They place great emphasis on enjoying life. This is what they reckon to be one of their fundamental objectives and many of them achieve it. They work because they have to earn a living. In general terms, they try to aspire to upward social mobility in their way of life. In politics, religion and social views they are flexible, and even evince considerable liberalism. They accept current social reality, while thinking that society has not evolved fast enough. In fact, they believe that the future will bring them stability or improvement.
Their main interests are frequent cinema visits and going out to enjoy nightlife. Their favourite types of music are pop, rock, disco and ballads. They are also noticeably not attracted to cultural activities such as exhibitions or touring monuments.
They read sports and car magazines but not newspapers of any kind. As for the other media, on television they mostly watch films and on the radio they listen to sports programmes and current affairs.
Short trips (under four nights). This segment is the one involved in the largest number of weekend and long weekend or public holiday trips, showing a preference for city destinations, whether in Spain or outside. The type of accommodation they usually look for would be hotels. They normally travel in the company of friends.
Long journeys (four or more nights). This group is characterized by spending
holiday time in coastal areas, especially in small villages. They go with friends for a one-week stay, using low-priced accommodation, but not the homes of family or friends.
Marketing implications. Holiday companies interested in meeting the needs of this market segment should give greatest prominence to weekend and public holiday trips.
In order to stress the sensation of independence and freedom characteristic of this grouping, there could be an option for mixing and matching the cities to be visited, whether within or outside Spain, with accommodation in hotels offering special prices for small groups. This would suit its members’ autonomous lifestyle, improve their social image and allow them to become acquainted with new places, so that they grow personally.
For destinations where a longer time will be spent, a week or thereabouts, very different options could be offered. One possibility would be traditional seaside resorts that have a lot of atmosphere and night life, with lower prices because of being a part of mass tourism. The group travelling, made up of a number of friends, could make its own transport arrangements in this case. The distribution channel could be intermediaries, such as travel agencies. The best advertising to use would be either direct marketing or adverts in any of the media connected with sport. The image to be presented would be a group of people who get on very well together, laughing and enjoying themselves in various situations. These could include eating out with a fashionable destination as the background, having a drink, or any other context of relaxed enjoyment. Hedonistic (fourth segment)
Lifestyle. The factors this group see as the symbol of success relate to two fields: human relationships, in which they think themselves already wellplaced, and work, where they want to have an interesting and successful job, allowing them to fulfil themselves, even though they do not feel attracted to managerial posts. They accept life as it comes and enjoy it. They are tolerant with regard to discipline, politics and law and order.
These are people attracted by products and services that have newly arrived
on the market. They are also the group with the greatest ecological leanings. Other activities which round out the lifestyle of this segment are, in particular: listening to music, reading national and local newspapers as well as professional and business magazines, but not those related to fashion and home. Finally, the individuals belonging to this group are not great enthusiasts regarding the other media, being interested only in news and films. Tourist behaviour:
Short trips (under four nights). This is the group that more than any other selects large cities, whether inland or on the coast. As a rule members of this group travel in the company of friends. The kind of accommodation they use for weekends and short breaks is chiefly high-class hotels.
Long journeys (four or more nights). This is the segment constituting the main group of travellers who select destinations outside Spain, principally in Europe. For travel within Spain they are attracted mostly to major cities, provincial capitals and coastal towns, also often to touring. The length of their holiday is usually two to three weeks, with accommodation in medium to high category hotels, apartments or serviced flats.
Marketing implications. This segment is composed of people attracted to
pleasure and enjoying life. They lay great stress on personal relationships, and so they tend to travel in the company of friends.
They are active and innovators, like success and try to realize themselves personally. All this is noticeable in the way they behave when travelling on holiday. The tour operator can offer them a range of choices, with the principal destinations being combinations of major cities within Spain, or outside the country when longer trips are involved. There should be attractive tour packages incorporating stays in hotels of a certain charm, of medium to high category. These could be rounded out with a list of cultural events taking place at the destinations, so that at the moment of booking the trip arrangements could also be made to attend any that took their fancy.
Distribution and marketing to this grouping could be not just via traditional channels, such as travel agencies working out a tailor-made package, but also through new technologies like the Internet. Here not just options for destinations could be set out, but also virtual visits to hotels and information relating to complementary items of any kind, so that bookings and packages could be handled directly from home or from the workplace.
Other possible means of communication would be the national and financial press, or professional journals linked to the employment of members of this grouping. Sales representatives could also go to workplaces. Conservative (fifth segment)
Lifestyle. This is a home-loving segment, members of which are focused on the wellbeing of their family and coping with day-to-day life, although they do not fulfil their expectations. When they have a problem, they turn to family or friends for help in solving it.
They work because they have to in order to earn a living. Their jobs are not stimulating, but this does not worry them unduly, as they tend to identify success simply with doing their work efficiently. This is the segment setting greatest store by managerial positions that would allow upward
social mobility. However, most of them do not achieve such posts. They are the most materialistic group.
Members of this grouping are in general pessimistic about modern society. They believe that it is important to practise their religion, but are tolerant on this point. They are stricter when it comes to law and order. Their pastimes include enjoying visits to areas of outstanding beauty, but their most noteworthy characteristic is a dislike of nightlife, modern music and the cinema.
With regard to the media they are keen on both television and radio. Their preference in television goes to watching local news, reality-shows, game shows and gossip programmes. As for radio, they listen to news, sports and humour among others.
Short trips (under four nights). This segment undertakes few weekend trips. Destinations chosen are principally within the members’ own region, especially rural areas. Here they normally stay with family or with friends, or in a second home they themselves own. They travel accompanied by their family or by family and friends.
Long journeys (more than four nights). They are attracted to traditional domestic seaside destinations, especially country villages on the coast. They usually stay with family or with friends or in a house they own themselves or rent. The most frequent length of their holidays is from one to two weeks, spent together with their immediate family. Marketing implications. This is the most conventional segment among Spanish holiday-makers. The grouping is tolerant, though it has very traditional habits. It is of relevance, when considering the tourist behaviour of members of this segment, to note that they are not keen to vary the type of resort used, preferring guaranteed sun and sea for one or two weeks, so that there is no point in trying to sell them weekend breaks. Another relevant
fact is that they try to keep up appearances, and so choose well-known destinations, mostly seaside, though usually smaller, rural resorts, which to some extent marks them off from the masses.
They often travel together with both family and friends, creating their own organized travel party. They do not want to go without holidays; nor, however, are they willing to spend a high proportion of their income on them, so they should be offered accommodation options that turn out to be relatively cheap for the group as a whole, such as renting a whole house or modest serviced flats. In addition, there is no need to offer them any great number of complementary choices, since what they enjoy is principally the beach. They would be interested in a half-board or full-board option, if this provided good value for money and was based on typical regional foods.
They would arrange their trip themselves or with a travel agent. The ways of communicating with them would be the mass media, television and radio, especially through sponsorship of programmes such as reality shows, quizzes, or gossip shows, currently extremely popular in Spain.
The conclusions that can be drawn include:
The swift and wide-ranging changes that present-day society is undergoing, together with the characteristics of the demand for tourism, are leading to an increasing personalization in the way this service is consumed. This is increasingly less well explained by sociodemographic and economic criteria. A review of the current literature on lifestyle leads to the choice of this criterion as of importance for segmenting the market for tourism and in explaining tourist behaviour patterns.
The results of this research point to the use of general lifestyle variables in order to segment the market and to predict the behaviour of leisure travellers. This is in contrast with the use of variables measuring the style of tourism adopted by a given traveller, as is more usual in work related to tourism.
This study offers certain improvements in the measurement of lifestyles. These relate to the use by most previous studies of exclusively ordinal scales, where here the measurement technique combines two types of scales, ordinal and nominal. Ordinal scales cover interests and opinions of respondents, as it is relevant what degree of importance they assign to each item. Nominal scales are used for activities, the third variable in this approach to lifestyles, and also for questions about travel. This combination allows improved efficiency in data collection, since the same amount of information is gathered, but the time needed to reply to the questionnaire is considerably reduced. Hence, the quality of the information obtained is enhanced, since bias caused by excessive length is avoided. Moreover, the cost of fieldwork can be cut substantially, which renders it more affordable for businesses with tight budgets.
The use of a combination of multivariate analytical techniques makes it possible to attain the final objective of the study, segmentation of the holiday market, on the basis of the several variables measured, as indicated above. This is because the use of factor analysis of principal components and of multiple correspondences is principally intended to homogenize information, so as to permit later application of a splitting technique, cluster analysis, without loss of statistical information. Most previous research has employed only principal component analysis, as all its data are gathered using ordinal scales. The tool is applied to reducing the volume of data, and this leads by the end of the investigation to a large amount of unused residual information, since work is ever more distant from the raw data and increasingly impoverished.
A follow-up use of discriminant analysis first allowed a check showing that 94 percent of the individuals were correctly classified and, second, provided a set of key predictor variables for use in segmenting the market for holidays. These were: self-realization at work, enterprising attitude, fashion, independence, concern for the environment, conformism, value for money, responsibility at work, emancipation, novelty, liberalism, hedonism, safety on the streets, development of society, pragmatism, solidarity, caution, attachment to home, familiarity, materialism, ambition, and conservatism.
The strong dependency relationship between tourists’ behaviours and lifestyle, as defined in this study using an activities, interests, opinions approach, has been shown once more. Consequently, it proved possible to see this method as applied to a specific sector and a particular European country, Spain.
The segmentation of the market that emerged makes a division into five
clusters: home-loving, idealistic, independent, hedonistic and conservative. These are labels trying to give a general idea of the lifestyle of each segment. However, the characteristics of each are much more extensive than the label defining them. It is quite likely that these labels and characteristics would not correspond precisely to those to be found in other countries. Nevertheless, the methodology could be extrapolated to any location. In this way businesses in the holiday trade can become acquainted with and predict the behaviour of their potential customers by gaining an in-depth view of their lifestyles as explained in this study, and furthermore could use this criterion in defining their business strategies.
1. In respect of the classification of trips in terms of length, in Spain, tourism is one strategic sector and it has had an important role at international level for a relatively long time. In this country, the variable “trip length” is divided into two types: Short trips, fewer than four nights, and long trips, four or more nights; as a consequence important differences are noticed in tourists’ behaviour as a function of this classification. This was stated by the researches of the Institute of Tourist Studies of the General Secretariat for Tourism and the Departments of Industry, Commerce and Tourism of self-governing communities. Owing to the significant wealth of experience of this country, so far as domestic tourists are concerned, it is important to extend this approach to other countries (Instituto de Estudios Turõsticos, 1998a, b; 1999a, b; Direccion General de Turismo, 1998, Â
2. The aim of choosing the capitals with a population greater than 100,000 habitants as the universe of the study arose from the positive relationship among the variables (place of residence and travels) in these town councils as for tourist behaviour (Esteban Talaya, 1987; Centro de Investigaciones Sociologicas, 1993, 1995).
3. There are many previous studies which used scales measuring the degree of agreement and disagreement, related to the assumption on lifestyle, either in the tourist field (Davis et al., 1988; Schewe and Calantone, 1978; Schul and Crompton, 1983; Silverberg et al., 1996; the first and the last ones used scales of five points and the rest of six points); or in other fields (Cosmas, 1982; Gunter and Furnham, 1992; Valette-Florence, 1985; Wind, 1978; these authors used scales between five, six and seven points). However, one of the precursors of this method, Wind, already advised to use other scales different from the classic ones about the degree of agreement. In spite of all, most researchers, who work with lifestyles, have used grading scales.
Thinking that this could give a new bulkiness to lifestyle analysis, it is therefore important to use other scales and other statistical techniques apart from the factorial analysis of principal components in this study. Consequently, there would be an important drop in the time spent on answering the questionnaire, if it was proved that tourist behaviour depended on lifestyle.
4. When the relevant information is whether this activity, travel, is done or not.
Abbey, J.R. (1979), “Does life-style profiling work?”, Journal of Travel Research, Vol. 18 No. 1, pp. 8-14.
Allport, G. (1935), Handbook of Social Psychology, Clark University Press, Worcester, MA. Alonso Rivas, J. (1997), Comportamiento del Consumidor, ESIC Editorial (Coleccion Universidad), Â
Baudrillard, J. (1970), La Societe de Consommation, Gallimard, Paris. ÂÂ
Benzecri, J.P. (1979), “Sur le calcul des taux d’inertie dans l’analyse d’un questionnaire”, Cahiers Â
de l’Analyse des Donnees, Vol. 3, pp. 377-8.
Bull, A. (1996), “Economõa del sector turõstico”, ICE (Informacion Comercial Espanola), No. 749, Â
Burnett, J.J. (1981), “Psychographic and demographic characteristics of blood donors”, Journal of Consumer Research, Vol. 8 No. 1, pp. 62-6.
Burtschy, B. (1994), “Traitements d’enquetes par modeles”, in Grange, D. and Lebart, L. (Eds), Ã
Traitements statistiques des enquetes, Dunod, Paris, pp. 197-226. Ã
Centro de Investigaciones Sociologicas (1993), Comportamiento de los Espanoles ante las Â
Vacaciones, No. 2069, Octubre.
Centro de Investigaciones Sociologicas (1995), Comportamiento de los Espanoles ante las Â
Vacaciones, No. 2193, Octubre.
Cosmas, C. (1982), “Life-styles and consumption patterns”, Journal of Consumer Research, Vol. 8 No. 4, pp. 453-5.
Crask, M.R. (1981), “Segmenting the vacationer market: identifying the vacation preferences, demographics and magazine readership of each group”, Journal of Travel Research, Vol. 20 No. 2, pp. 29-34.
Davis, D., Allen, J. and Cosenza, R.M. (1988), “Segmenting local residents by their attitudes, interests and opinions toward tourism”, Journal of Travel Research, Vol. 27 No. 2, Autumn, pp. 2-8.
Direccion General de Turismo (1998), El Turismo en Castilla y Leon, Junta de Castilla y Leon, Â
Direccion General de Turismo (1999), El Turismo en Castilla y Leon, Junta de Castilla y Leon, Â
Engel, J., Blackwell, R. and Kollat, D. (1978), Consumer Behavior, Holt, Rinehart and Winston, New York, NY.
Engel, J., Blackwell, R. and Miniard (1990), Consumer Behavior, 6th ed., The Dryden Press, Chicago, IL.
Esteban Talaya, A. (1987), “Analisis de la Demanda: Aplicacion a la Actividad Turõstica de las Â
Tecnicas de Prediccion”, Tesis Doctoral de la Facultad de Ciencias Economicas y Â
Empresariales, Universidad Complutense de Madrid.
Feldman, L. and Hornik, J. (1981), “The use of time: an integrated conceptual model”, Journal of Consumer Research, Vol. 7 No. 4, pp. 407-19.
Fisher, A.B. (1990), “What consumers want in the 1990s”, Fortune, Vol. 29, January, pp. 108-12. Gonzalez, A.M. (1998), “El estilo de vida como criterio de segmentacion del mercado turõstico: Â
propuesta metodologica y contrastacion empõrica”, Doctoral thesis presented in the Â
Faculty of Economics and Business Studies of the University of Leon. Â
Gunter, B. and Furnham, A. (1992), Consumer Profiles: An Introduction to Psychographics, Routledge, New York, NY.
Gutman, J. and Mills, M.K. (1982), “Fashion lifestyle, self-concept, shopping orientation and store patronage: an integrative analysis”, Journal of Retailing, Vol. LVIII No. 2, pp. 64-86. Hair, J.F. Jr, Anderson, R.E., Tatham, R.L. and Black, W.C. (1998), Multivariate Data Analysis, 5th ed., Prentice-Hall, Englewood Cliffs, NJ.
Hawes, D.K. (1988), “Travel-related lifestyle profiles of older women”, Journal of Travel Research, Vol. 27 No. 2, pp. 22-32.
Hay Associates (1978), United Kingdom: A Study of the International Travel Market, United States Travel Service, Department of Commerce, Washington, DC. Howard, J.A. (1993), El comportamiento del consumidor en la estrategia de Marketing, Dõaz de Â
Howard, J.A. and Sheth, J. (1969), The Theory of Buyer Behavior, John Wiley & Sons, New York, NY.
Hustad, P. and Pessemier, E. (1974), “The development and application of psychographic life-style and associated activity and attitude measures”, in Wells, W. (Eds), Life-Style and Psychographics, American Marketing Association, Chicago, IL, pp. 32-70. Instituto de Estudios Turõsticos (1998a), Familitur, Estadõstica sobre el turismo de los espanoles, Â
Ministerio de Economõa y Hacienda, Madrid.
Instituto de Estudios Turõsticos (1998b), Frontur, Estadõstica de Movimientos Turõsticos en Â
Fronteras, Ministerio de Economõa y Hacienda, Madrid.
Instituto de Estudios Turõsticos (1999a), Familitur, Estadõstica sobre el turismo de los espanoles, Â
Ed. Ministerio de Economõa y Hacienda, Madrid.
Instituto de Estudios Turõsticos (1999b), Frontur, Estadõstica de Movimientos Turõsticos en Â
Fronteras, Ministerio de Economõa y Hacienda, Madrid.
Jackson, G. (1974), “A preliminary bicultural study of value orientations and leisure attitudes”, Journal of Leisure Research, Vol. 5 No. 1, pp. 10-22.
Jackson, R.W., McDaniel, S.W. and Rao, C.P. (1985), “Food shopping and preparation: psychographic differences of working wives and housewives”, Journal of Consumer Research, Vol. 12 No. 1, pp. 110-13.
Kaiser, H.F. (1961), “A note on Guttman’s lower bound for the number of common factors”, Brit. J. Statist. Psychol., Vol. 14, pp. 1-2.
Kamakura, W.P. and Wendel, M. (1995), “Life-style segmentation with tailored interviewing”, Journal of Marketing Research, Vol. 32 No. 3, pp. 308-17.
Lambin, J.J. (1995), Marketing estrategico, 3rd ed., McGraw-Hill, Madrid. Â
Lastovicka, J., Murry, J. and Joachimsthaler, E. (1990), “Evaluating the measurement validity of lifestyle typologies with qualitative measures and multiplicative factoring”, Journal of Marketing Research, Vol. 27 No. 1, pp. 11-23.
Lawson, R. (1991), “What is psychographic segmentation?: A comparison between general consumer analysis and product specific analysis”, Conference: New Horizons in Tourism and Hospitality Education, Training and Research, The University of Calgary, Alberta, 2-5 July, pp. 445-55. Lebart, I., Morineau, A. and Piron, M. (1997), Statistique exploratoire
multidimensionnelle, 2nd ed., Dunod, Paris.
Lesser, J.A. and Hughes, M.A. (1986), “The generalizability of psychographic market segments across geographic locations”, Journal of Marketing, Vol. 50 No. 1, pp. 18-27. Mayo, E. (1975), “Tourism and the national parks: a psychographic and attitudinal study”, Journal of Travel Research, Vol. 14 No. 14, pp. 14-18.
Mitchman, R. (1991), Lifestyle Market Segmentation, Praeger, New York, NY.
Pearce, P.L. (1988), The Ulysses Factor: Evaluating Visitors in Tourist Settings, Springer Verlag, New York, NY.
Perreault, W.D., Darden, D.K. and Darden, W.R. (1977), “A psychographic classification of vacation life styles”, Journal of Leisure Research, Vol. 9 No. 3, pp. 208-24.
Pearce, P.L. (1991), “Fundamentals of tourist motivation”, in Pearce, D.G. and Sutter, R.W. (Eds), Tourism Research: Critiques and Challenges, Routledge, London. Pearce, P.L. and Caltabiano, M.L. (1983), “Inferring travel motivation from travelers’ experiences”, Journal of Travel Research, Vol. 20 No. 2, pp. 16-20.
Pessemier, E. and Handelsman, M. (1984), “Temporal variety in consumer behaviour”, Journal of Marketing Research, Vol. 21 No. 4, pp. 435-44.
Pessemier, E. and Tigert, D. (1966), “Personality, activity and attitude predictors of consumer behavior”, New Ideas for Successful Marketing, American Marketing Association, Chicago, IL, pp. 332-47.
Reynolds, F.D. (1972), “Intermarket patronage: a psychographic study of consumer outshoppers”, Journal of Marketing, Vol. 36 No. 1, pp. 50-4.
Reynolds, F.D. and Darden, W.R. (1971), “Mutually adaptive effects of interpersonal communication”, Journal of Marketing Research, Vol. 8 No. 4, pp. 449-54.
Ritchie, J.R.B. and Goeldner, C.R. (1987), Travel, Tourism and Hospitality Research: A Handbook for Managers and Researchers, Wiley, New York, NY.
Roberts, M.L. and Wortzel, L.H. (1979), “New life style determinants of women’s food shopping behaviour”, Journal of Marketing, Vol. 43 No. 1, pp. 28-39. Robertson, T. and Wind, Y. (1980), “Organizational psychographics and innovativeness”, Journal of Consumer Research, Vol. 7 No. 1, pp. 48-70.
Sachs, L. (1978), Estadõstica aplicada, Labor SA, Madrid.
Schewe, C. and Calantone, R. (1978), “Psychographic segmentation of tourists”, Journal of Travel Research, Vol. 16 No. 3, pp. 14-20.
Schul, P. and Crompton, J.L. (1983), “Search behavior of international vacationers: travel-specific lifestyles and sociodemographic variables”, Journal of Travel Research, Vol. 22 No. 2, pp. 25-31.
Silverberg, K.E., Backman, S.J. and Backman, K.F. (1996), “A preliminary investigation into the psychographics of nature-based travelers to the Southeastern United States”, Journal of Travel Research, Vol. 35 No. 2, pp. 19-28.
Teel, E., Bearden, O. and Durand, T. (1979), “Psychographics of radio and television audiences”, Journal of Advertising Research, Vol. 19 No. 1, pp. 53-6.
Thompson, A.M. and Kaminski, P.F. (1993), “Psychographics and lifestyle
antecedents of service quality expectations: a segmentation approach”, Journal of Services Marketing, Vol. 7 No. 4, pp. 53-61.
Tigert, D. (1969), “Psychographics: a test-retest reliability analysis”, in McDonald (Ed.), Marketing Involvement in Society and the Economy, American Marketing Association, Chicago, IL, pp. 310-15.
Tigert, D. (1971), “A research project in creative advertising through life-style analysis”, Attitudes Research Conference, Attitudes Research Reaches New Heights, American Marketing Association, Chicago, IL, pp. 223-7.
Triandis, H. (1979), Attitude and Attitude Change, Wiley, New York, NY. Unger, L. (1982), “Better knowledge of the consumer through market segmentation”, European Research, Vol. 10 No. 2, pp. 81-7.
Valette-Florence, P. (1985), “Une evaluation empirique des effets des styles de vie sur la Â
consommation”, These de Troisieme Cycle en Sciences de Gestion, Institut d’Etudes Á
Commerciales, Universite des Sciences Sociales de Grenoble.
Valette-Florence, P. (1994), “Introduction a l’analyse des chaõnages cognitifs”, Recherche et Á
Applications en Marketing, Vol. 9 No. 1, pp. 93-117.
Valette-Florence, P. and Jolibert, A. (1987), “Life-style: an empirical investigation of the relationships Â
between its indicators”, Publication de recherche du CERAG, July, Ecole Superieure des Â
Affaires, Universite des Sciences Sociales de Grenoble.
van Raaij, F. (1986), “Consumer research on tourism, mental and behaviour constructs”, Annals of Tourism Research, Vol. 13 No. 1, pp. 1-9.
Vazquez Casielles, R., Trespalacios Gutierrez, J.A. et al. (1994), Marketing: Estrategias y Â
Aplicaciones Sectoriales, Civitas, Madrid.
Venkatesh, A. (1980), “Changing roles of women ± lifestyle analysis”, Journal of Consumer Research, No. 6, pp. 189-97.
Villani, K. (1975), “Personality, life-style and television viewing behavior”, Journal of Marketing Research, Vol. 12 No. 4, pp. 432-9.
Wahlers, R.G. and Etzel, M.J. (1985), “Vacation preference as a manifestation of optimal stimulation and lifestyle experience”, Journal of Leisure Research, Vol. 17 No. 4, pp. 283-95. Wells, W.D. and Cosmas, S.C. (1977), “Life styles”, in Ferber, R. (Ed.), Selected Aspects of Consumer Behavior, National Science Foundation, Washington, DC.
Wells, W.D. and Tigert, D.J. (1971), “Activities, interests and opinions”, Journal of Advertising Research, Vol. 11 No. 4, pp. 27-35.
Wind, Y. (1978), “Issues and advances in segmentation research”, Journal of Marketing Research, Vol. 15 No. 3, pp. 317-37.
Wind, Y. and Green, P. (1974), “Some conceptual, measurement and analytical problems in life style research”, in Wells, W. (Ed.), Lifestyle and Psychographics, American Marketing Association, Chicago, IL, pp. 99-126.
Witt, S.F. and Moutinho, L. (1994), Tourism Marketing and Management Handbook, 2nd ed., Prentice-Hall, London.
Woodside, A. and Pitts, R. (1976), “Effects of consumer life styles, demographics and travel activities on foreign and domestic travel behavior”, Journal of Travel Research, Vol. 14, pp. 13-15.
World Tourism Organization (1995), Concepts, Definitions and Classifications for Tourism Statistics, WTO, Madrid.
World Tourism Organization (1999), International Tourism: A Global Perspective, 2nd ed., WTO, Madrid.
Ziff, R. (1974), “The role of psychographics in the development of advertising strategy and copy”, in Wells, W. (Ed.), Lifestyle and Psychographics, American Marketing Association, Chicago, IL, pp. 127-66.