Destination Image Components and Word-of-Mouth Intentions in Urban Tourism: A Multigroup Approach

Abstract

The purpose of this study was to explore differences among three distinct groups, namely local residents, past tourists, and prospective tourists, in their perceptions of cognitive, affective, and overall image of a city destination and their future behavior. Analysis of data generally confirmed previously established structural relationships of cognitive and affective image, overall destination image, and word-of-mouth intentions. However, differences were identified among the three groups in terms of their destination image perceptions and their behavioral intentions to engage in word-of-mouth communications. Specifically, residents who engaged in word-of-mouth were primarily influenced by the cognitive and affective destination image components, while tourists relied on overall image perceptions.

Keywords

urban tourism destination image cognitive affective word-of-mouth WOM intentions residents past tourists prospective tourists

Introduction

There are a number of stakeholders that play a role in the formation process of a destination’s image, including local residents, existing and future tourists (Bornhorst, Ritchie, & Sheehan, 2010; Merrilees, Miller, & Herington, 2009), who are usually influenced by the levels of familiarity with the destination (Baloglu, 2001). Taken together, the perceptions of these stakeholders morph the brand of a destination (Hankinson, 2004) and instigate a number of branding initiatives for destinations to respond to intense competition in the tourism industry (Ashworth & Kavaratzis, 2009). However, literature is lacking in direct comparisons of local residents, who can speak to the details of the image of the destination, with those perceptions of tourists or potential tourists. The focus is usually on either the residents’ destination perceptions (e.g., Merrilees et al., 2009) or the potential and future tourists (Bonn, Joseph, & Dai, 2005) or a combination of current and past residents (e.g., Novcic, Damnjanovic, & Popesku, 2012; Phillips & Schofield, 2007). This lack of research is more evident for destinations with an urban feel because of the potential multiple attributes, identities, and personality an urban place (i.e., city) can have compared with a rural place.

Destination image perceptions relate to destination branding, which can prove to be a particularly complex task given that destinations such as cities are multifaceted entities and inevitably mean different things to different consumers (Ashworth & Kavaratzis, 2009; Ashworth & Page, 2011). As Freire (2009, p. 420) stresses, even in cases where they are not managed as brands, cities are inherently “embedded with meaning” that leads to the creation of a city destination image. An equally important challenge is how the destination image of a city and the images circulated by the tourism industry relate to the real brand image and its potential to associate with a positive tourist experience offered by the city (Anholt, 2004). The roots of these arguments are in the absence of an adequately developed understanding of how people process perceptions and images about the city as a tourism destination and a place to live. Although there is consistent acknowledgment for the importance of destination image for tourism development, what is lacking is a comprehensive understanding of its components and interrelationships among various “consumers” such as local residents, past tourists, and potential tourists. Such understanding would empower destination experts to exploit the character of the city for tourism development (Elliot, Papadopoulos, & Kim, 2011).

Two key issues, rather underresearched, can be underlined in this direction. The first is that tourism image change depends on the accurate assessment of extant perceived images (Cai, 2002) including images held by locals as they constitute a significant stakeholder of a destination (Garrod, Fyall, Leask, & Reid, 2012). The second relates to the supply-oriented approach of destination image, meaning that branding is related to the development and promotion of a projected destination image with competitive advantage, or brand identity (Cai, 2002), where locals play a critical role. By personally reflecting the core values highlighted about a destination, local residents can directly influence the experience of tourists and, subsequently, tourists’ future consumption behavior (O’Leary & Deegan, 2003). Soliciting input from local residents, along with tourists, and engaging them equally in a destination’s tourism development efforts and branding is critical (Bramwell & Rawding, 1996) as it can enable building nodes to the real destination brand image characteristics compared with perceived brand image characteristics (Simpson & Siguaw, 2008). In fact, a few researchers (Ryan & Aicken 2010) have implicitly or explicitly suggested the simultaneous examination of image perceptions of residents and tourists, linking this approach to effective city branding.

Thus, the present study proposes and tests how image perceptions of a city destination (cognitive, affective, and overall image) influence word-of-mouth (WOM) intentions for three distinct groups: local city residents, past tourists, and prospective tourists of the destination. More specifically, the following research questions are explored in this study: (1) Are there differences among local city residents, past tourists, and prospective tourists in their destination image perceptions (cognitive, affective, overall) and WOM behaviors? (2) How do locals, past tourists and prospective tourists evaluate the structure of the relationship between the cognitive and affective destination image components and WOM intentions? In other words, which destination image component has higher predictive validity for each group? Following a holistic approach that involves groups with varying destination knowledge and experiences not only contributes to the development of place image theory (Elliot et al., 2011) but also enables better informed decisions about how to develop and promote consistent and realistic city images in order to advance a strong city brand for tourism purposes (Cai, 2002).

Literature Review

Destination Image Definition, Components, and Their Relationships

Extensive research has been performed in the area of destination image exploring topics such as the process through which destination image is formed (e.g., Gallarza, Saura, & García, 2002; Gartner, 1993); the identification of destination image components and their measurement (e.g., Baloglu & Mangaloglu, 2001; Baloglu & McCleary, 1999a, 1999b; Qu, Kim, & Im, 2011; Tasci, Gartner, & Cavusgil, 2007); the relationship between destination image and variables such as tourists’ sociodemographic characteristics, their motivation, and psychological connection with the destination, destination attractiveness, and future behavior (e.g., Baloglu & McCleary, 1999a; Beerli & Martín, 2004; Chi, 2011; King, Chen, & Funk, 2015; Prayag, 2009); and branding aspects of destination image management, including brand equity and brand personality (e.g., Ekinci & Hosany, 2006; Konecnik & Gartner, 2007; Papadimitriou, Apostolopoulou, & Kaplanidou, 2013).

Most frequently, the concept of destination image has been operationalized as consisting of two components: a perceptual-cognitive component that captures knowledge and beliefs about a destination’s attributes and an affective component that describes feelings toward a destination (Beerli & Martín, 2004; Konecnik & Gartner, 2007). Both the cognitive and affective components work to influence the overall image of a particular destination in the mind of past or prospective tourists (Baloglu & McCleary, 1999a). Interestingly, research has shown that, in addition to its direct effect, one’s knowledge about a destination (cognitive component) influences their overall perception about that destination indirectly through the affective component of destination image (Baloglu & McCleary, 1999a; Beerli & Martín, 2004).

Strong evidence on the relationship between the cognitive and affective components of image has come from Baloglu and McCleary (1999a) who examined four Mediterranean countries’ destinations (Turkey, Egypt, Greece, and Italy) and showed that the cognitive component of image has a direct and positive influence on the affective component and that the latter serves as a mediating variable between cognitive image and overall destination image. Baloglu (2000) further tested these relationships in another study where some of the components of cognitive destination image influenced affective destination image perceptions and visitation intentions, whereas affective image had a direct influence on visitation intentions as well. Beerli and Martín (2004) extended the above studies by incorporating a comparison between first time and repeat visitors in the relationship of cognitive and affective destination image perceptions. They found a structural relationship between cognitive and affective image, with some of the cognitive components influencing the affective and overall image perceptions, suggesting that cognitive image components are not evaluated all the same. They also found partial mediation of the affective image variable on WOM. Lin, Morais, Kerstetter, and Hou (2007) further supported Baloglu and McCleary’s results as they found that cognitive image influences affective image perceptions much more heavily than overall image, suggesting a partial mediation effect of affective images on overall image and eventually tourist behaviors.

Thus, the following hypotheses arise:

Hypothesis 1: The cognitive component of image of an urban destination has a direct and positive influence on the affective component of destination image.

Hypothesis 2: The cognitive component of image of an urban destination has a direct and positive influence on overall destination image evaluations.

Hypothesis 3: The affective component of image of an urban destination has a direct and positive effect on overall destination image evaluations.

Destination Image and Word-of-Mouth Behaviors

Ultimately, it is tourists’ future behavior that is of interest to destination marketers. In existing literature, future behavior has mainly been operationalized as intention to revisit a destination and/or a willingness to recommend the destination to others (WOM communication; e.g., Hosany, Ekinci, and Uysal, 2007; Prayag, 2009). The intention to engage in positive WOM, which can be powerful in generating new tourists, stems from an overall positive evaluation of a destination and reflects high levels of attitudinal loyalty (Konecnik & Gartner, 2007). Interestingly, the intention to recommend the destination to others, or WOM communication, is thought to be a better indicator of favorable image and a positive experience with a destination than one’s intention to revisit. This is because variety-seeking tourists might not return to the same destination, even if they are fully satisfied with their experience (Ekinci & Hosany, 2006; Kozak & Rimmington, 2000) but they can spread positive WOM.

Prayag (2009) explored relationships among the destination’s cognitive image (attributes), its overall image (holistic), tourists’ satisfaction, and future behavior in a study of tourists to the island of Mauritius, and found a direct and positive impact of both cognitive image and overall image on tourists’ intentions to revisit the island and recommend it to others. Baloglu (2000) found a direct and indirect relationship of cognitive image with visitation intentions, whereas affective image perceptions had a direct influence on intentions to visit the destination. On a different context examining the impact of mega sport events on destination image and future tourism development, Kaplanidou (2007, 2009) found that affective and cognitive destination image perceptions influence intentions to revisit the destination, while Li and Kaplanidou (2013) discussed how the cognitive and affective image components of China after the 2008 Olympic Games were perceived differently among prospective American tourists. More important, a study by del Bosque and Martín (2008) showed that affective images influence WOM as an outcome of brand loyalty to a destination suggesting a direct connection between affective image and WOM. Taken together, the above studies suggest a differential influence of cognitive and affective destination image components on tourists’ behaviors, such as WOM. Thus, we hypothesized the following:

Hypothesis 4: The cognitive component of image of an urban destination has a direct and positive effect on intention to recommend the destination to others.

Hypothesis 5: Affective image evaluations of an urban destination have a direct and positive effect on intention to recommend the destination to others.

Hypothesis 6: Overall image evaluations of an urban destination have a direct and positive effect on intention to recommend the destination to others.

Exposure to and Experience With a Destination: Local Residents Versus Past Tourists Versus Prospective Tourists

As would be expected, beliefs and feelings about a destination will differ depending on consumers’ past experience with the destination and tourism in general, their exposure to primary and secondary information sources, their motives and purpose for traveling, and their sociopsychological characteristics (e.g., Baloglu & McCleary, 1999a; Beerli & Martín, 2004; King et al., 2015). Those who have not visited a destination are influenced by informative tourism promotion efforts and often form more positive but unrealistic views about the destination (Fakeye & Crompton, 1991). Actual visitation, on the other hand, increases a visitor’s knowledge of a destination and provides a more realistic understanding of the attributes and offerings of that location (Baloglu & McCleary, 1999b). Not surprisingly, those closer to the destination, specifically the local residents, have a more intimate view of the destination and its attributes (Walmsley & Young, 1998). It stands to reason that local residents will have a different understanding of how the various destination image components work since their experience with the destination encompasses many activities and happens daily.

Prior research (e.g., Freire, 2009; Garrod et al., 2012; O’Leary & Deegan, 2003) has identified multiple benefits of engaging local residents in tourism development efforts. Local residents can become ambassadors for their destination by highlighting the destination’s positive attributes and unique offerings and by encouraging friends and family to visit (Simpson & Siguaw, 2008). Furthermore, they can contribute to tourists’ positive experience during their trip either through personal or professional interactions with them. Schroeder (1996) argued toward making residents more aware of the positive attributes and offerings of their area because they can directly influence nonresidents’ organic image through interactions and communication. Agapito, Mendes and Valle (2010) and Freire (2009) provided evidence for the critical importance of “local people” as a distinctive image component, while Merrilees et al. (2009) found that different destination elements carry more weight for local residents’ attitudes toward their own city brand.

Although the importance of locals as a stakeholder in tourism development has been identified (Garrod et al., 2012), studies that have compared various types of destination image consumers (e.g., locals, prospective, and past tourists) on their cognitive, affective, and overall image perceptions are lacking in the literature. Instead, fragmented approaches have been presented in terms of understanding one or two tourism “markets” at a time and one or two destination image components at a time. For example, Simpson and Siguaw (2008) found distinct differences in the perceptions of local people compared with tourists alluding but not testing different destination images held by the groups examined, while Phillips and Schofield (2007) along with Choo, Park, and Petrick (2011) focused on residents only and found that positive destination image perceptions influence WOM activity among residents. Similarly, Baloglu (2000) and Lin et al. (2007) focused only on tourists’ perceptions of cognitive and affective destination image perceptions and their influence on relevant behaviors. Fakeye and Crompton (1991) compared first-time and repeat tourists and prospective tourists of a region in southern Texas (USA) and found significant differences in their perceptions of image of that destination, suggesting that prospective tourists’ evaluations could have been a misconception. They also provided evidence for the more comprehensive (“complex, differentiated”) image a longer stay at a destination creates (Fakeye & Crompton, 1991, p. 15). Bonn et al. (2005) reinforced the above findings by illustrating how in-state, versus out-of-state versus international visitors differ in their destination image perceptions.

The reviewed literature solidifies the need to examine different consumer perceptions of tourist images to accurately understand how the components of a destination’s image influence tourism-related behaviors such as WOM. However, no study has looked at local residents, past tourists, and prospective tourists simultaneously to assess how destination image components interrelate and influence WOM intentions for each group. The previous discussion leads to the final hypothesis:

Hypothesis 7: Different interrelationships are expected among locals, past tourists, and prospective tourists in their destination image perceptions and word-of-mouth behaviors.

All hypotheses are portrayed in Figure 1.

Figure 1

Proposed Model Testing the Hypotheses of the Study for the Three Groups of Destination Image “Consumers”: Residents, Past Visitors, Prospective Visitors

Method

The Setting

In the most recent National Tourism Strategy (2011-2021), Greece defined 2 international (i.e., Athens and Thessaloniki) and 12 domestic city tourism destinations with strong potential in urban tourism. The city of Patras (with a population of 250,000) was one of these urban destinations as the capital of the Western Greece region and the main gateway to Italy. With its old and newly constructed ports, its three universities and major hospitals, and the variety of cultural attractions and events (e.g., the Carnival of Patras, Patras’ International Cultural Festival, 2006 European Capital for Culture), the city is particularly suited for domestic urban tourism.

Sampling and Data Collection

This study used the city of Patras as the main destination. The targeted sample consisted of adults aged 18 years and older who had at least one urban tourism experience in a Greek city over the past 2 years. To avoid mixing up the term city destination with other types of destinations (i.e., islands, sites known for religious attractions, etc.), only the Greek cities designated as urban tourism destinations in the Greek National Tourism Strategy (2011-2021) were accepted as valid answers. The sample population for this study was composed of adult citizens of the cities of Athens and Patras with recent experience in city tourism. Three distinctive groups made up the final sample of this research. The first group included residents of the city of Patras who had been tourists of other Greek cities at least once during the past 2 years and thus have knowledge of and personal experience with urban tourism in Greece. The second group, past tourists, consisted of people who had visited the city of Patras as tourists in the 2 years before their participation to the study. Because of their personal experience with Patras and its tourism product they would be able to reflect on their primary images. This group of domestic city tourists was selected with the intention to capture images about the city of Patras without the presence of personal experience. The third group, prospective tourists, consisted of individuals with experience from other Greek city destinations but no prior visitation to Patras. This group of domestic city tourists was selected with the intention to capture images about the city without the presence of personal experience. All three groups were asked to respond to the same set of questions in terms of evaluating the cognitive and affective images of the city of Patras.

The second and third groups of respondents were recruited from Athens, the city that is the largest pool for domestic tourists in Greece. The three samples were selected via the use of a computer-generated random sample that drew candidates from a voter registration list available for both cities (Athens and Patras). The sampling frame for Athens was 650,000 residents and for Patras 276,000 residents. A telephone-administered survey was used to collect data for this research across the three subsamples. A systematic sampling technique was followed to select respondents from the list (every 100th entry of the above list). At the beginning of each phone call respondents were screened by asking them whether they had visited a Greek city for tourism purposes over the past 2 years and had stayed overnight. If the respondents’ reply was positive, interviewers invited them to respond to the survey questions. A total sample of 1,125 respondents who had prior experience in urban domestic tourism in Greece was invited to take part in this survey. The number of the respondents per group was based on an initial number required for the project as determined by funding and expected statistical analysis. Once this sample size was reached, there were no further recruitment efforts from the company who solicited participants. From the recruiting process, a total of 540 individuals met the requirements and were willing to offer responses to the entire survey (48% response rate). This sample consisted of 207 local residents (38.3%), 158 past tourists (29.3%) and 175 prospective tourists (32.4%) of the city of Patras.

Measurement

This study adopts the definition of destination image as “mental portrayal or prototype” that depicts a knowledge structure in a person’s mind that is formed based on beliefs and emotional pieces of information (Govers, Go, & Kumar, 2007, p. 978). Reliable and valid measures from previously published empirical works were used to operationalize each variable of the present study. The cognitive image items were gathered from relevant literature and were modified to fit the context of this study. Baloglu and McCleary (1999b) developed a 14-item instrument to measure cognitive image perceptions of four country destinations, which loaded on three factors: quality of experience, attractions, and value/environment. From that scale, nine items were selected as fitting to the research context of an urban destination (i.e., suitable accommodation, appealing local food, friendly people, cleanness and hygiene, nightlife and entertainment, good value for money, interesting cultural attractions, interesting historical attractions, beautiful scenery/natural attractions). The remaining measurement items were generated from previous empirical works on cognitive images for state or national destinations (Murphy, Pritchard, & Smith, 2000; O’Leary & Deegan, 2003; Uysal, Chen, & Williams, 2000). Study participants were asked to rate the city as a short-break urban destination on each of the 20 items using a 5-point Likert-type scale ranging from 1 = offers very little to 5 = offers very much.

The items for measuring the cognitive image construct were adopted from different sources. This strategy is quite common for capturing the particular variable since the literature is lacking a universally accepted scale for the cognitive image of urban destinations (Baloglu & McCleary, 1999b; Song, Su, & Li, 2013). Given that the urban tourism product is multifaceted and quite unexplored, we also added items reflecting images for business activities, leisure, and sports activities unique to an urban space, choices offered for family and kids, accessibility, and so on that were not captured in the study of Baloglu and McCleary as it was intended to measure images at a country level.

Affective images of the destination were measured with four items: unpleasant/pleasant, distressing/relaxing, ugly/pretty, and gloomy/exciting based on existing literature (e.g., Baloglu & McCleary, 1999b; Hosany et al., 2007). These four items were measured on a 5-point bipolar scale. Perceived overall image of the destination was measured on a 5-point Likert-type scale with a single item requiring respondents to share their perceptions of the overall image of the city as a destination. Similar measures have also been used by Baloglu and McCleary (1999b) and Bigne, Sanchez, and Sanchez (2001). The anchors of the item were 1 = very negative and 5 = very positive. Future WOM behavior was measured with three items reflecting WOM communication. The three selected items were adopted from previous research (Bosnjak, Sirgy, Hellriegel, & Maurer, 2011; Lee, Petrick, & Crompton, 2007). Respondents were asked to report the likelihood of saying positive things about the city destination to other people, recommending the place for visit, and encouraging friends or relatives to visit the city. The anchors of the items were 1 = not at all likely and 5 = extremely likely. Items measuring revisit intentions were not considered due to the sample of residents included in this study.

The psychometric properties of the destination image variables and WOM used in this study were tested for face validity and internal consistency. The initial version of the scales was piloted with a sample of 15 tourism experts who lived in Patras and worked in the tourism industry of the city. Based on their comments, the survey was enhanced in terms of clarity and content validity. Internal consistency measures were estimated using Cronbach’s alpha and compared with the widely accepted rule of thumb of .7 (Nunnally & Bernstein, 1994). Reliability scores were .89 for the cognitive scale, .70 for the affective scale, and .94 for the WOM scale, supporting the scales’ good reliability.

Data Analysis and Results

Profile of Respondents

The demographic profile of study participants is presented in Table 1. The sample consisted of 38.2% males and 61.8% females. The majority of respondents were older than 35 years, with 50 years and older and 35 years to 44 years comprising the two largest age groups (24.9% and 24.1%, respectively). More than half of the respondents (63.4%) were married and reported having a university degree (53.7%).

Table 1

Profile of Respondents (N = 540)

Demographic Variables	Percentage
Gender
Male	38.2
Female	61.8
Age (years)
18-24	10.5
25-34	17.7
35-44	24.1
45-54	22.7
≥55	24.9
Marital status
Single	30.8
Married	63.4
Other (divorced, separated, etc.)	5.7
Education
Compulsory education	4.6
High school (Lykeio)	41.7
University level	53.7

Factor Analysis and Variable Preparation

The initial analysis of the data included factor analysis by using principal component extraction method with varimax rotation to explore the underlying dimensions of the 20-item cognitive image scale, the 4-item affective image scale, and the 3-item scale measuring intentions to engage in WOM communication. The appropriateness of the factor analysis was explored by two tests: the Kaiser–Meyer–Olkin of sampling adequacy that was .901, well above the recommended threshold of .6 (Kaiser, 1974), and Bartlett’s test of sphericity, which produced statistical significance (p = .000) indicating that the correlations were sufficiently large for factor analysis. To explore potential differences in the factor structure of the cognitive image scale based on the exposure of the sample to the city destination, initially four different principal component analyses were undertaken; for locals (N = 207), past tourists (N = 158), prospective tourists (N = 175), and for the total sample (N = 540). In all four analyses the optimal factor solution was based on a combination of criteria including Cattell’s scree plot (Cattell & Vogelmann, 1977), Kaiser’s eigenvalue greater than 1 (Fabrigar, MacCallum, Wegener, & Strahan, 1999), and the cumulative variance criterion (Hair, Black, Babin, Anderson, & Tatham, 2006). The results consistently produced a three-factor solution with very similar distribution of items per dimension, which accounted for 52% to 54% of the initial variance of the respective samples. However, three cognitive items had to be deleted due to small loadings on one of the three dimensions (i.e., local wine, good shopping, major development projects). The analysis was conducted again on the retained 17 items of the cognitive scale using the pooled data set (N = 540) and varimax rotation methods resulting to a solution that explained 53.40% of the total variance.

Testing the Measurement Model

The proposed model as depicted in Figure 1 was tested following a two-stage procedure: measurement model and structural model estimation. First, CFA was used to test the measurement model with the three latent constructs (i.e., cognitive image, affective image, and WOM) allowed to be correlated in order to estimate overall fit, validity, and reliability values for each construct. The statistical software package SPSS Amos 21 was used to test the measures on a confirmatory factor model with each scale item to be constrained to load on only one factor. The measurement model fit was examined using fit indices such as the chi-square statistic (χ²), comparative fit index (CFI), incremental fit index (IFI), and root mean square error of approximation (RMSEA; Hu & Bentler, 1998).

First, the conceptual measurement model was tested on each group separately in an effort to explore model fit per group. These initial CFAs demonstrated good fit with the majority of the standardized regression weights above .5 and significant loadings on all items. Only one item (i.e., “cultural activities and events”) produced lower than .5 standardized regression weight and it was dropped from further analysis. The respective CFAs were reestimated for the three groups and the fit indices were equally good for local residents (χ² = 382.17, degrees of freedom [df] = 220; p < .001; CFI = .90; IFI = .91; and RMSEA = .06), for past tourists (χ² = 305.77, df = 220; p < .001; CFI = .94; IFI = .94; and RMSEA = .05), and prospective tourists (χ² = 398.78, df = 220; p < .001; CFI = .92; IFI = .92; and RMSEA = .06). The next step was to test the CFA model on the three groups simultaneously. These results demonstrated that the fit of the model was applicable for the three groups (χ² = 1086.72.01, df = 660; p < .001; CFI = .92; IFI = .92; and RMSEA = .06) implying that the measurement properties of the model fits all three subgroups well. Table 2 presents the standardized factor loadings for the five indicators, Cronbach’s alpha values, average variance extracted (AVE), and composite reliability (CR) measures for each latent factor. All factor loadings were statistically significant and above .5 with the exception of one item (i.e., “cultural activities and events”), and Cronbach’s alpha values were above the recommended value .70 (Nunnally & Bernstein, 1994).

Table 2

Confirmatory Factor Analysis for Cognitive and Affective Destination Image and Behavioral Intentions (N = 540)

Scales/Items	Standardized Factor Loadings
A. Cognitive Image
Factor 1: Services, Experience, and Atmosphere (α^a = .87, CR = .86, AVE = .47)
Unique atmosphere and lifestyle	.73
City as famous destination	.63
Good value for money	.64
Suitable accommodations	.65
Offers good entertainment	.68
Offers appealing local food	.79
Clean and well maintained city	.65
Friendly local people	.64
Offers choices for family and kids	.67
Accessible city to the tourists	.69
Factor 2: Unique City Attractions (α = .74, CR = .78, AVE = .48)
Significant historical attractions	.71
Significant religious attractions	.65
Beautiful scenery/natural attractions	.70
Interesting built architecture	.69
Factor 3: Activities and Events (α = .74, CR = .84, AVE = .75)
Cultural activities and events	deleted
Business activities and conferences	.82
Leisure and sports activities	.88
B. Affective Image (α = .70, CR = .78, AVE = .50)
Unpleasant/pleasant	.72
Distressing/relaxing	.54
Ugly/pretty	.88
Gloomy/exciting	.62
C. Word-of-Mouth Intentions (α = .94, CR = .94, AVE = .85)
Say positive things about the city to other people	.86
Recommend the city to others as place to visit	.94
Encourage friends or relatives to visit the city	.95

Cronbach’s alpha; AVE = average variance extracted; CR = composite reliability.

Two types of construct validity were examined: convergent and discriminant validity. Convergent validity refers to the degree to which indicators of a construct converge or share a good percentage of variance (Hair et al., 2006). The construct AVEs ranged from .47 and .85 supporting convergent validity for all scales (Fornell & Larcker, 1981), with the exception of the “services, experiences and atmosphere” and “unique city attractions,” which were threshold cases (.47 and .48, respectively). It was decided to maintain these two cognitive factors in the study based on their estimates of composite reliability (.86 and .78) and internal consistency values (.87 and .74), which were satisfactory. In addition, these two factors explained a large amount of variance in the factor analysis procedure. Finally, discriminant validity was tested, following the guidelines by Fornell and Larcker (1981), which indicate that the squared correlation between two constructs should be less than the AVEs of each construct. As shown in Table 2, evidence of discriminant validity was provided as all AVEs exceeded respective squared factor correlations. Since all scales in this study were evaluated and deemed as having adequate measurement properties, the testing of the hypotheses was the next step in the analysis. The mean scores of each of the three cognitive image factors were created as new input variables for the latent factor of cognitive image for the next stages of the structural model analysis, creating a more parsimonious model for further analysis.

Testing the Structural Model

To test the hypotheses that related to the structural relationship of the destination image components and WOM intentions the multigroup approach was employed using structural equation modeling analysis with the AMOS 21 software. Following Kline (1998), we tested the conceptualized model by estimating group differences on a model level and among the path coefficients. The three subsamples of respondents were local residents, past tourists, and prospective tourists and represented different levels of experience with the city as the destination.

Assumptions of multivariate normality were met across the three samples through the evaluation of Mardia’s coefficient that was 0.96 (past tourists), 2.8 (residents), and 4.24 (prospective tourists). Correlation matrices along with the means and standard deviations of the model for each group are presented in Table 3. The hypothesized model was assessed by examining the fit statistics, along with the t values of the paths and their regression weights. Specifically, the indices used to evaluate the fit of the model were IFI, CFI, and the RMSEA (Bentler, 1990).

Table 3

Correlation Matrix of the Model Variables Tested in the Three Samples

Residents	Overall Image	Cog 1	Cog 2	Cog 3	Affective	WOM
Overall image	1
Cog 1	.57	1
Cog 2	.50	.47	1
Cog 3	.30	.46	.44	1
Affective	.47	.44	.32	.10	1
WOM	.55	.61	.45	.30	.34	1
Mean	3.51	3.00	3.32	2.90	3.53	3.52
SD	.84	.67	.80	.81	.78	1.13
Past Tourists	Overall Image	Cog 1	Cog 2	Cog 3	Affective	WOM
Overall image	1
Cog 1	.69	1
Cog 2	.48	.54	1
Cog 3	.44	.56	.47	1
Affective	.58	.59	.38	.34	1
WOM	.69	.72	.72	.41	.49	1
Mean	3.86	3.37	3.09	3.31	3.64	3.40
SD	.71	.60	.73	.80	.81	1.14
Prospective tourists	Overall image	Cog 1	Cog 2	Cog 3	Affective	WOM
Overall image	1
Cog 1	.78	1
Cog 2	.57	.63	1
Cog 3	.43	.48	.43	1
Affective	.68	.67	.48	.32	1
WOM	.75	.69	.56	.36	.55	1
Mean	3.69	3.29	2.95	3.34	3.51	3.11
SD	.78	.70	.74	.84	.93	1.16
Aggregate Sample	Overall Image	Cog 1	Cog 2	Cog 3	Affective	WOM
Overall image	1
Cog 1	.54	1
Cog 2	.34	.47	1
Cog 3	.30	.53	.38	1
Affective	.51	.55	.36	.24	1
WOM	.52	.61	.50	.30	.59	1
Mean	3.67	3.29	3.13	3.16	3.56	3.36
SD	.80	.68	.78	.84	.84	1.16

Note: WOM = word of mouth.

The results of the structural analysis for each of the groups are presented in Table 4 along with the fit statistics (χ² = 165.954, df = 117; p < .001; CFI = .98; IFI = .98; and RMSEA = .03) which implied good model fit across all three groups simultaneously. The majority of the paths were significant at p < .5 or higher, and had small to strong standardized regression weights (.17-.76).

Table 4

Comparative Fit Measures for Assessment of Measurement and Structural Model Invariance Tests

Comparative Fit Measures	χ²	df	Sig.	CMIN/DF	CFI	IFI	RMSEA
Model 1: Residents	382.17	220	.001	1.81	.90	.91	.06
Model 2: Past visitors	305.77	220	.001	1.39	.94	.94	.05
Model 3: Prospect visitors	398.78	220	.001	1.73	.92	.92	.06
Model 4: Combined measurement model (residents, past visitors, and prospect visitors)—Baseline model	1086.72	660	.001	1,64	.92	.92	.03
Model 5: Combined measurement model (residents, past visitors, and prospect visitors)—Constrained model	1149.05	716	.001	1.60	.92	.92	.03
Model 6: Baseline structural	165.954	117	.001	1.41	.98	.98	.03
Model 7: Constrained structural	261.718	171	.001	1.53	.97	.97	.03

Note: CFI = comparative fit index; IFI = incremental fit index; RMSEA = root mean square error of approximation; df = degrees of freedom; Sig. = significance.

At first glance (see Figure 2), the results suggest that the same destination image components are significant in predicting overall image perceptions for past tourists, prospective tourists, and local residents. More specifically, for all three groups the cognitive destination image positively influenced the affective destination image perceptions (β_residents = .57, p < .05; β_{past tourists} = .76, p < .05; β_{prospect tourists} = .76, p < .05). The effect of cognitive image on overall image perceptions was confirmed for residents (β_residents = .42, p < .05) but not for past or prospective tourists (β_{past tourists} = .27, p > .05; β_{prospect tourists} = .14, p > .05). Significant effects were found for cognitive image to intentions for WOM communication across all three groups (β_residents = .48, p < .05; β_{past tourists} = .44, p < .05; β_{prospect tourists} = .41, p < .05). Therefore, Hypotheses 1 and 4 were supported for all three samples, but Hypothesis 2 was only partially supported.

Figure 2

Estimated Structural Equation Modeling Model for the Hypotheses of the Study

The path from affective destination image to overall image perceptions was also significant across all three groups (β_residents = .25, p < .05; β_{past tourists} = .42, p < .05; β_{prospect tourists} = .57, p < .05), in support for Hypothseis 3. There was also a significant effect from affective image to intentions for WOM communication (β_residents = .37, p < .05; β_{past tourists} = .32, p < .05; β_{prospect tourists} = .29, p < .05) and thus Hypothesis 5 was supported. Moreover, for prospective visitors the overall image component positively and significantly influenced the WOM communication intentions (β_{prospect tourists} = .17, p < .05), but not in the case of residents or past tourists (β_residents = .05, p > .05; β_{past tourists} = .11, p > .05) leading to partial support for Hypothesis 6. The cognitive image construct explained 33% of the variance in affective image for residents, 58% for past tourists, and 58% prospective tourists, whereas substantial amount of variance in overall image (37% for residents, 43% for past tourists, and 49% prospective tourists) was explained by the cognitive and affective image variables. Finally, a large amount of variance in intentions for WOM communication (63% for residents, 65% for past tourists, and 63% prospect tourists) was explained by cognitive and affective factors and overall image. These results are summarized in Table 5.

Table 5

Results From Structural Analysis

Path Estimates	Residents	Past Tourists	Prospective Tourists	Hypothesis
Cognitive → Affective	.57*	.76*	.76*	Hypothesis 1 supported
Cognitive → Overall image	.42*	.27	.14	Hypothesis 2 supported
Affective → Overall image	.25*	.42*	.57*	Hypothesis 3 supported
Cognitive → WOM	.48*	.44*	.41*	Hypothesis 4 supported
Affective → WOM	.37*	.32*	.29*	Hypothesis 5 supported
Overall image → WOM	.05	.11	.17*	Hypothesis 6 partially supported
R² (on affective image)	.33	.58	.58
R² (on overall image)	.37	.43	.49
R² (on WOM)	.63	.65	.63
Fit indices	χ² = 145.954, df = 117; IFI = .98, CFI = .98, RMSEA = .03
Analysis sample size	N = 207	N = 144	N = 164

Note: WOM = word of mouth.

Significant path coefficients at p < .05.

On examination of the standardized loadings of the three cognitive image components it could be observed that each of the three factors exerted different influence on the latent variable of cognitive image across the three groups: “Services, experience and atmosphere” contributes more to the formation of cognitive image (β = .84 for residents, β = .91 for past tourists, and β = .93 for prospective tourists) followed by “unique city attractions” (β = .61 for residents, β = .62 for past tourists, and β = .67 for prospective tourists) and “activities and events” (β = .52 for residents, β = .60 for past tourists, and β = .53 for prospective tourists).

To test Hypothesis 7 of the study, a multisample approach was used in which the model is estimated simultaneously for the three groups that represent different levels of experience with the city destination (i.e., residents, past tourists, and prospective tourists). To examine variation in the measurement and the structural models for the three groups, the invariance procedure recommended by Byrne (2004) was followed. First, the baseline model was estimated for each group separately to identify group differences in the operation of the model. The goodness-of-fit indices suggested that the model had acceptable fit to the three sets of data (see entries Model 1, Model 2, and Model 3 in Table 4). Then, configural and metric invariance was tested to ensure that the model and its factor structure operate in the same way across the three groups under study. Initially the baseline measurement model (i.e., configural) was built with five latent variables (i.e., service/experience, city attractions, activities/events, affective image, and WOM) and 23 items without any invariance constraints, and was tested simultaneously for the three groups. Goodness-of-fit indices related to the three groups produced a good fit (χ² = 1086.72, df = 660; p < .001; CFI = .92; IFI = .92; and RMSEA = .03) indicating configural invariance, which implied that the factor structure was identical across the three groups (see entry Model 4 in Table 4). Next, metric invariance was tested by fixing all factor loadings, factor variances, and factor covariances of the measurement model to be equal across the three groups (constrained model). The testing of the three-group model produced good goodness-of-fit statistics (χ² = 1149.05, df = 716; p < .001; CFI = .92; IFI = .92; and RMSEA = .03) to the data (entry Model 5 in Table 4). For invariance testing, the chi-square value of Model 4 was compared with that of Model 5. This comparison yielded a chi-square difference (Δχ²) of 62.25 with 56 Δdf that was not statistically significant (p = .261) indicating that the three groups of this study are not invariant in terms of the item factor relationships.

The final step of this analysis concerned the testing of differences in the structural paths among the three groups. Both the unconstrained (Model 6) and the constrained structural model, with maintaining the factor loadings fixed across the three groups (Model 7), were tested and produced good fit to the data set (entries Models 6 and 7 in Table 4). The comparison of the chi-square values of these two models yielded a difference (Δχ²) of 96.764 with 54 Δdf, which was statistically significant (p = .001). This finding implies significant differences in the structural paths of the tested model among the three groups, leading to acceptance of Hypothesis 7.

Further analysis was undertaken to explore the specific significant path differences between pair of groups (e.g., residents vs. past tourists, residents vs. prospective tourists, and past tourists vs. prospective tourists) by imposing constraints on regression paths, one at a time, and exploring differences in the chi-square value of each model. These results revealed three regression paths that differed significantly across the groups. More specifically, significant differences were identified between residents and prospective tourists (β_residents = .57, β_{prospective tourists} = .76) and residents and past tourists (β_residents = .57, β_{past tourists} = .76) for the effect of cognitive image on affective image, and between residents and prospective tourists (β_residents = .42, β_{prospective tourists} = .14) for the relationship between cognitive and overall image.

Discussion

Theoretical Implications

The purpose of the study was to examine how the components of destination image about an urban destination interact and differ among three distinct groups of tourism product consumers based on their different levels of experience with the destination. More specifically, the study explored the influence of cognitive, affective, and overall image perceptions on intentions for WOM communication about the destination among local residents, past tourists, and prospective tourists. The results contribute to the literature in three ways. First, at the model level all groups differed in terms of how they processed some of the destination image components to engage in positive WOM. Second, for the residents and past tourists (i.e., people with direct experience of the destination) the cognitive and affective image components influenced directly positive WOM, whereas for prospective tourists, all destination image components proved significant in influencing their WOM suggesting that they need more information to draw their WOM recommendations from. Third, a stable measurement structure of the concept of destination image components was unveiled among the three groups based on the confirmatory factor analysis for the three groups. From a theoretical standpoint, the study shows that experience with a destination influences the manner in which people arrive to a WOM communication. The presence of past experience allows a more attribute-based processing stemming from cognitive and affective images, whereas for people who have not experienced the destination all pieces of destination image information can influence independently the WOM communication or converge to a holistic image and create positive WOM communications.

The first contribution of the study was about the significant differences among local city residents, past tourists, and prospective tourists in some of their destination image perceptions and their WOM intentions. The results revealed differences among the three groups in the manner in which they perceived the influence of cognitive image on affective image and overall image. More specifically, cognitive image was the pool of information for the influence of affective images for all groups, but played a more important role for past tourists and prospective tourists. Thus, we can discuss that exposure to the destination creates a differential outcome in affective destination image perceptions (Baloglu & McCleary, 1999b) across multiple groups with different destination experiences. We believe that the cognitive dimensions are subject to the level of knowledge a person has about a destination and thus are more likely to fluctuate among groups with different levels of knowledge. Affective and overall images have more enduring features, and their predictive ability was confirmed in our study across the three groups. Overall images are holistic and, as discussed in Baloglu and Brinberg (1997), can be lasting and stable.

The second contribution of the study, which builds on previous research (e.g., Baloglu & McCleary, 1999a; Beerli & Martín, 2004; Ekinci & Hosany, 2006; Hosany et al., 2007; Qu et al., 2011), involves the differential role of cognitive and affective image components on the overall image perceptions across the three groups. In detail, for local residents overall image perceptions relied on cognitive components, but the same did not apply for the other two groups. It appears that the extent of past experience with a place (extensive when you are a resident) underlines the role of cognitive information. Furthermore, direct experience influences memory structures and schemas about the encountered product or service (Braun, 1999) creating more complex image perceptions (Fakeye & Crompton, 1991). Direct experiences with destination “products” can also influence WOM activity (Westbrook, 1987) and decision-making processes due to realistic and direct understanding of their components by the consumer (Gartner, 1993).

Also, significant differences among the three groups were found in terms of how the overall destination image perceptions influenced WOM. Interestingly, prospective tourists were more prone to be influenced by holistic images. Equally interestingly, the propensity of local residents and past tourists to recommend the destination for visitation relying on the various cognitive components found in services, amenities, and attractions as well as affective components suggests a saliency of elements that encourage WOM communications irrespective of the holistic images (Simpson & Siguaw, 2008).

In line with previous research (e.g., Baloglu & McCleary, 1999a; Beerli & Martín, 2004; Ekinci & Hosany, 2006; Qu et al., 2011), the study found that affective image had a significant direct effect on overall destination image for all three groups. Thus, the critical role of emotions and overall perceptions toward a destination is underlined by the results of this study. The empirical validation of these associations across the three different path models provides a more comprehensive understanding of the factors that lead to positive holistic destination image perceptions and favorable future behavior as it examines consumers with varying degrees of personal experience with a destination. It also fills a current gap in the literature dealing specifically with mid-/major-sized cities competing in domestic tourism markets.

With regard to the third contribution of the study, results showed a stable factor structure of the destination image components across the three groups and reinforced the importance of a new cognitive image component capturing the city’s activities and events. In accordance with previous literature proposing cognitive image as a multidimensional construct (Baloglu & McLeary, 1999a; Dobni & Zinkhan, 1990), the principal component analysis performed in the sample of short break domestic Greek tourists produced three factors, namely services, experience and atmosphere, unique city attractions, and activities and events.

The characteristics of the first two dimensions are to some extent consistent with image measures of regions, countries, or islands identified in past studies (e.g., Baloglu & Mangaloglu, 2001; Beerli & Martín, 2004; Qu et al., 2011). The present study supports the relevance and applicability of the image dimensions in a city context. Furthermore, the emergence of the cognitive image dimension of activities and events, which is almost nonexistent in available measures, suggests that urban contexts may uniquely feature this component. The present case showed that cities, due to the extensive availability of infrastructure and amenities, provide an attractive platform for featuring various types of tourism events. As a result, a more comprehensive measure of cognitive image, particularly suitable for the study of city destinations, is proposed by this study.

Practical Implications

From a practical standpoint, the findings of this research could offer valuable insight to destination marketers, especially those who promote urban centers and are interested in developing urban tourism programs and attracting first-time and repeat tourists. The strength of the services, experience and atmosphere factor in predicting affective destination images and WOM communication provides clear direction to city officials in charge of branding and tourism efforts; strengthening the tourism experience in the city and raising the level of quality across all services can provide a competitive advantage to the city’s tourism development efforts. Regarding the role of affective images in future behavior it may be desirable to develop tourism campaigns intended to stimulate positive feelings about a destination among past and prospective tourists.

The consistently significant and positive influence of the cognitive and affective destination image factor on WOM across all three groups suggests that the elements that make urban destinations desirable and somewhat unique is emotional attachment and evaluations about the feel and atmosphere of a city, the quality of accommodations and dining in reasonable pricing, the variety in entertainment options, the attitude of local people, and the accessibility of the destination. These aspects could be effectively leveraged for positioning purposes and for the creation of an emotional connection with the city. Given the important role of affective destination image perceptions for all groups, city officials and tourism managers should invest in developing and promoting a rich portfolio of emotional products related perhaps to cultural and leisure events that garner excitement.

This study provides evidence that local residents should not be overlooked. So far the development of a destination’s image was seen as a process that involved mainly target tourist markets. An equally beneficial and rather inexpensive strategy would be for a city destination to engage local residents in tourism development efforts by exploring their evaluations of the city’s image. Findings of this study highlight the potential of locals to become significant image formation agents and ambassadors for their city, especially if their image of the city is favorable. On that note, we have to acknowledge that, purely descriptively, the results of this study with regard to mean scores of the items used in the model are slightly above the midpoint of the scale. This finding suggests that there is room for growth with regard to the destination image perceptions for all groups.

Given that locals are important not only as a visitor segment in their own right but also as promoters of the city through WOM recommendations (Garrod et al., 2012), their perceptions are of critical importance in the image formation process of an urban destination and the branding of the city. In fact, Beerli and Martín (2004) propose that WOM constitutes the most trustful and believable communication channel, which usually projects images very close to destination reality. For city tourism, a place-marketing framework needs to be implemented taking into consideration views of all three destination image consumer groups examined in this research effort as a basis for setting marketing and product development objectives.

It is appropriate to note that past experience with urban tourism was the only criterion used to screen study participants. No other variables were included in the model, for example sociodemographic characteristics of respondents, their motivations, sources through which they received their information, or satisfaction levels, all of which have been examined in past literature (e.g., Baloglu & McCleary, 1999a; Beerli & Martín, 2004; King et al., 2015; Prayag, 2009). Future research could incorporate certain demographic or psychological factors in the study of image evaluations and future visitor behavior.

The present research offers insights into the salient role of the cognitive and affective components in forming residents’ overall image and, most important, their intention to engage in WOM communication. This finding suggests several directions for further research. Clearly, the long exposure of the locals to the city contributes to a more differentiated image, which is enduring and less misleading. Future studies may explore this further to uncover not only the role of locals’ WOM communication in city tourists’ decision-making behavior and information process but also the differentiated effect of each of the image components as projected by locals on tourist’s behavior. Among others, this study has tested the model on locals who themselves are tourists for other urban destinations at the domestic level. A more rigorous comparative study is required to be conducted into the relationships of the model between business and leisure tourism segments of locals as these represent important markets in city tourism. Finally, extending this comparative study to other cities and even other tourism contexts could offer additional support on how destination image perceptions as well as behavioral intentions are formed across distinct consumer groups.

Footnotes

Dimitra Papadimitriou, PhD (e-mail: dpapad@upatras.gr), is an assistant professor in the Department of Business Administration at the University of Patras, Greece. Kyriaki (Kiki) Kaplanidou, PhD (e-mail: kkaplanidou@hhp.ufl.edu), is an associate professor in the Department of Tourism, Recreation and Sport Management at the University of Florida, Gainesville. Artemisia Apostolopoulou, PhD (e-mail: apostolopoulou@rmu.edu), is a university professor in the Department of Sport Management at Robert Morris University, Moon Township, Pennsylvania.

References

Agapito

Mendes

J. D.-C.

Valle

P. O. D.

(2010). Destination image: Perspectives of tourists versus residents. European Journal of Tourism, Hospitality and Recreation, 1, 90-101.

Anholt

(2004). Nation-brands and the value of provenance. In Morgan

Pritchard

Pride

(Eds.), Destination branding: Creating the unique destination proposition (2nd ed., pp. 26-39). Oxford, England: Elsevier Butterworth-Heinemann.

Ashworth

Kavaratzis

(2009). Beyond the logo: Brand management for cities. Brand Management, 16, 520-531.

Ashworth

Page

S. J. M.

(2011). Urban tourism research: Recent progress and current paradoxes. Tourism Management, 32, 1-15.

Baloglu

(2000). A path analytic model of visitation intention involving information sources, socio-psychological motivations, and destination image. Journal of Travel & Tourism Marketing, 8(3), 81-90.

Baloglu

(2001). Image variations of Turkey by familiarity index: Informational and experiential dimensions. Tourism Management, 22, 127-133.

Baloglu

Brinberg

(1997). Affective images of tourism destinations. Journal of Travel Research, 35, 11-15.

Baloglu

Mangaloglu

(2001). Tourism destination images of Turkey, Egypt, Greece, and Italy as perceived by US-based tour operators and travel agents. Tourism Management, 22, 1-9.

Baloglu

McCleary

K. W.

(1999a). A model of destination image formation. Annals of Tourism Research, 26, 868-897.

10.

Baloglu

McCleary

K. W.

(1999b). U.S. international pleasure travelers’ images of four Mediterranean destinations: A comparison of visitors and nonvisitors. Journal of Travel Research, 38, 144-152.

11.

Beerli

Martín

J. D.

(2004). Factors influencing destination image. Annals of Tourism Research, 31, 657-681.

12.

Bentler

P. M.

(1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238-246.

13.

Bigne

Sanchez

M. I.

Sanchez

(2001). Tourism image, evaluation variables and after purchase behaviour: Inter-relationship. Tourism Management, 22, 607-616.

14.

Bonn

M. A.

Joseph

S. M.

Dai

(2005). International versus domestic visitors: An examination of destination image perceptions. Journal of Travel Research, 43, 294-301.

15.

Bornhorst

Ritchie

B. J. R.

Sheehan

(2010). Determinants of tourism success for DMOs & destinations: An empirical examination of stakeholders’ perspectives. Tourism Management, 31, 572-589.

16.

Bosnjak

Sirgy

M. J.

Hellriegel

Maurer

(2011). Post-visit destination loyalty judgments: Developing and testing a comprehensive congruity model. Journal of Travel Research, 50, 496-508.

17.

Bramwell

Rawding

(1996). Tourism marketing images of industrial cities. Annals of Tourism Research, 23, 201-221.

18.

Braun

(1999). Postexperience advertising effects on consumer memory. Journal of Consumer Research, 25, 319-334.

19.

Byrne

B. M.

(2004). Testing for multigroup invariance using AMOS graphics: A road less traveled. Structural Equation Modeling, 11, 272-300.

20.

Cai

L. A.

(2002). Cooperative branding for rural destinations. Annals of Tourism Research, 29, 720-742.

21.

Cattell

R. B.

Vogelmann

(1977). A comprehensive trial of the scree and KG criteria for determining the number of factors. Journal of Multivariate Behavioral Research, 12, 289-325.

22.

Chi

C. G.

(2011). Destination loyalty formation and travelers’ demographic characteristics: A multiple group analysis approach. Journal of Hospitality & Tourism Research, 35, 191-212.

23.

Choo

Park

S.-Y.

Petrick

J. F.

(2011). The influence of the resident’s identification with a tourism destination brand on their behavior. Journal of Hospitality Marketing & Management, 20, 198-216.

24.

del Bosque

I. R.

Martín

H. S.

(2008). Tourist satisfaction a cognitive-affective model. Annals of Tourism Research, 35, 551-573.

25.

Dobni

Zinkhan

G. M.

(1990). In search of brand image: A foundation analysis. Advances in Consumer Research, 17, 110-119.

26.

Ekinci

Hosany

(2006). Destination personality: An application of brand personality to tourism destinations. Journal of Travel Research, 45, 127-139.

27.

Elliot

Papadopoulos

Kim

S. S.

(2011). An integrative model of place images: Exploring relationships between destination, product, and country image. Journal of Travel Research, 55, 520-534.

28.

Fabrigar

L. R.

MacCallum

R. C.

Wegener

D. T.

Strahan

(1999). Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4, 272-299.

29.

Fakeye

Crompton

(1991). Image differences between prospective, first-time, and repeat visitors to the lower Rio Grande Valley. Journal of Travel Research, 30, 10-16.

30.

Fornell

Larcker

D. F.

(1981). Evaluating structural equation models with unobserved variables and measurement error. Journal of Marketing Research, 18, 39-50.

31.

Freire

J. R.

(2009). Local people as critical dimension for place brands. Brand Management, 16, 420-438.

32.

Gallarza

M. G.

Saura

I. G.

García

H. C.

(2002). Destination image: Towards a conceptual framework. Annals of Tourism Research, 29, 56-78.

33.

Garrod

Fyall

Leask

Reid

(2012). Engaging residents as stakeholders of the visitor attraction. Tourism Management, 33, 1159-1173.

34.

Gartner

W. C.

(1993). Image formation process. Journal of Travel & Tourism Marketing, 2, 191-215.

35.

Govers

F. M.

Kumar

(2007). Virtual destination image a new measurement approach. Annals of Tourism Research, 34, 977-997.

36.

Hair

Black

Babin

Anderson

Tatham

(2006). Multivariate data analysis (6th ed.). Upper Saddle River, NJ: Prentice Hall.

37.

Hankinson

(2004). Relational network brands: Towards a conceptual model of place brands. Journal of Vacation Marketing, 10, 109-121.

38.

Hosany

Ekinci

Uysal

(2007). Destination image and destination personality. International Journal of Culture, Tourism and Hospitality Research, 1, 62-81.

39.

Bentler

P. M.

(1998). Fit indices in covariance structural equation modeling. Sensitivity to underparameterized model misspecification. Psychological Methods, 3, 424-453.

40.

Kaplanidou

(2007). Affective event and destination image: Their influence on Olympic travelers’ behavioral intentions. Event Management, 10, 159-173.

41.

Kaplanidou

(2009). Relationships among behavioral intentions, cognitive event and destination images among different geographic regions of Olympic Games spectators. Journal of Sport & Tourism, 14, 249-272.

42.

King

Chen

Funk

D. C.

(2015). Exploring destination image decay: A study of sport tourists’ destination image change after event participation. Journal of Hospitality & Tourism Research, 39, 3-31.

43.

Kline

R. B.

(1998). Principles and practice of structural equation modeling. New York, NY: Guilford Press.

44.

Konecnik

Gartner

W. C.

(2007). Customer-based brand equity for a destination. Annals of Tourism Research, 34, 400-421.

45.

Kozak

Rimmington

(2000). Tourist satisfaction with Mallorca, Spain, as an off season holiday destination. Journal of Travel Research, 38, 260-269.

46.

Lee

S. Y.

Petrick

J. F.

Crompton

J. L.

(2007). The roles of quality and intermediary constructs in determining festival attendees’ behavioral intention. Journal of Travel Research, 45, 402-412.

47.

Kaplanidou

(2013). The impact of the 2008 Beijing Olympic Games on China’s destination brand: A U.S.-based examination. Journal of Hospitality & Tourism Research, 37, 237-261.

48.

Lin

C.-H.

Morais

D. B.

Kerstetter

D. L.

Hou

J.-S.

(2007). Examining the role of cognitive and affective image in predicting choice across natural, developed, and theme-park destinations. Journal of Travel Research, 46, 183-194.

49.

Merrilees

Miller

Herington

(2009). Antecedents of residents’ city brand attitudes. Journal of Business Research, 62, 362-367.

50.

Murphy

Pritchard

M. P.

Smith

(2000). The distinction product and its impact on traveler perceptions. Tourism Management, 21, 43-52.

51.

Novcic

Damnjanovic

Popesku

(2012). Serbia brand identity: Perspectives of residents and diaspora. EuroMed Journal of Business, 7, 256-267.

52.

Nunnally

J. C.

Bernstein

I. H.

(1994). Psychometric theory. New York, NY: McGraw-Hill.

53.

O’Leary

Deegan

(2003). People, pace, place: Qualitative and quantitative images of Ireland as a tourism destination in France. Journal of Vacation Marketing, 9, 213-226.

54.

Papadimitriou

Apostolopoulou

Kaplanidou

(2013). Destination personality, affective image, and behavioral intentions in domestic urban tourism. Journal of Travel Research, 54, 302-315. doi:10.1177/ 0047287513516389

55.

Phillips

Schofield

(2007). Pottery, pride and prejudice: Assessing resident images for city branding. Tourism Analysis, 11, 397-407.

56.

Prayag

(2009). Tourists’ evaluations of destination image, satisfaction, and future behavior intentions—The case of Mauritius. Journal of Travel & Tourism Marketing, 26, 836-853.

57.

Kim

L. H.

H. H.

(2011). A model of destination branding: Integrating the concepts of the branding and destination image. Tourism Management, 32, 465-476.

58.

Ryan

Aicken

(2010). The destination image gap—Visitors’ and residents’ perceptions of place: Evidence from Waiheke Island, New Zealand. Current Issues in Tourism, 13, 541-561.

59.

Schroeder

(1996). The relationship of residents’ image of their state as a tourist destination and their support for tourism. Journal of Travel Research, 34, 71-73.

60.

Simpson

P. M.

Siguaw

J. A.

(2008). Destination word of mouth. Journal of Travel Research, 47, 167-182.

61.

Song

(2013). The indirect effects of destination image on destination loyalty intention through tourist satisfaction and perceived value: The bootstrap approach. Journal of Travel & Tourism Marketing, 30, 386-409.

62.

Tasci

A. D. A.

Gartner

W. C.

Cavusgil

S. T.

(2007). Conceptualization and operationalization of destination image. Journal of Hospitality & Tourism Research, 31, 194-223.

63.

Uysal

Chen

Williams

(2000). Increasing state market share through a regional positioning. Tourism Management, 21, 89-96.

64.

Walmsley

D. J.

Young

(1998). Evaluative images and tourism: The use of personal constructs to describe the structure of destination images. Journal of Travel Research, 36, 65-69.

65.

Westbrook

R. A.

(1987). Product/consumption-based affective responses and postpurchase processes. Journal of Marketing Research, 24, 258-270.