Understanding memorable tourism experiences and behavioural intentions of heritage tourists

This study aims to investigate the interplay of visitor engagement, authenticity, and destination image in driving revisit and electronic word of mouth (eWOM) intentions of heritage tourists through the mediating role of Memorable Tourism Experiences (MTE). The data for this research were collected from tourists in the UNESCOlisted heritage city of Kashan, Iran. Using a convergent parallel mixed methods approach, the study’s findings highlighted the importance of MTE as a mediator of these interrelationships. The results also identified the positive direct and indirect effects of visitor engagement on revisit and eWOM intentions. The indirect effects of authenticity on revisit and eWOM intentions through MTE were also significant. The findings also showed the positive direct and indirect effects of destination image on eWOM intention, with the indirect effect on revisit intention being significant. The practical implications of the study and potential future directions for research are also discussed in the conclusion section.


Introduction
Heritage tourism offers experiences that involve visiting or engaging with "places, artefacts and activities that authentically represent the stories and people of the past and present" (Hargrove, 2002, p. 10). Many people search for unique travel experiences that blend culture, education, entertainment, and authenticity (Garrod & Fyall, 2000, 2001Hall & Zeppel, 1990). Cultural heritage tourism is a significant global tourism market (Poria, Butler, & Airey, 2003;Richards, 2018). The United Nation World Tourism Organization (UNWTO) (2015) regards this segment of tourism as 'a major element of international tourism consumption' and estimates that four out of ten tourists select their destination based on its cultural heritage offering. Richards (2018) states that there has been a recent shift in the nature of cultural tourism demand from purely quantitative growth of cultural tourism demand toward qualitative shifts in the nature of that demand, with a particular emphasis on the increasing quest for 'cultural experiences'. Thus, cultural heritage tourism experiences has potentially emerged as a significant element of tourists' memorability (Lee, 2015) which makes heritage tourism, like many other recreational and tourism practices, a form of experiential consumption (Garrod & Fyall, 2000, 2001Hall & Zeppel, 1990;Richards, 2018). An improved understanding of the tourist experience and behavioral intentions at heritage sites and destinations is therefore essential to better meet the expectation of this market (Richards, 2018;Wu & Li, 2017).
As an overarching concept in the literature on tourist experience over the recent years, Memorable Tourism Experiences (MTE) has received limited attention within the backdrop of heritage tourism. The experiences of heritage tourists and its structure are not adequately addressed in the extant literature. Additionally, as Chen and Rahman (2018) noted, the knowledge on the potential influencers of MTE in the context of heritage tourism is limited. This emphasizes the need for further inquiry to get a better and broader understanding of heritage tourists' experiences and to advance understanding of MTE in a heritage tourism setting.
To address these gaps in the literature, this study develops and empirically examines an integrated model of visitor engagement, authenticity, and destination image in driving behavioural intentions of heritage tourists which is represented by revisit and electronic word of mouth (eWOM) intentions. In addition, the study tests the mediating effect of MTE in such interrelationships. Given the experiential nature of heritage tourism (Richards, 2018) and that tourists willing for generating memorable experiences visiting heritage attractions at destination, this study therefore sought to identify MTE as a mediator in enhancing behavioural intentions.
This research was conducted using data collected from tourists in the UNESCO listed heritage city of Kashan, Iran. In Iran, as in most other parts of the developing countries, cultural heritage tourism research has mainly focused on product-driven (Seyfi & Hall, 2018), with little research conducted on visitor experience in general and on memorability in particular. The city of Kashan is a significant heritage destination and has three UNESCO-listed tangible and intangible heritage listings as well as numerous historical buildings (Gannon, Rasoolimanesh, & Taheri, 2020). The city is a major holiday destination for domestic and international tourists (Seyfi & Hall, 2018) and provides an appropriate context for exploring visitor's behaviour in a heritage tourism context.
Given the exploratory nature of the study and the gaps identified in the literature which illustrated the limited knowledge on the structure of heritage tourism experiences and the potential influencers of MTE in the context of heritage tourism, this study follows a parallel database variant of the convergent parallel mixed-methods design with a combination of quantitative (questionnaire) and qualitative (interviews) data collection methods (Creswell & Clark, 2018) to gain a comprehensive understanding of the MTE in the backdrop of heritage tourism. The purpose of the convergent design is "to obtain different but complementary data on the same topic" (Morse, 1991, p. 122) to best understand the research problem. The selection of this approach for our study was not for the purposes of comparing results, instead in this study, our aim was to use the complementary transformed data to gain a more complete understanding of the facets of the MTE in the heritage tourism context.
One of the strengths of this approach is the sperate analysis of both quantitative and qualitative datasets before synthesising the independent quantitative and qualitative results in the final discussion. This variant is particularly useful as all data strands are used in tandem to gain a deeper understanding of a phenomenon (Creswell & Clark, 2018).

Heritage tourism
"Heritage is our legacy from the past, what we live with today, and what we pass on to future generations" (Rodwell, 2008, p. 7). Heritage tourism refers to travel with the main goal of learning about a place's culture and heritage (Poria et al., 2003;Timothy, 2011). Heritage tourism typically relies on "living and built elements of culture and refers to the use of the tangible and intangible past as a tourism resource" (Timothy & Nyaupane, 2009, p. 4). Today, culture presents a key resource upon which travel draws, and "the majority of tourism attractions and destinations around the world are focused on cultural heritage elements" (Timothy, 2011, p. 3). Heritage tourism therefore constitutes a large share of cultural tourism (Seyfi, Hall, & Fagnoni, 2019), and is "one of the largest, most pervasive, and fastest growing sectors of the tourism industry today" (Timothy & Nyaupane, 2009, p. 4). In addition to enabling tourists to generate and strengthen a sense of identity for themselves (or others), heritage tourism is of great significance for many countries particularly in the developing world as an important potential tool for poverty alleviation and community economic development (Garrod & Fyall, 2000;Poria et al., 2003;Seyfi et al., 2019;UNWTO, 2015). Such interest could be seen in the growing number of sites receiving World Heritage Site recognition by UNESCO. Park (2013) state that heritage tourism has gained special interest in countries that have historically relied on 3s offerings (sun, sand, and sea) and are attempting to diversify their offerings. The UNWTO (2015) considers this segment of tourism to be a significant element of international tourism consumption, accounting for around 40% of global tourism. Timothy (2011) even goes further and states that approximately 85% of the general population are estimated to be considered current or potential heritage tourists. Thus, many destinations have paid special interest to their tangible and inflexible cultural assets to gain a competitive advantage in the growingly fierce tourism marketplace. Enhancing cultural heritage tourists' experiences and satisfaction is therefore significant and as Seyfi, Hall, and Rasoolimanesh (2020) argue understanding how tourists produce positive feelings which potentially increase the likelihood of gaining MTE (Zhang, Wu, & Buhalis, 2018). While a number of researchers have elucidated the importance of MTE for the competitive advantage of tourism destinations, there is a limited research on the potential influencers of MTE in the context of heritage tourism. This bias in extant knowledge point to gaps which the present study addresses.

Electronic word of mouth (eWOM) intention
Word of mouth, including its electronic form has been identified as a major driver in decision making (Filieri & McLeay, 2014;Litvin, Goldsmith, & Pan, 2018). Electronic word of mouth (eWOM) is defined as "any positive or negative statements made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the Internet" (Henning--Thurau et al., 2004, p. 39). While the traditional offline WOM depends heavily on face-to-face information exchange, eWOM shapes through sharing experiences regarding products and services through the internet (Litvin et al., 2018). With the advancement of Internet technology and the advent of Web 2.0, eWOM has become increasingly significant. In contrast to traditional WOM, eWOM is "more influential due to its speed, convenience, one-to-many reach, and its absence of face-to-face human pressure" (Sun, Youn, Wu, & Kuntaraporn, 2006, p. 1106. The role of eWOM has been widely recognized in the post-consumption intention (Litvin et al., 2018). According to information adoption theories, consumers change their behaviour in response to comments shared on the Internet (Filieri & McLeay, 2014).
Tourism is regarded as the biggest experience generator in which tourists construct their own specific narratives (Litvin et al., 2018). eWOM is of paramount significance in the tourism industry in general, and particularly in the hospitality industry (Binkhorst & Den Dekker, 2009;Litvin et al., 2018). According to Chevalier and Mayzlin (2006), eWOM is an important tool for people to learn about their own preferences, such as service quality, travel experiences, and food. This emphasizes the significance of memorable experience, which potentially lead tourists to evaluate their experience through social media or other online platforms where experiences are compared, evaluated, defined and exchanged. Rivera, Semrad, and Croes (2015) observed that a memorable experience generates various forms of eWOM which echoes the finding of Semrad and Rivera (2018). Similarly, Serra-Cantallops, Ramon-Cardona, and Salvi (2018) also found that positive emotional experiences positively influence satisfaction and eWOM generation.

Revisit intention
Experience is one of the most powerful predictors of the behavioural intent of a tourist (Tung & Ritchie, 2011) and has a profound effect on memory generation, the primary aim of tourism practitioners (Tsai, 2016). Ajzen and Fishbein (2000) refers to behavioral intentions as perceptions of individuals about what they expect to do in a given situation. Tourist impressions of the travel experience and destination perceptions are the most reliable source of information when it comes to revisit and eWOM intentions . Experiences of positive emotions and mood of a previous trip and feelings of joy can affect an individual's future decision and behaviour (Prayag, 2009;Tsai, 2016). Revisit intention, which refers to an individual's readiness and willingness to make a repeat visit (Prayag, 2009) is an important behavioral intention, i.e. the degree of deliberate commitment that a person can make to execute an action (Tsai, 2016). Tourists' behaviors including their selection of destinations to visit, subsequent evaluation of destination decisions and future behavioral intentions, are related to their willingness to revisit a destination . Many destinations also rely heavily on repeat visits as it commonly costs much less in marketing terms to retain repeat visitors than to attract new ones (Tsai, 2016).

Influencing factors on memorable tourism experiences
The concept of MTE was initially proposed by Kim, Ritchie, and McCormick (2012) and has been defined as "a tourism experience positively remembered and recalled after the event has occurred [that is] selectively constructed from tourism experiences based on the individual's assessment of the experience" (Kim et al., 2012, p. 13). The MTE scale proposed by Kim et al. (2012) includes seven components (hedonism, novelty, local culture, refreshment, meaningfulness, involvement, and knowledge) that are seen as important elements of the tourism experience that affect its memorability for an individual. Understanding and strengthening the recollection of the positive memories of tourists is regarded as a competitive advantage in the contemporary tourism marketplace (Kim et al., 2012;Zhang et al., 2018). This seminal scale has been confirmed in other studies and is widely used to measure tourist experiences Gohary, Pourazizi, Madani, & Chan, 2020;Kim & Ritchie, 2014;Sthapit, Björk, & Coudounaris, 2017;Tsai, 2016;Yu, Chang, & Ramanpong, 2019). This study therefore adopts this framework.

Visitor engagement and MTE
The concept of consumer engagement (CE) has been examined and conceptualized in several disciplines such as psychology, sociology and organizational behavior and has gained popularity in marketing studies in recent time (Brodie, Hollebeek, Jurić, & Ilić, 2011). Yet, its conceptualization has been subjected to varying interpretations (Rasoolimanesh et al., 2019;So, King, & Sparks, 2014). For Brodie et al. (2011, p.9) the CE is "a multidimensional concept subject to a context-and/or stakeholder-specific expression of relevant cognitive, emotional and/or behavioral dimensions". CE has also been viewed as a unidimensional (predominantly focus on only behavioral aspect of customer engagement or multidimensional construct that encompassed both psychological and behavioral dimensions (Brodie et al., 2011;Hollebeek, Conduit, & Brodie, 2016;Rasoolimanesh et al., 2019).
This study adopts the multidimensional conceptualization of visitor engagement proposed by So et al. (2014) which comprises identification, enthusiasm, attention, absorption, and interaction on hospitality services. Customer engagement is a highly important and an essential component of a tourism experience (Rasoolimanesh et al., 2019;So et al., 2014), including enhanced customers' interaction as co-creators of products and services that further enhances customers service experience (So et al., 2014) and optimizes the overall tourist experience (So et al., 2014). Accordingly, empirical studies in tourism suggest that better visitor engagement positively affects the overall visitor experience and can create a higher level of MTE (Chen & Rahman, 2018;Taheri, Jafari, & O'Gorman, 2014). A positive relationship is suggested between the engagement during the visit to cultural attractions and sites and the enhanced memorable experiences (Chen & Rahman, 2018). In their study, Seyfi, Hall, and Rasoolimanesh (2020) also found that visitor engagement with tourist attractions lead to positively formed MTE.
The positive effect of visitor engagement on MTE is in line with the self-determination theory (Ryan & Deci, 2002) which explains travellers' motivation and can be applied to cultural tourism (Chen & Rahman, 2018). Applying this theory, Chen and Rahman (2018) argue that different forms of motivations to participate in cultural tourism result in varying degrees of visitor engagement in cultural tourism. They go on to say that while a higher degree of involvement and awareness of a destination's culture leads to a higher level of MTE, a higher level of visitor engagement is likely to lead to a higher level of MTE.

Authenticity and MTE
Given the experiential nature of tourist services, authenticity is seen as a key construct in studying tourist experiences (Kolar & Ž abkar, 2010;MacCannell, 1976). Similarly, Hargrove (2002) argues that authenticity is a central component of a meaningful heritage experience. Waitt (2000) defined authenticity in terms of feelings of historic and cultural connectedness to the past. From a heritage perspective authenticity is linked to the understanding of specific cultures, places and communities (Hall, 2007), what may be termed intangible heritage, as well the tangible heritage qualities that can be found in an object or structure (MacCannell, 1976). Authenticity is considered as an important antecedent of tourist experience (Chen & Chen, 2010). Empirical studies in tourism also suggest that there is a positive relationship between authenticity and experience when it comes to the post-consumption stage when tourists evaluate their travel experiences Ramkissoon & Uysal, 2010). In their study on dining context, Antón, Camarero, Laguna, and Buhalis (2019) also found that there is a positive link between authenticity and travelers' memorable gastronomic experiences.

Destination image and MTE
Destination image is the sum of the knowledge, beliefs, ideas and overall perceptions that a tourist has about a destination (Baloglu & Brinberg, 1997;Tasci & Gartner, 2007). As a multidimensional construct, destination image is formed by three distinct but hierarchically interrelated components: the cognitive, affective, and conative (Dann, 1996;Tasci & Gartner, 2007). The cognitive dimension refers to tourists' beliefs and knowledge about the attributes or features of the destination (Baloglu & Brinberg, 1997), whereas the affective component refers to the evaluation stage and tourists' feelings about a destination (Tasci & Gartner, 2007). Lastly, the conative dimension involves action, i.e. the actual actions or intention of the tourists to make a repeat visit and willingness to recommend the destination to others on the basis of their cognitive and affective images (Prayag, 2009). This study focuses on the cognitive component of destination image because it is explicitly visible, concise, and tangible (Tasci & Gartner, 2007) and offers knowledge regarding a destination's uniqueness. Studies have explored the influence of pre-visit images on decision making (e.g. Baloglu & Brinberg, 1997;Tasci & Gartner, 2007) and holistic impression of an experience. Of significance to the current study, is evidence of a link between destination image and direct impacts on other critical tourism constructs such satisfaction, destination choice and behavioral intent (Tasci & Gartner, 2007). Previous studies have identified that destination image affects tourist's satisfaction and experience significantly. Kim and Ritchie (2014) explored the destination attributes associated with MTE. Kim (2018) and Zhang et al. (2018) also reported that destination image influences the MTE.

Direct effects
Marketing literature has evidenced that consumer behaviour is mediated by memory, which influences return decisions (Tung & Ritchie, 2011). In tourism, empirical studies have reported that MTE significantly affect individuals' future behavioral intentions (e.g. revisit, re-practice, and WOM communications) (Kim et al., 2012;Kim & Ritchie, 2014). The study of Adongo, Anuga, and Dayour (2015) and Tsai (2016) found that MTE significantly influences behavioural intentions to revisit and recommend. Marschall (2012) states that tourists are more probable to revisit a destination where they have had memorable experiences. Previous studies have examined the effect of MTE on revisit intention in different contexts. For instance, Yu et al. (2019) found tourists intend to revisit places they have good and positive memories of, while Adongo et al. (2015) also suggest that memorable local food experiences affect tourists' satisfaction and intention to recommend. The study of Kim and Ritchie (2014) on tourists visiting Taiwan also highlighted the influence of MTE on tourists' behaviour. Barnes, Mattsson, and Sørensen (2016) and Kim et al. (2012) suggest that longer-term remembered experiences have the strongest impact on tourists' intention to make a repeat visit. Prior studies also suggest that MTE significantly affect individuals' behavioral intentions to spread positive WOM (e.g. Adongo et al., 2015;Kim & Ritchie, 2014;Tsai, 2016). Similarly, Adongo et al. (2015) reported that memorable local food experiences influence tourists' intentions to engage in WOM communications.
Additionally, the favorable links between MTEs and behavioural intentions as represented by revisit and eWOM intentions are supported by the theory of the tourism consumption system. This theory asserts that tourists' evaluation of the destination experience influences their overall destination assessment and behaviour (Woodside & Dubelaar, 2002). Thus, obtaining support from the above discussed literature and the theory of the tourism consumption systems, the following hypotheses are therefore proposed:

H5. MTE has a positive effect on revisit intention
In the marketing literature, consumer engagement has long been viewed as predictive of the intention to repurchase (e.g. Hollebeek, 2011). Studies in the tourism literature also posits that higher levels of tourist engagement with a destination positively affects tourist attitude and intention to re-visit and loyalty towards a destination (e.g. Chen & Rahman, 2018;Rasoolimanesh et al., 2019). This has also been reported in various empirical studies. For instance, Rasoolimanesh et al., 2019 found that tourist engagement has a strong influence on satisfaction and tourists' destination loyalty and revisit intention. A similar observation is also reported by Alrawadieh, Prayag, Alrawadieh, and Alsalameen (2019) who found that tourists engagement strongly influence their eWOM intention.

H7. Visitor engagement has a positive direct effect on revisit intention
Prior empirical studies in the marketing literature have shown that brand authenticity positively influences consumer reactions to brands and consumers' behavioral intent (Oh, Prado, Korelo, & Frizzo, 2019). If a brand perceived as authentic is evaluated more positively it should positively affect the behavioral intent of consumers towards the brand (e.g. repurchase intention, recommending the brand to other people) (Oh et al., 2019). In the tourism literature, authenticity has been viewed as an antecedent of tourism experience and satisfaction (Ramkissoon & Uysal, 2010). Empirical studies have reported a positive relationship between authenticity and tourists experience in the post-consumption stage (e.g. Ramkissoon & Uysal, 2010). For instance, Zhang, Chen, and Hu (2019) reported that a high level of authenticity leads to high quality attributes of food tourism, which in turn contributes to higher degree of tourist satisfaction and loyalty. In their study on dining context, Antón et al. (2019) also found that authenticity significantly influence future tourist' intention to revisit and get engaged in the eWOM activity.
H8. Authenticity has a positive direct effect on eWOM intention H9. Authenticity has a positive direct effect on revisit intention Destination image is regarded as the most essential antecedent of travels' post-consumption behaviour (Tasci & Gartner, 2007). Wu and Li (2017) argue that tourists having a favorable image of a destination will perceive their experiential quality positively. Several studies indicate that the likelihood to recommend a destination for a satisfied tourist is high (e.g. Tasci & Gartner, 2007;Zhang et al., 2018). Empirical studies also suggest a positive link between a favorable destination image and higher revisit and eWOM intentions (Zhang et al., 2018). Today, many travelers share their impressions of a destination through eWOM, which has a positive impact and eventually contributes to increased revisit and eWOM intentions.
H10. Destination image has a positive direct effect on eWOM intention H11. Destination image has a positive direct effect on revisit intention

Indirect effects (mediator)
Given the lack of theoretical unanimity among researchers regarding the components and antecedents of MTE and the complex nature of experience, further research is suggested to explore and enrich understanding of MTE in other contexts (Chandralal & Valenzuela, 2015;Kim, 2018). Sthapit and Coudounaris (2018) asserted that the MTE framework is context-based. Numerous works support the effects of various antecedents, including visitor engagement, authenticity, and destination image on MTE (e.g. Antón et al., 2019;Chen & Rahman, 2018;Kim et al., 2012;Seyfi, Hall, & Rasoolimanesh, 2020;Zhang et al., 2018). Prior studies also acknowledge the effect of MTE on behavioral intentions (e.g. Chandralal & Valenzuela, 2015;Kim, 2018;Kim et al., 2012;Kim & Ritchie, 2014;Tung & Ritchie, 2011). Binkhorst and Den Dekker (2009) argue that a meaningful experience will lead tourists to evaluate their experiences through the eWOM process. Obtaining support from the above discussed literature, Thus, the following additional hypotheses are proposed:

H12. Visitor engagement has a positive indirect effect on eWOM intention through MTE
H13. Authenticity has a positive indirect effect on eWOM intention through MTE H14. Destination image has a positive indirect effect on eWOM intention through MTE As noted above, given the context-based nature of MTE, further scholarly research is suggested to better understand its antecedents and potential influencers in other contexts (Chandralal & Valenzuela, 2015;Kim, 2018) and their behavioral consequences. The interplay between the antecedents of MTE in a heritage tourism context and subsequent effect on the revisit intention. The direct effects of various antecedents, including visitor engagement, authenticity, and destination image have been hypothesized in the literature, suggesting how each influences MTE (e.g. Antón et al., 2019;Kim, 2014;Kim et al., 2012;Seyfi, Hall, & Rasoolimanesh, 2020;Zhang et al., 2018). Furthermore, as noted above, prior studies acknowledge the direct effect of MTE on revisit intention (Kim et al., 2012;Kim & Ritchie, 2014;Yu et al., 2019;Zhang et al., 2018). Thus, the following additional hypotheses are proposed:

Research methodology
A parallel database variant of the convergent parallel mixed-methods was adopted for this study (Creswell & Clark, 2018). The combination of quantitative and qualitative methods offers greater insight into the phenomena under study "when the researcher collects and analyses both quantitative and qualitative data during the same phase of the research process and then merges the two sets of results into an overall interpretation" (Creswell & Clark, 2018, p. 77). Two types of data are used "to examine facets of a phenomenon, and the two sets of independent results are then synthesized or compared during the discussion" (Creswell & Clark, 2018, p. 81). In practice, a series of semi-structured interviews were undertaken in parallel with a self-administered questionnaire.

Study 1: quantitative strand
This study employs a quantitative method and a self-administered questionnaire to collect data. The questionnaire for this study was developed on the basis of previous studies that measured visitor engagement and has five reflective dimensions (Rasoolimanesh et al., 2019So et al., 2014), authenticity (nine items) (Kolar & Ž abkar, 2010), destination image (four items) (Rindell, 2013;Wu & Li, 2017), MTE including seven reflective dimensions (Kim et al., 2012;Kim & Ritchie, 2014), revisit intention (three items) (Chen & Chen, 2010), and eWOM intention (four items) (Pandey & Sahu, 2020;Yen & Tang, 2015) (See Appendix 1 for full name of items). This study involves four reflective constructs and two reflective-formative (composite) second-order constructs. The visitor engagement construct includes five dimensions, and MTE contains of seven dimensions, which these dimensions represent different aspects of these constructs. According to MacKenzie, Podsakoff, and Jarvis (2005, p.715), a higher-order formative construct "faithfully represents all of the conceptual distinctions that the researcher believes are important and provides the most powerful means of testing and evaluating the construct". In a second-order formative or composite construct, all dimensions together establish the construct, and one dimension cannot represent the meaning of construct . Thus, the visitor engagement and MTE should be considered and assessed as reflective-formative second order constructs.
The data for this study were collected from May to August 2019 using purposive sampling. Five highly visited heritage sites in Kashan were selected for data collection. Trained research assistants approached tourists and asked about their experiences of visiting heritage sites in Kashan and explained about the purpose of research. If tourists had already visited the heritage site and were willing to participate, the questionnaire was given to them, and after completing, the questionnaires were collected by the research assistants. A total of 409 questionnaires were collected which, after screening for completeness, left 350 completed questionnaires for use in this study. Out of 350 respondents, 204 (58.3%) respondents were male and 146 (41.7%) female. Among the 350 respondents, the majority (63.4%) were between 18 and 38 years old and the majority of respondents had graduated either from college or university (55.4%). A total number of 173 (49.4%) respondents had visited Kashan before, and the rest (50.6%) were visiting Kashan for the first time. Due to the self-administered nature of data collection, non-response bias was checked by comparing early and late collected data (Gannon et al., 2020), with no significant differences detected. Due to collecting data from a single source, the Common Method Variance (CMV) was assessed using Harman's single-factor test (Rasoolimanesh et al., 2019) and a full collinearity VIF (variance inflation factor) approach was used (Kock, 2015). Factor analysis was performed using unrotated principal component analysis and the results showed the explained variance of 33.1% for the first component. Thus, the results satisfy Harman's single-factor test. Moreover, the results show the full collinearity VIF of all constructs lower than 3.3 (Kock, 2015) indicating the non-existence of CMV issues for the data collected for this study.
This study employs partial least squaresstructural equation modeling (PLS-SEM) using WarpPLS 7.0 software package (Kock, 2019) to analyze the quantitative data and assess the measurement model and structural model. PLS-SEM is the preferred approach of this study because of complexity of a model including two reflective-formative second-order constructs and four reflective constructs (Ali, Rasoolimanesh, Sarstedt, Ringle, & Ryu, 2018;Hair, Risher, Sarstedt, & Ringle, 2019). WarpPLS 7.0 software has been used for this study, because it provides some features such as full collinearity VIF including both lateral and vertical collinearity to assess discriminant validity of formative constructs, and effect size of indirect effects, with these criteria not available in other PLS software packages (Rasoolimanesh, Jaafar, & Barghi, 2017). To ensure the adequacy of data, using G*Power, the minimum sample size was calculated (Faul, Erdfelder, Buchner, & Lang, 2009). The results of G*Power suggested the minimum sample size of 151 to achieve the power of 0.8 for this study. In addition, Reinartz et al.'s (2009) study recommended a sample size of 100 to conduct PLS-SEM. Therefore, a sample size of 350 is more than enough to perform data analysis for this study.

Study 2: qualitative strand
Overall, 20 face-to-face interviews were carried out with the visitors in the five well-known heritage sites in Kashan. A 'first past the interviewer' basis was used as the sampling method which is to suggest that the next person to walk past was approached after the interviewer had finished an interview. Before each interview, the objectives of the study, and timing were explained and asked if the person was a visitor to Kashan and those who agreed to participate were asked to answer to the interview questions. One of the researchers who lives in the Isfahan province conducted all interviews. This helped to reduce bias and limit variation in interview technique (Creswell & Clark, 2018). Interviews were conducted and analysed in Farsi, transcribed verbatim, and translated into English retroactively to ensure consistency in meaning (Creswell & Clark, 2018). Interviews lasted about an average 30 min. Questions were framed according to the focus of the research and were underpinned by relevant related studies (Chandralal & Valenzuela, 2015;Kim et al., 2012;Seyfi, Hall, & Rasoolimanesh, 2020;Tung & Ritchie, 2011). The participants were first asked to provide in detail an account of their experiences during their visit to Kashan (e.g. How would you evaluate your experience of visiting Kashan? What factors you feel contribute to your memorable experience? How was your feeling when you were visiting heritage sites in Kashan?). Next, they were asked to identify those elements that they may recall when they back home and may share with friends and family (e.g. What elements of this trip you would remember or recall while you talk about your trip to family or friends?). Finally, they were asked to explain why their travel experiences in Kashan are different comparing with their previous travels to other destinations (e.g. How this trip to Kashan was different from your previous experiences?) and to talk about their future travel behavioral intention (e.g. What is your plan for future visit in Kashan? What will you do to encourage your friends and relatives to visit Kashan?).
The questions were open-ended in order to gain more spontaneous opinions and avoid the potential bias of limiting responses to the researcher's fixed categories. In qualitative research, the sample size depends primarily on data saturation, target sample access, resources, and available time (Patton, 2002). Interviews were conducted until theoretical saturation. After interviewing 20 participants, the level of data saturation was achieved as the final interviews did not yield new information. Thematic analysis was used for analysing the data (Braun & Clarke, 2006).

Assessment of measurement model using PLS-SEM
The framework of this study contains six constructs, of which four (authenticity (AUT), destination image (DEI), revisit intention (INT), and eWOM (eWOM) intention are reflective constructs, whereas the other two constructs (visitor engagement (VE) and memorable tourism experiences (MTE) are second-order reflective-formative constructs. The visitor engagement construct includes five dimensions; enthusiasm (VEN), attention (VAT), absorption (VAB), interaction (VIN), and identification (VID). These five dimensions are reflective and formatively (composite) establish the visitor engagement. Moreover, the MTE is formatively (composite) established with seven reflective dimensions including hedonism (MHE), novelty (MNO), local culture (MLC), refreshment (MRE), meaningfulness (MME), involvement (MIN), and knowledge (MKN). In order to assess the measurement model, in the first stage the reliability and validity of all reflective constructs should be evaluated and established (Rasoolimanesh et al., 2019). In the second stage, using the score of associated dimensions of visitor engagement and MTE, the 2 s-order composite constructs are established (Gannon et al., 2020).
To assess the reliability and convergent validity of reflective constructs in the first stage, the outer loadings of the associated items of each construct should be greater than 0.7, and the composite reliability (CR), rho_A, and average variance extracted (AVE) of each construct should be greater than 0.7, 0.7, and 0.5 respectively Hair et al., 2019). Nonetheless, outer loadings greater than 0.5 are acceptable if the CR and AVE meet the threshold (Hair, Matthews, Matthews, & Sarstedt, 2017). Table 1 shows acceptable reliability and convergent validity for all first-order reflective constructs including the dimensions of visitor engagement and MTE.
To establish discriminant validity in the first stage, two conservative approaches, the Fornell-Larcker criterion and the heterotrait-monotrait (HTMT) ration, have been applied to this study . To establish discriminant validity using Fornell-Larcker criterion, the square root of AVE of each construct should be greater than the correlation with any other constructs, whereas the HTMT value for each construct should be lower than either 0.85 or 0.9 . Tables 2 and 3 show the acceptable discriminant validity for all reflective constructs in first stage using Fornell-Larcker criterion and the HTMT approach.
To assess 2 s-order reflective-formative (composite) constructs (visitor engagement and MTE) in the second stage, the multicollinearity, via variance inflation factors (VIF) should be lower than 5, and the outer weights of associated items should be significant . Table 1 shows that the VIF are lower than 5 and the outer weights are significant for the associated items of both visitor engagement and MTE constructs (Hair et al., 2017). In addition, to assess the discriminant validity of formative (composite) and reflective constructs in the second stage, the full collinearity VIF should be checked, and the value of VIF should be lower than 3.3 to establish discriminant validity (Rasoolimanesh et al., 2017). The results show that the full collinearity VIF for all constructs in second stage is lower than 3.3 indicating acceptable discriminant validity in second stage.

Assessment of structural model and hypothesis testing
In order to assess the structural model, the R 2 and Q 2 values for endogenous constructs have been assessed and reported. The results show the values of 0.50, 0.21, and 0.45 for the R 2 of MTE, eWOM intention, and revisit intention respectively, which these values are considered high in behavioral studies (Hair et al., 2017;Rasoolimanesh et al., 2017). In addition, the values of Q 2 are 0.50, 0.21, and 0.45 for MTE, eWOM intention, and revisit intention respectively, which are considered very high to assess predictive reliance of research model (Hair et al., 2017). To test the hypotheses, the significance of the direct and indirect effect using the product coefficients approach (Nitzl, Roldan, & Cepeda, 2016), and effect size (f 2 ) for both direct and indirect effects have been reported. Table 4  The results show that the direct effect and indirect effect of authenticity on eWOM are competitive (Nitzl et al., 2016), and the MTE plays a significant mediation role in transferring the effect of authenticity to eWOM. The results indicate the positive and significant direct and indirect effects of destination image on eWOM creation (H10 & H14).
The results of this study support the positive and significant direct and indirect effects of visitor engagement on revisit intention (H7 & H15), as well as the positive indirect effects of authenticity and destination image on revisit intention (H16 & H17). The results do not support the direct effects of authenticity and destination image on revisit intention (H9 & H11), highlighting the importance of mediation role of MTE to transfer the effects of authenticity and destination image on revisit intention.

Study 2: qualitative findings
Overall, the analysis of interviews showed that visitor engagement, authenticity, and destination image positively influence the memorability of visitor's experiences. As the intercepts shows engagement with cultural attractions and sites during the visit is a significant contributing factor to the forming of memorable experiences. This reflects the findings of previous studies that reported a positive relationship between increasing engagement and satisfying consumption experiences (e.g. Brodie et al., 2011;Hollebeek, 2011). The literature evidenced that better engagement with an attraction's context and contents optimizes the overall visitor experience (So et al., 2014) and positively creates a higher level of MTE (Chen & Rahman, 2018). This is also consistent with the self-determination theory (Ryan & Deci, 2002) which posits that a higher level of visitor engagement is likely to lead to a higher level of MTE.
During the visit I felt related and connected to the history. The design of buildings such as Tabatabaei House with its magnificent architecture and Tepe Sialk as the most historical site in Kashan, always brings me to the past and makes me to have a nice feeling of my visit (P.4).
For me, the authenticity of traditional houses in Kashan is unique and I can definitely say that the traditional houses in Kashan have a unique architecture and interior design and I feel like I have immersed myself in the local culture and you observe how luxurious life looked like in the      past. This has made my visit memorable (P.8).
… traditional houses along with their courtyards, wall paintings, elegant stained-glass windows are excellent features of traditional Persian residential architecture. you can hardly see such places in other cities in Iran. When you come to Kashan and walks on the stoned pavements, it seems you are walking in a history book. I believe the experience of visiting Kashan is the most unique experience I have ever had in all my life (P.17).
Other respondents also noted how their perceived image of cultural attractions in Kashan contributed to the memorability of their experience of visiting Kashan. This finds support in the literature which commonly reflects that destination image influences holistic impression of an experience (e.g. Tasci & Gartner, 2007;Zhang et al., 2018).
Fin Garden with hundreds of years of history is the symbol of the Persian Gardens. The inscription of this garden on the UNESCO heritage sites has given a nice reputation to Kashan. I have always wanted to visit this beautiful garden and the famous Bathhouse where one of the icons of contemporary Iran, Amir Kabir, was murdered (P.14).
A number of beautiful historical houses and the Tepe Sialk with over 7000 years history and the traditional bazaar of Kashan reflects the historical atmosphere and cultural blend of the city. These are on the top list of any visitor coming to Kashan (P.11).
The analysis of the interviews showed the significant effects of visitor engagement and authenticity on their revisit and eWOM intentions. The tourism literature has also identified authenticity and engagement as significant antecedents of the memorability of visitor's experiences and satisfaction Ramkissoon & Uysal, 2010;Zhang et al., 2019). As Richards (2007) notes, cultural visitors' quest for authenticity and an in-depth understanding of destinations in their travel experience is integral to the evaluation of tourist experience and subsequent behavioral intentions (Chen & Rahman, 2018).
The overall architecture and impression of the traditional houses in Kashan and the old history of ancient hill of Tepe Sialk inspired me and during the visit I felt connected with the history and historical personalities … I think this deserves attention by the visitors although I believe these have not been well introduced to tourists. I will share my experiences on social media to let other people know of this mixture of tradition and authenticity (P.3).
People here are very friendly. They like to talk with visitors and exchanging ideas. I wanted to buy a carpet and the seller was so patient and gave me a lot of good details about the carpet. He invited me to drink tea and we talked a lot about Kashan its people and attractions … I have decided to write about this experience and this carpet seller shop on my Instagram so that my friends also know more about Kashan (P.20).
I had participated in the rose picking ceremony in Kashan together with Kashani people last year. I still believe that it was a unique chance for me to see local people and participate in this traditional local event. I spent the whole day in this ceremony and did not even realise how time past. It is on my priority list to participate in next year's festival with my family (P.14). The unique attractions in Kashan are the features of the positive image of the city in the mind of tourists. I have seen many foreign tourists in different attractions in Kashan. If they write about their visit and share their experiences on international websites such as Tri-pAdvisor, people from other countries know more about the ancient treasures of Kashan and they will come to visit this city (P.2).
I think people often have a positive image of Kashan after their visit. At least this has been the case for me and my family. I remember a friend of mine who had visited Kashan suggested us to come here to visit the city. We will also share our good experiences with other people to encourage them to come to Kashan (P.19).
The effects of MTE on both revisit and eWOM intentions were also reflected by the interviewees. Overall, tourists are more likely to revisit a place where they have had good and meaningful memories. The literature has also reported a favorable link between MTE and behavioural intentions (e.g. Adongo et al., 2015;Kim et al., 2012;Tsai, 2016). This is also supported by the theory of the tourism consumption system which asserts that tourists' evaluation and impressions of the destination experience affect their overall destination appraisal and behavior (Woodside & Dubelaar, 2002).
For me, the Spring of Kashan and rose water festival was awesome, and I had never seen such rose-picking ceremonies before. I will definitely come again next year to see this festival and will encourage my friends to come to see such a dream-like festival (P.12).
My wife and I both study architecture and Kashan has always been our top visit. Traveling to Kashan is full of differences. Kashan has always been full of beautiful contradictions between tradition and modernity for us. We often come to visit this city and always share positive things with our friends (P.16).

Discussions and conclusions
Using a convergent parallel design, this study investigated the effects of visitor engagement, authenticity, and destination image on revisit and eWOM intentions directly and indirectly through MTE. The findings of this study demonstrate the positive effects of these antecedents on eWOM and revisit intentions either directly or indirectly through MTE. The analysis of the interviews also revealed similar observation. Several interviewees mentioned their inspiration and feeling during their visit of heritage sites in Kashan and traditional guesthouses which have unique and magnificent architecture. Some others narrated their immersion in the local culture during the different interactions with local people and particularly through the participation in the rose water festival. These findings are also consistent with previous studies that identified the positive effects of visitor engagement (Brodie et al., 2011;Rasoolimanesh et al., 2019), authenticity (Oh et al., 2019), and destination image (Wu & Li, 2017;Zhang et al., 2018)

on tourists' behavioral intentions.
This study also demonstrated the importance of MTE for the relationships between antecedents (e.g. visitor engagement, authenticity, and destination image) and revisit and eWOM intentions. For all three antecedents the MTE is a significant mediator and indirect effects are positive and significant, whereas the results could not support the direct effects of authenticity on revisit and eWOM intentions as well as destination image on revisit intention. This finding suggests that tourists with vivid and meaningful memories of their recent travels were more likely to engage in eWOM behavior. This post-experience reflection was also noted by several interviewees who narrated their memorable moments dusting the visit of Kashan and mentioned their intention to share their travel experiences with others on social media platforms and to let other people know of the mixture of tradition and authenticity in Kashan. Some other believed that unique architecture and authentic heritage sites in Kashan have not been well introduced to visitors and particularly international tourists. They reflected on their intention to share their experiences on international travel platforms such as TripAdvisor to let people from other countries know more about the cultural heritage offerings of Kashan and encourage visits to the city. They believe that shared information on social media could influence travel decision making for potential international travelers. Others commented on their memorable moments during interactions with local people and their participation in the rose water festival. They mentioned their intention to revisit Kashan again and share positive things and stories on social media to promote the city of Kashan for domestic visitors. The results are consistent with several previous empirical studies which identified the importance of MTE for engaging tourists in the eWOM communications and encourage them to revisit a destination (Adongo et al., 2015;Barnes et al., 2016;Gohary et al., 2020;Kim, 2018;Kim & Ritchie, 2014).
This study also identified the strong and significant effects of visitor engagement, authenticity and destination image on MTE and is consistent with previous studies identifying the positive and strong effect of visitor engagement (Chen & Rahman, 2018;Seyfi, Hall, & Rasoolimanesh, 2020;Taheri et al., 2014), authenticity (Antón et al., 2019;Ramkissoon & Uysal, 2010) and destination image (Kim, 2018;Zhang et al., 2018) on improving tourists' experiences and creating MTE. The findings showed that authenticity had the highest effect on the MTE, indicating that by visiting an authentic heritage site a tourist can better enjoy higher quality experiences during their visit. A similar observation was also reflected by the interviewees who mentioned the architecture and impression of the traditional houses in Kashan and the old history of the heritage sites in the city such as Tepe Sialk which provided them with authentic and memorable experiences and talked about their intention to share such experiences with others on social media. Such on-line experience sharing has an important influence on tourists' perceptions of the destination's trustworthiness among prospective travelers. This also finds supports in prior studies that reported that authenticity has a positive impact on the perception of MTE (Antón et al., 2019), and significantly influence future tourist behaviour by increasing the intention to make a repeat visit or recommend it to others (Kolar & Ž abkar, 2010).

Theoretical implications
Overall, this study focuses on a post experience behavior (eWOM) and contributes to the extant literature on the relationship between MTE and behavioural intentions in the backdrop of heritage tourism. This study illustrated that MTE could be a mediator for the effects of visitor engagement, authenticity, and destination image on revisit and eWOM intentions in a heritage tourism context. This also provides for a better understanding of the context-based focus of the MTE by focusing on a less studied heritage tourism context. By testing the MTE scale in a heritage tourism context, this research has advanced the heritage tourism experience literature and raise intriguing questions for future research, particularly with respect to the potential influencers of MTE. Additionally, The MTE scale developed by Kim et al. (2012) and visitor engagement developed by So et al. (2014) have been examined as reflective-formative (composite) second-order constructs. Recent literature identified that misspecification of measurement type of constructs in the model can lead to biased results at the structural level, indicating the importance of correct identification of the nature of constructs in the research model (Mikulić & Ryan, 2018). The findings confirmed these constructs as multi-dimensional reflective-formative (composite) second-order constructs in the context of the heritage setting, which can be considered a unique methodological contribution of the current paper.

Practical implications
The findings of this study also provide practical recommendations for destination managers in service adjustments, recognizing and sustaining MTE among visitors and contribute to the competitive advantage of the cultural tourism destinations. As the findings showed, to improve revisit and eWOM intentions of heritage tourists, the MTE has critical direct effect and important mediating role for the effects of visitor engagement, authenticity, and destination image. As Pine and Gilmore (1998) stated, the richest experiences create a "sweet spot" in consumers and affect their future decision and behaviour and are often the most reliable source of information in revisit intention and eWOM communications. This is important, as established tourism destinations have mainly prioritized product-oriented marketing and management practices, and often neglected the needs and experiences of tourists (Zhang et al., 2018). As a result, they may fail to design and deliver those experiences that can be selectively recalled while tourists talk about their travel and therefore negatively affect revisit-intention as well as the generation of positive eWOM on travel platforms and particularly on social media.
In addition to MTE, our findings demonstrate the positive effects of visitor engagement, heritage destination image and authenticity on destination revisit and eWOM intentions. Such findings are helpful for DMOs to understand how by improving visitor engagement using different approaches, as well as destination image and perceived authenticity of heritage, can increase tourists revisit intention, and tourists' involvement in destination eWOM. This is especially important in the COVID-19 travel environment in which destinations that have been visited previously and provided authentic and memorable experiences to the tourist may be perceived more positively in consumer travel planning (Gössling, Scott, & Hall, 2020). Given tourist need for reassurance for safety and security in travel as a result of COVID-19 (Gössling et al., 2020;Rastegar, Seyfi, & Rasoolimanesh, 2021;Seyfi, Hall, & Shabani, 2020), the results of the present research suggest that DMOs and heritage tourism businesses should potentially focus their marketing on potential repeat visitors reinforcing the authentic and unique nature of previous experiences and the capacity to experience them again. This is especially important given the close relationships that exist between positive perceptions of authenticity and feelings of security and trust (Wickham, 2013). DMOs should seek to enable more effective engagement of visitors in the destination as well as on-site and aim to increase perceived authenticity in order to improve tourist experiences, destination image and revisit intention, with destination image being an important factor in affecting the formation of MTEs in a heritage tourism context.

Limitations and directions for future research
The study has limitations which provide avenues for future research. First, in this study, the questionnaires were distributed to onsite visitors. In contrast to previous studies (Chandralal & Valenzuela, 2015;Kim & Ritchie, 2014;Tung & Ritchie, 2011) that asked respondents to recall their most memorable recent tourism experience, respondents were immediately surveyed after or during their visit to Kashan. Although, such an approach probably allowed respondents to have a vivid and better memory of their experience and increase recall precision, compared with the post-visit approach, it is suggested for future studies to focus on the former approach to capture the post-trip behaviour, ideally within an extended research framework. Second, to measure MTE, this study was built on the framework developed by Kim et al. (2012) and tourists' positive MTE were only investigated, and negative MTE dimensions were not reflected. Negative experiences can also generate distinct and memorable experiences (Kim, Wang, & Song, 2020). Thus, a comprehensive study covering positive and negative dimensions related to experiences is suggested for future research. Third, this study focuses on visitors in a heritage site. Future studies are suggested to investigate our integrated framework of influencing factors on revisit and eWOM intentions in other tourism contexts. Finally, we have suggested that perceptions of authenticity will also influence revisit intentions in a COVID-19 tourism environment. Further research is required to assess the extent of this relationship and its connections to feeling of security among tourists.

VE_Absorption
When I am interacting with the tourism site, I forget everything else around me. Time flies when I am interacting with the tourism site. When interacting with the tourism site, it is difficult to detach myself. VE_Interaction I am someone who enjoys interacting with like-minded community in the tourism site. I am someone who likes to actively participate in the tourism site community discussions. In general, I thoroughly enjoy exchanging ideas with other people in the tourism site community.

VE_Identification
When someone criticizes this tourism site, it feels like a personal insult. When I talk about this tourism site, I usually say 'we' rather than 'they' because the identity of the site suites me. When someone praises this tourism site, it feels like a personal compliment. Authenticity (continued on next page) S.M. Rasoolimanesh et al.