An integrative model of destination brand equity and tourist satisfaction

This paper develops an integrative model that includes the dimensions of destination brand equity (i.e. awareness, image, quality and loyalty), as well as a key behavioural variable that is tourist satisfaction. In particular, our paper aims to contribute to the literature: (1) by examining the relationships between the dimensions of destination brand equity, which is not a deeply investigated issue so far; and (2) by adding tourist satisfaction, a key concept in loyalty formation, to the variables usually considered in destination brand equity models. The model was tested in two samples, national and international tourists visiting a destination in Spain, in order to also explore the role of the geographical and cultural distance between tourist and destination. Our results from a multi-group analysis indicate: (1) a robust link between “quality-satisfaction-loyalty” in both samples of tourists; (2) a chain of effects among awareness, image (separately considering the cognitive and affective dimensions) and perceived quality, although with the presence of some significant difference between both samples of tourists; and (3) a positive influence of cognitive image on affective image in both cases.


Introduction
Tourist destinations can be considered products (Boo, Busser, & Baloglu, 2009) which, in a global environment, compete to attract tourists by enhancing their differentiators and competitive advantages (Pike, 2009). With this in mind, one key strategy for achieving positive returns is to develop destination brand. In particular, destination branding is considered a powerful marketing tool to build a positive image of the place and develop emotional links with their visitors (Barnes, Mattsson, & Sørensen, 2014). Although branding emerged in the marketing literature in the fifties (Gardner & Levy, 1955) and has an extensive background (Aaker & Joachimsthaler, 2000), this topic has only attracted the attention of researchers in the field of tourist destinations in recent years (Boo et al., 2009;Kladou, Giannopoulos, & Mavragani, 2015;Pike, , 2010Pike & Bianchi, 2016;Yang, Liu, & Li, 2015). Consequently, we still have a long way to go to understand the implications of brand equity in the tourism field.
Nowadays, there is an agreement that the general principles of branding regarding goods and services can be applied in a similar way in the field of tourist destinations (Caldwell & Freire, 2004;Dioko & So, 2012). This strategy leads to "brand equity", a concept that can be defined as the overall value created by a brand (Bailey & Ball, 2006). Taking into account its potential, this paper takes as a starting point the dimensions proposed for customer-based brand equity (i.e. awareness, image, quality perceived and loyalty) and examines the relationships among them in the context of a tourist destination. In this sense, it is important to highlight that, although the dimensions of destination brand equity are examined in more or less extent in some previous studies, the link between each other remains unexplored up to now (Kladou et al., 2015). Furthermore, our model contains two more contributions to the academic literature in this field. First, destination image is analyzed by taking into account their two dimensions, the affective and cognitive (Baloglu & McCleary, 1999;Walmsley & Young, 1998). While the majority of models published up to now consider only one dimension of image, mainly related to social image and self-image (Bigné, Andreu, & Zanfardini, 2013;Boo et al., 2009;Pike & Bianchi, 2016;Pike, Bianchi, Kerr, & Patti, 2010), or the image as a second order construct composed of several dimensions (Gartner & Konecnik, 2011), this paper adopts a different approach in order to better understand the role of destination image in destination brand equity. Specifically, we establish that a destination evokes cognitions and emotions among tourists and, consequently, it is necessary to examine the effects of these two dimensions of image (i.e. cognitive image and affective image) separately, as well as the possible link between them. Second, the model is completed with the inclusion of tourist satisfaction as a direct antecedent of loyalty toward the destination. As established by Van Raaij, Van Veldhoven, and Wärneryd (2013), satisfaction is a behavioural phenomenon that is extremely important since it represents the main objective of marketing activities and, in general, a keystone for the well-being of people. Under these circumstances, extensive research on the relationship between consumer satisfaction and loyalty has been previously conducted in the marketing literature and, particularly, in tourism research. However, this causal relationship, which is important to offer an integral view of the study of brand equity, has been only included in destination brand equity models very recently (Bigné et al., 2013;Fuchs, Chekalina, & Lexhagen, 2012;Ghafari, Ranjbarian, & Fathi, 2017).
It should be highlighted that our model was empirically tested by carrying out a quantitative research focussed on tourists visiting a Spanish destination. Particularly, the model was evaluated in two different samples: national tourists and international tourists. Since the different geographical and cultural origins of tourists may lead to different perceptions and mental structures regarding the destination concerned, a multi-group analysis was conducted in order to identify if the chain of effects included in our model of destination brand equity is significantly different or not between the two groups of tourists.
In brief, the overall objective of our paper is to generate new knowledge about destination brand equity by adopting a new approach, which is focussed on: (1) the interrelationships among brand equity dimensions, (2) the tourist satisfaction as a key behavioural variable in loyalty formation, and (3) the geographical and cultural distance between tourist and destination as a potential variable influencing the chain of effects in the model. With this in mind, the present paper is organized as follows; first, we provide an overview of the theoretical basis of the paper, justifying our hypotheses and theoretical model. Second, we explain the methodology of this empirical research. Third, we present the findings of the causal model and the possible differences among the groups of tourists. Finally, we summarize the most relevant theoretical and managerial implications, as well as the limitations and future lines of research.
2. Literature review 2.1. Brand equity for a tourist destination "Place branding" is becoming a growing area in the field of destination and tourism marketing (Kaplan, Yurt, Guneri, & Kurtulus, 2010). Much work has been done in the area of analyzing the image of destinations (Baloglu & McCleary, 1999;Kladou et al., 2015;Pike & Ryan, 2004;, but place branding is a complex and extensive field of research, and it is accepted that it cannot be limited to image studies only (Barnes et al., 2014;Kaplan et al., 2010). In general terms, efforts to measure the value of a brand are becoming increasingly important with the need of firms to compete globally (Hsu, Oh, & Assaf, 2012;Keller, 2003). This value has been conceptualized as brand equity, a multidimensional construct initially proposed by Keller (1993) and Aaker (1996) in the field of goods and services and now extended to places (Bigné et al., 2013;Im, Kim, Elliot, & Han, 2012;Konecnik, 2006;Konecnik & Gartner, 2007;Pike et al., 2010;Yang et al., 2015).
Although there are a variety of conceptual approaches, customer-based brand equity can be defined as a measure of the strength of consumers' attachment to a brand or a description of the associations and beliefs the consumer has about the brand (Feldwick, 1996). In this context, destination image is a key variable in the study of brand equity, but other dimensions are also necessary to truly measure this construct (Boo et al., 2009;Konecnik & Gartner, 2007).
It is accepted that customer-based brand equity encompasses four dimensions: brand awareness, brand image, brand quality and loyalty (Barnes et al., 2014). In this paper, we add a new variable, satisfaction, already included in recent work about destination brand equity (Bigné et al., 2013;Fuchs et al., 2012). In line with previous studies, brand awareness would reflect the tourist's knowledge of a particular destination or the presence of a destination in the minds of tourists when a given travel context is considered (Pike & Bianchi, 2016). Brand image, often interchangeably referred to as brand associations, would represent the set of associations attached to the destination, composed of a variety of individual perceptions relating to various attributes of the destination that may or may not reflect objective reality (Aaker, 1996). In this sense, it is important to differentiate the cognitive image and the affective image. The first one refers to the individual's own knowledge and beliefs about the destination while the affective image is associated with emotions and feelings about it (Baloglu & McCleary, 1999;Walmsley & Young, 1998). Brand quality is a holistic judgment made based on the excellence or overall superiority (Bigné, Sánchez, & Sanz, 2005). Satisfaction is a tourist's cognitive-affective state derived from his/her experience at the destination . Finally, loyalty represents the core dimension of brand equity (Keller, 2003). In tourism, loyalty is usually considered as the intention to revisit the destination and word-ofmouth intentions (Chen & Myagmarsuren, 2010;Prayag, 2012).

2.2.
Chain of effects among the dimensions of brand equity: awareness, image and quality First, we analyze the awareness-image relationship. According to different approaches, awareness is a first and necessary step to loyalty (Konecnik, 2006). On the one hand, it is accepted that the higher the level of awareness the more dominant is the brand. This will increase the likelihood that the brand will be a member of the consideration set and will receive serious consideration for purchase (Keller, 1993;Yasin, Noor, & Mohamad, 2007). On the other hand, and according to the associative network model, memory consists of nodes or units of information, defined as stored information connected by links that vary in strength (Anderson, 1983;Keller, 1993). A destination brand represents a potential node, to which a variety of associations is linked, forming a knowledge structure (Pike & Bianchi, 2016;Pike et al., 2010). According to Keller (2003), once a brand is identified, the customer tends to proffer a meaning to the brand, giving rise to brand associations. Brand awareness influences the formation of these associations, so it is to be expected that a greater awareness of a destination will enhance the perception of its brand image (Bigné et al., 2013;Ghafari et al., 2017;Liu & Fang, 2016;Pike et al., 2010).
In tourism research, destination image (i.e. brand image) is defined as the set of impressions, beliefs, ideas, expectations and feelings accumulated towards a tourist destination over time (Kim & Richardson, 2003). This approach to destination image, which includes both cognitive and affective associations, has been supported by recent studies about destination image (e.g. Han & Hwang, 2016;Lim & Weaver, 2014;Lin, Morais, Kerstetter, & Hou, 2007;Papadimitriou, Kaplanidou, & Apostolopoulou, 2015;Smith, Li, Pan, Witte, & Doherty, 2015;Yacout & Hefny, 2015). In this sense, since a tourist destination is capable of evoking different emotions, such as pleasure or excitement, among visitors (Walmsley & Young, 1998), destination image would be represented not only by the beliefs or knowledge an individual has of the attributes of the destination (Pike & Ryan, 2004) but also by his/her feelings toward the place (Chen & Uysal, 2002;Kim & Richardson, 2003).
Accordingly, this study adopts a cognitive-affective approach of destination image when establishing the interrelationships between this construct and other variables such as awareness and perceived quality. Therefore, based on the brand equity theory, this research establishes that a greater awareness of the destination will enhance not only the cognitive associations linked to the place by tourists, but also the affective ones. Consequently, the following hypotheses are established: H1. Destination awareness will have a direct and positive influence on the cognitive image of the destination. H2. Destination awareness will have a direct and positive influence on the affective image of the destination.
Second, previous research on consumer behaviour has widely recognized that perceptions of quality are influenced by the perceived image of a product or service (Bloemer, De Ruyter, & Peeters, 1998). In tourism, the relationship between image and perceived quality, which can be defined as tourists' evaluation of a destination's offerings (Zabkar, Brencic, & Dmitrovic, 2010), has been confirmed in several previous works (Bigné, Sánchez, & Sánchez, 2001;Bigné et al., 2005;Chen & Tsai, 2007;Hankinson, 2005;Kim, Holland, & Han, 2013). The image that tourists form of a destination influences the way they perceive the destination's quality. In particular, given that destination image is formed based on a tourist's feelings (affective associations) and perceptions (cognitive associations) concerning a destination's resources and attractions, an improved image will reinforce the quality of the destination as perceived by visitors. With this in mind, the third and fourth hypotheses are established: H3. The cognitive image of the destination will have a direct and positive influence on its perceived quality. H4. The affective image of the destination will have a direct and positive influence on its perceived quality.
Finally, and as previously established in the introduction, it is necessary to examine not only the influence of these two dimensions of destination image on other constructs, but also the possible link between them. According to the traditional attitude models (Bagozzi, 1982;Fishbein & Ajzen, 1975), the affective evaluations of an attitude object are influenced by the cognitive evaluations of that object, so both dimensions of attitude are closely related. In a similar way, Russell (1980) establishes that information is initially interpreted by individuals and endowed with meaning, and that it subsequently contributes to forming their emotional states. Based on these theoretical arguments, several studies in tourism research have postulated, and empirically confirmed, that the cognitive image of a tourist destination positively influences its affective image (Baloglu & McCleary, 1999;Beerli & Martín, 2004;Han & Hwang, 2016;Lin et al., 2007;Papadimitriou et al., 2015). Consequently, we aim to enhance our model of destination brand equity by including the following hypothesis: H5. The cognitive image of the destination will have a direct and positive influence on its affective image.

The role of tourist satisfaction in loyalty formation
Customer-based brand equity implies a direct relationship between perceived quality and consumer loyalty. However, according to consumer behaviour literature (Gounaris, Dimitriadis, & Stathakopoulos, 2010;Srivastava & Sharma, 2013), if individuals have direct experience with a product (e.g. destination), their future behaviour will also be affected by their satisfaction with the consumption experience. Thus, we include this variable in our theoretical model, and we examine the role of satisfaction considering its antecedents and its influence on loyalty towards the destination.
Since previous research on consumer behaviour usually establishes that perceived quality is an antecedent of satisfaction (e.g. Cronin, Brady, & Hult, 2000;Gounaris et al., 2010;Olsen, 2002;Srivastava & Sharma, 2013), our study postulates a positive relationship between perceived quality and tourist satisfaction. Therefore, satisfaction, which can be defined as a tourist's cognitive-affective state derived from his/her experience at a destination , is considered a behavioural variable that plays a key role in the formation of loyalty toward the destination. The relationship qualitysatisfaction can be based on the theoretical arguments proposed by Lazarus (1991) and Oliver (1997). According to these authors, consumers develop a behavioural sequence initiated by a cognitive phase (a component that acquires a greater importance in quality judgments) and followed by an emotional phase (a component that shows a greater relevance in satisfaction states). In tourism research, several studies show empirical evidence of a positive relationship between quality and satisfaction (Baker & Crompton, 2000;Bigné et al., 2001;Jin, Lee, & Lee, 2015;Lee, Jeon, & Kim, 2011;Petrick, 2004). Thus, the fifth hypothesis is established: H6. Destination quality will have a direct and positive influence on tourist satisfaction.
Finally, satisfaction plays a critical role in predicting and understanding an individual's responses after a consumption experience. In this respect, the relationship between satisfaction and consumer loyalty has been widely explored and confirmed (e.g. Miguel-Dávila, Cabeza-García, Valdunciel, & Flórez, 2010;Nam, Ekinci, & Whyatt, 2011;Pleshko & Heiens, 2015;Wu, Zhou, & Wu, 2012). Two main dimensions contribute to loyalty, considered as a consumer response based on brand commitment : the intention to repurchase and the willingness to recommend the brand (Andreassen & Lindestad, 1998). The first one can be defined as the likelihood that consumers will buy the product or service again (Szymanski & Henard, 2001) while the second is not only an indicator of the user's intention to continue the relationship with a firm but also a credible source of information for potential users (Maxham, 2001).
Our study uses this attitudinal approach to loyalty instead of a behavioural focus (i.e. loyalty as a repeated behaviour over time) for two reasons. On the one hand, the desire of individuals to seek variety in their destination experiences (Jang & Feng, 2007) justifies the exploration of future intentions instead of past behaviours. On the other hand, the behavioural approach cannot distinguish between true loyalty and spurious loyalty (Chen & Gursoy, 2001). In tourism, several studies have empirically confirmed that the intention to revisit a destination and the willingness to recommend it to other people are positively affected by tourist satisfaction (Bigné et al., 2001;Lee et al., 2011;Ozdemir, Aksu, Ehtiyar, Çizel, & Içigen, 2012;Prayag & Ryan, 2012;San Martín, Collado, & Rodríguez, 2013;Song, Su, & Li, 2013;Yoon & Uysal, 2005). In accordance with this evidence, the last hypothesis of this study is established: H7. Tourist satisfaction will have a direct and positive influence on loyalty toward the destination.
Next, Figure 1 summarizes the research hypotheses formulated in this study.

Cultural distance and destination brand equity
Once the theoretical model has been established, we consider it is worth investigating in an exploratory way the role of the cultural distance between tourists and destinations in the study of destination brand equity. Cultural distance measures the extent to which consumers' origin cultures are different from or similar to the culture of the host (Ahn & McKercher, 2015). According to the study of MacKay and Fesenmaier (2000), the cultural background of individuals filters their perceptions of a tourist destination. In this regard, several previous works have empirically demonstrated that the cultural distance with the destination influences the tourists' destination choice (Bi & Lehto, 2018) and the perceived image that tourists have of the place (Huang, Chen, & Lin, 2013;Kastenholz, 2010;. Cultural distance can interplay with travel motivations. Specifically, novelty seeking and escape are two important motivational drivers for international travel; people may travel because they want to experience something completely new and different (Lee & Crompton, 1992). In this sense, cultural distance promises opportunities for novelty for travellers (Bi & Lehto, 2018). On the opposite, commonplace or familiar trips fulfil other social demands such as kinship or social interactions (Snepenger, 1987).
Additionally, tourists typically purchase and consume a whole range of services, which together make up the "holiday or vacation experience", and they tend to base their judgements on the quality of and satisfaction with a vacation experience on all components of this complex tourism system (Klaus, 2000). With regard to this, Klaus (2000) and Lee and Lee (2009) highlight that the cultural distance may influence the overall valuation of the place and the tourism experience. In this line, our study establishes that the different cultural distance with the destination that have national and international tourists would lead to different cognitive and emotional connections with the place. In consequence, the origin of tourists may influence the intensity of the chain of effects proposed in our theoretical model. Specifically, taking into account the cultural distance as a control variable, we establish the following research question to be explored in our empirical research: RQ. Is the cultural distance between tourists and destination a variable that influence the chain of effects among the dimensions of destination brand equity?

Methodology
Quantitative research was carried out to test the factorial structure proposed in the theoretical model and to confirm our hypotheses. Data was collected using a questionnaire that included: (1) the dimensions of brand equity for the destination, (2) satisfaction during the stay at the destination, and (3) the socio-demographic profile of respondents. Particularly, the variables of the model were all measured using multi-attribute instruments through 10-point Likert scales (see Appendix). In addition, several recommendations made by Sekaran (1983) and Churchill (1991) about research conducted in different languages were taken into account in order to assure the validity of our questionnaire. More concretely, the questionnaire was initially translated from Spanish to English by the researchers. Subsequently, one expert who is a native English speaker checked both accuracy and meaning of the translated items. Finally, we did a back translation into Spanish in order to ensure that both versions of the questionnaire were equivalent (several minor amendments were included in this process).
The target population for quantitative research consisted of national and international tourists, above 18 years of age, who were visiting the region of Cantabria in the north of Spain. The samples were selected by using the methods of quotas and convenience. In a first stage, we used a quota sampling method; particularly, according to the statistics provided by the Spanish Institute of Tourism Studies and the Cantabrian Institute of Statistics in terms of gender and age of the target population, we build a profile of national and international tourists to be surveyed. In a second phase, we used a convenience method where the main tourist attractions and infrastructures of the region of Cantabria were selected to collect empirical data in an efficient way. In this sense, data were gathered through a personal survey that was conveniently administered at the International Airport of Santander and the international tourist attractions of Cabárceno Wildlife Park and Altamira Caves during the summer of 2013. Taking into account that non-responses may provide a bias in the estimation of results (Lynn, 1996), we adopted the following strategy to avoid this problem: the interviewers were informed that, if during the data collection a questionnaire included a non-response for any item measuring the variables of the model, they had to repeat the survey with another tourist with similar characteristics.
With the aim of ensuring an adequate sample size, we considered two types of criteria previously established in Structural Equation Modeling Research. In this sense, Nunnally (1967) originally indicated that at least 10 observations per indicator are necessary in SEM estimation; several more recent studies have suggested that the ratio (r) of indicators to latent variables is a better option (Boomsma, 1982;Marsh & Bailey, 1991). In particular, Marsh and Hau (1999) suggest a sample of at least 400 if r = 2.0, at least 200 if r = 3.0, and at least 100 if r = 4.0. Taking into account that our model includes six latent variables and 21 indicators (r = 3.5), the lower bound is 210 tourists if we consider the first criterion and a sample of between 100 and 200 tourists if we used the second one. Finally, 667 responses (251 international tourists and 416 national tourists) were obtained since a significant effort was made to obtain a more representative sample in overall terms. The stratification and the socio-demographic characteristics of respondents are indicated in Tables 1 and 2, respectively. Particularly, the dominant profile are tourists aged 25-44 years (in coherence with the distribution of the population under investigation), workers and with university studies.

Results
Considering that method bias negatively affects the item validities and reliabilities, as well as the covariations between latent constructs (MacKenzie & Podsakoff, 2012), we initially performed Harman's single-factor test in order to check for the presence of the so-called "common method variance -CMV" (Chang, Van Witteloostuijn, & Eden, 2010). The results of the exploratory factor analysis executed in IBM-SPSS software indicate that all the items measuring the variables load into more than one factor; in particular, six factors (almost 70.0% of explained variance) were obtained in the analyses of both samples of tourists. Therefore, since these items are not concentrated in any one general factor, CMV is not expected to influence the results from quantitative research.
Subsequently, a covariance based Structural Equations Model (CB-SEM) approach is used to test the model. The choice of this statistical analysis is justified on a number of grounds (Hair, Ringle, & Sarstedt, 2011;Hair, Sarstedt, Ringle, & Mena, 2012): first, CB-SEM is particularly suited for testing and confirming well-founded theoretical Although international tourists only represent around the 10.0% of the tourists visiting Cantabria every year, this group is overrepresented in the sample to obtain a more reliable estimation of the research model. models, as is the case in this research; second, it allows one to conduct confirmatory factor analysis (CFA), to test the psychometric properties of the measurement instruments (reliability, and convergent and discriminant validity) according to the two-stage procedure proposed by Anderson and Gerbing (1988); third, it allows one to analyze the fit between the theoretical model proposed and the data (goodness-of-fit indexes); fourth, CB-SEM can be used to develop multi-group analysis, thus allowing one to compare the results obtained for different samples, as is the case in this research (national tourists and international tourists). Particularly, CB-SEM analyses were run using a robust maximum-likelihood estimation procedure using EQS 6.1 software, in order to avoid problems of non-normality of the data. First, the measurement model was estimated with CFA to test the psychometric properties of the measurement scales (reliability and validity). Next, the structural model was estimated in order to contrast the research hypotheses. Lastly, the effect of cultural distance as a control variable is tested through a multi-group analysis (national tourists versus international tourists), in order to check if the relationships established in the research model are affected by the cultural distance between tourists and destination.

Confirmatory factor analysis
The results obtained in the CFA support the reliability and convergent validity of the measurement scales in both samples of tourists (see Tables 3 and 4). The fit criteria indicate the extent to which the factorial model fits the empirical data. In particular, there are three main classes of fit criteria: measures of absolute fit, measures of incremental fit, and measures of parsimonious fit (Hair, Anderson, Tatham, & Black, 2010). In this case, the statistics, given by EQS 6.1 software, are widely used in the SEM literature (Hair et al., 2010): Bentler-Bonett Non-Normed Fit Index (BBNNFI) and Root Mean Square Error of Approximation (RMSEA) are used for the measurement of overall model fit; Incremental  Tables 3 and 4 confirm that, for both samples of tourists, the BBNNFI, IFI, and CFI statistics exceed or are very close to the recommended minimum value of 0.9. Similarly, in both cases RMSEA is located within the maximum limit of 0.08, and Normed χ 2 takes a value clearly under the recommended value of 3.0 (Hair et al., 2010). The reliability of the measurement scales is evaluated using Cronbach's alpha and composite reliability coefficients (Bagozzi & Yi, 1988). In general, the values of these statistics are clearly above the required minimum of 0.7 (Hair et al., 2010), thus supporting the inner reliability of the constructs (Tables 3 and 4). Only in the case of the affective image scale for the sample of national tourists are the coefficient values slightly under the recommended levels. However, given the good values obtained in the sample of international tourists, we have decided to maintain the same scale of affective image for both samples of tourists in order to allow an exact comparison between the results obtained.
Finally, none of the confidence intervals for pairs of latent constructs include 1.0 (Anderson & Gerbing, 1988), thus supporting the discriminant validity of the model in both samples (see Tables 5 and 6).

Hypothesis testing
The model was estimated in both samples of tourists in order to test the research hypotheses. A first estimation of the structural model showed that some relationships proposed in the research model were not significant for each of the samples considered (i.e. the relationship between cognitive image and quality in the sample of national tourists (hypothesis H3), and the relationship between awareness and affective image in the sample of international tourists (hypothesis H2). Accordingly, and following the model development approach proposed by Hair et al. (2010), the original model was reformulated to exclude the non-significant relationships for each sample. The results obtained for the respective re-specified structural models are summarized in Figures 2  and 3. The empirical results indicate that the goodness-of-fit indexes obtained are within or very near to the recommended values, thus confirming that the model adequately fits the data in both samples. In particular, for both samples the BBNNFI, IFI, and CFI statistics exceed or are very close to the recommended minimum value  of 0.9, RMSEA is located within the maximum limit of 0.08, and Normed χ 2 takes a value clearly under the recommended value of 3.0 (Hair et al., 2010).
Regarding the research hypotheses, the relationship between quality and satisfaction (standardized coefficient = 0.61, p-value < 0.05 for national tourists, and standardized coefficient = 0.63, p-value < 0.05 for international tourists), as well as the relationship between satisfaction and loyalty (standardized coefficient = 0.51, p-value < 0.05 for national tourists, and standardized coefficient = 0.64, p-value < 0.05 for international tourists), are positive and statistically significant in both samples, thus supporting hypotheses H6 and H7. Therefore, a strong link between "quality-satisfactionloyalty" is empirically demonstrated in this research. In addition, concerning the drivers of perceived quality, it is found that the influence of affective image on quality is positive and significant in both samples (standardized coefficient = 0.93, pvalue < 0.05 for national tourists, and standardized coefficient = 0.58, p-value < 0.05 for international tourists), which support hypothesis H4. However, as previously established, the effect of cognitive image on quality is significant only in the sample of international tourists (standardized coefficient = 0.32, p-value < 0.05), so hypothesis H3 is only partially confirmed. Next, the hypothesis H5 is supported since cognitive image positively influence on affective image in both samples (standardized coefficient = 0.62, p-value < 0.05 for national tourists, and standardized coefficient = 0.65, p-value < 0.05 for international tourists). Finally, it is found that the relationship between Table 6. Confidence interval for the correlations between pairs of latent variables (international tourists). awareness and cognitive image is positive in both samples (standardized coefficient = 0.32, p-value < 0.05 for national tourists, and standardized coefficient = 0.32, p-value < 0.05 for international tourists), supporting hypothesis H1. However, as previously established, the effect of awareness on affective image is significant only in the sample of national tourists (standardized coefficient = 0.26, p-value < 0.05) so hypothesis H2 is only partially confirmed.

Multi-group analysis: national vs international tourists
A multi-group model was applied to analyze if all the causal effects included in the theoretical model are statistically different or not depending on the cultural distance between tourists and destination. Specifically, the multi-group models allow analyzing the equality of the parameters considered in a structural model between different groups (Bentler, 2006;Byrne, 1994). Thus, the multi-group models allow evaluation of the structural parameters for each sub-group, as well as compare the obtained value for each causal relation using the LM Test. Such a statistic requires the causal model to include aset of restrictions to equal the structural parameters of all groups being analyzed. In particular, the differences among groups using the statistics χ 2 with a degree of freedom are analyzed comparing the restricted and non-restricted models. The results of the multi-group analysis are summarized in Table 7.
The goodness-of-fit indexes of the multi-group model are clearly within the recommended values, thus confirming the adequacy of the research model to the data for the two samples analyzed. Once the goodness of fit of the model was confirmed, we analyzed if the moderating effect of cultural distance between tourists and destination on the causal effects included in the theoretical model is statistically significant. Thus, the structural coefficients obtained for the two groups identifiednational tourists vs international touristswere compared. Likewise, the results of the LM Test were analyzed for all causal relationships by considering the equality restriction among the parameters applied to each group.
The results from the Lagrange Multiplier test (LM test) show that the only significant difference between national and international tourists is related to the causal effect of tourist satisfaction on loyalty (p-value < 0.05). Specifically, the satisfaction with the destination experience has a stronger influence on intention to revisit the destination and willingness to recommend it to other people for the group of international tourists. In contrast, the chain of effects that lead to their satisfaction with the destination is very similar for both groups.

Conclusions
On the basis of the theory of customer-based brand equity, we establish a model that include the causal interrelationships among the dimensions of destination brand equity (i.e. awareness, image, perceived quality and loyalty) and the role of tourist satisfaction within this chain of effects. Additionally, in the development of our research model, we consider that destination image is integrated not only by cognitive associations, but also by affective ones. Consequently, we examine the effects of both dimensions of destination image separately as well as the interrelation between them. Finally, it is very important to emphasize that the theoretical model is empirically tested in two samples, national and international tourists. This approach allows identifying possible differences in the interrelations between the dimensions of brand equity because of the different cultural distance with the tourist destination.

Theoretical implications
Our findings can be discussed by comparing and contrasting some related works. First, in line with Keller (2001) and the evidence recently obtained in tourism research, our study confirms the existence of a chain of effects among the dimensions of destination brand equity. However, this chain of effects is narrower and longer that the ones found in previous studies. More concretely, in contrast to several previous papers which test the interrelationships among the proposed dimensions (Bigné et al., 2013;Kladou & Kehagias, 2014;Pike et al., 2010), our study demonstrates that the causal relationships between destination awareness, image, perceived quality and loyalty is sequential, so that awareness has a direct effect on destination image, which, in turn, is the direct determinant of perceived quality. In consequence, a higher brand awareness attributed by tourists to the destination will give place to a better image, which will then lead to a higher perceived quality of the destination, and consequently to a higher loyalty.
Second, one of the most important dimensions of destination brand equity is image (Konecnik & Gartner, 2007). With regard to this, and in contrast to previous works that have measured destination image as a single constructi.e. social image and self-image (Bigné et al., 2013;Boo et al., 2009;Kladou & Kehagias, 2014;Pike & Bianchi, 2016;Pike et al., 2010)we considered two types of associations or image in our model of destination brand equity: cognitive image (i.e. beliefs regarding the characteristics of the destination), and affective image (i.e. feelings toward the place). This approach let us delve further into the multidimensional nature of destination image and, consequently, provide more exact knowledge about the influence of this variable on the formation of destination brand equity. Thus, we find that both dimensions of destination image are interrelated (in particular, affective image is positively influenced by cognitive image), so it can be concluded that the affective and cognitive associations together contribute to the representation of the destination in tourists' mind. However, our empirical evidence shows that, in the formation of quality judgements during the destination experience, the affective image is a more powerful driver than the cognitive image. This finding may be because the tourist experience, in contrast to other consumption contexts, has a much higher emotional content for individuals, which leads to a greater role of the emotions or feelings in their subjective evaluations. Third, other theoretical contribution of this research is related to the effect of tourist satisfaction within the chain of effects among the dimensions of destination brand equity. Thus, while satisfaction has been widely studied in the literature on marketing and tourism, the effect of this variable had been very scarcely analyzed in the field of brand equity in tourism (Bigné et al., 2013). In this sense, the evidence obtained supports that satisfaction influences on loyalty towards the destination. Thus, once tourists have visited the destination, satisfaction appears as a key variable for brand equity, acting as a strong driver of loyalty in terms of the intention to revisit the destination and the willingness to recommend it to other people.
Finally, it is necessary to emphasize that the theoretical model was tested by considering two different samples: national and international tourists. In particular, our results offer support for the relationships of "quality-satisfaction" and "satisfaction-loyalty" in both samples. Consequently, a strong link between "quality-satisfaction-loyalty" is evidenced regardless of the origin of tourists. However, in the comparison between the two types of tourists, the difference regarding the role of the cognitive and affective image in the formation of quality judgments is remarkable. For international tourists, the cognitive associations significantly have a higher influence on the perceived quality of the destination, but the affective ones have a lesser effect in the formation of this judgement (in comparison to national tourists). This finding may be due to the international tourists' reduced emotional connection with the destination, which would be a consequence of an increased cultural distance with the place in comparison to the national tourists. Therefore, an increased distance between the tourist's cultural values and the culture that defines the destination could provoke a diminished contribution of affective image to the perceived quality of the tourist destination.

Managerial implications
These findings have important implications for attracting and fostering the loyalty of tourists. In this sense, given that brand awareness seems to be the "the ticket to enter the market"  and the first stage in the formation of brand equity, destination marketing organizations should first focus their efforts on fostering their brand awareness. Particularly, if a tourist destination is little known in the target market, traditional campaigns and social communication conducted by the destination should stress the brand name to make it more recognizable and renowned. This strategy is especially relevant in promoting a destination in international markets where tourist destinations are usually less well known to individuals. According to our findings, international tourists will form in part their quality judgements based on their cognitive associations with the place, which are positively influenced by their awareness of the destination.
Destination communication should also contribute to build a positive and consistent image (brand associations) based both on cognitive and affective issues. Cognitive image is usually based on the main resources and attractions of the destination that are demanded by target segments, and for which the destination has competitive advantages. For example, Cantabria, a small coastal region in the north of Spain with an Atlantic mild weather, would focus on natural resources, heritage and gastronomy, instead of the traditional "sun and beach" image predominating in Spain. According to our findings, affective associations would have a greater role in the sequence "quality-satisfaction-loyalty" than cognitive ones, so destination-marketing organizations should also focus on creating and/or reinforcing the affective image of the place within its positioning strategy. In consequence, positioning should be largely based on an amalgam of feelings and emotions (for example, pleasure, fun, excitement or amazement) that the destination is able to evoke among tourists on the basis of its resources and capabilities, and also taking into account the motivations of their different types of visitors.
Given that perceived quality and tourist satisfaction are key determinants of loyalty towards a destination, it is also very important that destination-marketing organizations conduct proactive and long-term management of both variables. In this sense, a good starting point would be to make a continuous assessment of the perceived quality of the destination and tourist satisfaction, with periodical research to control that these variables are within the desired levels. Additionally, communication campaigns should make a destination attractive for tourist but without generating unreachable expectations, which are based on an unreal image of the destination. Thus, according to the performance expectations theory, high perceived quality will positively influence tourist satisfaction as far as it matches or exceeds expectations. Finally, given the importance of perceived quality and satisfaction as determinants of tourist loyalty, it is recommended to implement customer recovery mechanisms in case of low perceived quality and satisfaction. Thus, destination marketing organizations and tourism companies should build the appropriate channels (e.g. physical offices, telephone service or web pages) to facilitate tourists' complaints and to provide a quick response in order to alleviate or solve quality problems or other causes of tourist dissatisfaction.
Finally, destination managers can increase the value of loyal tourists if they explicitly encourage them to revisit the destination and to recommend it to other people. On the one hand, future revisits of satisfied tourists can be fostered with direct marketing campaigns, reminding them of their previous experiences and informing them of new attractions and experiences. In some cases, promotional incentives such as discount coupons or special attention can be used to encourage revisits. On the other hand, loyal tourists can be a great source of positive word-of-mouth as they are willing to recommend a destination to other people. While, in the past, this phenomenon was limited to friends and family, information and communication technologies now allow electronic word-of-mouth, which has a global impact. Thus, destination marketing organizations and tourism companies can take advantage of these technologies, encouraging loyal tourists to post opinions and ratings on recommendation websites (for example, TripAdvisor) and to upload contents on social networks (for example, Facebook) and specialized platforms (for example, YouTube or Instagram). This effect can be amplified if destination marketing organizations and firms have their own profiles and sites on recommendation websites and social networks and use them to share and promote the electronic word-of-mouth from loyal tourists.

Limitations and further research
Despite the rigorous methodology used in the design and development of the empirical research, this study has several limitations. The fact that this research focuses on a small region in the north of Spain could limit the generalization of the results. Therefore, it would be very interesting to test the model in other destinations with different degrees of brand awareness, cognitive and affective images and perceived quality. Nevertheless, Spain should be a relevant benchmark for the understanding of the relationship among the dimensions of destination brand equity as it is one of the major tourist destinations in the world.
Additionally, it would be interesting to continue studying the dimensionality of the cognitive and affective images of a destination, trying to isolate specific image dimensions applicable to any place but that could have a different effect depending on the type of tourist destination. For example, natural resources or heritage could be dimensions of cognitive image present in any destination but that would have a different impact on perceived quality or satisfaction for city tourism and mountain tourism. Similarly, the affective perception that a destination is quiet versus exciting could have a positive or negative effect on tourist satisfaction depending on travel motivations.

Disclosure statement
No potential conflict of interest was reported by the authors.