IDENTIFYING LEISURE TRAVEL MARKET SEGMENTS BASED ON PREFERENCE FOR NOVELTY

Abstract

The motivation and knowledge of travelers affects their choice of destination and selection of activities while on vacation. Understanding motivation is important for all stakeholders in the tourism industry. The search for novelty is one motivating factor when planning leisure trips and vacation activities. A foreign destination offers a form of novelty to the traveler because it can provide experiences that are not of the everyday variety. The chosen location may have specific geographical features, a different cultural background, special shopping opportunities, a different language, new foods, and unique leisure activities that can provide a novel experience. The purpose of this study is to identify segments of potential U.S. leisure travelers based on their level of novelty seeking and to investigate the relationship between travelers' levels of novelty seeking and their knowledge of a destination and preferred travel style.

Keywords:

• Novelty seeking
• leisure travel
• segmentation
• cluster analysis

INTRODUCTION

Researchers have often employed psychographic information to help define market segments and explain consumer behavior. In travel and tourism, the psychographic variable—novelty seeking—was proposed several decades ago as a useful dimension to focus on for understanding travel behavior (Cohen, 1972; Bello & Etzel, 1985). More recent studies have explored the role of novelty seeking and its impact on repeat visitation (Baloglu & Erickson, 1998; Jang & Feng, 2007). Other studies have looked at novelty as one component in identifying specific market segments (Crompton, 1979; Petrick, 2002).

The purpose of our study is to explore the use of novelty seeking as a means of segmenting the leisure travel market. Specifically, the study examines novelty seeking to determine whether distinct novelty segments exist and if they can be identified by their demographic characteristics. While other studies have related demographics to novelty‐seeking behavior, our study also explores whether novelty seekers have greater familiarity with and knowledge of foreign destinations. Previous studies have suggested that familiarity is negatively related to novelty (Jang & Feng, 2007), but it is proposed here that novelty seekers will have a greater familiarity with foreign destinations in general, reflecting their desire for more allocentric experiences. Another unique aspect of our study is the use of two international destinations as reference points. Most studies of novelty seeking have focused on a single country or activity (Bello & Etzel, 1985; Petrick, 2002; Chang & Chiang, 2006; Ariffin, 2008; Kim, Kim, Park, & Guo, 2008) Finally, novelty seekers are examined to see if their preferred type of travel differs from travelers less concerned with novelty.

NOVELTY SEEKING

Plog (2002) used the idea of psychocentric and allocentric travelers (or “dependables” and “venturers”) to examine travel patterns based on the traveler's novelty seeking, or “venturesomeness.” Venturesomeness was found to be a predictor of the amount of leisure travel and the types of activities engaged in while traveling. Allocentric travelers are more adventurous in their pursuit of novel travel experiences than psychocentric travelers.

Cohen (1972) was instrumental in defining traveler types and investigating the concept of novelty seeking. He developed four roles of tourists while incorporating novelty seeking: organized mass tourists, individual mass tourists, explorers, and drifters. Organized mass tourists show up at a destination in large numbers by taking part in group tours and allowing others to plan the trip for them. Individual mass tourists also come to a destination in large numbers, but they play a larger role in planning their own trip than the organized mass tourist. The explorer plans his/her own trip and does not travel as part of a group tour, but still relies on tourism services and infrastructure. Finally, the drifter is the most likely to travel without an itinerary, may not utilize tourist services and infrastructure, and will likely avoid popular tourist locations or go off the beaten path. Explorers and drifters are more likely to pursue novel experiences than organized and individual mass tourists. Cohen's tourist typology has been utilized in a number of other tourism studies. Snepenger (1987) used the typology in studying Alaskan tourists, but he could not identify the drifter role, possibly due to the use of secondary data.

Basala and Klenosky (2001) used conjoint analysis to compare travelers with higher novelty‐seeking tendencies to those who preferred familiarity with regard to communication, travel companions, and accommodations. They found that those who preferred novelty seeking were more likely to visit countries with a different native language than their own and were more likely to stay in locally owned accommodations (although they did prefer modern amenities in those accommodations). A surprising outcome of the study was that although they were willing to travel alone, novelty seekers still preferred not to travel alone. The authors suggested the reason for this finding was that respondents were provided with a hypothetical travel scenario in which the unspecified destination had experienced recent instability and terrorism so respondents may have felt less comfortable traveling alone in that scenario.

Lepp and Gibson (2003) used Cohen's typology to see if novelty or familiarity affected the way young travelers viewed risk. They found support for Cohen's typology and showed that there was a negative relationship between novelty seeking and risk. High desire for novelty could be correlated with a low perception of risk in international travel and vice versa. Based on Cohen's typology, explorers and drifters would perceive less risk in international travel than the organized or individual mass tourists because explorers and drifters have a greater desire for novel experiences.

A study of British travelers to Turkey identified five tourist types: fuzzy tourists, active tourists, recreational type, escape, and relax‐quiet tourists (Andreu, Kozak, Avci, & Cifter, 2005). Although the authors did not specifically use novelty seeking as a segmentation base, some of the items were related to novelty seeking such as seeking adventure, sports participation, getting away from home, and increasing knowledge of new places.

Laing and Crouch (2005) found novelty to be a motivating force behind tourists who sought travel adventures on the “frontier.” In a series of in‐depth interviews, the authors explored the motivations of travelers who seek unique and rarely experienced forms of tourism such as space tourism or visits to remote or dangerous locations. Novelty was high on the list of motivations for pursuing these extreme experiences. These tourism activities are often seen as prestigious or elite, and may result in increased public attention because of their uniqueness and can involve significant costs. While not everyone can participate in these activities, the authors contend that these “outliers” can be said to be exploring the next frontiers of travel.

Chang, Wall, and Chu (2006) identified novelty seeking as a significant characteristic of tourists who visited aboriginal sites in Taiwan. They found that novelty seekers were more concerned with thrill seeking, change from routine, relief from boredom, and experiencing surprises than other types of market segments.

Tse and Crotts (2004) looked at the role of novelty seeking in the culinary exploration of visitors to Hong Kong. They found that foreign visitors come to the destination with different levels of tolerance for risk (uncertainty avoidance) and different desires for novelty in culinary exploration. Destinations should recognize those differences and provide choices that appeal to all levels of visitors. In another study, novelty seeking, referred to as “lifestyle,” was found not to have a significant influence on travel risk, anxiety, or the intent to visit a foreign destination (Reisinger & Mavondo, 2005). This finding is in opposition to most of the previously mentioned studies and emphasizes the fact that novelty seeking, despite the amount of attention it has received, continues to deserve exploration.

A 21‐item instrument for measuring novelty seeking was developed by Lee and Crompton (1992). Originally, they proposed six dimensions, but ended up with four: “thrill”, “change from routine,” “boredom alleviation,” and “surprise”. Thrill incorporates an experience where excitement is the essential element. Change from routine provides travelers with something new and different than their ordinary experiences. Boredom alleviation involves stimulating activities that satisfy an individual's need for something out of the ordinary. Surprise relates to the unexpected difference between perception and reality. Items relating to the dimensions were measured on a 5‐point scale; the higher a person's response, the higher the level of novelty seeking and the lower the response, the more familiarity is desired. Validity was determined by expert examination and by comparing their results to those of other novelty‐seeking scales. The instrument had a high internal consistency (Cronbach's alpha ranged between .89 and .91 across four independent samples). This scale has served as the base for other studies on novelty seeking which found similar dimensions when subjected to factor analysis (Jeong & Park, 1997; Chang et al. 2006). Because of the strength of this scale and its representativeness of novelty seeking, it was used for the current study as well.

Based on the previous discussion of novelty seeking, it appears possible to identify segments of potential travelers based on their desire for novelty during leisure travel. The following hypotheses were developed:

H1: Specific groups of potential vacation travelers can be segmented using the four dimensions of novelty seeking identified by Lee and Crompton (1992) .

H2: Demographic characteristics such as age, gender, marital status, number of children, education level, and annual household income level can be identified within specific novelty‐seeking segments of potential leisure travelers.

DESTINATION KNOWLEDGE/FAMILIARITY

The concept of familiarity is often discussed along with novelty seeking. A person's knowledge of a destination can be influenced by their familiarity with the destination and their prior experience (Cho, 2001). Cho found a relationship between familiarity/expertise and past experience and used the concepts together to measure the prior knowledge of travelers. Plog's (2002) psychocentric‐allocentric model can be looked at as a continuum with high novelty seeking on one end (allocentric) and high familiarity on the other end (psychocentric). The psychocentric travelers seek more familiar forms of travel and destinations. Cohen (1972) discussed the idea of an environmental bubble that provides a realm of familiarity to travelers. His explorers provide a good example of the familiarity bubble: they look for novel experiences, but return to familiar amenities (such as accommodations, food, etc.) while on vacation. Familiarity is also inversely related to lower perceived risk in revisiting a destination (Moutinho, 1987).

Familiarity is generally considered to be the opposite of novelty (Jang & Feng, 2007). Therefore, a novelty seeker who is familiar with a destination should be less likely to visit. However, familiarity can take several forms. Familiarity can be actual physical experience with a destination gained through visitation or as exposure derived through some form of communication such as advertising or information provided by friends or family members (Gursoy & McCleary, 2004a and 2004b). A large number of studies have examined the connection between the messages sent to consumers, the various media used to send those messages, and the impact that the message has on the consumer (Fodness & Murray, 1998, 1999; Chen & Gursoy, 2000; Desai & Hoyer, 2000; Gursoy, 2003). The desired impact of these marketing messages is to increase the potential traveler's level of knowledge of a destination, including the features and activities at that destination, the availability of accommodations, and/or the means of getting to that destination.

Past experience can be a determinant in revisiting the same destination or going to other destinations (Sonmez & Graefe, 1998). Lehto, O'Leary, and Morrison (2004) found that the amount of past experience influences the choice of destinations or activities while on vacation. March and Woodside (2005) found that the number of activities engaged in during a vacation exceeds the number that were planned. They also recognize the role that product [prior] experience plays in the selection and number of activities.

Crompton (1992) discussed the development of choice sets in the destination selection process. An awareness set is first developed of all potential vacation destinations. This initial list is evaluated and reduced to a late consideration (evoked) set, and eventually a destination is selected. In the process of narrowing down the potential destination choices, both internal and external forces help to shape the decision. Internal forces are related to a person's motives for travel, which could include the desire for novelty seeking. External forces can be described as previous experience, prior knowledge of the destination, and information obtained from a variety of sources about the destination (Gursoy & McCleary, 2004a).

Japan and Australia were used in the current study to provide a point of comparison between two international destinations which have both common and contrasting characteristics, although other destinations could have been used. The rationale behind assessing knowledge by using two distant potential tourist destinations was that the knowledge of two such distant places could be used as a proxy for general travel knowledge. Using specific destinations gave respondents a frame of reference in which to assess general knowledge about distant destinations and also to provide a self‐measure of their travel knowledge relative to the general population. In the case of Japan and Australia, both are modern countries located a substantial distance from the United States, would require similar means of travel to reach, and offer unique physical and cultural opportunities. The two destinations also have significant differences. For example, Japan and Australia differ from each other in terms of race, native language, prevailing religion, and cultural background. To measure how “knowledgeable” potential vacation travelers are with the reference destination or the amount of “knowledge” they have about the potential vacation destination, the questions used in our study were modified from the instrument used by Cho (2001). The questions used to determine the knowledge of the destination were: “How familiar are you with [Japan or Australia] as a vacation destination?”; “How interested are you in [Japan or Australia] as a vacation destination?”; “How much do you know about [Japan or Australia] as a vacation destination?”; and “How knowledgeable are you about vacation travel in [Japan or Australia] relative to the rest of the U.S. population?” Each of the above questions were stated twice; once using “Japan as the country of interest” and the second time using “Australia as the country of interest.” In addition, two questions were asked to assess travelers' past experience with Japan and Australia as vacation destinations.

The following hypotheses were added to explore the relationships between novelty seeking, prior experience with, and knowledge of a destination:

H3: Travelers with a high level of novelty seeking will differ from those with a low level of novelty seeking in regard to how knowledgeable they are about a foreign destination.

H4: Travelers who have been to one international destination will have a higher level of knowledge regarding travel to another international destination than those who have not been to an international destination.

H5: An interaction exists between foreign destination visitation and novelty seeking with regard to how knowledgeable individuals are with another international destination.

TRAVEL STYLE

A person's decision on how to travel may vary based on the location of the destination they select, the amount of time they have to spend, their budget for travel, the distance to the destination from their place of residence, and geographical features. Taylor (1998) identified three types of travel: planned travel (those who use package travel options), independent travel (those who arrange travel for themselves), and reluctant travel (those who prefer not to travel). Plog (2004) also classified travel style and identified the styles as independent travel, inclusive package, escorted travel, and partial package. Chang et al. (2006) noted that although previous research found a difference in preferred travel style between novelty seekers and other psychographic segments, their research on travel to aboriginal sites found no difference between independent and group travelers on novelty‐seeking dimensions.

Basala and Klenosky (2001) compared three groups of travelers (novelty seekers, average travelers, and familiarity seekers) in regard to the type of accommodations they preferred, the number and type of travel companions, and the language of the host community. However, although the categories of group tours and traveling alone were included in their concept of travel companions, there was no mixture of the two such as in Plog's (2004) partial package.

For the current study, questions from the Hong Kong International Visitor Survey (McCleary, Weaver, & Hsu, 2006) were adopted. The four choices were: “a fully packaged tour,” “a partially packaged tour with transport and accommodation only,” “non‐packaged/independent travel,” and “I would never go.” Technically, “I would never go” is not a travel style, but it is included here as a choice which can be used to help identify respondents who may have a low preference for international travel.

The following hypothesis regarding travel style and novelty seeking was considered:

H6: Travelers who are classified as having different levels of novelty seeking will prefer different travel styles when planning a vacation to a foreign destination.

RESEARCH METHOD

A pretest of the self‐administered questionnaire was distributed to a convenience sample of 249 students in one large undergraduate class at a southeastern university. This sample selection was deemed appropriate for the pretest because its main purpose was to evaluate items for comprehension and clarity. The 21 Novelty‐Seeking items were not included in the pretest due to the previous purification of the scale by Lee and Crompton (1992). In addition to the demographic information, four familiarity questions were asked of students for both Australia and Japan. The coefficient alphas for the four Familiarity items for both Japan and Australia were .83, verifying internal consistency.

The final questionnaire was sent to a random sample of 2000 Virginia residents 18 years of age or older. The mailing list was obtained from ReferenceUSA database (http://www.referenceusa.com), which provides residential information. The number of respondents selected from each county and city was determined by a stratified sampling method; a random sample was utilized to select the assigned respondents from each city and county. Only one mailing was necessary to generate the desired minimum number of responses. The sample size utilized for this study was 358 for a response rate of 17.9%. An examination of early and late respondents was undertaken to evaluate non‐response bias as suggested by Armstrong and Overton (1977). Although this method to test non‐response bias has its drawbacks (Blair & Zinkhan, 2006), it was felt that the considerable time lag between first and final respondents provided a reasonable assessment that late respondents would approximate non‐respondents, particularly given the sample size. Chi‐square and t tests for the various scales indicated no statistically significant differences between the respondents at the .01 level and they did not differ from each other regarding visiting Australia/Japan nor did they differ on their demographic makeup.

RESULTS

General Information

Of those who responded to individual questions, 62% were male, 69% were between 31 and 60 years of age (with 30% between 31 and 45 and 39% between 46 and 60), and 78% were married. Fifty‐ nine percent of the respondents did not have children living at home; however, of those with children living at home, 76% had one or two children currently at home. Over half of the sample had at least a bachelor's degree (63%) and an annual income of at least $80,000 (52%). Please refer to Table 1 for a complete listing of common demographics.

Table 1. Demographic Profile

Category		%
Gender	Female	38.4
(n  =  354)	Male	61.6
Age	17–30	7.9
(n  =  343)	31–45	29.4
	46–60	39.1
	61–75	17.8
	⩾ 76	5.8
Marital Status	Never Married	9.2
(n  =  348)	Now Married	78.2
	Divorced	4.9
	Widowed	5.2
	Living with Partner	2.6
Number of Children Presently	1	35.6
Living in House	2	41.1
(n  =  146)	3	19.2
	4	2.7
	5	1.4
Education Level	< High School	3.1
(n  =  351)	High School	11.4
	Some College	14.8
	2 yr./Vocational	8.0
	Bachelors degree	31.1
	Graduate degree	31.6
Annual Household Income Level	Less than $40,000	14.7
(n  =  333)	$40,000–$59,999	18.6
	$60,000–$79,999	15.6
	$80,000–$99,999	14.7
	$100,000–$119,999	10.8
	$120,000–$139,999	7.5
	$140,000–$159,999	5.1
	$160,000 or more	12.9

Twenty‐eight (7.8%) respondents had visited Australia and 53 (14.8%) had visited Japan. Four questions were asked of respondents to determine their familiarity with Japan and Australia (a scale of 1 thru 5 with 1  =  not at all, 2  =  a little, 3  =  somewhat, 4  =  quite, 5  =  extremely). Approximately, 33% of the respondents were not at all familiar with Australia as a vacation destination, while 53% were not familiar with Japan. While 38.4% were quite or extremely interested in Australia as a vacation destination, only 14.1% felt likewise about Japan. Twenty‐six percent felt they knew nothing about Australia as a vacation destination, and 52.5% knew nothing about Japan as a vacation destination. With regard to knowledge about each of the countries when it comes to vacation travel; 32.3% felt they had no knowledge of Australia, while 53.5% felt they had no knowledge of Japan (Table 2).

Table 2. Frequency of General Interest/Visitation Questions

Australia	Frequency	Yes	No
Have you ever visited Australia? (n  =  358)		28	330
	%	7.8	92.2
		Not at all	A little	Somewhat	Quite	Extremely
How familiar are you with Australia as a vacation destination? (n  =  358)	Frequency	118	57	136	32	15
	%	33.0	15.9	38.0	8.9	4.2
How interested are you in Australia as a vacation destination? (n  =  357)	Frequency	77	42	101	80	57
	%	21.6	11.8	28.3	22.4	16.0
How much do you know about Australia as a vacation destination? (n  =  357)	Frequency	93	91	131	32	10
	%	26.1	25.5	36.7	9.0	2.8
How knowledgeable are you about vacation travel in Australia relative to the rest of the U.S. population? (n  =  356)	Frequency	115	96	92	33	20
	%	32.3	27.0	25.8	9.3	5.6
Japan	Frequency	Yes	No
Have you ever visited Japan? (n  =  358)		53	305
	%	14.8	85.2
		Not at all	A little	Somewhat	Quite	Extremely
How familiar are you with Japan as a vacation destination? (n  =  356)	Frequency	190	73	73	12	8
	%	53.4	20.5	20.5	3.4	2.2
How interested are you in Japan as a vacation destination? (n  =  356)	Frequency	163	62	81	32	18
	%	45.8	17.4	22.8	9.0	5.1
How much do you know about Japan as a vacation destination (n  =  356)	Frequency	187	91	50	22	6
	%	52.5	25.6	14.0	6.2	1.7
How knowledgeable are you about vacation travel in Japan relative to the rest of the U.S. population? (n  =  355)	Frequency	190	74	58	16	17
	%	53.5	20.8	16.3	4.5	4.8

Factor Verification

Exploratory factor analysis was first performed on the novelty‐seeking scale to verify the four factors found by Lee and Crompton (1992) and to generate factor scores for use in a subsequent cluster analysis. Factor analysis verified the original four factors identified by Lee and Crompton. All four factors had eigenvalues larger than 1.00, the pre‐specified cutoff value. Together, the four factors explained 74.4% of the total variance. One concern emerged, however. The question “I seek adventure on my vacation” had a factor loading of .517 on the factor “Change from Routine” and .607 on the factor “Thrill.” The correlations showed that this question was strongly correlated with questions in both factors. This item was deleted and the analysis was rerun using the same technique.

The second factor analysis identified the same four factors (see Table 3). The reliability analysis showed all four factors exceeded the guidelines for Cronbach's alpha (Hair, Black, Babin, Anderson, & Tatham, 2005). Thrill consisted of six items (the deleted question originally loaded highest on this factor) and had a reliability score of .937; Change from Routine consisted of eight items with a reliability score of .926; Boredom Alleviation and Surprise had three items each and reliability scores of .899 and .825, respectively. Overall, the scale of 20 items had a Cronbach's alpha of .917. The Kaiser‐Meyer‐Olkin measure of sampling adequacy for the scale was .893, indicating that the data were appropriate for factor analysis; this is well above the .50 minimum and falls into the “meritorious” level (Hair et al., 2005). Overall, the four factors explained 75.0% of the total variance. With the original four factors identified by Lee and Crompton (1992) now verified the factor scores were saved for further use.

Table 3. actor Analysis of Novelty‐Seeking Items

	Factor Loading	Eigenvalue	Variance Explained	Cronbach Alpha
Factor 1: Thrill		8.26	41.3%	0.937
I sometimes like to do things on vacation that are a little frightening.	0.84
I enjoy doing “daring” activities while on vacation.	0.86
Sometimes it is fun to be a little scared on vacation.	0.85
I enjoy experiencing a sense of danger on a vacation trip.	0.87
I would like to be on a raft in the middle of a wild river at the time of the spring floodwaters.	0.78
I enjoy activities that offer thrills.	0.83
Factor 2: Change from Routine		2.9	14.5%	0.926
I like to find myself at destinations where I can explore new things.	0.78
On my vacation, I want to experience new and different things.	0.8
On vacation I want to experience customs and cultures different from those in my own environment.	0.81
I enjoy the change of environment which allows me to experience something new on vacation.	0.88
My ideal vacation involves looking at things I have not seen before.	0.77
I want there to be a sense of discovery involved as part of my vacation.	0.84
I like to travel to adventurous places.	0.72
I feel powerful urge to explore the unknown on vacation.	0.63
Factor 3: Boredom Alleviation		2.23	11.2%	0.899
I want to travel to relieve boredom.	0.88
I have to go on vacation from time to time to avoid getting into a rut.	0.91
I like to travel because the same routine work bores me.	0.89
Factor 4: Surprise		1.60	8.0%	0.825
I don't like to plan a vacation trip in detail because it takes away some of the unexpectedness.	0.83
I like vacations that are unpredictable.	0.82
I would like to take off on a trip with no pre‐planned routes in my mind.	0.81
Total Variance Explained			75.0%

Cluster Analysis of Novelty‐Seeking Factors

Mo, Howard, and Havitz (1993) suggested the use of cluster analysis to identify various combinations of novelty and familiarity seeking. Following this line of thought, cluster analysis was performed on the factor scores of the aforementioned four factors to identify specific groupings of respondents. A hierarchical cluster method with a randomly generated sample of respondents was first performed to aid in identifying the ideal number of clusters to analyze. Thirty‐six cases (roughly 10%) were randomly selected by SPSS for use in the initial cluster analysis; a smaller sample is acceptable to identify large clusters of cases (Hair et al., 2005).

After several iterations of cluster analysis, the furthest neighbor, or complete linkage method with the squared Euclidean distance measure was used in this research. Hair et al. (2005) suggests that this method often generates the most compact clustering solutions. This time the agglomeration schedule suggested three clusters to five clusters as possibilities. With the goal of identifying a broad means of clustering groups based on novelty seeking scores and a perusal of the various solutions, a three‐cluster group was selected. Each respondent was assigned to one of three clusters and the seeds were generated for the next step.

A non‐hierarchical cluster (or k‐means) method was performed on the entire sample (n  =  343) with three clusters specified. The final cluster assignments for cluster one, two, and three contained 157, 142, and 44 members, respectively. The small number in the third cluster is not a major concern; the dendrograms in the earlier hierarchical cluster analysis showed the presence of a small cluster. An examination of the distances between clusters showed that clusters one and two are the most alike with a standardized distance score of 1.801. Cluster three is very different from cluster one (2.525) and cluster two (2.412). This third cluster could represent a unique market segment not interested in novelty. Analysis of Variance showed statistical significance at the .05 level for all four factors across the three clusters.

The final cluster results show that cluster one is most distinct with high values across all four factors of novelty seeking. By examining the descriptive statistics of the 20 novelty‐seeking items for these groups (Table 4), it is seen that cluster one scored all of the items on the novelty‐seeking scale highly, but the Change from Routine items and Thrill items received the highest ratings. Cluster one was titled “Thrill Seekers.” Cluster two was titled “Change Seekers,” and rated the Change from Routine factor highest. Cluster three, titled “Homebodies,” did not score any factors highly, but Boredom Alleviation was the highest scoring factor for this group. Only one item—“my ideal vacation involves looking at things I have not seen before”—had a mean higher than 3.0 (on a 5‐point Likert scale) for “Homebodies.”

Table 4. luster Scores on Novelty‐Seeking Factors

Novelty‐Seeking Items	Thrill Seeker		Change Seeker		Homebodies
	N  =  157		N  =  142		N  =  44
	Mean	SD	Mean	SD	Mean	SD
Thrill
I sometimes like to do things on vacation that are a little frightening.	3.67	0.737	1.69	0.947	1.39	0.722
I enjoy doing “daring” activities while on vacation.	3.71	0.708	1.75	0.879	1.34	0.776
Sometimes it is fun to be a little scared on vacation.	3.38	0.909	1.53	0.760	1.27	0.624
I enjoy experiencing a sense of danger on a vacation trip.	3.01	0.980	1.21	0.443	1.18	0.495
I would like to be on a raft in the middle of a wild river at the time of the spring floodwaters.	2.70	1.293	1.17	0.394	1.18	0.540
I enjoy activities that offer thrills.	3.78	0.745	1.83	1.045	1.41	0.757
Change From Routine
I like to find myself at destinations where I can explore new things.	4.38	0.664	4.15	0.891	1.86	1.069
I want to experience new and different things on my vacation.	4.48	0.584	4.29	0.795	1.98	1.045
On vacation, I want to experience customs and cultures different from those in my own environment.	4.10	0.833	4.31	0.835	2.18	1.040
On vacation, I enjoy the change of environment which allows me to experience something new.	4.37	0.653	4.49	0.555	2.59	1.127
My ideal vacation involves looking at things I have not seen before.	4.43	0.643	4.54	0.591	3.18	1.206
I want there to be a sense of discovery involved as part of my vacation.	4.33	0.674	4.27	0.773	2.25	1.037
I like to travel to adventurous places.	4.22	0.663	3.73	0.981	2.05	1.056
I feel a powerful urge to explore the unknown on vacation.	3.82	0.953	3.19	1.160	1.66	0.834
Boredom Alleviation
I want to travel to relieve boredom.	3.14	1.118	2.50	1.330	2.48	1.355
I have to go on vacation from time to time to avoid getting into a rut.	3.31	1.114	2.69	1.430	2.84	1.478
I like to travel because the same routine work bores me.	3.15	1.148	2.52	1.362	2.61	1.262
Surprise
I don't like to plan a vacation trip in detail because it takes away some of the unexpectedness.	3.28	1.126	2.70	1.220	2.32	1.308
I like vacations that are unpredictable.	3.15	1.012	2.56	1.176	1.77	1.031
I would like to take off on a trip with no pre‐planned routes in my mind.	3.15	1.229	2.42	1.395	2.05	1.293
Note. *5‐point Likert scale: 1  =  strongly disagree; 5  =  strongly agree.

Hypotheses Testing

H1: Specific groups of potential vacation travelers can be segmented using the four dimensions of novelty seeking identified by Lee and Crompton (1992) .

Results of the cluster analysis provide strong support for Hypothesis 1.

H2: Common demographic characteristics such as age, gender, marital status, number of children, education level, and annual household income level can be identified within specific novelty seeking‐segments of potential leisure travelers.

To further identify features of these three novelty clusters, crosstabs were performed on demographic and descriptive questions. The Thrill Seeker and the Change Seeker Cluster are more highly educated than the Homebodies. Thrill Seekers tended to be younger than both Change Seekers and Homebodies, with Homebodies being the oldest cluster. In regard to gender, both Novelty Seekers and Homebodies were predominantly male, while Change Seekers are more evenly dispersed. Marital status, income level, and number of children at home were not statistically significant. Refer to Table 5 for the crosstab analysis of the demographic breakdown by cluster.

Table 5. ovelty‐Seeking Clusters by Demographics

Demographics	Novelty‐Seeking Clusters			Chi‐Square	Sig.
	Thrill Seeker	Change Seeker	Homebodies
Education Level
Less than high school (%)	2.0	4.3	0.0	χ ^2  =  18.611	.045
High school (%)	8.3	11.4	22.7
Some College (%)	16.6	10.7	20.5
2‐yr. or vocational degree (%)	5.8	9.3	13.6
Bachelor’s degree (%)	34.6	31.4	18.2
Graduate degree (%)	32.7	32.9	25.0
Household Income
Less than $40,000	10.8	15.7	25.0	χ ^2  =  21.081	.100
$40,000–$59,999	14.9	20.9	27.5
$60,000–$79,999	16.9	14.2	17.5
$80,000–$99,999	14.2	17.9	7.5
$100,000–$119,999	10.8	9.7	12.5
$120,000–$139,999	11.5	4.5	2.5
$140,000–$159,999	4.7	6.0	2.5
$160,000 or more	16.2	11.2	5.0
Number of Children Presently Living in the Household
1	40.6	32.3	20.0	χ ^2  =  13.971	.083
2	29.7	48.4	60.0
3	26.6	16.1	6.7
4	3.1	1.6	6.7
5	0.0	1.6	6.7
Marital Status
Never married	13.5	4.3	11.4	χ ^2  =  14.038	.081
Now married	74.8	82.6	79.5
Divorced	3.2	5.8	4.5
Widowed	3.9	6.5	2.3
Living with a partner	4.5	0.7	2.3
Gender
Female	29.3	48.6	38.6	χ ^2  =  11.723	.003
Male	70.7	51.4	61.4
Age
17–30	12.7	3.7	4.7	χ ^2  =  28.744	.000
31–45	33.1	27.9	25.6
46–60	41.4	36	30.2
61–75	11.5	23.5	30.2
76 and older	1.3	8.8	9.3

The crosstabs analysis provides partial support for Hypothesis 2.

H3: Travelers with a high level of novelty seeking will differ from those with a low level of novelty seeking in regard to how knowledgeable they are about a foreign destination.

In order to test Hypotheses 3 through 5 a Multivariate Analysis of Variance (MANOVA) was used with “Novelty Seeking Clusters” and “Visited Japan” as the two independent variables (see Tables 6 and 7). “Novelty Seeking Clusters” consisted of three levels (the three separate clusters) and “Visited Japan” consisted of two levels (yes/no). “Visit Japan” was selected as the independent variable because of the higher number of respondents who had visited Japan (52) versus Australia (28). The larger number was necessary as a requirement for running MANOVA. The four familiarity questions regarding Australia were selected as dependent variables. (Refer to Table 2 for summary information.) Despite the selection of the country that more respondents had traveled to, the cell sizes were not large and one cell in the 3 × 2 MANOVA analysis did not meet a desired level of 20 cases (only four respondents were assigned to Novelty Seeking Cluster three, the Homebodies, and had visited Japan). However, the assumption of equality of covariance matrixes was met utilizing Box's M test. The size of this sample is considered in the discussion of the Homebodies cluster.

Table 6. ANOVA Multivariate Results: Main Effects and Interaction Effect Between Japan Visitation Item and Three Novelty‐Seeking Clusters Using Wilks' Lambda

	Wilks' Lambda	F‐value	Sig.
Japan Visitation	0.955	3.910	0.004
Novelty‐Seeking Clusters	0.944	2.407	0.015
Japan Visitation by Novelty‐Seeking Clusters	0.970	1.263	0.260

Table 7. ANOVA Univariate Results: Main Effects—Japan Visitation and Novelty‐Seeking Clusters (Dependent Variable—Familiarity)

Familiarity Questions	Japan Visitation		F‐value	Sig.
	Yes	No
How familiar are you with Australia as a vacation destination?	2.720	2.207	5.025	0.026
How interested are you in Australia as a vacation destination?	3.093	2.769	1.523	0.218
How much do you know about Australia as a vacation destination?	2.720	2.163	7.253	0.007
How knowledgeable are you about vacation travel in Australia relative to the rest of the U.S. population?	2.861	2.000	15.360	0.000
*5‐point scale: 1  =  not at all; 2  =  a little; 3  =  somewhat; 4  =  quite; 5  =  extremely.
Familiarity Questions	Novelty‐Seeking Clusters			F‐value	Sig.
	Thrill Seeker	Change Seeker	Homebodies
How familiar are you with Australia as a vacation destination?	2.627	2.514	2.250	0.729	0.483
How interested are you in Australia as a vacation destination?	3.405	2.925	2.463	4.636	0.010
How much do you know about Australia as a vacation destination?	2.762	2.500	2.063	3.353	0.036
How knowledgeable are you about vacation travel in Australia relative to the rest of the U.S. population?	2.835	2.631	1.825	5.383	0.005

Reviewing the multivariate model, the interaction effect was not significant but both main effects for the two independent variables were statistically significant (Novelty Seeking Cluster assignment α < .05; Visit to Japan α < .01). Analyzing the individual dependent variables showed that with regard to Cluster assignment, three of the four dependent variables were significant (α < .05). However; “how familiar are you with Australia?” was not significant from a univariate standpoint. For the three significant variables, the summary data is highest for “Thrill seekers” and lowest for the “Homebodies.”

Because the main effect, “Cluster,” was significant, Hypothesis 3 is supported.

H4: Travelers who have been to one international destination will have a higher level of knowledge regarding travel to another international destination than those who have not been to an international destination.

Visit to Japan was statistically significant from a multivariate perspective. From a univariate perspective, three of the four familiarity variables were significant (α < .05). “How interested are you in Australia as a vacation destination” was not significant from a univariate standpoint.

Because the multivariate main effect was significant, Hypothesis 4 is supported.

H5: An interaction exists between foreign destination visitation and novelty seeking with regard to how knowledgeable individuals are with another international destination.

Because there was no statistically significant interaction effect, Hypothesis 5 was not supported.

H6: Travelers who are classified as having different levels of novelty seeking will prefer different travel styles when planning a vacation to a foreign destination.

The final analysis included two crosstabs analyses—Novelty Seeking Cluster Assignment by Preference of Travel Style for Visiting Australia and Cluster Assignment by Preference of Travel Style for Visiting Japan. The Preference of Travel Style variable included four options: Fully packaged tour, partially packaged tour with transport and accommodation, Non‐packaged independent travel, I would never go. Both chi‐square results were statistically significant at a p  =  .000. (Please refer to Table 8.)

Table 8. Preference of Travel Style for Three Novelty‐Seeking Clusters Broken Out by Japan and Australia

Travel Style Variables	Novelty‐Seeking Clusters			Chi‐Square	Sig.
	Thrill Seeker	Change Seeker	Homebodies
Preference of travel style to Australia
Fully packaged tour (%)	22.9	41.2	33.3	χ ^2  =  51.636	.000
Partially packaged tour with transport & accommodation (%)	45.8	24.3	25.6
Non‐packaged, independent travel (%)	29.4	21.3	10.3
I would never go (%)	2.0	13.2	30.8
Preference of travel style to Japan
Fully packaged tour (%)	37.7	43.5	39.5	χ ^2  =  24.894	.000
Partially packaged tour with transport & accommodation (%)	29.2	16.7	10.5
Non‐packaged, independent travel (%)	16.2	12.3	2.6
I would never go (%)	16.9	27.5	47.4

Results for both countries were statistically significant (α < .001). Thrill Seekers are the most likely to select either a non‐packaged/independent travel option or a partially packaged tour (transport and accommodations only) for either Australia or Japan. They were not likely to select “I would never go.” The Change Seekers were more evenly dispersed, but preferred the fully packaged option. Homebodies scored low on the independent travel and partially packaged options and preferred to not go at all or utilize a fully packaged tour.

Hypothesis 6 is supported.

DISCUSSION AND IMPLICATIONS

The novelty‐seeking scale developed by Lee and Crompton (1992) is a useful tool in segmenting potential travelers. The factor analysis of the scale yielded results similar to Lee and Crompton's research, as well as other studies such as Chang et al. (2006) with four factors being identified. Using the original four factors of Thrill, Change from Routine, Boredom Alleviation, and Surprise, it was determined that three specific groups of travelers could be identified based on their levels of novelty seeking. This finding is similar to the findings in Chang and Chiang's (2006) study of Japanese and American tourists in that three distinct novelty‐seeking segments were identified for Japanese tourists. However, American tourists were only classified into two groups. While Chang and Chiang found differences based on several demographic variables (gender, marital status, age), the current study found gender and education to be significantly different between clusters in our study. Chang et al. also used cluster analysis and identified three clusters based on Lee and Crompton's scale and found differences in marital status, age, education, and occupation, but did not find differences based on income or gender. These inconsistent results point out the need for caution when basing strategy decisions on demographics alone. In our study, novelty‐seeking clusters one (Thriller Seekers) and two (Change Seekers) are the most similar, but cluster one, Thrill Seekers, desires higher levels of novelty in their vacations. They rated all of the items in the novelty‐seeking scale higher than the other two clusters. Thrill Seekers tend to be almost three‐fourths male, more highly educated, and also younger than the other two clusters.

Thrill Seekers are also more likely to plan their own vacations or look for packages that offer basic travel and accommodations while leaving the choice of activities to the traveler. The significant finding that the segments identified in the cluster analysis differed on preferred travel style is in contrast to Chang et al. (2006) which found no difference in preferred travel style. Our finding does, however, support the research of Keng and Cheng (1999) who found that their segment identified as novelty seekers preferred self‐guided travel while familiarity seekers overwhelmingly preferred packaged tours. The finding also supports Cohen's (1972) tourist role typology.

Promotional campaigns to Thrill Seekers should emphasize destinations and activities that are out of the ordinary, and which may even be thought of by average travelers to be “dangerous.” Travel agencies may want to provide basic travel packages with lots of room for the traveler to select their own activities and destinations should provide plenty of information and direct purchase options for those who do not use a travel agent or select even partially packaged trips. This is not a small cluster, which indicates that a large number of potential travelers desire novel activities in their vacations or would be willing to visit unique destinations. Thrill Seekers are more likely to have experienced or desire to experience international travel.

Although familiarity might be considered the opposite of novelty (Jang & Feng, 2007), in our study the segments seeking the greatest amount of novelty also claimed the most knowledge and expressed the most interest in visiting Australia. While familiarity is gained by knowledge from both visitation and various forms of communication such as advertising, travel magazines, and word‐of‐mouth from friends and relatives (Gursoy & McCleary, 2004a and 2004b), knowledge obtained from sources other than visitation may actually encourage novelty‐seekers' visitation (Chen & Gursoy, 2000) while knowledge based on previous visitation can discourage additional visits (Sonmez & Graefe, 1998). To complicate the matter even further, there seems to be some relationship between visiting one foreign destination and knowledge of another. Our results show that there was a significant difference between those who had visited Japan and their familiarity with Australia. More research needs to be conducted on these two types of knowledge acquisition and their impact on novelty seekers.

Novelty Seeking Cluster two, the Change Seekers, want to have new and different experiences on their vacation, but not at the same level of novelty as the Thrill Seekers. They are interested in variety and the opportunity to see new places and learn while on vacation. They prefer travel packages that are more structured and which offer convenience. This group tends to have a balance of males and females. Travel agencies providing full and partial service packages to novel destinations could target this cluster. Similar to the Thrill Seekers, the Change Seekers market makes up a much larger percentage of travelers than the Homebodies. Marketers should showcase the uniqueness of a destination and promote attractions and activities that stimulate the mind while at the same time offering convenient services for the traveler.

Novelty Seeking Cluster three, the “Homebodies,” is distinctly different from the other two groups as indicated by their low scores on all aspects of the novelty‐seeking scale. Perhaps routine and familiar vacations suit this group the best. They seem to equate to Taylor's (1998) reluctant travelers. International destinations are not likely to see many of these travelers and should concentrate their efforts on attracting the other two. The low number of Homebodies who had been to Japan provided a limitation in the statistical analysis for that group, and interpretation of the Homebodies, must bear that in mind. It appears likely from their scores on the novelty‐seeking scale, however, that international travel is not the most likely choice for this group when selecting a vacation destination.

LIMITATIONS AND FUTURE STUDIES

A limitation of this study is the selection of a sample from one region of the United States. Cross‐cultural differences may exist as travelers from around the globe routinely visit destinations away from their home country. Only one form of motivation, a desire to seek novel experiences, is examined here. Other travel motivations and purposes could provide differences; for example, business travel or travel to visit family and friends.

This study also only looked at novelty seeking in regard to visiting international destinations. Certainly, many opportunities can exist for novel experiences within the traveler's home country. Future studies could examine the levels of novelty seeking in regard to activities and destinations that are closer to the home environment. Evaluating the interacting effects the respondent's travel partners has on the novelty‐seeking behavior of the traveling unit would also be of interest. Further study of the respondent's travel partners with regard to relationship to the respondent, age, and gender could also aid in the identification of leisure travel market segments. Finally, as mentioned earlier, the role of knowledge and familiarity for novelty seekers, particularly for travel to foreign destinations, should be examined more closely.

CONCLUSION

This study identified three distinct types of travelers based upon their level of novelty seeking. The three Novelty Seeking groups, Thrill Seekers, Change Seekers, and Homebodies have unique characteristics that can be used by travel and tourism providers to promote different types of destinations with different forms of travel style. The Change Seekers and Thrill Seekers make up a large portion of the market of potential travelers and they are likely to consider international travel as an option. They are both higher in their level of novelty seeking than the Homebodies and share some common characteristics. They differ, however, in regard to their desire to experience destinations and activities that offer extremely novel experiences and in their preference for travel style, two characteristics that should be kept in mind by destinations and by marketers of travel and tourism and services.