Impact of Samarkand’s Destination Attributes on International Tourists’ Revisit and Word-of-Mouth Intention

: Samarkand, Uzbekistan is a relatively unknown but emerging tourism destination. The city was once a major trading hub along the Great Silk Road. The once great city of Samarkand was hidden from the world tourism market, only to emerge after the country gained its independence in 1991. Since then, the artistically beautiful and historically-rich monuments have been attracting tourists from around the world. However, Samarkand has not been gaining much research attention at the same rate as international tourist arrivals. Thus, this study aims to comprehensively assess the city’s destination attributes and their impact on international visitor loyalty behaviors. The literature review process identiﬁed 16 destination attribute dimensions. Additionally, cognitive and a ﬀ ective evaluation, satisfaction, revisit intention, and word-of-mouth intention were identiﬁed and added to form the study model. A total of 223 international tourist samples were collected in Samarkand to validate the proposed model. The ﬁndings revealed 10 attributes to be consistently rated by the samples. Subsequently, the attributes were signiﬁcantly inﬂuencing all subsequent hypothesized relationships. Both theoretical and practical implications are also discussed.


Introduction
Although Uzbekistan can proudly profess a millennia's old historical heritage, and a long pilgrimage tourism history, the modern development of its tourism industry is relatively young when compared to other countries [1][2][3]. This is because Uzbekistan only started to develop and promote its own tourism industry after the country gained its independence in 1991 [2]. Despite that, the tourism industry has been growing rapidly in the last years [4]. In one year, from 2017 to 2018, international arrivals increased by 2.3 times, from 2.8 million to 6.4 million [5]. In the latest statistics, foreign visitor arrivals for January 2020 was 538,360, a 23.1% year-on-year increase [6]. Given the benefits tourists may bring, the Uzbek government has implemented various policies to ensure a continued and sustained development of its tourism industry [4]. The majority of tourists are clustered within the four major touristic destinations. They are Samarkand, Bukhara, Khiva, and Tashkent [2]. Among the four destinations, Samarkand plays a major role, both in terms of its touristic appeal and as a transportation hub linking Tashkent and the further two historical cities of Bukhara and Khiva. Figure 1 shows the map of Uzbekistan and the location of Samarkand [7]. The national tourism authority, Uzbektourism, estimated 37.7% of Uzbekistan's touristic resources are located within Samarkand city and the Samarkand region [4]. Specifically, 780 architectural, 1530 archeological, 36 historical, 225 monumental art, and 71 sculptural resources. Thus, it is no surprise that Samarkand is known for its cultural-historical tourism [1,4,8]. Most of Samarkand's historical resources are linked to how the city was one of the major commercial centers along the Great Silk Road [8]. An analysis of packaged tours offered by travel agencies also revealed that almost all promotions are linked to the Great Silk Road [3]. However, Samarkand is believed to have a wider range of resources and tourism products on offer [2]. Complementing the culturalhistorical tourism, Samarkand also offers appealing ecotourism, pilgrim tourism, and culinary tourism products to visitors, both domestic and international [3,4]. Today, the tourism industry is undoubtedly the most important industry of Samarkand city [8,9]. The tourism industry is often described as complex, in that it needs all stakeholders to individually perform, but tourists perceived the overall experience holistically [10][11][12]. Therefore, a comprehensive analysis of the various attributes of a destination would offer critical insights for effective management and marketing, as well as long-term sustainable development.
The term 'destination' in the tourism context often describes more than just a geographic location or its boundaries. A destination then refers to an area that includes a collection of different natural attributes, features, attractions, and services that tourists may choose to visit and experience [13][14][15]. Thus, the given definition implies that within a destination, there are different attributes a tourist perceived during a visit [13]. Additionally, different experience providers collectively contribute to the overall experience of a tourist [11]. Although the overall satisfaction or dissatisfaction is often evaluated holistically, an individual tourist still perceives and evaluates the different cues, or attributes, individually during the trip [16,17]. Moreover, tourists generally look to experience more than one product at a destination [18][19][20]. For example, hotels, activities, restaurants, and attractions are some of the typical experiential providers at a tourist destination [11,21,22]. As a result, each provider relies on one another to generate satisfactory experience to the same tourist [13]. When one experience provider fails to produce a quality experience, the overall satisfaction level or image perception may be drastically diminished [23,24]. Consequently, dissatisfaction may negatively impact future intentions [12,25,26]. Therefore, it is critically important to measure, not just the overall satisfaction level, but also the various destination attributes. Assessing each component or attribute's performance level may help demonstrate the destination's specific area of strengths and weaknesses [13].
Moreover, the currently available research on tourist behaviors in Samarkand, the leading destination within Uzbekistan, is lacking. In general, the few available research on the tourism industry of Uzbekistan generally looks at the entire country rather than specific destinations [2, [27][28][29][30][31][32]. Among them, the focus has mostly been on the macro view of tourism development [3,4,[28][29][30]. Furthermore, two available streams of research on Samarkand has either been related to the conservation and management of heritage assets [1,33,34] or religious pilgrimage [1,32,35]. However, The national tourism authority, Uzbektourism, estimated 37.7% of Uzbekistan's touristic resources are located within Samarkand city and the Samarkand region [4]. Specifically, 780 architectural, 1530 archeological, 36 historical, 225 monumental art, and 71 sculptural resources. Thus, it is no surprise that Samarkand is known for its cultural-historical tourism [1,4,8]. Most of Samarkand's historical resources are linked to how the city was one of the major commercial centers along the Great Silk Road [8]. An analysis of packaged tours offered by travel agencies also revealed that almost all promotions are linked to the Great Silk Road [3]. However, Samarkand is believed to have a wider range of resources and tourism products on offer [2]. Complementing the cultural-historical tourism, Samarkand also offers appealing ecotourism, pilgrim tourism, and culinary tourism products to visitors, both domestic and international [3,4]. Today, the tourism industry is undoubtedly the most important industry of Samarkand city [8,9]. The tourism industry is often described as complex, in that it needs all stakeholders to individually perform, but tourists perceived the overall experience holistically [10][11][12]. Therefore, a comprehensive analysis of the various attributes of a destination would offer critical insights for effective management and marketing, as well as long-term sustainable development.
The term 'destination' in the tourism context often describes more than just a geographic location or its boundaries. A destination then refers to an area that includes a collection of different natural attributes, features, attractions, and services that tourists may choose to visit and experience [13][14][15]. Thus, the given definition implies that within a destination, there are different attributes a tourist perceived during a visit [13]. Additionally, different experience providers collectively contribute to the overall experience of a tourist [11]. Although the overall satisfaction or dissatisfaction is often evaluated holistically, an individual tourist still perceives and evaluates the different cues, or attributes, individually during the trip [16,17]. Moreover, tourists generally look to experience more than one product at a destination [18][19][20]. For example, hotels, activities, restaurants, and attractions are some of the typical experiential providers at a tourist destination [11,21,22]. As a result, each provider relies on one another to generate satisfactory experience to the same tourist [13]. When one experience provider fails to produce a quality experience, the overall satisfaction level or image perception may be drastically diminished [23,24]. Consequently, dissatisfaction may negatively impact future intentions [12,25,26]. Therefore, it is critically important to measure, not just the overall satisfaction level, but also the various destination attributes. Assessing each component or attribute's performance level may help demonstrate the destination's specific area of strengths and weaknesses [13].
Moreover, the currently available research on tourist behaviors in Samarkand, the leading destination within Uzbekistan, is lacking. In general, the few available research on the tourism industry of Uzbekistan generally looks at the entire country rather than specific destinations [2, [27][28][29][30][31][32]. Among them, the focus has mostly been on the macro view of tourism development [3,4,[28][29][30]. Furthermore, two available streams of research on Samarkand has either been related to the conservation and management of heritage assets [1,33,34] or religious pilgrimage [1,32,35]. However, the tourism industry requires not only heritage attractions to create a memorable tourism experience but all experience facilitators to collectively perform [36,37]. Hence, there is a lack of available research on the destination attributes of Samarkand. At this current stage of development, and given the available tourism assets, the current state of destination attributes of Samarkand is unique. The findings of destination attributes in other destinations may not necessarily be transferable to Samarkand. Thus, there is the need to consider Samarkand's distinctive attributes and verify them using the most current samples of international tourists. As a result, this present research project aims to contribute to the research pool by comprehensively assessing the dimensions of tourism attributes of Samarkand and their impact on tourist loyalty.
The following objectives were designed to help achieve the research aim. First, the dimensions of destination attributes relevant to the current tourism conditions of Samarkand would be identified via a thorough review of the literature. The second objective was to develop a study model to include mediating constructs critical in explaining the hierarchical flow from destination attributes to the loyalty behaviors. Specifically, the study identified the cognitive-affective evaluation constructs and satisfaction as mediators. The cognitive evaluation dimension was identified to represent how tourists are generally understood to be data-and knowledge-driven [38]. Whereas, the affective evaluation variable measures the emotional response at the individual level [39,40]. Furthermore, satisfaction measures the holistic evaluation of the tourism experience [11,16]. The last objective then empirically validated the proposed study model with samples of international visitors to Samarkand.

Destination Attributes
Measuring the various attributes of a destination has matured to include an extensive list of attributes. However, it is still being argued that a universal attribute measurement scale does not exist due to the diverse nature and composition of each destination [11,19]. In practice, researchers often prefer a shorter list of attributes due to the ease of research execution [41]. Whereas, a more comprehensive list of destination attributes can provide more conclusive findings [10,42]. The conceptual model of destination competitiveness developed and introduced by Crouch and Ritchie [10] has been accepted as one of the most comprehensive lists of destination attributes to date [11,43]. The model's development focused on the relationships and dependencies between the stakeholders involved in creating the complete tourism experience at a destination. Hence, the model was developed to measure the competitiveness from the supply perspective [44]. Thus, some attributes are not suitable to be assessed by the tourists, or the demand side. In this present study, the attribute dimensions are based on Kiatkawsin and Han's [11] study which assessed international tourists' perception of destination attributes in Thailand. The study used a scale which assessed all aspects of the tourism experience at a destination.
The first group's attributes are classified as inherited or endowed resources. They include climate, nature, heritage, and local cuisine. The inherited resources are usually the main attractions that pull tourists to the destination in the first place [36,43]. The climate of the destination has always been a key consideration for tourists when choosing a destination. A comfortable climate, in most cases, refers to a warm climate that can dictate the range of tourism activities, preparation, and overall competitiveness of the destination's tourism industry [45]. Nature referred to the natural scenery, including both flora and fauna, of the destination. Natural resources have always been treated as the foundation of most tourism destinations. Natural assets are non-renewable, at the same time, many tourist attractions and products rely on such resources [26]. Thus, a plethora of research has focused on sustainable development and conservation of the natural environment [11]. Heritage and local cuisine are resources that have been manifested by the local people living in the community [42]. They are not just historical resources, but also living elements of the culture [37]. Samarkand is most well-known for its rich historical heritage linked to the Great Silk Road [4,8]. However, Samarkand's local cuisine is often neglected but research has found culinary offerings to be a strong advantage of the city [3].
The second group of attributes is created resources. Although natural resources are fundamentals to popular and successful destinations, it is also possible for destinations to overcome the perceived disadvantage in natural resources by creating or building resources that are still integral to the tourism experience [43,46,47]. Note, created resources are not the same as man-made. Historical buildings, such as the Pyramids of Giza or the Great Wall of China, are man-made but not necessarily created to be tourist attractions. The attribute dimensions under created resources are accommodation, transportation, tourist information, tourism activities, shopping, and entertainment. Accommodation is generally one of the first components when planning a trip, and the quality can play a significant role in creating a memorable tourism experience [41]. Among international travelers, transportation options and quality within the destination country, between different cities, and within the destination may determine the length of visit and the depth of the overall experience [13]. Tourism information covers not only information about the attractions and activities but also critical necessities such as communication (telephones and internet connections) and banks (ATMs, currency exchange). Tourism activities have become much more important today than before. Activities can provide a more immersive experience [12]. Thus, it can leave longer-lasting memories about the trip [26,41]. In recent years, tourists no longer seek just sightseeing and seeing different cultures, and shopping has emerged as an important determinant when choosing a destination [48][49][50]. Entertainment options such as cultural shows and performances can be viewed as an effective tool to add some element of learning to tourists [41,51].
The next group of attributes is supporting factors and resources. Attributes in this group are not necessarily present to directly serve tourists but are general characteristics or skills that may aid or enhance the tourism experience [10,42,46]. In other words, the attributes in this group concern with 'how' the tourism experience is provided and perceived. The four attributes are service quality, accessibility, hospitality, and communication. Service quality reflects how the products and experiences are being delivered to tourists and it is often found to be a key determinant of overall satisfaction [52,53]. The accessibility attribute measures the perceived ease of travel from their home location to the destination [54]. Research has found that accessibility significantly impacts the total number of tourist arrivals and subsequently, the popularity among the mass market [55]. Hospitality refers to the general friendliness towards tourists among the local residents [10]. Local residents are among the three main stakeholders of a successful destination, and the community support of tourism activities is key to a warm and hospitable tourist destination [41,56]. Lastly, communication is how easy the tourists find interacting with people or navigating within the destination. This dimension does not only refers to the language ability of service staff but also covers signs and signages at hotspots [46,47].
The last group of attributes is situational conditions. It consists of perceived value and safety and security attributes. Situational conditions are elements that have the potential to either enhance or reduce the overall satisfaction level [42]. Both attributes are perceived to be outside the control of stakeholders on the supply side as both the perception of value and safety is dependent on the individual tourists [43,47,57]. The actual price itself can hardly be a useful indication of satisfaction as prices are driven by many socioeconomic factors within a destination [58]. However, the concept of perceived value is a result of the difference between the evaluation of services an individual received against the costs of the services [17]. The final attribute, safety and security, is also a factor that measures the psychological perception of risks and danger of a destination. The attribute does not necessarily consider actual safety and security concerns such as crime rates at a destination [43,59]. Although, indicators such as news on political unrests, crime rates, and other related news and information may serve as important cues to the perception of safety and security [60].

Cognitive and Affective Evaluation
Since the conceptualization of satisfaction has been expanded to include an affective element to the traditional cognition-only model, most research today have always operationalized satisfaction as a cognitive-affective system [38,39,61,62]. An evaluation based on cognition was introduced based on the expectation-disconfirmation paradigm. Specifically, consumers evaluate their post-consumption experience and compare that against their pre-consumption expectations. If the experience meets or exceeds expectations, satisfaction is formed [38,39]. The cognition-based model assumes that tourists or consumers are rational individuals whereby they rely on external information to form their own beliefs and judgment [38]. On the contrary, an emotional or affective element was entirely neglected. However, the contemporary pool of research has found that consumers are also evaluating experiences based on their feelings such as pleasure, enjoyment, and excitement [11,26,61].
Several previous studies have validated both cognitive and affective evaluation constructs as a mediator between attributes and loyalty behaviors [11,63,64]. A study involving international tourists in Thailand found a significant mediating effect of positive-negative emotions between destination attributes and revisit intention [11]. In research on wellness spa tourism, the multiple dimension of wellness spa performance was found to have positively and significantly influenced positive affective experience, and negatively impacted negative affective experience [64]. Similarly, halal-friendly attributes of a destination were critical antecedents of affective destination image that eventually impacted revisit and recommendation intention among Muslim tourists in a non-Muslim destination [63]. Based on the conceptualization of both cognitive-affective evaluations and findings of previous studies, hypotheses 1 and 2 were developed.

Satisfaction, Revisit Intention, and Word-of-Mouth Intention
Satisfaction has always been at the center of many consumer behavior research. That is due to its critically important role as an indicator of the destination's success [13,[17][18][19]65]. Although satisfaction is only a holistic measurement of the overall experience, the level of satisfaction can still be useful information to researchers [25,65]. Due to its simplistic form, satisfaction can be used as an indicator of performance, and when comparing between two units of evaluation [66,67]. The level of satisfaction is often determined by using expectations against perceived performance [19]. However, satisfaction by itself offers limited insights into the performance being evaluated [16]. Therefore, satisfaction is often used as a mediator between a more detailed set of performance indicators, such as destination attributes, and post-evaluation behaviors, such as loyalty intentions [11,12,20,25].
Customer loyalty is especially important to businesses due to it being a strong determinant of customer retention and eventual profitability [68,69]. Satisfaction alone does not necessarily lead to customer retention because of interferences such as switching intention and alternative attractiveness [70][71][72]. Therefore, many researchers always measure the level of loyalty in addition to satisfaction. Research on customer loyalty found that different behaviors indicate loyalty; most notable are repurchase/revisit, positive word-of-mouth/recommendation, willingness to pay a premium, and brand advocacy [53,68,70,71,73,74]. However, not all behaviors are applicable to loyalty to a tourism destination context [25,64]. Hence, this present study measures tourist loyalty using two constructs: revisit intention and word-of-mouth intention. The relationships between cognitive-affective evaluation, satisfaction, and loyalty intentions are demonstrated in several previous studies [11,12,18]. In a rural tourism context, a study found significant relationships of tourists' perceived quality to satisfaction and loyalty behaviors [12]. In sum, this present study proposed a hierarchical flow from destination attributes to cognitive-affective evaluation, then to satisfaction, and eventually tourist loyalty. Hypotheses 3, 4, 5, and 6 were developed to reflect the proposed conceptual framework. Additionally, Figure 2 illustrates the study model.

Measurement Items and Survey Development
The study identified 16 dimensions of destination attributes based on previous literature. Thus, the measurement items for the destination attributes have been consolidated and adapted from previous studies [10,11,41,43,75]. Each dimension consists of three items, as previously used in a previous study in which the research samples were also international visitors [11]. Subsequently, three items were adopted to measure cognitive evaluation and also three items for affective evaluation [40]. The overall satisfaction of tourists in Samarkand was also measured using three items [40]. Lastly, revisit intention and word-of-mouth intention were measured using previously validated items [40,76]. All items were measured using 7-point Likert-type scales. All measurement items used in the study can be seen in Appendix A (table A1).
The initial set of measurement items were reviewed by senior academics to ensure the appropriateness of the items to the research context. Then, demographic questions were included. A brief cover letter was also added to introduce the research topic and to inform the participants of the research's purpose and confidentiality agreement. A pre-test was conducted by members of the general public, industry professionals, and academics. Slight adjustments were made based on their comments, and the survey was finalized. The survey was developed in English, and no translation was done.

Sampling and Data Collection
The research population was international tourists in Samarkand, Uzbekistan. A convenient sampling method was deployed for this study. As Samarkand is a well-established destination among foreign tourists visiting Uzbekistan, there are approachable international tourists throughout the city. Moreover, the data collection period was chosen to coincide with Sharkh Taronalari, an international music festival. During this period, the number of foreign tourists was higher than the average. The data collection period was the summer months of 2019.
The survey was distributed mostly as self-administered questionnaires at tourist hotspots and transportation hubs. For example, the Registan Square, Guri Amir (Tomb of Timur), Siyab Bazaar, Shakhi Zinda Necropolis, Samarkand Airport, and Railway Station. Additionally, after being granted permission, the surveys were distributed to guests at restaurants, cafes, and malls. Most participants completed the surveys without any assistance. However, a small group of participants was aided with the translation by their tour guides. All survey participants were not offered any incentives, and their participation was voluntary. They were also informed they could stop completing the survey at any time.

Sample Profiles
A total of 223 completed responses were collected. However, one response was removed due to

Measurement Items and Survey Development
The study identified 16 dimensions of destination attributes based on previous literature. Thus, the measurement items for the destination attributes have been consolidated and adapted from previous studies [10,11,41,43,75]. Each dimension consists of three items, as previously used in a previous study in which the research samples were also international visitors [11]. Subsequently, three items were adopted to measure cognitive evaluation and also three items for affective evaluation [40]. The overall satisfaction of tourists in Samarkand was also measured using three items [40]. Lastly, revisit intention and word-of-mouth intention were measured using previously validated items [40,76]. All items were measured using 7-point Likert-type scales. All measurement items used in the study can be seen in Appendix A (Table A1).
The initial set of measurement items were reviewed by senior academics to ensure the appropriateness of the items to the research context. Then, demographic questions were included. A brief cover letter was also added to introduce the research topic and to inform the participants of the research's purpose and confidentiality agreement. A pre-test was conducted by members of the general public, industry professionals, and academics. Slight adjustments were made based on their comments, and the survey was finalized. The survey was developed in English, and no translation was done.

Sampling and Data Collection
The research population was international tourists in Samarkand, Uzbekistan. A convenient sampling method was deployed for this study. As Samarkand is a well-established destination among foreign tourists visiting Uzbekistan, there are approachable international tourists throughout the city. Moreover, the data collection period was chosen to coincide with Sharkh Taronalari, an international music festival. During this period, the number of foreign tourists was higher than the average. The data collection period was the summer months of 2019.
The survey was distributed mostly as self-administered questionnaires at tourist hotspots and transportation hubs. For example, the Registan Square, Guri Amir (Tomb of Timur), Siyab Bazaar, Shakhi Zinda Necropolis, Samarkand Airport, and Railway Station. Additionally, after being granted permission, the surveys were distributed to guests at restaurants, cafes, and malls. Most participants completed the surveys without any assistance. However, a small group of participants was aided with the translation by their tour guides. All survey participants were not offered any incentives, and their participation was voluntary. They were also informed they could stop completing the survey at any time.

Sample Profiles
A total of 223 completed responses were collected. However, one response was removed due to exhibiting evidence of unengagement. Thus, the final number of responses was 222. Out of the 222 respondents, males accounted for 62.6% (139) and 37.4% (83) for females. The mean age of all participants was 44.6 years old. The respondents reported coming from 38 different countries. Notable nations included France (26), Spain (25), Germany (21), Japan (28), and Italy (13). Most of the participants reported annual household income to be between USD 25,000 and 39,999 (33.8%), followed by between USD 40,000 and 54,999 (24.9%). The majority of the samples hold a higher-education degree. Bachelor's degree and master's degree holders accounted for 41.6% and 32.1%, respectively. An overwhelming majority of the samples are full-time employees (70%), followed by full-time self-employed (13.7%).
In terms of their travel characteristics, more than half reported having visited Samarkand for the first time (64.4%). Their visit companionship was mostly family/relatives (41%) and friends (28.8%). The vast majority visited Samarkand for shorter than seven days. The respondents reported to be spending less than three nights the most (43.2%) and then between four to six nights (38.7%). During the visit, most stay at budget hotels (1-3 star) (43.2%), followed by hostel or guesthouse (30.6%). A full summary of the demographic information can be seen in Table 1.

Data Screening
After the preliminary examination of the collected data by identifying missing values and unengaged responses, the 222 samples were tested for any potential outliers. The produced boxplots did not show any outliers. The test of multivariate normality has shown the skewness to range between −1.405 and 1.346. The kurtosis results ranged from −0.987 to 3.405. Both ranged below the recommended thresholds of absolute value 3 and 5, respectively [77]. Thus, no violation of multivariate normality was detected. The study used IBM SPSS version 23 and IBM SPSS AMOS version 23 for data analysis.

Confirmatory Factor Analysis
The path analysis follows the recommended two-step approach of structural equation modeling (SEM) technique by Anderson and Gerbing (1998) [78]. In the first step, a confirmatory factor analysis (CFA) is conducted to test the proposed study model's goodness-of-fit statistics, validity, and reliability at the measurement model. The goodness-of-fit statistics have shown that at the measurement model, the data has a satisfactory fit to the proposed model (χ 2 = 1338.811, df = 917, p < 0.001, χ 2 /df = 1.560, RMSEA = 0.46, CFI = 0.943, IFI = 0.944) when compared against the recommended thresholds [79,80]. Next, composite reliability was calculated for all latent variables, and the results (ranging from 0.821 to 0.912) were all higher than the recommended threshold of 0.7, providing evidence of reliability [79]. Convergence validity was assessed using the average variance extracted (AVE) scores, which should be higher than the recommended 0.5 [80]. During this process, six attribute dimensions were removed due to low loadings or failed to produce satisfactory AVE. They were climates, safety and security, heritage, perceived value, communication, and hospitality. The AVE values from all constructs ranged from 0.517 to 0.696. Hence, all exceeded the recommended threshold. Discriminant validity can be assessed by comparing the squared correlations against the AVE. The results showed that evidence of discriminant validity generally exists. However, the squared correlation of cognitive evaluation and affective evaluation, as well as affective evaluation and satisfaction were higher than the AVE [80]. Therefore, an additional step is required to further assess discriminant validity by combining items between both latent variables, then compare the chi-square against the base model. The test produced changes in chi-square difference of 39.397 and 69.627. The degree of freedom were both 5, and the p-values were less than 0.001. Hence, evidence of discriminant validity was established. A summary of the confirmatory factor analysis can be seen in Table 2.

Structural Equation Modeling
After establishing the fit, validity, and reliability of the measurement model, the second stage involves testing the proposed hypotheses. The goodness-of-fit statistics of the structural model have also been satisfactory (χ 2 = 1517.722, df = 926, χ 2 /df = 1.639, RMSEA = 0.054, CFI = 0.921, IFI = 0.921, TLI = 0.915, NFI = 0.820, PGFI = 0.690). The proposed model indicated that the antecedents strongly predicted the dependent variables. Revisit intention and word-of-mouth intention were predicted by 42.3% and 57.4% of the total variance, respectively. Satisfaction was explained by 80.5% of its antecedents. Both cognitive and affective were explained by 29.3% and 42.5% of the variance in destination attributes, respectively. Interestingly, revisit intention and word-of-mouth were most influenced by different antecedent variables. Satisfaction produced the largest total impact on revision intention (65%), whereas affective evaluation had the largest total impact on word-of-mouth (83%). The cognitive evaluation had the least total impact on both of the final constructs. The relatively strong total impact and variance explained suggests that the proposed variables were effective in predicting revisit and word-of-mouth intention among the samples.
In terms of the proposed hypotheses, the first hypothesis was supposed by the results. The standardized estimate from destination attributes to cognitive evaluation was 0.532 and the relationship was significant at the 0.001 level. Likewise, destination attribute also significantly predicted affective evaluation (β = 0.653, p ≤ 0.001). Both cognitive (β = 0.158, p ≤ 0.01) and affective (β = 0.830, p ≤ 0.001) evaluation subsequently significantly predicted satisfaction. Lastly, satisfaction was significantly predicting both revisit intention (β = 0.650, p ≤ 0.001) and word-of-mouth intention (β = 0.758, p ≤ 0.001). Thus, all six proposed hypotheses were supported by the results. The summary of the SEM results is presented in Table 3. Figure 3 illustrates the proposed study model and the SEM results.

Indirect-Impact Assessment
The indirect-impact assessment was conducted to test the mediation effect between the variables within the study model. Bootstrapping method with 2000 samples and a 99% confidence level was conducted. The results have shown that the indirect effect of destination attributes on satisfaction (β = 1.139, p ≤ 0.01), revisit intention (β = 0.910, p ≤ 0.01), and word-of-mouth intention (β = 0.887, p ≤ 0.01) were all significant. Hence, the results indicate that cognitive evaluation, affective evaluation, and satisfaction are all partial mediators between destination attributes and the final two constructs. The indirect impact of cognitive evaluation on revisit intention (β = 0.131, p ≤ 0.01) and word-of-mouth intention (β = 0.128, p ≤ 0.01) were also significant. The results imply that satisfaction is also a partial mediator between cognitive evaluation and the two dependent variables. Lastly, the affective evaluation did not produce any significant indirect impact on revisit intention and word-of-mouth intention. Therefore, the results indicated that satisfaction was the complete or full mediator between affective evaluation and both revisit and word-of-mouth intention. Table 4 summarized the results of the indirect-impact assessment.  Note = *p < 0.001, **p < 0.01

General Discussions
The first area of discussion is the attribute dimensions. From the initial 16 attribute dimensions, 10 attributes were found to have produced adequate loadings to form a factor and retained in the final model. The absence of proposed attributes does not imply the negative rating of the attributes but rather reflects the inconsistency in the ratings by the research samples [11]. The inconsistency in ratings could be influenced by a number of factors. For example, it could be that not all tourists in the sample experienced an identical amount of services or visited the same attractions. In other words, due to varying itineraries, not all were exposed to the same products and services. Secondly, inconsistent ratings may reflect the volatility of service quality. Third, the evaluation of the experience is agreed to be subjective on each individual background. Although attributes were removed from the final, it does not necessarily mean they were rated negatively. Negative but consistent results would still form a factor, albeit the loading would be negative. Nevertheless, inconsistencies of attribute ratings still highlight areas requiring management of attention.
The first most notable attribute that did not form a factor was heritage. Given that previous research had universally identified cultural heritage to be the primary attraction of Samarkand [2, 3,8], it is surprising to find heritage attributes to be omitted. Besides, the climate was also removed from the final model. The data collection was completed within a period of two weeks. Hence, seasonality should not be the cause. Still, among the inherited resources, nature and local cuisine were both positively rated. The results should not be surprising due to the relative lack of urbanization; the natural environment may still be in pristine condition. The local cuisine has been mentioned by previous researches as a potential area for development and promotion [3,4]. These findings already show a promise for future development in this area. In terms of the created resources, all six attributes were retained in the final model, implying the generally satisfactory evaluation of tourism-related services.
On the other hand, two supporting factors were retained: service quality and accessibility. Samarkand city is one of, if not the most visited, tourist destination in Uzbekistan. Therefore, transportation to and from the city has already been well established. Additionally, service quality was also rated consistently by the samples, indicating the relatively high standard of training among the service staff. The indication of high-quality training may explain the absence of hospitality and communication attributes. Hospitality measured the perceived friendliness of locals toward tourists, whereas communication measured the perceived ease of communication; while at the destination, both were rated inconsistently. However, service quality was consistent and positive. The findings may imply that although the locals may not always present themselves to be friendly and able to

General Discussions
The first area of discussion is the attribute dimensions. From the initial 16 attribute dimensions, 10 attributes were found to have produced adequate loadings to form a factor and retained in the final model. The absence of proposed attributes does not imply the negative rating of the attributes but rather reflects the inconsistency in the ratings by the research samples [11]. The inconsistency in ratings could be influenced by a number of factors. For example, it could be that not all tourists in the sample experienced an identical amount of services or visited the same attractions. In other words, due to varying itineraries, not all were exposed to the same products and services. Secondly, inconsistent ratings may reflect the volatility of service quality. Third, the evaluation of the experience is agreed to be subjective on each individual background. Although attributes were removed from the final, it does not necessarily mean they were rated negatively. Negative but consistent results would still form a factor, albeit the loading would be negative. Nevertheless, inconsistencies of attribute ratings still highlight areas requiring management of attention.
The first most notable attribute that did not form a factor was heritage. Given that previous research had universally identified cultural heritage to be the primary attraction of Samarkand [2, 3,8], it is surprising to find heritage attributes to be omitted. Besides, the climate was also removed from the final model. The data collection was completed within a period of two weeks. Hence, seasonality should not be the cause. Still, among the inherited resources, nature and local cuisine were both positively rated. The results should not be surprising due to the relative lack of urbanization; the natural environment may still be in pristine condition. The local cuisine has been mentioned by previous researches as a potential area for development and promotion [3,4]. These findings already show a promise for future development in this area. In terms of the created resources, all six attributes were retained in the final model, implying the generally satisfactory evaluation of tourism-related services.
On the other hand, two supporting factors were retained: service quality and accessibility. Samarkand city is one of, if not the most visited, tourist destination in Uzbekistan. Therefore, transportation to and from the city has already been well established. Additionally, service quality was also rated consistently by the samples, indicating the relatively high standard of training among the service staff. The indication of high-quality training may explain the absence of hospitality and communication attributes. Hospitality measured the perceived friendliness of locals toward tourists, whereas communication measured the perceived ease of communication; while at the destination, both were rated inconsistently. However, service quality was consistent and positive. The findings may imply that although the locals may not always present themselves to be friendly and able to communicate well, the effort and goodwill were still perceived by the tourists. Lastly, situational conditions were both removed from the final model, indicating that both perceived value, and safety and security, were evaluated differently among the research samples. Given that perceived value can be highly subjective depending on the individual tourists' background, the results were not surprising.
Furthermore, the destination attribute was a significant antecedent to both cognitive and affective evaluations. Interestingly, the variance explained score was higher for affective and cognitive as well as receiving a stronger correlation score. Moreover, affective evaluation produced approximately four times higher amplitude of influence on satisfaction than cognitive evaluation. The results imply that among the international tourists in Samarkand, the emotional response is a much more important determinant of satisfaction. In other words, tourists rely more on their feeling at an individual level when evaluating their experience. Given that tourism experience can be classified as an elective expenditure, it is expected to be evaluated similarly to luxury consumption. Thus, the utilitarian functional value becomes less important compared to hedonic value. Both revisit intention and word-of-mouth intention were strongly predicted by satisfaction and all other antecedents. Nonetheless, word-of-mouth intention still produced a slightly higher total variance explained score than revisit intention. Due to the higher costs and barriers associated with revisit and other psychological influences such as variety-seeking, it is expected that intention to recommend and sharing the experience were more likely.

Implications
Practically, the research findings have illustrated a number of noteworthy areas for interested stakeholders within the tourism sector in Samarkand. Many businesses and service providers measure the level of satisfaction of their guests on a regular basis via either their own monitoring surveys or from online user-generated content. However, businesses usually lack access to research on the destination level. Therefore, research of this type not only provides an overview of the strengths and weaknesses but also provides an update on the current state of the destination. In general, attributes that were retained in the final model would be considered the strengths, whereas those that were removed require further attention. A business may identify attributes that their business falls under and check how those attributes have been rated by tourists; bearing in mind that the strengths or weaknesses may not be dependent solely on their performance.
Furthermore, managers can look to combine their offering with other attributes to create a more inclusive tourism experience. For example, local cuisine and shopping were identified as strengths of Samarkand. Local restaurants may want to package some of their famous dishes and food items for tourists to purchase and take home. Alternatively, businesses may want to address critical components to the overall tourism experience but did not produce consistent enough results. For instance, Samarkand's rich historical monuments were evaluated poorly and inconsistently due to poor interpretation or storytelling. Therefore, improving the quality of tour guides through extensive training, or setting up signs and signage explaining the history of such monuments may help improve tourists' evaluation of the heritage dimension.
Another critical finding for businesses and site managers was the relative importance of the affective evaluation construct. As a strong mediator between destination attributes and satisfaction, it highlights the role of emotional or hedonic value among travelers. Promotional messages and services that can persuade positive emotional benefits when visiting Samarkand should be effective to attract new and retain existing tourists. Therefore, it is recommended that marketers should focus on the affective motivation for traveling and memorable experiences. A study on memorable tourism experiences highlighted that not all experiences encountered by tourists are being remembered. Tourists choose to remember and form a longer-lasting memory than those they do not remember, many of which are affective-based.
Theoretically, this research project contributed in a few ways. First, this study was one of the only few available comprehensive assessments of a tourist destination in Uzbekistan. A study-specific to Samarkand city could hardly be found currently, especially in English. As an emerging destination in the global tourism market, Samarkand city and its tourism stakeholders would benefit from such an independent assessment. The study has also compiled an extensive list of destination attributes to be utilized in replica research in the future. Other constructs included in the study model have also been verified to explain the post-visit evaluation process of tourists. Hence, the conceptualized model should be able to provide a basis for future updates and extensions.

Limitations and Recommendations for Future Research
The study has a few limitations to be addressed. This research has only sampled a smaller number of international visitors relative to the total number of international arrivals. Thus, readers should view the findings of this research with caution regarding the generalizability of the entire population of international visitors. Secondly, tourists mostly have their unique trip itinerary, and not all visitors experienced the same amount. Therefore, the results should be interpreted with this limitation in mind. Although the study attempted to comply with a comprehensive list of destination attributes, the tourism industry is ever-evolving and new products and services are always being introduced as well as new attractions being developed. Hence, the study only represents a specific period of Samarkand. It is also recommended for future research projects to periodically measure tourist attribute evaluation for the most up-to-date results. The study model was first conceptualized for this present study and additional validation could only further strengthen its robustness. Lastly, the study only focused on international visitors. However, domestic visitors also account for a significant portion of the total tourists in Samarkand. Therefore, future research may like to compare tourist evaluations between domestic and international tourists to provide more useful knowledge to both businesses and site managers.
Author Contributions: Conceptualization, S.J. and K.K.; formal analysis, K.K.; data curation, S.J. and A.K.; writing-original draft preparation, S.J. and K.K.; writing-review and editing, K.K. and A.K.; supervision, K.K. All authors have read and agreed to the published version of the manuscript.  Table A1. Cont.

Tourist information
Availability of tourism guidance and information (0.846) Financial institutions and currency exchange facilities (0.908