 |
Patricia Oom do Valle, Faculty of Economics, University of Algarve, Faro, Portugal
Cláudia Moço, University of Algarve, Faro, Portugal
Design/methodology/approach – A structural model and a categorical principal component analysis are used to assess to what degree motivational factors may influence perceptions about a destination. The empirical study is supported through data from a sample of 1,097 individuals who travelled by plane to exotic places, during 2004.
Findings – The findings show that push and pull motivations influence the way tourists perceive the destination but the resulting image does not determined the intrinsic motivations.
Research limitations/implications – The study has the restriction of being limited to the Portuguese case. However, these findings open paths for further investigation, namely extending to other destinations and to tourists with different incentives and nationality.
Originality/value – This study contributes to the overall understanding of why tourists adopt specific behaviors. In specific, motivations and perceptions were combined in order to understand why people could be pushed to travel to exotic destinations and how they form their perceptions.
Introduction
Portugal is one of the most important receiving tourist destinations in Europe. The short break holiday tradition, the low cost of living of Portuguese citizens, and the country's economic development explains this position. Only recently have Portuguese citizens begun to travel abroad. Factors that account for this change are related to lower traveling costs, shorter holiday periods and higher socio-economic conditions. This study is a first attempt to explore what stimulates travel behavior of Portuguese tourists.
Considering human, the indication would be that research into why people travel and into what they intend to do on holidays, would generate a multiplicity of motives related to the tourist destinations. In this sense, motivations and perceptions are fundamental constructs in the definition and testing of a model that asks the question: why do people travel to exotic places?
Motivations are forces that influence and predispose to a specific behavior (Dann, 1981; Pearce, 1982). In tourism research, a perception is the image of a tourist destination that makes effective the behavior intentions (Gnoth, 1997). Baloglu and McCleary (1999) state that perceptions about the destination are a function of internal motivations (push motives) and external motivations (pull motives). Each tourist has his/her own internal and external motivations to travel which lead to different perceptions about a tourist destination.
Founded on Crompton's (1979) push-pull motive model and supported by the literature review, the present study proposes an integrated approach to understanding the motivations of tourists in relation to the underlying associations between push and pull motives and their contributions to the perception of the tourist destination as a whole. A structural equation modeling procedure (Joreskog and Sorbom, 1986) is used to test the theoretical model suggested. In order to further explore the links between the push motives, pull motives and perceptions, we use a categorical principal components analysis (CATPCA). This last statistical data analysis technique produces perceptual maps which allow a clearer understanding of the simultaneous connections among the observed variables considered to measure the latent constructs in the structural model. Since, CATPCA allows a joint analysis of a set of (categorical) observed variables, this statistical method represents a step forward in comparison with traditional correlation measures which only focus on pairs of variables. The implications of this study are apparent in terms of defining market strategies well-suited to the motivating factors which play a strong role in the general perception of a travel destination.
The paper is organized as follows. The first section provides the theoretical framework of the study and reviews the previous literature on the models and empirical results about the constructs of the proposed model. The following section presents the conceptual model and defines the set of research hypotheses. The study proceeds with a description of the methods which are applied, including information about the data and statistical procedures. Results are presented and some of their implications and limitations are discussed in the final section.
Literature review
“Motivation is the need that drives an individual to act in a certain way to achieve the desired satisfaction” (Beerli and Martín, 2004, p. 626). Many different reasons and motives compel people to travel. These forces are perceived as being able to decrease the condition of tension felt by the individual. The state of tension then gives way to the necessity that encourages an action or attitude (Fodness, 1994). Although the decision to satisfy needs may rely on other psychological variables, in reality, all human behavior is motivated (Crompton, 1979).
Most studies looking to explain the tourist decision are based on the expectancy value theory (Fishbein, 1967). This theory defines expectation as the probability that a certain attitude will lead to positive or negative benefits, thus allowing the isolation of determining factors of behavior and, furthermore, specifying how expectations and values can be combined in order for choices to be made. The possibilities for combining expectation with value are numerous. The motivation for adopting a certain form of behavior is determined by the value and by the expectation for each benefit. The greater or lesser tendency to adopt certain behavior depends on expectations and the consequent value of these attitudes for the individual.
In general, the main advantages of the expectancy value theory include the following points (Cohen et al., 1972). This theory enables the use of some of the concepts present in the same model; allows the integration of the emotional component in tourist motivation; the theory can incorporate all the reasons for travelling put forward in the studies on motivation; the expectancy value theory also enables the resolution of the problem of “push” and “pull” factors, as well as the evaluation of personality; the theory allows a more realistic and sophisticated view of tourist motivation.
Most tourism literature on the factors behind the tourist decision has focused on a single motivational construct, whether it be cognitive or emotional (Gnoth, 1997; McCabe, 2000). Research focusing fundamentally on cognitive aspects does not consider the relations between these factors and emotional aspects. In this field empirical evidence is lacking.
The existence of internal and external factors which motivate human behavior is assumed by several authors. Kotler (1982), for instance, states that motivations can be the result of internal and external stimuli. Internal stimuli arise from personal needs that can be physiological, social, egocentric, safety and self-actualization. External stimuli result from publicity and promotion. Motivations around traveling can be personal (personal training, compensation, rest and knowledge) and interpersonal (resulting from the social relations) (Crompton, 1979; Dann, 1977; Yoon and Uysal, 2005). Dann (1977) classifies personal motives as those that predispose the individual for traveling: escape from daily routines, the desire to escape from solitude. According to this study, interpersonal motives arise from the need to seek some form of social recognition that is obtained through travel.
Crompton (1979) refers to more specific and direct motives that can direct the tourist in his/her decision about the type of holiday or the travel destination. Psychological or social motives (push motives) which sustain the desire to travel are identified by the author. On the other hand, pull motives affect the travel decision and are associated to the destination's characteristics (Lundberg, 1990). Gnoth (1997) states that the necessity for holidays depends on desires such as self actualization, sense of self-esteem and social status. In the last case, the tourist produces his/her perceptions in accordance to his/her social group. In this sense, Cohen (1972) introduces sociological motives which direct the tourist to a socially accepted behavior.
Based on intrinsic and extrinsic motivations, the tourist builds his/her perceptions (Gartner, 1993; Dann, 1996; Baloglu, 1997). Perceptions can be different from the true attributes of the product depending on how the individual receives and processes information. In other words, perceptions focus on the attributes of products that affect behavior and not on the real attributes of products (Dann, 1981; Pearce, 1982). According to Morrison (1989), perceptions are a cognitive measure of tourism destination value. This value represents the opportunity cost of the product (value for money), that is, that perceptions are formed based on a cost benefit assessment.
Gnoth (1997) reports that the perception of a destination may be analysed from a cognitive or behavioral perspective. This author argues that perceptions are of several types: they can have a cognitive component (which results from the evaluation of the destination attributes) and a personal component (that depends on how the individual intends to perceive that destination). The cognitive structure comprises the shaping of a perception represents of internal and external stimuli into the “awareness set” which is, in fact, a cognitive structure (Woodside and Lysonski, 1989; Crompton, 1979).
A general conclusion can be drawn that personal motives (push motives), as well as the view of the characteristics of the tourism destination (pull motives), determine perceptions. These motives interact in a dynamic and evolving context (Correia, 2000). From another point of view, the tourist motivation should be seen as a multidimensional concept which explains the tourist decision (McCabe, 2000).
Most of the studies on motivations and perceptions formation rely on multivariate methods that allow us to determine multidimensional structures. The techniques more often applied are factor analysis, multidimensional scaling, principal components analysis (PCA), segmentation procedures, correlational tests, ANOVAS and MANOVAS.
Formation of perceptions, from a conceptual point of view, is presented by Kim and Yoon (2003) and also by Vogt and Andereck (2003). These authors propose structural equation models which look to analyse how emotions and cognitions can influence tourist destination perceptions. Seddighi and Theocharous (2002) use a conditional logit model to measure the perceptions/feelings about the characteristics of tourist destination. From this methodology, the authors detected the probabilities of revisiting a travel destination. Murphy et al. (2000) define a structural model that relates the tourist intention to return (as a proxy of satisfaction/quality) with his/her perceptions of the travel experience.
In general, these models combine motivations, expectations and choice and are estimated by using regressions techniques and structural models. Despite the power of these analyses, nothing has been done in order to deeply evaluate the relationships among the observed variables in the proposed models, which can be done by observing the graphical displays provided by CATPCA. The conceptualized model in the following section incorporates insights from the literature review. The objective of this study is to develop and test a structural equation model of overall perception of the exotic destinations which is a function of perceptions and motivations, in a cognitive and emotional perspective. Perceptual maps via CATPCA complement the conclusions from applying the proposed model.
Theoretical model and hypotheses
Figure 1 shows the hypothetical causal model that proposes interactions among the constructs of motivations and perceptions. The choice of each component of the model was based on the literature review. The model's contribution depends upon the identification of the relationship between the two constructs of motivation (push and pull motives).
The hypotheses presented in the conceptual model, which will be tested in this study, are the following:H1.Push motives lead to different pull motives. McCabe (2000) shows that the relationship between push and pull factors is the result of a sequential decision. Firstly, a tourist decides to travel depending on his/her emotional state (push-motives). Next, the tourist determines the place to visit and related activities (pull-motives). These choices are a function of push motives that determine the decision to travel. More specifically, Russel and Pratt (1980) suggest that the attributes of the destination are perceived by their ability to achieve affective motives:H2.The overall perception of the destination depends on the evaluation of the pull motives. According to Murphy et al. (2000), positive perceptions about a destination are related to a positive summary evaluation of tourist destination attributes:H3.Different push motives lead to different perceptions. Choy (1992) and Murphy and Pritchard (1997) find that more than the infrastructures of a tourist destination, emotional states affect the perceptions regarding that destination. In the same way, Beerli and Martín (2004) conclude that the emotional motives directly influence the affective perceptions of the destination.
Method
Survey and sample
A survey was developed in order to test the proposed hypotheses. The survey includes three sections: the first one tries to measure the push motives that may lead the tourist to the decision to travel. The second section presents the rate of destination attributes and landscape features (i.e. the pull motives) which represent what the tourist intends to do at the exotic destinations. The latter section considers the overall level of perceptions.
The set of push and pull motives considered in this work is the most quoted on literature (Uysal et al., 1996; Iso-Ahola and Mannel, 1987; Lundberg, 1990; Fodness, 1994; Holden, 2003; Mohsin and Ryan, 2003; Shoemaker, 1989; Cossens, 1989). Specifically, the following push motives were included: “relieving stress,” “escaping from the routine,” “physical relaxing,” “doing different things,” “stimulating emotions and sensations,” “being an adventurer,” “having fun,” “increasing knowledge,” “knowing different cultures and lifestyles,” “enriching myself intellectually,” “knowing new places,” “meeting interesting people,” “developing close friendships,” “going places my friends have not been,” “talking with my friends about the trip.” Pull factors considered were the following: “landscape,” “natural environment,” “cultural attractions,” “night-life,” “sports equipment,” “transport,” “lodging,” “weather,” “accessibility,” “beaches,” “gastronomy,” “security,” “distance,” “shopping facilities,” “relaxing atmosphere,” “social environment,” “hospitality,” “different ethnics,” “standard of living.”
As Maio and Olson (1994) propose, push and pull motives, as well as perceptions, were assessed by a seven point Likert-type scale. This scale ranged from “not important” (1) to “extremely important” (7) concerning motives and from “very low” (1) to “very high” (7) regarding perceptions. Owing to sampling convenience, the study took place on flights to the destination. The flights were on Air Luxor, SA, a Portuguese air travel company which has a significant market share as a tour operator.
A pre-test of the survey with a sample of 150 tourists enhances the validity and reliability of the questionnaire. After the pre-test, which resulted in only minor amendments, a sample of 1,097 Portuguese tourists was taken on flights to exotic places (such as, Brazil, Morocco, Egypt, Mexico, Sao Tome and Principe) during August and September, 2004. These are the months during which most Portuguese citizens have holidays. Respondents were invited to participate in the survey and asked to fill out the survey during the flight, and just before arriving at the airport. As reported in Table I, the larger proportion of respondents are married, have high level education, average age of 35 and belong to social middle class (Table I).
Statistical data analysis method
Three main stages of statistical data analysis follow in this report. The first stage entails the reduction of the data through the application of PCA to the set of push and pulls motives. To assess the pertinence of using PCA in this research, the Kaiser-Meyer-Olkin (KMO) statistic was computed and the Bartlett test was performed (Bartlett, 1947). The reliability of the obtained factors was measured by α Cronbach's coefficients (Cronbach, 1951). This analysis was carried out with statistical package for the social sciences (SPSS) software, version 12.
A structural equation model was estimated in the second stage in order to assess the underlined research hypotheses. The factors obtained with PCA were used as indicators of the constructs push and pull motives. The model was estimated with the software analysis of moment structures (AMOS 5) (Arbuckle and Wothke, 1999) and the weighted least squares estimation method was applied. This is an asymptotically distribution-free method, which is not sensitive to the non-normality of the data. The model fit was evaluated by following the approach suggested by Hair et al. (1998), which first requires the assessment of the overall model fit and then the measurement and structural models individually.
Measures of overall model fit include absolute, incremental and parsimonious fit measures. The χ2 goodness-of-fit test is the best known index of absolute fit. However, this index is quite sensitive in large samples, showing significant differences between the covariance matrices in any proposed model (García and Martinez, 2000). Therefore, besides the χ2 test, the goodness of fit index (GFI) (Joreskog and Sorbom, 1986), the root mean square residual (RMSR) and the root mean square residual of approximation (RMSEA) (Steiger, 1990) were used to evaluate the proposed model's overall absolute fit. Six incremental fit measures were used to evaluate the proposed model's fit: the adjusted goodness of fit index (AGFI) (Joreskog and Sorbom, 1986), the normed fit index (NFI) (Bentler and Bonnet, 1980), the Tucker and Lewis index (TLI) (Tucker and Lewis, 1973) and the incremental fit index (IFI) (Bollen, 1988), the relative fit index (RFI) (Bollen, 1986) and the comparative fit index (CFI) (Bentler, 1990). Parsimonious fit was measured by the normed χ2 (Joreskog, 1969) that should range from 1 to 5. Excluding RMSR and RMSEA, in which smaller values are better (zero indicates perfect fit) all other measures and indexes range from 0 (no fit) to 1 (perfect fit).
In evaluating the measurement model, each latent variable was assessed separately through examining the standardized loading, the construct reliability and the variance extracted. In analyzing the structural model fit, parameter estimates were examined in relation to their sign and statistical significance. Standardized estimates are useful in comparing the parameters' effect throughout the model since they remove scaling information. All proposed hypotheses were tested by observing the statistical significance of the corresponding paths in the structural model.
The final stage involved focusing on the relationship between push factors, pull factors and perceptions. The analysis began by recoding these variables in order to obtain categorical variables. Then, through the application of a CATPCA, it was possible to explore the nature of these relationships. This method allows us to represent the various relationships of categorical variables in a low-dimensional space, also known as a perceptual map. In this sense, CATPCA can make data interpretation easier since the relative positions of categories on this map represent the relations among them. Categories with similar distributions are represented as points on the map and the proximities means that they are associated. On the contrary, categories with quite different distributions will be positioned far apart on the map, suggesting that they are not related. Thus, in this study, the perceptual maps depicting the relationships among push and pull factors, push factors and perceptions and, lastly, pull factors and perceptions, were presented. These maps complete the analyses provided by the structural equation model since they allow a detailed evaluation of the relations among each pull and push factor and perceptions. As PCA, this analysis was carried out with SPSS, 12.
Results
Principal components analysis
Push factors
The reduction of the initial 15 push motives into three new factors, together accounting for 68.4 percent of the total variance (KMO =0.9; Bartlett test: p=0.05), was obtained through PCA. These findings are presented in Table II. Considering the meaning of the motives with higher loadings, the push factors were labelled knowledge, leisure and socialization. The internal consistency of these factors was measured by the corresponding Cronbach's α coefficients (they exceed 0.8 in all cases).
Knowledge is the first factor and is specifically related to the need to explore new cultures and places as well as to do and learn new things. In the Leisure factor the principal motives are connected to personal well-being. The third factor, socialization, is primarily the desire to share travel experiences, develop close friendships and go to destinations not yet visited by friends.
Table II also shows the mean importance level of each push motive. The importance ranking was based on the mean scores on the Likert seven-point scale. These results are in accordance with the push motives determined in previous research (Sefton, 1989; Loundsbury and Franz, 1990; Pearce, 1988; Laing, 1987; Smith, 1991). Most of these studies are based on the leisure motivation scale of Beard and Raghep (1983) and show that the main push motivations to travel are related to intellectual motives (learning, exploring, discovering), social motives (friendship and interpersonal relationships, need for esteem of the others) and relaxation motives (escape from daily routine, get-away).
Pull factors
Table III presents results from the application of PCA to the nineteen pull motives. As with push motives, three new factors were extracted, together accounting for 63.5 percent of the total variance (KMO = 0.9; Bartlett test: p=0.0). Pull factors were labelled facilities, core attractions and landscape features. Cronbach's α for these factors also exceeds 0.8, indicating an adequate degree of internal consistency.
Facilities of the travel destination such as lodging, weather, food, security, roads and hospitality are included in the facilities factor. Motives related to the social environment, shopping facilities, sports activities and night life are encompassed by the second factor, core attractions. Landscape features is the last pull factor and is linked to the natural and cultural environment of the holiday destination.
The last two columns of Table III prove that the social and sports motives are the least important in choosing a tourist destination. Inversely, the natural and regional component clearly determines the destination image. This especially includes natural environment, weather, beaches and hospitality and can be considered as an indicator of how the natural resources of a tourist destination were perceived as competitive factors par excellence. These results are also in accordance to previous studies in the field which have been quoted above.
However, despite the fact that these studies have provided important contributions in order to define push and pull motivations, they do not examine the interactions among them (Ryan and Glendon, 1998). Yoon and Uysal (2005) state that inherent push motivations are stimulated and reinforced by destination attributes. Sequentially, different perception levels of a tourist destination are expected to result from different combinations of push and pull factors.
Structural equation model
Examining the nature of the relationships among push and pull factors and overall perceptions of the destination includes estimating the structural equation model proposals in Figure 1. Figure 2 shows the estimated standardized path coefficients on the model itself. The values for the selected overall fit measures are reported in Table IV. As expected due to the large sample size, the observed value for the χ2 statistic is high and statistically significant, suggesting that the observed and predicted covariance matrices are not equal. In considering the dependence of the χ2-test on the sample size, the evaluation of the absolute fit of the model relied on the other absolute measures, which are very high. The proposed model also presents high values for the remaining measures, suggesting an adequate incremental and parsimonious fit.
A measurement model analysis follows the analysis. This model could be improved if all the observed variables had significant loadings in corresponding latent variables. This was true for all indicators except for Leisure (p>0.05). On the other hand, the constructs reliability and variance extracted for the latent variables push motives and pull motives do not exceed the desirable levels of 0.7 and 0.5, respectively. Owing to these results, the model was re-estimated by excluding the indicator leisure. The results of this model, concerning the GFI and the standardized loadings, were essentially the same but the measurement model has clearly improved. For this reason, we decided to go on considering the model as had been proposed initially.
Figure 2 also shows that two specified paths between the constructs are statistically significant which supports only research hypotheses H1 and H2. The same results, with similar path estimates, were obtained in the restricted model: H1 and H2 are still supported by the data and H3 is rejected.
Categorical principal components analysis
In order to simultaneously analyze the relationship among each push factor, pull factor and perceptions, these variables were transformed into categorical variables. Table V shows the relative distribution of respondents by the categories of the new variables.
Having transformed the pull and push factors into ordinal variables, the study includes applying CATPCA to complement the evaluation that the structural equation model provides. CATPCA helps to deeply analyze the relationship between push and pull categorized factors and, afterwards, to relate pull factors and perceptions. The decision was made not to present the perceptual map connecting push factors and perception because the estimated model showed a non-significant relationship between these constructs.
Figure 3 shows the combination of categories of each pull factor with categories of each push factor through a perceptual map. As can be observed, high valorization of pull factors (represented by the full lines) is related with high valorization of push factors (represented by the broken lines) and vice-versa. This map also suggests which push and pull factors are strongly related, which was not detected with the structural equation model: socialization and knowledge are particularly related to facilities; leisure is specially associated to core attractions.
Figure 4 shows a similar analysis but in what concerns the connections between perceptions and pull factors. The relationship appears to be more fragile than the one previously reported between pull and push factors. As can be observed on the perceptual map, landscape features is the pull factor least related to perceptions and there is also a very similar connection between perceptions and facilities, on the one hand, and perceptions and core attractions on the other hand.
Discussion and managerial implications
The theoretical model proposals offer a logical sequence of the formation of tourist perceptions. The empirical study proves that perceptions of tourism destinations are formed based on push and pull factors. The study supports the view that all destination attributes contribute to the perceived image of the destination. The formation of perceptions seems to happen in the following sequence: push factors are determinants of pull factors which, in turn, explain perceptions.
In the model, the relationship between push factors and perceptions is not significant, and that proves that the tourist decides to go on holiday because he/she needs to solve a conflict arousal (rest, social and intellectual rewards). Then, he/she decides where to go based on the destination attributes. The destination attributes (pull motives) are perceived as the way to solve intrinsic motives (push motives), but these constructs are not directly related to the overall perception of the destination because they are apparently solved when the tourist turns her/his attention to specific attributes. Figure 3 also supports the view that a high valorization of all pull factors is associated to a high valorization of all push factors. This evidence depends upon the very essence of humans' nature. Humans are continuously seeking and solving problems, with different levels of involvement with the decision. This conclusion also stresses the need for testing the inclusion of emotions and learning as constructs of the theoretical model.
Looking deeply into the factors identified for the pull and push attributes, this study proved that knowledge and socialization are the major motives which cause the need to travel abroad. This evidence is in accordance to Oppermann (1996), who stresses that travelling to unknown places (and this is the case of exotic places) gives rise to a variety seeking behavior, looking for social and intellectual rewards. Similarly, Gnoth (1997) highlights the difference between social trip and a variety seeking trip. While the first one represents the choice of a fashionable exotic resort, the second one is more related to savage and unexplored environments. So, the social trip is more associated to social status, self esteem and sense of belonging. The variety seeking trip is related to adventure and intellectual rewards.
The destination attributes highlighted by the PCA represent the core tourism product: facilities, core attractions and landscape features. Figure 4 shows that the facilities factor is the prior motive that contributes to overall perception. Although knowledge is one of the major push motives which generally determine the choice of an exotic destination, tourists assume that these places abound in landscape and scenic aspects and so they do not look for much information about these aspects (they are assumed to be intrinsic characteristics of the destination).
Considering the correlation between push and pull motives that Figure 3 shows, socialization and knowledge, the more reliable push motives in the structural model, stimulate the mind of the tourist for the need for attributes related to the facilities of the destination. The tourist who decides to travel in order to increase his/her knowledge would be expected to reveal more concern about the landscape features and sport activities. In turn, the tourist who travels in order to increase his/her social status would rank higher facilities and core attractions. However, the empirical results prove that the tourist who travels due to intellectual or social rewards decides where to go based on facilities. Even though this result could seem strange, the result reveals that tourists are conscious that they are traveling to destinations where human and social development is low and, therefore, main concerns turn to where they will sleep, eat as well as to how they will travel throughout the destination. This is an issue that needs more research.
Still concerning the relationship between push and pull motives, Figure 3 shows that leisure is mainly correlated with core attractions. This result can also seem strange if rest and relaxation are perceived in a physical perspective. Portuguese tourists seem to understand rest and relaxation from a different point of view: relaxation and rest does not mean “doing nothing” but, instead, “doing different things.” In this sense, in the case of Portuguese tourists, the statement that leisure is more mental than physical can be made.
The major findings of this study have significant managerial implications for destination marketing. Firstly, the exploratory analysis shows that the decision to travel to exotic places arises from the desire of knowledge, having social status and intellectual leisure and also that these factors determine the perceived pull motives: facilities, core attractions and landscape features. Since, tourists were found to be more aware of facilities and core attractions, marketing of these destinations must be especially focused on these factors rather than on beautiful images of natural landscapes. These empirical results could be fundamental in order to direct and target the destination marketing strategy to enhance the overall perception of these exotic places.
Secondly, special attention must be given to the type of activities which are offered to tourists. Even tourists who simply want to relax do not relate leisure to “doing nothing.” According to the structural model and CATPCA analyses, social and intellectual rewards are influencing factors in what pertains to the demand for exotic destinations. The offering of appropriate activities and attractions to the right tourists may be the result of the knowledge gained through identifying why people travel the way they do. These results indicate that marketers and managers of exotic destinations should concentrate on improving push factors in order to enhance the destination's competitiveness.
Thirdly, CAPTCA results suggest two levels of awareness in what concerns the perception of an exotic location: one with tourists who are highly involved and the other with tourists who are only moderately involved. To avoid dissatisfaction at the end of the trip the latter group should receive special attention from destination managers.
Conclusions
This study contributes to the overall understanding of why tourists behave the way they do from four different perspectives. First, despite the number of models which had considered the relations between push and pull motives, the inclusion of perceptions in the theoretical model is a reason to believe that this research provides a step forward in literature. Second, this was the first time that structural equation modelling and CATPCA were put together in order to prove the relationships between push, pull motives and overall perceptions. Third, only now has the outbound Portuguese tourist market been studied at a more scientific level. Lastly, this was the first time that motivations and perceptions were combined in order to understand why people could be pushed to travel to exotic destinations and how they form their perceptions.
Empirical evidence in this study opens paths for further investigation. The study does have some limitations, however, the main one being the domain which can be considered as restricted. Generalizing the model is an important step to better understand tourist behavior. Extending the study using other destinations and attributes, and tourists with different incentives would be a useful step.
 |
 |
 |
 |
 |
 |
 |
 |
 |
| References |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Antónia Correia can be contacted at: acorreia@ualg.pt |