Inferring Future Vacation Experience Preference from Past Vacation Choice

Abstract

Tourists make two fundamental decisions when they travel: where to go (destination) and what to do (experience/activity). Whereas the modeling of destination choice behavior has a substantial research history, there has been little research that has sought to model and explain the choice behavior of tourists when deciding what type of vacation experience they wish to undertake. To address this gap, this study investigates how experience types are related with regard to actual past tourism consumption and preferred future experience choice. In order to investigate these relationships, the authors employed latent class modeling on survey data from a sample of respondents who had reported their recent and intended future vacation travel choices and preferences. Despite the acknowledged importance of variety seeking in a tourism choice context, the study found strong evidence for the stability of preferred experiences. In addition, it found insightful patterns of association between experience types.

Keywords

tourism experiences experience preference preference stability latent class modeling

Introduction

When seeking to understand an individual’s tourism consumption behavior, we may wish to know answers to a variety of questions. How much did the person spend? Who did they travel with? How long were they away? Did they have a good time? What did they enjoy most? But above all else, the questions most often requiring answers are, Where did the person go and what did they do? Arguably nothing tells us more about a person’s vacation than the answers to these two questions. Knowing what destination or location a person has chosen and what experience (defined here as the type of vacation activity, with both terms used interchangeably) they engaged in reveals a great deal about that particular tourism consumption decision. Knowledge of the chosen destination and visitor experience, then, provides an important contextual basis for evaluating the other elements of consumption choice.

In addition to helping us better understand these other aspects of consumption choice, information about the destination and the vacation activity may also be pivotal in helping us better understand future vacation choices. Some tourists may seek to return to the same destination or undertake the same experience or to consider other but similar destinations or experiences. Other tourists may instead prefer rather different destinations or experiences, either because they seek something new or novel or because they have discovered that their travel tastes differ from what they thought they might enjoy. The choice of a future tourism destination or experience may also reflect a pattern of compatible or complementary alternatives that, although different from past choices, is indicative of an underlying preference structure connected with an individual’s personality or self-concept. For example, an outdoors type might prefer vacation destinations or experiences that focus on nature, sports, and adventure but have little interest in entertainment and nightlife.

While some tourists may possess such clear and seemingly compatible preference structures, the preferences of others may be so eclectic and diverse that using past destination or experience choice as a guide for future vacation choice is clouded by other (perhaps unobserved) choice factors. Nevertheless, each individual tourist may possess a travel preference structure, or preference “fingerprint,” and consequently market segments may be characterized by differences in these preference structures.

Given the importance of knowledge of destination and experience choice behavior, we would expect to find a great deal of research into both. However, while there is considerable evidence in the literature of the former, by comparison, the literature focused on the choice of experience is surprisingly limited and sketchy. The study of destination choice behavior has examined a wide range of topics including destination choice sets and choice modeling (e.g., Crompton and Ankomah 1993; Huybers 2003; Pike 2006; Seddighi and Theocharous 2002; Woodside and Lysonski 1989), destination image (e.g., Baloglu and McCleary 1999), destination loyalty and repeat visitation (e.g., Alegre and Juaneda 2006; Changuk 2001; Hong, Kim, and Lee 2006; Oppermann, 2000), econometric modeling of tourism flows to particular destinations (Eymann and Ronning 1997; Crouch 1995), destination variety seeking behavior (e.g., Bigné, Sánchez, and Andreu 2009), and destination promotion (Johnson and Messmer 1991). In contrast, an examination of the literature reveals only a few equivalent studies focused on tourism choice and decision-making behavior with regard to types of experiences or activities. We review the existing research in the next section. The point we wish to make, however, is that experience choice research has not enjoyed the same level of interest and attention as has destination choice research.

The limited attention in the literature to choice of experience type reflects the historical, general practice of tourism promotional activities aimed principally at attracting tourists to a particular place rather than enticing them to engage in particular activities or experiences at the destination. King (2002) argues that destination marketing bodies need to shift their focus from promoting the destination to creating and promoting available experiences as competition and available choices have expanded. “The destination marketer has a clear role in facilitating the connection between the customer and the experience they are seeking” (p. 108). Anecdotal evidence suggests that destination marketers have begun to recognize this need to emphasize the experiential features of alternative tourism products. While, in the past, destination promotional images tended to show attributes of the place itself, it has become more common to see images showing tourists engaging in activities enabled by that place. In recent times, destination promotions also usually go beyond triggering attention by attempting to create interest and arouse desire to additionally enabling action (e.g., through the use of coupons and links to Internet sites where information is available for planning and booking).

The available research to date on the tourist experience tends to be of a more recent origin. This coincides with the growth of both scholarly and practitioner interest in experiences in the consumer choice and satisfaction literature as exemplified in the more popular literature by Pine and Gilmore (1999), who talked of the experience economy. This research has extensively dealt with consumer experiences in a range of tourist, hospitality, and retail settings, including aspects of experience measurement and establishing the dimensionality of the experience (e.g., Kim, Ritchie, and McCormick 2012) and how it relates to consumer satisfaction and store choice (e.g., Sands, Oppewal, and Beverland 2009) and to value creation, with the latter research increasingly focusing on consumer co-creation (e.g., Carù and Cova 2003; see also Pine and Gilmore 2012). However, despite this attention to a range of experience facets, the research literature has paid little attention to preferences for vacation experience types.

The aim of the research reported in this article has been to help address this comparative neglect by examining some fundamental determinants and patterns of vacation experience preferences and choices and exploring the potential for developing analytical and predictive tools that can provide practical and insightful assistance to tourism marketers. More specifically, we seek to model how recent past tourism experience choice relates to future experience preference (Mazursky 1989) and how this relationship depends on tourist demographic factors and motivations. Our aim is to examine how such relationships can be effectively analyzed to derive segments that assist in marketing decision making.

Background and Theory

If it were possible to explain a significant portion of the observed heterogeneity of preferred tourism experience/activity types, that knowledge could be employed to design and develop tourism experiences and to target those experiences to the most interested market segments. Some humans have an innate restlessness and there is growing evidence to suggest that some of this urge is part of our human nature. Specific gene variants have now been linked to aspects of personality, including openness to new experiences, novelty, sensation seeking, and risk taking (Savitz and Ramear 2004; Schinka, Letsch, and Crawford 2002; Dobbs 2013 ; Wallace 1991). As tourism experiences are often rich in such qualities, it has been suggested that some of the variability in tourism choice behavior may be conditioned by such innate genetic predispositions (Crouch 2013). There is evidence that in the context of tourism consumption, travelers seek a degree of change, novelty, escape, exploration, sensation, and variety in unfamiliar (or less familiar) environments (Assaker and Hallak 2013; Chai 2012; Weaver et al. 2009; Fuchs 2013; Wymer, Self, and Findley 2010). So while some tourists may open themselves to the possibility of “new” experiences, others may display a preference to more stable choices. These considerations might suggest that the challenge of explaining preferred tourism experiences is made more difficult because of the intrinsically variable nature of tourism experience choices, at least for some. What is important is whether this variability is random or, instead, whether some of it can be understood and explained or predicted based on knowledge of stochastic patterns or relationships.

In order to investigate potential explanatory variables and relationships, we examine the available literature on tourist experience choice behavior and decision making. While over the last several years there has been a growing literature on the general topic of “tourism experience,” the literature shows that very few studies actually investigate how and why tourists choose or prefer particular types of tourist experiences. The majority of tourism research articles associated with the search word experience/experiences examine numerous and varied issues such as the meaning of tourism experiences to consumers (e.g., Lindberg, Hansen, and Eide 2014), the quality, emotions, motivations and satisfaction with destination/visitation experiences (e.g., Hosany and Gilbert 2010; Huang and Hsu 2009), the design and management of tourism experiences (e.g., Guthrie and Anderson 2010; Neuhofer, Buhalis, and Ladkin 2013), and experience cocreation (e.g., Sfandla and Björk 2013), among many other topics. Only a handful of studies have actually sought to model, explain, or predict tourism experience choice from among the different types of experience activities available.

While variety seeking appears an important determinant of tourist choice in some individuals, past behavior may nevertheless be useful for inferring or predicting future preference and interest. This is illustrated by Pearce and Kang (2009), who studied the effects of prior and recent experience on continuing interest in tourist settings. Their dependent variable was “future interest in the setting,” which specified visitor interest in the type of setting in general rather than an interest in returning to a specific business or location. So their study defines settings not at the level of the destination but at a more local level associated with particular experience types. Their results reported “a set of findings linking experience levels and future interests” (p. 180). The findings by Sönmez and Graefe (1998), in their study of traveler reactions to risk and safety issues, also confirm the value of using past travel experience as a predictor of future travel behavior. Bello and Etzel (1985) explored the behavioral and demographic differences associated with different levels of novelty in vacation experiences. They established that “high-novelty experiencers are more predisposed to take another similar vacation. However, they also indicate a lower likelihood of returning to the same destination” (p. 24). This suggests that greater novelty is sought more with a view to the vacation location than in relation to the overall type of vacation experience or activity; that is, vacation choice stability tends to occur in terms of tourists’ preferences for vacation experiences. This is supported by Lehto, O’Leary, and Morrison (2004) who examined the effect of prior experience on vacation choice behavior. Part of this study focused specifically on involvement in particular experience activity types. They concluded that activity participation patterns were strongly influenced by prior experience with an activity, a finding consistent with Bryan’s (1977) recreation specialization theory, which contends that people progress through an activity spectrum from being a generalist to a specialist. The latter is consistent with Mehmetoglu (2005) who found that distinguishing between activity generalists and specialists helped explain important differences in behavior in the context of the choice of nature-based tourism experiences. Thus, activity specialization may reinforce preference stability and may lead to positive associations between similar types of experience activities.

While there may be an element of stability in terms of the variation of experiences or activities that tourists seek, it is likely that such stability declines over the long term. Indeed, Crawford, Godbey, and Crouter (1986) found that preferences for leisure activities were significantly positively correlated over time but that a pattern of decay occurs. They note that “to the extent that they are stable over time, it may be possible to use the results of such surveys as a basis for planning for several years” (p. 96). With respect to the aims of our current study, to the extent that there is decay in preferences over time, if past tourism experiences help to predict future experience preference, relatively recent past experiences are likely to be more efficacious.

The above discussion suggests that there may be a degree of persistence between past and future vacation experience decisions. However, there may be other influencing factors that play a role such as tourists’ demographic characteristics. Crawford, Godbey, and Crouter (1986) observed that the stability of preferences over time is likely to be greater when driven by biologically-based traits (such as introversion-extraversion and sensation-seeking) than by socially determined traits. This observation is consistent with recent evidence from the field of genetic science which shows that, while environmental and situational factors may exert some influence, novelty-seeking behavior appears to be, to a significant extent, heritable (Crouch 2013). Crawford, Godbey, and Crouter (1986) also suggest that decay in leisure preferences differs by gender; in particular, decay is greater for men than for women. Similarly, Lehto et al. (2008) studied the link between the age of the tourism consumer and the stability of their preferences for vacation experiences. Consistent with Atchley’s continuity theory (1993) and the findings of Lohmann and Danielsson (2001), they found that age cohorts correlate with tourism experiences and activity participation. Bello and Etzel (1985), however, found that those who participate in novel vacation experiences and those who participate in “commonplace” experiences do not differ demographically.

In addition to demographic traits, overall travel motivations may affect the relationship between past vacation activities and intended future activities. Tourists repeating their previous vacation in terms of the actual destination or the vacation activity are often interpreted as a manifestation of loyalty. Backman and Crompton (1991) examined the usefulness of selected variables for predicting activity loyalty in the context of users of parks and recreation facilities (in particular, participation in golf and tennis). One of the five independent variables pertained to motivation defined by personal competence, mastery, and intrinsic/extrinsic motivation toward an activity. Their analysis demonstrated a significant effect of motivation on loyalty.

In summary, only a few studies have sought to explain or model decision making among discrete types of tourist experiences. This is perhaps surprising given the potential value of such knowledge from a marketing perspective. The limited research discussed above suggests, however, some insights that serve as a starting point for the present study and the formulation of the following research hypotheses. The research indicates that past vacation activity choices may help predict preferences for future experiences but that this relationship is likely to be moderated by a series of demographic and psychographic variables, in particular, travel motivations.

We therefore propose the following research hypotheses and will explore them in our empirical study reported hereafter:

Hypothesis 1: Preferences for future tourist experience types are overall positively correlated with actual past tourist experience choices.

Hypothesis 2: Individuals vary in their preference for similar future and past vacation experiences, such that groups with relatively homogeneous preference structures can be identified in the population of potential travelers.

Hypothesis 3: These groups relate to identifiable demographic and psychographic characteristics of travelers.

We now outline our research approach.

Research Design and Data

In order to address the research aim, data were collected in an online survey among a sample of residents of Melbourne, Australia. The survey was introduced as a study on vacation (or holiday) choices, where “a holiday trip is a trip mainly for pleasure and recreation involving travel away from home for at least one night. A trip for a holiday might include other purposes as well, such as business, visiting friends and relatives, etc., but the primary purpose is a holiday.”

Data on experience preferences was obtained by presenting respondents with a hypothetical travel opportunity and then asking them for their travel preferences. Specifically, they received a brief scenario as follows: “Suppose that, in a raffle, you have won a prize consisting of a holiday for two people in Australia or the Asia Pacific region. The prize includes travel and three to five nights accommodation and provides one of eight different holiday experiences.” We employed this technique of establishing a hypothetical scenario in order to assist in the reduction or control of other variables or differences between responses that may cloud the underlying relationships we wished to model.

Immediately after the prize setting had been introduced, respondents were asked, “Which of the following experiences would you be interested in choosing? (Select up to three).” The question provided eight experience categories from which respondents could choose, plus a ninth category (i.e., none of the other eight). These eight experience categories accorded with the definitions employed by Tourism Australia (Australian Experiences Toolkit 2007) and were as follows:

Arts, culture, history, and heritage;

Entertainment, nightlife, and shopping;

Festivals and events;

Food and wine;

Indigenous culture;

Nature (beaches, waterways, wilderness, and wildlife);

Relaxation, health, and indulgence; and

Sports, outdoors, and adventure.

Being asked to make up to three choices forced respondents to make a selection from the available eight but did not limit their choices to just one, allowing them to express a broader range of interests. Consequently, this data is a form of “pick any” choice data.

For information on previous actual travel experiences consumed, we next asked the question: “Over the last five years, have any of these “experiences” been a major part of any holiday trip you undertook? (Click any that apply)”—using the same eight options in a closed-response format. We chose a time frame of five years as we felt this would be long enough to provide for the possibility that respondents have had time to display some of their experience behaviors, but not so long that recall ability and memory become a problem. Respondents were then shown the same eight experiences and were asked to indicate one experience about which they would like to receive more information when choosing one of the experiences for their holiday prize.

Additional questions in the survey gathered demographic data, including gender, age, employment status, household dependents, marital status, education, and income. In order to obtain information on travel motivations, the survey included 24 further questions introduced as follows: “The following is a list of the most common holiday travel motives. Please indicate how important these motives are to you when you travel on a holiday.” These were answered using a four-point scale (1 = not at all important to 4 = very important). The 24 questions covered items such as to be close to nature, to engage in challenging physical activities, to learn about new things/places/cultures, to relax mentally, to master a skill, and to have social contact/meet new people. The motivation scale was adapted from Mehmetoglu (2005) and expanded using table 10 from Horneman et al. (2002). The total list of items can be found in the relevant tables below.

The data were collected via an Australian commercial online panel. The panel organization drew a random sample of residents from the greater Melbourne metropolitan area as available from their member register, which includes a large number of residents from across the population recruited to participate in online surveys for small incentives. Quotas were applied for income, education, and age groups to obtain a sample that would be broadly representative of the population. We used an online panel provider since we wished to investigate the relationships between past vacation experience and future experience preferences across a large sample from across the population, and the use of an online panel provides significant efficiencies for this purpose. Note the data collection was cross sectional; we relied on respondents’ recall of previous experiences and their stated preferences for future experiences as explained above. Hence, our use of an online panel should not to be confused with “panel data” as used for gathering longitudinal data. An initial sample of 100 respondents was run as a pilot test but as no problems were detected and no changes were required, we consequently combined the pilot sample with the main sample, resulting in a total sample size of 922 completed surveys. Table 1 summarizes the overall profile of all respondents.

Table 1.

Overall Profile of Respondents.

Characteristic	Percent	Characteristic	Percent
Gender		Employment
Female	59	Full-time employee	55.5
Male	41	Part-time employee	13.8
		Casual employee	4.3
Age (years)		Full-time home duties	5.2
18-24	16.0	Retired	4.5
25-34	17.2	Full-time student	8.6
35-44	30.0	Unemployed/seeking work	1.4
45-54	22.2	Employer/contractor	5.8
55-64	11.4	Other	0.9
≥65	3.2
		Household income ($)
		<20,000	3.2
Education		20,000-40,000	6.8
<Year 12	15.9	40,001-60,000	13.7
Year 12	20.9	60,001-90,000	20.3
Diploma/certificate	23.5	90,001-120,000	16.4
Undergraduate	22.6	120,001-150,000	9.2
Postgraduate	16.1	150,001-180,000	4.9
Other	0.9	>180,000	3.8
		Would not say	21.7

Analysis, Results and Discussion

Experience Choice and Future Interest

Table 2 shows the percentage and ranking of respondents who had actually consumed each experience in the past five years and the percentage interested in undertaking each type of tourism experience in the future.

Table 2.

Tourism Experiences—Past and Future Interest.

Experience	Past^a		Future Interest^b
Experience	%	Rank	%	Rank
Nature	56.5	1	65.8	1
Relaxation, health, and indulgence	53.4	2	53.3	2
Entertainment, nightlife, and shopping	39.5	3	45.3	3
Arts, culture, history, and heritage	32.9	4	41.6	4
Food and wine	30.6	5	31.9	5
Sports, outdoors, and adventure	25.1	6	31.1	6
Events and festivals	24.8	7	29.4	7
Indigenous culture	7.4	8	14.0	8
None	0.4	-	4.7	-

Undertaken in the past five years as a major part of any holiday.

A person could indicate a maximum of three interests; interests refer to experiences in context of a holiday voucher prize including travel and three to five nights accommodation somewhere in Australia or the Asia Pacific Region.

The three most favored experience types in terms of both recent past consumption and future interest in participation were nature; relaxation, health, and indulgence; and entertainment, nightlife, and shopping. The least favored experience concerned indigenous culture. Ranked from most to least favored, recent past consumption ranks are identical to future interest ranks. These figures point broadly to the likely significant link between past vacation experiences consumed and the persistence of preferences toward similar tourism experiences in the future.

Travel Motivations

Table 3 summarizes the sample mean ratings for each travel motive. Overall, the five strongest motives were to have fun (3.83), to visit new places (3.67), to relax mentally (3.61), to get away from everyday life (3.60), and to do/experience something new (3.53). In contrast, the five weakest motives were to master a skill (2.06), to engage in challenging physical activities (2.21), to improve self-confidence (2.47), to engage in nature-based activities (2.62), and to engage in nonchallenging physical activities (2.62). Factor analysis (in particular, principal components analysis with varimax rotation) was used to investigate the underlying response patterns across the 24 motivation items and allow incorporation of the travel motivation factors in the later modeling. Table 4 summarizes these findings.

Table 3.

Travel Motives Mean Importance Ratings.

Motive	Mean Rating^a	Rank
To have fun	3.83	1
To visit new places	3.67	2
To relax mentally	3.61	3
To get away from everyday life	3.60	4
To do/experience something new	3.53	5
To gain experience/knowledge	3.42	6
To learn about new things/places/cultures	3.40	7
To spend time with my family	3.38	8
To be with friends and relatives	3.23	9
To have time for myself	3.22	10
To have experiences to talk about	3.16	11
To visit natural attractions	3.10	12
To experience something adventurous	2.95	13
To have social contact/meet new people	2.86	14
To improve my health	2.81	15
To experience nature	2.79	16
To obtain some physical exercise	2.79	17
To obtain a feeling of achievement	2.72	18
To be close to nature	2.69	19
To engage in nonchallenging physical activities	2.62	20
To engage in nature-based activities	2.62	21
To improve my self-confidence	2.47	22
To engage in challenging physical activities	2.21	23
To master a skill	2.06	24

1 = not at all important to 4 = very important.

Table 4.

Factor Analysis of Travel Motives.

Motive	Factor
	1	2	3	4	5	6	7
	Nature	Social and Self-Enhancement	New Experiences and Knowledge	Health and Exercise	Relaxation and Escape	Adventure and Excitement	Family, Friends, and Fun
1. To be close to nature	.866	.013	.069	.064	.052	.089	−.041
2. To experience nature	.896	.029	.108	.051	.016	.077	−.016
3. To visit natural attractions	.777	−.002	.252	.005	.019	.001	.080
4. To engage in nature-based activities	.833	.036	.068	.078	−.052	.228	.037
5. To engage in nonchallenging physical activities	.500	.019	.066	.103	.105	−.002	−.033
6. To engage in challenging physical activities	.216	.069	−.005	.252	−.102	.765	.025
7. To do/experience something new	.134	.079	.714	−.051	.126	.373	.030
8. To visit new places	.090	.082	.848	−.072	.136	.124	.090
9. To learn about new things/places/cultures	.217	.145	.820	.087	−.017	−.142	.011
10. To gain experience/knowledge	.170	.230	.803	.132	−.046	−.023	−.018
11. To relax mentally	.122	−.031	.020	.168	.783	−.177	.066
12. To have time for myself	−.004	.193	−.001	.317	.702	−.080	−.052
13. To spend time with my family	.021	−.124	.039	.484	.225	−.191	.553
14. To obtain some physical exercise	.162	.078	.036	.759	.068	.326	.098
15. To improve my health	.141	.200	.044	.796	.203	.070	.076
16. To master a skill	.098	.468	.001	.459	−.051	.420	−.037
17. To get away from everyday life	.051	.104	.121	−.102	.702	.245	.191
18. To experience something adventurous	.090	.252	.168	.061	.071	.809	.035
19. To have fun	.004	.054	.171	−.136	.412	.199	.597
20. To be with friends and relatives	−.013	.292	−.037	.170	−.058	.022	.783
21. To have social contact/meet new people	−.050	.712	.153	.076	.015	.079	.237
22. To have experiences to talk about	−.031	.710	.214	−.074	.122	−.001	.229
23. To improve my self-confidence	.068	.836	.049	.155	.084	.111	−.057
24. To obtain a feeling of achievement	.078	.828	.122	.101	.057	.160	−.060

The analysis identified seven principal factors, labeled in the top row of Table 4 to reflect the essence of the loaded items. Because of the significant differences between study contexts and aims, it is difficult to make direct comparisons between these findings and the results of other studies that have factor-analyzed tourism motivations. However, in broad terms, these underlying factors have often arisen in studies that have examined core, underlying motivations. The desire to experience nature in one form or another has always been regarded as a strong pull motivation. The importance of social and self-development motivations find support in studies by Pan and Ryan (2007), Le and Pearce (2011), Sohn and Yuan (2013), Ai-ping (2009), Mak, Wong, and Chang (2009), and Park, Reisinger, and Kang (2008). The novelty of new experiences has been emphasized in studies by Le and Pearce (2011), Chen and Xiao (2013), and Park, Reisinger, and Kang (2008). Pan and Ryan (2007), mastery; Le and Pearce (2011), education; and Kim and Ritchie (2012), learning identify the importance of developing some skill or capability. The desire for relaxation and escape though is probably the most frequently noted underlying travel motive as evidenced in the studies by Pan and Ryan (2007), Kim and Ritchie (2012), Ai-ping (2009), Mak, Wong, and Chang (2009), Artuger and Kendir (2013), and Van der Merwe, Slabbert, and Saayman (2011). The need to enhance relationships with family and friends and to make new acquaintances are also typically found to be important (Pan and Ryan 2007; Kim and Ritchie 2012; Mak, Wong, and Chang 2009; Cohen and Ben-Nun 2009; Correia and Pimpao 2008; Park, Reisinger, and Kang 2008).

Factor analysis provides factor scores for each respondent. These scores serve as a measure of how well each respondent aligns with each of the seven motivation factors. The individual factor scores were used in the estimation of the latent class model discussed in the following section.

Latent Class Regression Modeling

We now proceed to account for the multivariate relationships and will apply latent class regression analysis to address (1) which interests are attractive to particular respondents; (2) which past experiences are associated with, and hence possibly can be used to infer, respondents’ interests; (3) whether subgroups or segments of respondents can be identified with different relationships between past experiences and future interests; and (4) whether these segments can be described by demographic or motivational characteristics.

Latent class regression assumes the possible existence of classes (or segments) of respondents who differ in their responses to a set of independent variables and who consequently will display different weights in a regression function that maps these variables onto the response variable. A multinomial logit function next predicts the class to which each respondent belongs. Latent class analysis specifies a likelihood function across these two functions and uses maximum likelihood estimation, as an iterative procedure, to derive the parameters that optimize the model fit across the two model components. That is, whereas traditional clustering methods first derive the number of clusters and determine the optimal cluster composition to then explain cluster membership, latent class analysis simultaneously determines the statistically optimal number of classes, the parameters of a multiple regression equation within each of these classes, and the parameters of a multinomial regression that predicts class membership from a set of background variables (e.g., Agresti 2002; Vermunt and Magidson 2005; Wedel and Kamakura 2000). This simultaneous optimization and the use of a statistical model instead of ad hoc rules as applicable in cluster analysis are among the approach’s main benefits over traditional multistage clustering methods. Van der Ark and Richards (2006), for example, used a latent class modeling approach to analyze cultural tourism behavior and destination preference for 19 European capital cities. In the present case, the approach allows us to determine how many segments should be defined to best describe respondents’ “pick any” interest choices as a function of their past experiences, demographic factors, and travel motives.

The latent class regression procedure consists of first defining the regression equation that is estimated within each class. Because our dependent variable is binary (each experience is either of interest or not), we apply a logistic regression routine. The probability of vacation experience option i being selected conditional on the respondent being in class c is specified using the logistic model as follows:

p (i, c) = \exp (u_{i, c}) / (1 + \exp (u_{i, c}))

where, u_i,_c is a linear function of the constants defined for the N available experiences and an error term:

u_{i, c} = b_{c 0} + b_{c 1} X_{1} + b_{c 2} X_{2} + \dots + b_{cN} X_{N} + ε_{j, c}

in which X₁, X₂,…, X_N are dummy variables for the presence or absence of each of the N possible experience types (note that there are N dummies, not N – 1 dummies, because the survey allowed for multiple choices per respondent), b_c0 is the intercept and b_c1, b_c2, etc. are the parameters to be estimated for each class c separately, and ϵ_i,c is an error term that follows a logistic distribution.

Because in our study, all experience types are always available and no experience attributes are specified, the logistic regression equation simply comprises a set of constants, one for each possible experience of interest (if there had been variation in the availability of experiences, this could have been accommodated by coding the dummies of nonavailable options as zero). The general constant (intercept) in the model captures the overall responsiveness in terms of whether one or multiple types were selected. This approach thus allows accommodating the multichoice (or “pick any that apply”) nature of the data as each respondent could select multiple experiences of interest (which is different from the typical multinomial case where each respondent chooses one option only).

The estimation of this regression equation across all data, without allowing for multiple classes, describes how often each experience on average was selected. For completeness and for illustration purposes, before going into the estimation method and multiclass details, we present the findings of the single-class estimation for our data in Table 5. The findings show that “nature” has the highest coefficient, and hence across the whole sample it is most likely to be selected as one of the future experiences of interest. The predicted choice probability for nature can be derived using equations (1) and (2), noting that for this outcome the dummy is 1 while for all other experiences the dummies are 0. Using equation (2), u_i is −0.397 + 0.968, which is 0.571, and using equation (1), the choice probability for nature is exp(0.571)/(1 + exp(0.571)), which is .639. This is close to but not equal to the observed marginal choice probability of .658 for nature in Table 2. Similar calculations for the other types reveal that most have larger discrepancies from the observed choice, indicating that the single-class model does not summarize the overall choice pattern particularly well. This low fit is further confirmed in the R-square value being only .099, or less than 10%.

Table 5.

Estimation Results for Single Class Regression Model.

		Class 1
R ²		0.099
Class size		100%
Model for Dependent		Class 1
	Intercept	−0.397
Future interest	Arts, culture, history, and heritage	0.128
	Entertainment, nightlife, and shopping	0.574
	Festivals and events	−0.225
	Food and wine	−0.108
	Indigenous culture	−1.817
	Nature	0.968
	Relaxation, health, and indulgence	0.836
	Sports, outdoors, and adventure	−0.355

The multiclass analysis consists of estimating, simultaneously, models for different sets of classes and determining the optimal number of classes. The solutions for the different numbers of classes are compared and typically the solution with the lowest Bayesian information criterion (BIC) value is deemed the optimal solution, as it achieves the highest statistical fit with the least numbers of parameters. While the parameter values of the binary logistic regression equation vary across classes, the model also includes class predictors. These are variables predicting to what class each respondent belongs. This component in fact comprises a multinomial logit model, with class membership as the dependent variable.

In case of two classes, the model comprises a binary logistic model with class membership as the dependent variable and class predictors as the independent variables. When there are more than two classes, the class membership estimates will be relative to one class that is set as a reference class. Formally, the log odds of a respondent being in class c instead of in the reference class c’ is

\begin{array}{l} Ln [p (c) / (p (c)] = g_{c 0} + \\ g_{c 1} X_{1} + g_{c2} X_{2} + g_{c 3} X_{3} + \dots \end{array}

where X₁, X₂, etc. are class predictors, g_c₀ is an intercept term and g_c₁, g_c₂ etc. are the parameters to be estimated for class c.

The final outcome of the analysis is a model that incorporates the logistic regression equation for predicting vacation experience interest choices for each class as well as the multinomial model that predicts class membership from a set of background variables. In the present case, the class membership equation includes the most relevant variables of the current analysis: whether an experience has been part of a recent vacation, selected demographic variables, and vacation motivations.

Model Findings

The estimation was conducted using the maximum likelihood estimation routine as implemented in the Latent Gold software (version 5 from Statistical Innovations). The final model is selected after estimating a series of models with an increasing number of classes. A model with five latent classes was optimal as it has the lowest BIC (as shown in Table 6). The model parameters are displayed in Table 7. The intercepts of the within-class regressions are constrained to be equal across classes in the final model as there were no significant differences between the constants when they were unconstrained (Wald = 1.589, p > .10) and the BIC was higher (8089.49, Npar = 128, R² = 0.380) than for the constrained model. This indicates that the classes do not differ in how many experiences a person selected on average. The other predictors in the model for the dependent variable (vacation experience interest) concern dummies for whether interest was expressed for an experience. As a group, the dummies are significant (Wald = 1216.108, p < .001) and they also differ significantly across classes (Wald = 741.693, p < .001). In the class prediction model, the results reveal significant effects for all past experiences, age (using number of age category, with youngest category coded 1, oldest 6), income group, and all motivational factors except “health and exercise.” Gender is marginally significant (p < .10) and education is not significant.

Table 6.

Latent Class Regression Performances.

Number of classes	LL^a	BIC^b(LL)	Npar^c	R ²
1	−4316.85	8688.30	8	0.10
2	−4053.89	8360.37	37	0.20
3	−3858.50	8167.56	66	0.31
4	−3717.34	8083.20	95	0.36
5	−3615.94	8078.37	124	0.38
6	−3543.72	8131.90	153	0.44

Log likelihood values.

Bayesian information criterion values.

Number of model parameters

Table 7.

Latent Class Estimation Results.

		Class 1	Class 2	Class 3	Class 4	Class 5	Overall
R ²		0.394	0.426	0.364	0.242	0.264	0.377
Class size		25.1%	20.2%	18.6%	18.1%	18.0%
Model for Dependent^a		Class 1	Class 2	Class 3	Class 4	Class 5	Wald	p Value
	Intercept	−0.546	−0.546	−0.546	−0.546	−0.546	449.298	.000
Future interest	Arts, culture, history, and heritage	−1.150	2.140	−0.596	0.513	−0.446	1216.108^c	.000
	Entertainment nightlife and shopping	0.217	−0.121	−0.670	0.710	1.296
	Festivals and events	0.072	−0.280	−0.392	0.056	0.342
	Food and wine	0.423	−0.179	−0.181	0.678	−0.502
	Indigenous culture	−0.847	−0.228	−0.875	−0.779	−1.816
	Nature	1.235	1.132	1.040	−0.904	0.263
	Relaxation, health, and indulgence	1.242	0.261	0.427	0.693	0.364
	Sports, outdoors, and adventure	−1.191	−2.726	1.246	−0.967	0.499
Model for Classes^b		Class 1	Class 2	Class 3	Class 4	Class 5	Wald	p Value
	Intercept	0.493	−1.650	−0.615	−0.364	2.137	16.500	.002
Past experience	Arts, culture, history, and heritage	−1.324	2.227	−1.095	2.010	−1.817	77.984	.000
	Entertainment, nightlife, and shopping	−1.123	−1.189	−2.158	1.798	2.672	23.609	.000
	Festivals and events	−0.377	−0.933	−1.990	1.959	1.341	21.863	.000
	Food and wine	−0.055	−1.471	−1.493	5.655	−2.636	34.196	.000
	Indigenous culture	1.227	1.542	−0.130	0.232	−2.870	16.176	.003
	Nature	2.731	2.749	2.854	−8.067	−0.267	26.432	.000
	Relaxation, health, and indulgence	0.960	−0.420	−0.245	−0.909	0.615	17.913	.001
	Sports, outdoors, and adventure	−0.775	−0.092	3.866	−3.537	0.539	68.541	.000
Gender	Female	0.212	0.249	−0.149	0.245	−0.558	8.485	.075
Age	(category number)^e	0.135	0.472	0.236	−0.003	−0.840	19.873	.001
Education	Up to Year 12/other	−0.267	−0.531	0.057	0.655	0.086	11.381^d	.180
	Diploma/trade certificate	0.424	0.186	0.117	−0.898	0.172
	University degree	−0.157	0.344	−0.174	0.244	−0.258
Income ($)	Up to 60k	−0.431	−0.595	0.414	−1.042	1.655	25.329^d	.001
	60k-120k	0.193	0.273	−0.843	0.698	−0.321
	120k and up	0.238	0.322	0.430	0.344	−1.334
Motivation factor	Nature	0.599	0.530	1.295	−0.933	−1.490	25.013	.000
	Social and self-enhancement	−0.467	−0.073	−0.436	0.509	0.468	14.520	.006
	New experiences and knowledge	−0.384	0.564	0.005	−0.291	0.106	21.939	.000
	Health and exercise	−0.347	0.035	−0.132	0.253	0.190	6.723	.150
	Relaxation and escape	0.617	0.059	−0.331	−0.186	−0.159	15.319	.004
	Adventure and excitement	−0.141	−0.421	1.147	−1.648	1.063	46.725	.000
	Family, friends, and fun	−0.369	−0.356	−0.209	0.884	0.050	9.655	.047

Binary logistic models for each separate class, predicting interest in choosing an experience conditional on its type (intercepts constrained to be equal across classes).

Multinomial logistic model predicting class membership conditional on past visits, demographics, and motivation.

Wald statistic testing significance of parameters jointly across all classes; Wald statistic for test of parameter differences across classes: Wald = 741.693, p < .001.

The multiple rows for these variables represent the categories of a single multinomial variable; hence, there is only a single Wald statistic.

For parsimony, age was entered as a numerical predictor using category number as a proxy for age, lowest category number is youngest group as in Table 1.

The five latent classes range in size from 18% to 25% of the sample. The characterization of each of the classes is summarized in Table 8. This table is based on the latent class modeling findings (Table 7) and the sample descriptive statistics for the relevant latent class descriptors (Table 9). The first, and largest, class (25%) displays relatively high levels of interest in nature and relaxation experiences while showing relatively little interest in arts and sport. Their past vacations consist of mostly nature and indigenous experiences while they have relatively little familiarity with arts and entertainment types of vacations. The class has a balanced age and income profile in terms of the breadth of age and income categories represented, but includes mostly females. They score high on motivation factors nature and relaxation. We label this class “nature-loving relaxation seekers.”

Table 8.

Latent Class Regression Interpretation.

Class		Future Interest		Past Experience		Gender	Age	Income	Motivation Factor
No.	Size %	High on	Low on	High on	Low on	Gender	Age	Income	Positive	Negative
1	25.1	Nature/Relaxation, health, and indulgence	Arts, culture, history, and heritage/Sports, outdoors, and adventure	Nature/Indigenous culture	Arts, culture, history and heritage/Entertainment, nightlife, and shopping	Female	Balanced	Higher	Nature/Relaxation and escape	Social and self-enhancement
2	20.2	Arts, culture, history, and heritage/Nature	Sports, outdoors, and adventure	Arts, culture, history, and heritage/Indigenous culture/Nature	Entertainment, nightlife, and shopping/Food and wine	Female	Oldest	Higher	Nature/New experiences and knowledge	Adventure and excitement/Family, friends and fun
3	18.6	Nature/Sports, outdoors, and adventure	Indigenous culture	Nature/Sports, outdoors, and adventure	Entertainment, nightlife, and shopping/Festivals and events/Food and wine	Male	Older	Lowest and highest	Nature/Adventure and excitement	Social and self-enhancement/Relaxation and escape
4	18.1	Food and wine/Arts, culture, history and heritage/Entertainment, nightlife, and shopping/Relaxation, health, and indulgence	Nature/Indigenous culture	Food and wine/Arts, culture, history, and heritage/Entertainment, nightlife, and shopping/Festivals and events	Nature/Sports, outdoors, and adventure	Female	Balanced	Higher	Social and self-enhancement/Family, friends, and fun	Nature/Adventure and excitement
5	18.0	Entertainment, nightlife, and shopping	Indigenous culture	Entertainment, nightlife, and shopping/Festivals and events	Arts, culture, history, and heritage/Food and wine/Indigenous culture	Male	Youngest	Lowest	Social and self-enhancement/Adventure and excitement	Nature

Table 9.

Sample Descriptors by Class.

	Class 1	Class 2	Class 3	Class 4	Class 5
Gender^a
Male	33.1	33.7	50.4	36.6	54.9
Female	66.2	66.3	49.4	63.4	45.0
Age (years)^a
18-24	11.7	9.0	38.4	10.1	12.6
25-34	13.2	10.3	24.8	14.5	25.0
35-44	40.0	26.7	21.8	29.4	28.7
45-54	25.2	27.3	12.0	23.4	21.3
55-64	6.5	20.7	1.3	16.8	12.5
65+	2.6	6.0	1.7	5.9	0.0
Income ($)^a
Up to 60k	22.8	17.8	21.1	17.9	36.9
60k-120k	40.0	37.3	27.4	38.9	33.1
120k and up	15.2	17.4	26.1	18.8	10.4
Motivation factor^b
Nature	0.29	0.41	0.56	−0.49	−0.60
Social and self-enhancement	−0.23	−0.09	−0.25	0.26	0.35
New experiences and knowledge	−0.25	0.46	−0.09	0.01	−0.17
Health and exercise	−0.07	0.08	0.13	−0.05	0.12
Relaxation and escape	0.33	−0.08	−0.23	−0.02	−0.09
Adventure and excitement	−0.23	−0.34	0.64	−0.53	0.68
Family, friends, and fun	0.01	−0.19	0.03	0.18	−0.03

Column percentages.

Mean motivation scores, a higher score means the class scored higher on that motivation factor.

Class 2 comprises 20% of the sample and displays an interest in arts- and nature-related vacation activities, and a relatively low interest in sports. Their past experiences center around arts and nature and they tend to have had more indigenous vacation experiences than the other classes but little evidence of past vacations in terms of entertainment and food experiences. The class contains a relatively large number of elderly people and has relatively more females; they are equally spread across the income groups. On the motivation factors, they score high on nature and seeking new experiences. We label this class “elderly arts and nature lovers.”

Class 3 makes up 19% of the sample and has the greatest interest in sport and nature, with little interest in arts and indigenous experiences. Sport and nature are also their most prominent past vacation activities with little experience in entertainment, events, and food-related activities. The group has a relatively high representation of elderly people and males but is balanced across income groups. Their motivation factor scores are highest on nature, and adventure and excitement. We call them “mature male sports and nature lovers.”

Class 4 is interested especially in food, arts, entertainment, and relaxation although none of these interests are particularly strong. They comprise 18% of the sample, have little to no past experience with nature and sport types of vacations and, instead, in the past, relatively often have engaged with art, entertainment, and food experiences. The class consists of a greater number of females and individuals who are more from the higher income groups. They score high on the motivation factors of social and self enhancement and on family and friends. We call this class “females interested in food and entertainment experiences.”

Finally, class 5 comprises 18% of the sample, is most keen on entertainment, and least interested of all in indigenous experiences. Their previous holiday experience centers on entertainment and events and shows a low presence of arts, food, and indigenous experiences. They have the highest proportion of younger people and of males and are overrepresented in the lowest income group. Their motivation scores are highest on social and self-enhancement and on adventure and excitement. We call them “young male entertainment seekers.”

Finally, we present descriptive class information for the interest respondents expressed in receiving further information about one particular experience, if they had the opportunity to travel. As shown in Table 10, the results are consistent with the earlier findings and reveal that class 1 is most interested in nature and relaxation information; class 2 is interested predominantly in arts and to a lesser extent in nature information; class 3 seeks nature- and sport-related information; class 4 opts for information about arts, entertainment, and relaxation; and class 5 is most interested in entertainment related information.

Table 10.

Information Interest for the Five Latent Classes (Column Percentages).

Information interest	Class 1	Class 2	Class 3	Class 4	Class 5
Arts, culture, history, and heritage	8.4	55.1	11.2	29.3	9.6
Entertainment nightlife, and shopping	10.3	3.6	4.9	19.8	34.4
Festivals and events	8.7	4.6	5.5	6.6	12.6
Food and wine	2.7	0.7	2.6	11.3	0.1
Indigenous culture	2.9	2.8	1.6	3.2	1.2
Nature	27.9	22.3	37.0	5.3	11.2
Relaxation, health, and indulgence	33.2	7.0	11.4	21.1	11.0
Sports, outdoors, and adventure	5.5	3.9	23.6	3.3	18.2

Discussion and Conclusion

While many studies have dealt with destination choice, only very few have addressed the issue of how tourists choose vacation experience/activity types. The present study explored how tourists’ experience choices depend on their past vacation experience choices, travel motives, and demographic characteristics.

Overall, the study findings suggest a strong positive relationship between past experience choices and future experience preferences that supports hypothesis 1. To analyze if homogenous groups of respondents with similar interests could be defined, a latent class regression analysis was conducted. Consistent with hypothesis 2, the analysis revealed that different experience preference groups can be distinguished. The five groups identified differ in their preferred future experience and in how the preference for future experiences relates to the respondents’ past experience choices as well as to their travel motives and selected demographics. The latter two aspects support hypothesis 3. The interest in future experiences also strongly carries through in expressed interest in receiving information about a particular holiday experience option.

Segments varied in size between 18% and 25% and were interpreted based on the prominence of choice patterns and on the respondents’ scores on the travel motivation scale and the demographic characteristics that were significant in separating segments. The largest segment was characterized as “nature loving relaxation seekers.” The other classes were labeled as “elderly arts and nature lovers,” “mature male sports and nature lovers,” “females interested in food and entertainment experiences,” and “young male entertainment seekers” (the smallest segment containing 18% of respondents).

The most striking finding of our study was the strong association between past travel and future intent. There are different possible explanations for this association. While tourists seek variety, they may not aspire to variety in experience type. Their variety seeking may focus on finding new destinations, but apparently they still have a strong desire to enjoy the same type of experience and activities as on previous trips.

The type of experience(s) sought related to various demographics, in particular age and income group, and to a lesser extent to gender. Elderly respondents were more likely to have engaged in and seek future arts- and nature-based experiences while the youngest respondents showed an overwhelming interest in and engagement with entertainment and events as past experiences. The latter group was also characterized as having a relatively lower income. Higher-income respondents were more interested in food-related experiences, and also in arts-, entertainment-, and relaxation-related vacation activities. They also had the highest levels of past engagement with these experiences.

In summary, respondents to the survey did not use the presented (unexpected) vacation option as an opportunity to try out or explore new types of activities. Instead, they seemed to know what they liked and stuck with their familiar vacation type. Hence, the broad experiences that tourists seek seem to be quite stable and can be inferred from past experience choices. This means the experience segments will tend to be stable over time. Our research did not seek specifically to estimate the time decay involved in the gradual shifting of experience preferences over time, but the findings do suggest that preferences are likely to shift with age, which is collinear with time. For tourism marketers, the implication of the key finding of the stability of experience preferences is that it is not an easy task to convince people to consider, let alone choose, vacation experiences that do not match their recent historical pattern.

In terms of research limitations and future research, our study is based on a hypothetical scenario. It is of course possible that respondents may respond differently in real-life settings; however, we do feel that the findings from this research provide an insightful examination of the relationship between past vacation choice and future experience preference as a basis for exploring this link in greater detail in future research. Of course, this relationship could also be different from the one observed here when studying travel preferences in a group context where the influence of other family members and children, friends, or travel companions would be important. Hence, there is scope for future research to investigate these contextual issues. Finally, our study applied a choice task that allowed respondents to choose multiple options of their interest. This allowed for a novel application of latent class regression to “pick any” choice data. This approach holds promise for future applications.

Footnotes

Acknowledgements

Sincere thanks are extended to the various members of the Industry Reference Group who assisted at various stages of this research.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The Sustainable Tourism Cooperative Research Centre, an Australian Government initiative, funded this current research.

Author Biographies

Geoffrey I. Crouch is the Professor of Tourism Policy and Marketing in the La Trobe Business School, La Trobe University, Melbourne, Australia. His research interests include tourism marketing, tourism psychology and consumer behavior, and destination competitiveness. He is a Fellow of the International Academy for the Study of Tourism, and a Fellow of the Council of Australasian Universities in Tourism and Hospitality Education.

Twan Huybers is Associate Professor in the School of Business at the University of New South Wales–Canberra, Australia. He is an economist and his broad research interest is in decision-making analysis using choice modeling approaches. His work focuses on the application of choice experiments to a range of decision contexts including vacation choices, performance-enhancing drug use choices and student evaluations of university teaching.

Harmen Oppewal is Professor of Marketing in the Monash Business School, Monash University, Melbourne, Australia. His research centers on modeling and understanding consumer decision-making behavior in marketing, retailing, and tourism. Topic areas include destination and channel choice, assortment perceptions, place attachment, information search and preference formation, and experiential retail. He is a Fellow of the Australian and New Zealand Academy of Marketing.

References

Agresti

(2002). Categorical Data Analysis, 2nd edition. New York: John Wiley.

Ai-ping

. (2009). “Motivations for Chinese Outbound Tourists.” The Business Review, Cambridge, 14 (1): 150-56.

Alegre

Joaquín

Juaneda

Catalina

. (2006). “Destination Loyalty: Consumers’ Economic Behavior.” Annals of Tourism Research, 33 (3): 684-706.

Artuger

Savas

Kendir

Hakan

. (2013). “Agritourist Motivations: The Case of Turkey.” International Journal of Business and Management, 8 (21): 63-69.

Assaker

Guy

Hallak

Rob

. (2013). “Moderating Effects of Tourists’ Novelty-Seeking Tendencies on Destination Image, Visitor Satisfaction, and Short- and Long-Term Revisit Intentions.” Journal of Travel Research, 52 (5): 600-13.

Atchley

R. C.

(1993). “Continuity Theory and the Evolution of Activity in Later Adulthood.” In Activity and Aging, edited by Kelly

Newbury Park, CA: Sage, pp. 5-16.

Backman

Sheila J.

Crompton

John L.

(1991). “The Usefulness of Selected Variables for Predicting Activity Loyalty.” Leisure Sciences, 13 (3): 205-20.

Baloglu

Seyhmus

McCleary

Ken W.

(1999). “A Model of Destination Image Formation.” Annals of Tourism Research, 26 (4): 868-97.

Bello

D. C.

Etzel

M. J.

(1985). “The Role of Novelty in the Pleasure Travel Experience.” Journal of Travel Research, 29 (1): 20-26.

10.

Bigné

J. Enrique

Sánchez

M. Isabel

Andreu

Luisa

. (2009). “The Role of Variety Seeking in Short and Long Run Revisit Intentions in Holiday Destinations.” International Journal of Culture, Tourism and Hospitality Research, 3 (2): 103-15.

11.

Bryan

(1977). “Leisure Value Systems and Recreational Specialization: The Case of Trout Fishermen.” Journal of Leisure Research, 9:74-187.

12.

Carù

Antonella

Cova

Bernard

. (2003). “Revisiting Consumption Experience: A More Humble But Complete View of the Concept.” Marketing Theory, 3 (2): 267-86.

13.

Chai

Andreas

. (2012). “Consumer Specialization and the Demand for Novelty: A Reconsideration of the Links and Implications for Studying Fashion Cycles in Tourism.” Jahrbücher für Nationalökonomie und Statistik, 232 (6): 678-701.

14.

Changuk Lee

(2001). “Predicting Tourist Attachment to Destinations.” Annals of Tourism Research, 28 (1): 229-32.

15.

Chen

Ganghua

Xiao

Honggen

. (2013). “Motivations of Repeat Visits: A Longitudinal Study in Xiamen, China.” Journal of Travel and Tourism Marketing, 30: 350-64.

16.

Cohen

Eli

Ben-Nun

Livnat

. (2009). “The Important Dimensions of Wine Tourism Experience from Potential Visitors’ Perception.” Tourism and Hospitality Research, 9 (1): 20-31.

17.

Correia

Antonia

Pimpao

Adriano

. (2008). “Decision-Making Processes of Portuguese Tourist Travelling to South America and Africa.” International Journal of Culture, Tourism and Hospitality Research, 2 (4): 330-73.

18.

Crawford

John L.

Godbey

Crouter

(1986). “The Stability of Leisure Preferences.” Journal of Leisure Research, 19:96-115.

19.

Crompton

John L.

Ankomah

Paul K.

(1993). “Choice Set Propositions in Destination Decisions.” Annals of Tourism Research, 20 (3): 461-76.

20.

Crouch

Geoffrey I.

(1995). “A Meta-Analysis of Tourism Demand.” Annals of Tourism Research, 22 (1): 103-18.

21.

Crouch

Geoffrey I.

(2013). “Homo Sapiens on Vacation: What Can We Learn from Darwin?” Journal of Travel Research, 52 (5): 575-90.

22.

Dobbs

David

(2013). “Restless Genes”, National Geographic, January 2013, 45-57.

23.

Eymann

Ronning

(1997). “Microeconometric Models of Tourists’ Destination Choice.” Regional Science and Urban Economics, 27:735-61.

24.

Fuchs

Galia

. (2013). “Low versus High Sensation-Seeking Tourists: A Study of Backpackers’ Experience Risk Perception.” International Journal of Tourism Research, 15 (1): 81-92.

25.

Guthrie

Cathy

Anderson

Alistair

. (2010). “Visitor Narratives: Researching and Illuminating Actual Destination Experience.” Qualitative Market Research, 13 (2): 110-29.

26.

Hong

Sung-kwon

Kim

Jae-hyun

Lee

Seokho

. (2006). “The Roles of Categorization, Affective Image and Constraints on Destination Choice: An Application of the NMNL Model.” Tourism Management, 27 (5): 750-61.

27.

Horneman

Louise

Carter

R. W.

Wei

Sherrie

Ruys

Hein

. (2002). “Profiling the Senior Traveler: An Australian Perspective.” Journal of Travel Research, 41 (1): 23-37.

28.

Hosany

Sameer

Gilbert

David

. (2010). “Measuring Tourists’ Emotional Experiences toward Hedonic Holiday Destinations.” Journal of Travel Research, 49 (4): 513-26.

29.

Huang

Songshan

Hsu

Cathy H. C.

(2009). “Effects of Travel Motivations, Past Experience, Perceived Constraints, and Attitude on Revisit Intention.” Journal of Travel Research, 48 (1): 29-44.

30.

Huybers

Twan

. (2003). “Domestic Tourism Destination Choices—A Choice Modelling Analysis.” International Journal of Tourism Research, 5 (6): 445-59.

31.

Johnson

Robert R.

Messmer

Donald J.

(1991). “The Effect of Advertising on Hierarchical Stages in Vacation Destination Choice.” Journal of Advertising Research, 31 (6): 18-24.

32.

Kim

Jae Hak

Ritchie

Brent W.

(2012). “Motivation-Based Typology: An Empirical Study of Golf Tourists.” Journal of Hospitality and Tourism Research, 36 (2): 251-80.

33.

Kim

Jong-Hyeong

Ritchie

J. R. Brent

McCormick

Bryan

. (2012). “Development of a Scale to Measure Memorable Tourism Experiences.” Journal of Travel Research, 51 (1): 12-25.

34.

King

(2002). “Destination Marketing Organizations—Connecting the Experience Rather Than Promoting the Place.” Journal of Vacation Marketing, 8 (2): 105-8.

35.

Lehto

O’Leary

J. T.

Morrison

A. M.

(2004). “The Effect of Prior Experience on Vacation Behavior.” Annals of Tourism Research, 31 (4): 801-18.

36.

Lehto

Xinran Y.

Jang

SooCheong (Shawn)

Achana

Francis T.

O’Leary

Joseph T.

(2008). “Exploring Tourism Experience Sought: A Cohort Comparison of Baby Boomers and the Silent Generation.” Journal of Vacation Marketing, 14 (3): 237-52.

37.

Diem-Trinh Thi

Pearce

Douglas G.

(2011). “Segmenting Visitors to Battlefield Sites: International Visitors to the Former Demilitarized Zone in Vietnam.” Journal of Travel and Tourism Marketing, 28:451-63.

38.

Lindberg

Frank

Hansen

Ann Heidi

Eide

Dorthe

. (2014). “A Multirelational Approach for Understanding Consumer Experiences within Tourism.” Journal of Hospitality Marketing and Management, 23 (5): 487-512.

39.

Lohmann

Danielsson

(2001). “Predicting Travel Patterns of Senior Citizens: How the Past May Provide a Key to the Future.” Journal of Vacation Marketing, 7 (4): 357-66.

40.

Mak

Athena H. N.

Wong

Kevin K. F.

Chang

Richard C. Y.

(2009). “Health or Self-Indulgence? The Motivations and Characteristics of Spa-goers.” International Journal of Tourism Research, 11:185-99.

41.

Mazursky

(1989). “Past Experience and Future Tourism Decisions.” Annals of Tourism Research, 16 (3): 333-44.

42.

Mehmetoglu

Mehmet

. (2005). “A Case-Study of Nature-Based Tourists: Specialists versus Generalists.” Journal of Vacation Marketing, 11 (4): 357-69.

43.

Neuhofer

Barbara

Buhalis

Dimitrios

Ladkin

Adele

. (2013). “A Typology of Technology-Enhanced Tourism Experiences.” International Journal of Tourism Research, 16 (4): 340-50.

44.

Oppermann

(2000). “Tourism Destination Loyalty.” Journal of Travel Research, 39 (1): 78-84.

45.

Pan

Steve

Ryan

Chris

. (2007). “Mountain Areas and Visitor Usage—Motivations and Determinants of Satisfaction: The Case of Pirongia Forest Park, New Zealand.” Journal of Sustainable Tourism, 15 (3): 288-308.

46.

Park

Kwang-Soo

Reisinger

Yvette

Kang

Hyun-Jung

. (2008). “Visitors’ Motivation for Attending the South Beach Wine and Food Festival, Miami Beach, Florida.” Journal of Travel and Tourism Marketing, 25 (2): 161-81.

47.

Pearce

Kang

(2009). “The Effects of Prior and Recent Experience on Continuing Interest in Tourist Settings.” Annals of Tourism Research, 36 (2): 172–90.

48.

Pike

Steven

. (2006). “Destination Decision Sets: A Longitudinal Comparison of Stated Destination Preferences and Actual Travel.” Journal of Vacation Marketing, 12 (4): 319-28.

49.

Pine

Joseph B.

II Gilmore

James H.

(1999). The Experience Economy: Work Is Theatre and Every Business a Stage. Boston, MA: Harvard Business School Press.

50.

Pine

Joseph B.

II Gilmore

James H.

(2012). “The Experience Economy: Past, Present and Future.” In Handbook on the Experience Economy, edited by Sundbo

Jon

Sørenson

Flemming

. Northampton, MA: Edward Elgar, pp. 21-44.

51.

Sands

Sean

Oppewal

Harmen

Beverland

Michael B.

(2009). “The Effects of In-Store Themed Events on Consumer Store Choice Decisions.” Journal of Retailing and Consumer Services, 16:386-95.

52.

Savitz

J. B.

Ramear

R. S.

(2004). “Genetic Variants Implicated in Personality: A Review of the More Promising Candidates.” American Journal of Medical Genetics (Neuropsychiatry Genetics), 131B (1): 20-32.

53.

Schinka

J. A.

Letsch

E. A.

Crawford

F. C.

(2002). “DRD4 and Novelty Seeking: Results of Meta Analyses.” American Journal of Medical Genetics (Neuropsychiatry Genetics), 114 (6): 643-48.

54.

Seddighi

H. R.

Theocharous

A. L.

(2002). “A Model of Tourism Destination Choice: A Theoretical and Empirical Analysis.” Tourism Management, 23 (5): 475-87.

55.

Sfandla

Chouki

Björk

Peter

. (2013). “Tourism Experience Network: Co-creation of Experiences in Interactive Processes.” International Journal of Tourism Research, 15 (5): 495-506.

56.

Sohn

Eunmi

Yuan

Jingxue

. (2013). “Who Are the Culinary Tourists? An Observation at a Food and Wine Festival.” International Journal of Culture, Tourism and Hospitality Research, 7 (2): 118-31.

57.

Sönmez

S. F.

Graefe

A. R.

(1998). “Determining Future Travel Behavior from Past Travel Experience and Perceptions of Risk and Safety.” Journal of Travel Research, 37 (2): 171-77.

58.

Tourism Australia. (2007). Australian Experiences Toolkit. Sydney, Australia.

59.

Van der Ark

L. Andries

Richards

Greg

. (2006). “Attractiveness of Cultural Activities in European Cities: A Latent Class Approach.” Tourism Management, 27: 1408-13.

60.

Van der Merwe

Petrus

Slabbert

Elmarie

Saayman

Melville

. (2011). “Travel Motivations of Tourists to Selected Marine Destinations.” International Journal of Tourism Research, 13:457-67.

61.

Vermunt

J. K.

Magidson

(2005). Technical Guide for Latent GOLD 4.0 Basic and Advanced. Belmont, MA: Statistical Innovations.

62.

Wallace

(1991). The Tribal Self. Lanham, MD: University Press of America.

63.

Weaver

Pamela A.

McCleary

Ken W.

Han

Jiho

Blosser

Phillip E.

(2009). “Indentifying Leisure Travel Market Segments Based on Preference Novelty.” Journal of Travel and Tourism Marketing, 26 (5/6): 568-84.

64.

Wedel

Kamakura

(2000). Market Segmentation: Conceptual and Methodological Foundations, 2nd edition. Boston, MA: Kluwer Academic.

65.

Woodside

Arch G.

Lysonski

Steven

. (1989). “A General Model of Traveler Destination Choice.” Journal of Travel Research, 27 (4): 8-14.

66.

Wymer

Walter W.

Jr. Self

Donald R.

Findley

Carolyn Sara

. (2010). “Sensation Seekers as a Target Market for Volunteer Tourism.” Services Marketing Quarterly, 31 (3): 348-62.