Destination brand experience and visitor behavior

Annals of Tourism Research 48 (2014) 121–139

Contents lists available at ScienceDirect

Annals of Tourism Research journal homepage: www.elsevier.com/locate/atoures

Destination brand experience and visitor behavior: Testing a scale in the tourism context Stuart J. Barnes a,⇑, Jan Mattsson b,1, Flemming Sørensen b,2 a b

University of Kent, United Kingdom Roskilde University, Denmark

a r t i c l e

i n f o

Article history: Received 27 December 2013 Revised 4 June 2014 Accepted 11 June 2014 Available online 5 July 2014 Coordinating Editor: Metin Kozak Keywords: Destination brand experience PLS path modeling Satisfaction Loyalty

a b s t r a c t Destination branding has developed considerably as a topic area in the last decade with numerous conceptualizations focusing on different aspects of the brand. However, a unified view has not yet emerged. This paper examines destination branding via a new conceptualization, destination brand experience (DBE), which provides a more holistic and unified view of the destination brand. It examines the direct and mediated role of DBE components in determining revisit intentions and word-of-mouth recommendations. The findings suggest that DBE is an important determinant of all study outcomes, but that there is a strong mediating role of satisfaction and distinct variation in significant DBE components, with sensory DBE playing a leading role. The paper concludes with implications for research and practice. Ó 2014 Elsevier Ltd. All rights reserved.

Introduction A brand can provide a significant means of differentiation and thus competitive advantage for products and services (Aaker, 1991, 1996; Aaker & Joachimsthaler, 2000; Gardner & Levy, 1955; Keller, 1993). This is important whether the product or service is a razor, breakfast cereal, insurance policy, or a tourist destination. While branded consumer products have a longstanding academic ⇑ Corresponding author. Tel.: +44 1634 888839; fax: +44 1634 888890. 1 2

E-mail addresses: [email protected] (S.J. Barnes), [email protected] (J. Mattsson), fl[email protected] (F. Sørensen). Tel.: +45 46742506; fax: +45 46743081. Tel.: +45 46743312; fax: +45 46743081.

http://dx.doi.org/10.1016/j.annals.2014.06.002 0160-7383/Ó 2014 Elsevier Ltd. All rights reserved.

122

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139

literature, the literature on destination branding is of much more recent origin. The destination is also a complex product, not least because it is an amalgam of different tourist products and is ‘‘also a perceptual concept, which can be interpreted subjectively by consumers, depending amongst other things, on their travel itinerary, cultural background, purpose of visit . . .’’ (Buhalis, 2000, p. 97). More generally, as noted by Govers, Go, and Kumar (2007), tourism is a consumption experience that ‘‘designates those facets of consumer behavior that relate to the multisensory, fantasy and emotive aspects of one’s experience of products’’ (Hirschman & Holbrook, 1982, p. 92). Destination branding is a powerful marketing tool that can build a positive image and emotional links with visitors and underpinning destination branding are the perceived experiences that a visitor will have at a destination (Morgan, Pritchard, & Pride, 2004). For the purposes of this study, destinations are considered as ‘‘a defined geographical region which is understood by its visitors as a unique entity, with a political and legislative framework for tourism marketing and planning’’ (ibid, p. 98). To understand, explain and predict consumer behavior with respect to destination brands, many mainstream marketing concepts, theories and frameworks have been applied. These have included those on: destination image (Baloglu & McCleary, 1999; Kotler & Gertner, 2004), customer-based brand equity (Boo, Busser, & Baloglu, 2009; Konecnik & Gartner, 2007), destination brand personality (Ekinci & Hosany, 2006; Murphy, Moscardo, & Benckendorff, 2007), and destination brand identity (Cai, 2002; Govers & Go, 2004). However, these brand conceptualizations have been criticized for being incomplete, generalized, and for not capturing the full range of experiences impacting on the consumer from brand stimuli (Brakus, Schmitt, & Zarantonello, 2009). Recently, a new conceptualization has emerged—that of brand experience, which provides a more complete evaluation based on sensory, affective, intellectual and behavioral dimensions of the brand (Brakus et al., 2009). While brand attitudes are general evaluations, brand experiences ‘‘include specific sensations, feelings, cognitions, and behavioral responses triggered by specific brand stimuli. For example, experiences may include specific feelings, not just an overall liking.’’ (Brakus et al., 2009, p. 53). This broader conceptualization has yet to be applied to tourism research. The contribution of this paper is to apply this new theoretical concept to destination brands to focus on the concept of destination brand experience (DBE). The purpose of this study is to determine whether DBE is a significant determinant of visitor outcomes including satisfaction, intention to recommend and intention to revisit a destination. In particular, we are interested to determine which elements of DBE impact on visitor outcomes by using an experience model and a range of destinations. The study focuses on purposive sampling of three locations (a semi-rural region in Denmark and two very different cities in Sweden) for data collection and utilizes partial least squares path modeling for its analysis.

Literature and hypotheses Approaches to destination branding Within the burgeoning literature on tourism marketing, numerous conceptualizations have been applied to understand and explain tourist behavior, most of them adapted from the mainstream marketing literature (e.g. Aaker, 1997; Keller, 1993). Within this developing literature base, one key strand of literature is that on destination branding. Table 1 provides an overview of key branding concepts adapted from the marketing literature and applied to the tourism context. In each case the original concept name and early marketing sources are provided along with the name and a selected definition of the adapted concept in the tourism research literature. Much of the research in the area of destination branding has focused on destination image, but not necessarily traditional branding (Baloglu, 1996; Baloglu & Brinberg, 1997; Bigné, Sánchez, & Sánchez, 2001), although more recently some authors have attempted to integrate destination image with other branding concepts (Hosany, Ekinci, & Uysal, 2007; Nandan, 2005; Qu, Kim, & Im, 2011). Recent evidence suggests complex, nested brand associations in tourism. For example, Deng and Li (2014) use image transfer theory to empirically demonstrate that brand associations from events transfer to destination image, while Nicolau and Mas (2014) use associative network theory to show that some

123

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139 Table 1 Overview of Key Branding Concepts Applied in the Tourism Context. Name of Marketing Concept

Marketing Source

Name of Marketing Concept in Tourism

Tourism Source

Brand Image

Newman (1957), Dichter (1985), Aaker (1991), Keller (1993)

Destination Image

Lawson and Baud-Bovy (1977), Crompton (1979)

Brand Identity

Kapferer (1998), Aaker and Joachimsthaler (2000)

Customer-Based Brand Equity

Aaker (1991), Keller (1993)

Brand Personality

Aaker (1997)

Brand Experience

Brakus et al. (2009)

Definition

‘‘ . . . the sum of beliefs and impressions people hold about place. Images represent a simplification of a larger number of associations and pieces of information connected to a place. They are a product of the mind trying to process and pick out essential information from huge amounts of data about a place’’ A general definition Destination Brand Cai (2002), from marketing that is Identity Govers and relevant in the tourism Go (2004) context: ‘‘. . . a set of associations the brand strategist seeks to create or maintain’’ Kim and Kim ‘‘Brand equity . . . Consumer-Based encompasses brand Brand Equity (for a (2005), Konecnik and image (e.g., perception Destination) of service quality) and Gartner (2007), Boo at brand familiarity . . . brand equity entails al. (2009) favourable disposition that may not necessarily result in purchasing behaviour’’ ‘‘. . . the set of human Ekinci and Destination characteristics Hosany (Brand) associated with a (2006), Personality Murphy et al. destination’’ (2007) Not yet applied None Not applicable

Definition Source

Kotler and Gertner (2004, p. 42)

Aaker and Joachimsthaler (2000, p. 40)

Nam et al. (2011, p. 1011)

Ekinci and Hosany (2006, p. 127)

None

destinations might free-ride on collective national brands. Stepchenkova and Li (2014) further demonstrate that such brand associations can be captured though top-of-mind evaluations. Overall, the destination branding literature has developed substantially over the past decade and a half and has provided significant implications for travel destination management, developing into a rich stream of practical research (Boo et al., 2009; Buhalis, 2000; Konecnik & Gartner, 2007). The general marketing literature contains a number of models which authors suggest cover the essential characteristics of brands such as customer-based brand equity (Keller, 1993; Yoo & Donthu, 2001), brand personality (Aaker, 1997), brand image (Aaker, 1991), and brand identity (Aaker & Joachimsthaler, 2000; Kapferer, 1998). These models have different foci. For example, whereas customer-based brand equity attempts assess the differential effects of brand knowledge to the marketing of the brand, via sub-constructs such as brand loyalty, brand awareness, perceived quality of the brand and brand associations, brand personality strives to catch the personality traits that consumers see in the brand.

124

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139

One of the most recent conceptualizations for measuring the response to a brand is that of brand experience. Brakus et al. (2009) criticize the existing brand constructs for providing incomplete and generalized conceptualizations that do not capture the full range of experiences impacting on the consumer from brand stimuli. Brand experience, while related, is quite different to concepts such as brand attitudes, brand personality, brand involvement and brand attachment. Brand experience provides a more holistic evaluation of a brand—a gestalt of experiences. While brand attitudes are general assessments, brand experiences involve particular sensations, feelings, cognitions, and behavioral responses as a result of particular brand stimuli (Brakus et al., 2009, p. 53). Similarly, brand personality is an inferential type of brand association that does not imply that consumers experience actual sensations, feelings, cognitions and behavioral responses; in essence, customers merely project particular characteristics of brand personality onto a specific brand. Brakus et al. (2009) point out that much of the research on experiences has focused on utilitarian product attributes and category experiences rather than brands and as such there is a dearth of literature on service brand experiences. They examine philosophical, cognitive and marketing experience literature with a view to broadening the concept and measurement of brand experience. Drawing on the work of the philosopher Dewey (1925) and later research by Dubé and LeBel (2003), Brakus et al. (2009) suggest that experience goes beyond knowledge to include ‘‘intellectual experiences resulting from knowledge, experiences also include perceiving (through the senses), feeling, and doing’’ (p. 54). They further cite the support from Pinker (1997) in cognitive science for these specific human mental responses to environmental cues. From the experience marketing literature, Brakus et al. (2009) draw on the work of Pine and Gilmore (1999) in retail environments and Schmitt (1999) for a variety of consumer product brands to identify five key areas of consumer experience: sense, feel, think, act and relate, stating that ‘‘the feel experience includes moods and emotions. The think experience includes convergent/analytical and divergent/imaginative thinking. The act experience refers to motor actions and behavioral experiences. Finally, the relate experience refers to social experiences’’ (p. 54). Brakus et al. (2009) propose and rigorously test a new scale for brand experience. The scale is developed based on a comprehensive set of six studies to develop and test the scale items and final scale. Dozens of brands were examined, including LEGO, iPod, Clinique, Sudoku, W Hotels and Tropicana, although none of the brands examined in the study were specifically tourist destinations. The final brand experience scale is composed of four aspects: sensory, affective, intellectual and behavioral brand experience. Sensory brand experience refers to bodily experiences based on visual, aural, olfactory, gustatory and tactile experiences, e.g. the sound of activity in the open-air market, the taste of the food, the feel of the sand beneath one’s feet, the beauty of the city’s skyline, or the smell of the forest. Affective brand experience refers to feelings, sentiments and emotions, e.g. feeling welcome in the hotel or a love of the city’s architecture. Behavioral brand experience refers to physical actions, bodily experiences and behaviors, e.g. being incited to dance in a nightclub or windsurf in the ocean. Finally, intellectual brand experience refers to thought, stimulation of curiosity and problem-solving, e.g. a thought-provoking museum exhibit or economical souvenir purchase. While other authors in tourism marketing have utilized constructs that may be construed as sensory, affective, behavioral or intellectual (e.g. Bigné, Andreu, & Gnoth, 2005; de Rojas & Camarero, 2008; del Bosque & San Martin, 2008), the Brakus et al. (2009) model provides the first to formally, rigorously and systematically create a unified model that captures four key dimensions of DBE. The model appears particularly pertinent for destination branding and this study provided an opportunity to test the relevance of the model in this context. Research model and hypotheses Brand experience applies to all kinds of products and services, including complex experiential products such as are found in tourism (Buhalis, 2000). There is a developed literature that links aspects of destination branding, including the scales discussed above, to satisfaction and brand loyalty (e.g. Chi & Qu, 2008; Nam, Ekinci, & Whyatt, 2011). Similarly, Brakus et al. (2009), drawing on the earlier work of Pine and Gilmore (1999), state that ‘‘. . .because experience provides value, we expect that the more a brand evokes multiple experience dimensions, and therefore has a higher overall score on the scale, the more satisfied a consumer will be with the brand’’ (p. 63). They argue that since brand experiences create pleasurable outcomes, these will affect future-oriented decision making: consum-

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139

125

ers are more likely to make repeat purchases and recommendations to others (Brakus et al., 2009). Brakus et al. (2009) further find empirical support for the relationship between the overall brand experience scale and both customer satisfaction and customer loyalty in the sixth of their studies, examining brands in a variety of categories including computers, clothing, water, cars, newspapers and sneakers (shoes). However, the study did not examine the relationship between the individual brand experience scale components and both satisfaction and brand loyalty or examine these relationships in the context of tourism. Notwithstanding, some of these relationships have been alluded to in tourism studies. For example, del Bosque & San Martin (2008) examine the cognitive and affective models of tourist behavior and point out that in the cognitive model ‘‘the key elements are the mental representations of objects such as knowledge and beliefs, i.e. cognitions. Individuals would process external information of the tourist experience in order to form their own beliefs and judgments’’ (p. 553) while the emotional model ‘‘is based on the assumption that feelings are an important component of the experience since destinations are considered to include, for example, sensory pleasures, daydreams and enjoyments’’ (p. 553). The brand experience scale, by its very nature, integrates both of these approaches through affective (emotional model) and intellectual (cognitive model) components, while also including sensory and behavioral components. The components are considered important for explaining decision-making and behavior processes, including satisfaction and loyalty. We believe that this study provides an important opportunity to both understand the individual components of brand experience (via the use of a four-dimensional brand experience model) and to test the relationships in a new context, destination branding. The degree of visitor loyalty is typically measured via two key constructs: a visitor’s intention to revisit a destination and a visitor’s willingness to recommend a destination to others (Chi & Qu, 2008; del Bosque & San Martín, 2008; Oppermann, 2000; Simpson & Siquaw, 2008). It is important to note that the pattern of significant brand experience constructs may vary according to the specific tourism context and characteristics (in the same way that Gentile, Spiller, & Noci, 2007, find that dimensions of experience vary depending on product characteristics for consumer goods). Thus, we propose that: H1. A positive (H2a. sensory, H2b. affective, H2c. behavioral and H2d. intellectual) DBE will increase visitor satisfaction with the destination.

H2. A positive (H2a. sensory, H2b. affective, H2c. behavioral and H2d. intellectual) DBE will increase visitor intention to revisit the destination. H3. A positive (H2a. sensory, H2b. affective, H2c. behavioral and H2d. intellectual) DBE will increase visitor intention to recommend the destination. There is a strong body of literature linking visitor satisfaction, typically seen as a post-purchase assessment (Oliver, 1981), and destination loyalty in the tourism context (Alexandris, Kouthouris, & Meligdis, 2006; Bramwell, 1998; del Bosque & San Martín, 2008; Nam et al., 2011; Pritchard & Howard, 1997; Yoon & Uysal, 2005). Building on research in marketing, evidence has shown that more satisfied customers tend to purchase more products or services and to spread positive word-of-mouth that is likely to influence others to become customers (Maxham, 2001; Mittal & Kamakura, 2001; Oliver, 1997; Reicheld, 1996), and this has been shown to hold in many different contexts, including retail banking, mobile commerce, hotels and hairdressing (Lin & Wang, 2006; Maxham, 2001; Rust & Zahoric, 1993). In the tourism context, the same relationships have been empirically supported in numerous studies (Assaker, Esponsito Vinzi, & O’Connor, 2011; Campo-Martínez, Garau-Vadell, & Martínez-Ruiz, 2009; Chi & Qu, 2008; del Bosque & San Martín, 2008; Nam et al., 2011; Yuksel, Yuksel, & Bilim, 2010). Thus, in this study we posit that: H4. Higher visitor satisfaction will positively influence visitor intention to revisit the destination.

H5. Higher visitor satisfaction will positively influence visitor intention to recommend the destination.

126

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139

Evidence suggests that satisfaction typically plays a mediating role between many branding constructs and loyalty. For example, Chi and Qu (2008) found that satisfaction fully mediated the relationship between destination image and destination loyalty. Nam et al. (2011) found that satisfaction partially mediated several relationships between components of a decomposed consumer-based brand equity model and brand loyalty. Similarly, Osman and Sentosa (2013) found a significant mediating effect for satisfaction between service quality and visitor loyalty. Brakus et al. (2009) state that: ‘‘brand experience is also likely to result in further processing and thus affect satisfaction and loyalty indirectly’’ (p. 63) and demonstrate empirical evidence for roughly equal direct and indirect effects of brand experience on loyalty (through satisfaction). However, as yet, the role of satisfaction in mediated relationships between individual brand experience components and brand loyalty has not been explored. We therefore intend to examine the mediating role of satisfaction between a model of individual DBE constructs and brand loyalty components. The testing of this model in the context of tourism is also a new dimension to the examination of these hypotheses. Thus we posit: H6. The relationship between (H6a. sensory, H6b. affective, H6c. behavioral and H6d. intellectual) DBE and visitor intention to revisit the destination will be positively mediated by visitor satisfaction with the destination. H7. The relationship between (H6a. sensory, H6b. affective, H6c. behavioral and H6d. intellectual) DBE and visitor intention to recommend the destination will be positively mediated by visitor satisfaction with the destination. To provide a comprehensive test of the research model, illustrated in Fig. 1, in the tourism context, we took a purposive approach to selecting destination brand locations. Thus, to test the research model, we developed three different studies in different parts of Scandinavia, each with a very different tourist experience. For study one, we wished to select an area which is rural and peripheral, but that still has a clear name, location, and demarcation as islands. The area selected was Lolland-Falster, an area of Denmark typically associated with beach holidays. As a contrast, we wished to focus on two geographically close but very different cities in another country, Sweden. The first is a small city that is both a cultural and religious centre. It is perceived as traditional and academic, and hosts the largest university in Scandinavia. The second city is ‘‘in transition’’ from heavy industry to services and high

Sensory Desnaon Brand Experience Intenon to Revisit

Affecve Desnaon Brand Experience

Sasfacon Behavioral Desnaon Brand Experience Intenon to Recommend Intellectual Desnaon Brand Experience Destination Brand Experience Fig. 1. Research Model.

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139

127

technology. The city is nearly three times as large, the third largest city in Sweden, with a substantial migrant population and considerable investment in new infrastructure. We selected these two cities in order to triangulate country against two cities that are geographically close but very different in size and composition. Study one: the Lolland-Falster region of Denmark Background to study one Tourism is a very developed industry in Denmark, with 4.5 million visitors in 2008 (UNWTO, 2010). However, its contribution to GDP is relatively small at US$6.2 billion (1.8%) in 2011, ranking it 44th in the world for absolute value and 152nd in the world for proportion of GDP (World Tourism and Trade Council, 2012). The study focuses on the Lolland-Falster region of Denmark (http://www.visitlollandfalster.com/). This comprises two islands in the south-east of Denmark that are a popular tourist destination: Falster (which is known for its beaches along the coast of the Baltic Sea and Denmark’s southernmost point, Gedser, which has ferry links to Germany) and Lolland (a flat island which includes, among other things, Knuthenborg Safari Park, holiday resorts, cafés, restaurants, discotheques, small villages, and the Fuglsang Art Museum). The islands have a population of 107,000 people and the largest town is Nykøbing Falster with approximately 20,000 inhabitants (Danish Statistical Institute, 2012). The dominant types of accommodation in the destination are self catering vacation houses and apartments, as well as camping parks. The area enjoys a significant number of tourists, primarily repeat visitors of families with children attracted to the beaches; overnight stays plus day visitors totalled more than 4.5 million in 2008, 70–80% as repeat visitors (Brandt & Thessen, 2011). The destination attracts mainly Danish and German tourists and to a lesser degree Dutch, Norwegian and Swedish tourists. Due to the nature of the tourist activity—family beach holidays—we anticipate that the sensory element will play an important role in determining the three outcome measures and that this will be more important than the other experience measures. Method for study one Data were collected via a survey at two tourist offices in Lolland-Falster, Marielyst and Nykøbing Falster, between the 12th and 18th of July 2010. The questionnaire was respondent-completed, but assistance was available when requested. The survey utilized the recent brand experience scale of Brakus et al. (2009), modified to reflect the specific domain under investigation. The scale is composed of four aspects of brand experience: sensory, affective, intellectual and behavioral. In addition, the survey also collected single-item measures of satisfaction (‘‘My stay on Lolland-Falster has been satisfactory’’), intention to recommend (‘‘I will recommend others to visit Lolland-Falster’’), and revisit intention (‘‘I would like to return to Lolland-Falster again another time’’) (adapted from Oliver, 1997). Within the tourism literature, the disconfirmation paradigm has been widely employed but heavily criticized due to various issues, including those surrounding the chain of causality and in the complex and ambiguous nature of products that are hard to evaluate (Assaker, Esposito Vinzi, & O’Connor, 2011). Consequently, in line with the suggestion of Tse and Wilton (1988), the use of a global, single-item measure of tourists’ satisfaction is recommended (Assaker, Esposito Vinzi, & O’Connor, 2011; Kozak & Rimmington, 2000). Items were measured using 7-point Likert scales ranging from 1 = strongly disagree to 7 = strongly agree, where 4 = neutral. In all, 360 usable responses were received, 176 (48.9%) from Marielyst and 184 (51.1%) from Nykøbing Falster. The majority of respondents were German (50.3%), Danish (29.7%) or Norwegian (7.8%). All were tourists. Respondents typically stayed in the location for 1–2 days (30.6%) or 3–5 days (31.3%). Tables 2 and 3 present the reliability and validity of the DBE construct based on a confirmatory factor analysis using XLSTAT. Overall, the composite reliability (CR) for the sub-constructs ranged from 0.791 to 0.851, well above the recommended level of 0.7 (Nunnally, 1978). All items loaded on their expected constructs very strongly at p < .001. Convergent validity was measured by average variance extracted (AVE) and ranged from 0.580 to 0.669, again above the recommend level of 0.50 (Fornell &

128

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139

Table 2 Psychometric analysis of the brand experience construct in study one. Sub-construct

Items

Mean

Std. Dev.

Loading

Std. Error

Critical Ratio (tvalue)

Sensory DBE (AVE = 0.669; CR = 0.851)

1: Lolland-Falster makes a strong impression on my senses, visually and in other ways 2: I find Lolland-Falster interesting in a sensory way 3: Lolland-Falster does not appeal to my senses (r)

5.488

1.227

0.863

0.020

44.085⁄⁄⁄

5.127 5.751

1.383 1.639

0.849 0.736

0.027 0.052

31.653⁄⁄⁄ 14.219⁄⁄⁄

Affective DBE (AVE = 0.601; CR = 0.815)

4: Lolland-Falster induces feelings and sentiments 5: I do not have strong emotions for Lolland-Falster (r) 6: Lolland-Falster is an emotional area

4.651 5.053

1.671 1.826

0.822 0.729

0.043 0.077

19.083⁄⁄⁄ 9.456⁄⁄⁄

4.463

1.568

0.772

0.050

15.550⁄⁄⁄

Behavioral DBE (AVE = 0.583; CR = 0.791)

7: I engage in physical activities and behaviors when I am on Lolland-Falster 8: Lolland-Falster gives me bodily experiences 9: Lolland-Falster is not activity oriented (r)

4.990

1.595

0.839

0.064

13.159⁄⁄⁄

4.792 5.278

1.548 1.780

0.825 0.605

0.062 0.119

13.327⁄⁄⁄ 5.069⁄⁄⁄

Intellectual DBE (AVE = 0.580; CR = 0.814)

10: I engage in a lot of thinking when I am on Lolland-Falster 11: Lolland-Falster does not make me think (r) 12: Lolland-Falster stimulates my curiosity and problem solving

4.778

1.658

0.752

0.048

15.520⁄⁄⁄

5.475 4.935

1.628 1.523

0.669 0.853

0.073 0.041

9.101⁄⁄⁄ 20.667⁄⁄⁄

Note:

⁄⁄⁄

denotes p < .001; (r) indicates reverse-coded items.

Table 3 Test for discriminant validity in study one (Squared correlations < AVE). SEN

AFF

BEH

INT

SAT

WOM

REV

AVE

Sensory DBE (SEN) Affective DBE (AFF)) Behavioural DBE (BEH) Intellectual DBE (INT) Satisfaction (SAT) Intention to Recommend (WOM) Intention to Revisit (REV)

1 0.420 0.168 0.342 0.193 0.231 0.218

0.420 1 0.139 0.412 0.085 0.094 0.074

0.168 0.139 1 0.143 0.063 0.062 0.079

0.342 0.412 0.143 1 0.093 0.117 0.085

0.193 0.085 0.063 0.093 1 0.521 0.417

0.231 0.094 0.062 0.117 0.521 1 0.595

0.218 0.074 0.079 0.085 0.417 0.595 1

0.669 0.601 0.583 0.580 – – –

AVE

0.669

0.601

0.583

0.580

–

–

–

Bold values indicate significance at p < .05.

Larcker, 1981). Similarly, all items loaded more strongly on their own construct (Chin, 1998) and the AVEs for constructs were considerably larger than squared intercorrelations (Fornell & Larcker, 1981). Results of study one The results of testing the research model using PLS path modeling in XLSTAT are presented in Table 4 (testing Hypotheses 1 to 5). The results of mediation tests using Preacher and Hayes (2008) are shown in Table 5 (testing Hypothesis 6 and 7). Overall, the research model has a good fit with the data. The goodness of fit indices shown in Table 4 appear acceptable, with the relative goodness of fit above the 0.90 level (Esposito Vinzi, Chin, Henseler, & Wang, 2010; Henseler & Sarstedt, 2013). As we can see from Tables 4 and 5, there is demonstrable support for each of the hypotheses. In particular, in Table 4 there is very strong support for the sensory DBE component of the DBE scale in significantly determining satisfaction (H1a: b = 0.347, p < .001), intention to revisit (H2a: b = 0.227, p < .001) and intention to recommend (H3a: b = 0.259, p < .001). None of the other DBE components were significant in determining the outcome measures. Similarly, there is strong support for the relationships between satisfaction and both intention to revisit (H4: b = 0.492, p < .001) and

129

S.J. Barnes et al. / Annals of Tourism Research 48 (2014) 121–139 Table 4 Test of research model in study one. Relationship

Path Coeff.

St. error

t

Pr > |t|

f2

Sensory DBE ? Satisfaction Affective DBE ? Satisfaction Behavioural DBE ? Satisfaction Intellectual DBE ? Satisfaction Satisfaction: R2 = 0.200; F = 31.052; p < .001

0.347 0.054 0.081 0.089

0.061 0.065 0.050 0.063

5.684 0.836 1.613 1.409