Drivers of Revisit Intention in a Sacred Heritage Site: An Integrated Theory of Planned Behavior, Attribution Theory, and Elaboration Likelihood Model Approach at Mount Wutai

Abstract

As a representative case that embodies both the attributes of a Buddhist sacred site and those of a UNESCO World Heritage site, Mount Wutai provides a distinctive research setting for examining behavioral mechanisms in temple tourism. This study aims to construct an integrated model to systematically test the effects of enjoyment, memorability, attitude, subjective norm, and perceived behavioral control (PBC) on revisit intention (RI), while incorporating social media exposure as a moderating variable. Based on data collected through a two-wave on-site survey, this study analyzed 617 tourists in Mount Wutai and employed structural equation model to examine the relationships among the variables. The results indicate that all five psychological antecedents exert significant positive effects on revisit intention, among them, PBC demonstrating the most substantial impact. Further analysis reveals that social media exposure significantly moderates the relationships among enjoyment, memorability, attitude, subjective norm, and revisit intention, most notably in the “memorability–RI” relationship, whereas its moderating effect on the “PBC–RI” relationship is not significant. These findings not only enrich the theoretical framework by integrating emotional attribution, behavioral cognition, and digital media engagement but also provide practical implications for sacred tourism destinations, enabling them to enhance visitor loyalty through digital communication and experience optimization.

1. Introduction

Tourism, as one of the most dynamic and multidimensional industries today, plays a vital role in stimulating global economic growth, generating employment opportunities, and fostering cultural identity (Liang et al., 2022). With the continuous expansion of cultural and religious tourism, temple tourism, characterized by ritualistic religious experiences and deep cultural immersion, has increasingly drawn scholarly attention (Chanapal et al., 2025; George, 2025). In China, a country with profound religious traditions and rich cultural heritage, Mount Wutai holds dual status as both a sacred Buddhist site and a UNESCO World Heritage Site, making it an ideal setting for examining mechanisms of visitor behavior in temple tourism.

Although Mount Wutai has long maintained a large and stable tourist population, the traditional, pilgrimage-centered tourism model has become less able to meet the diverse expectations of younger generations. Contemporary tourism increasingly exhibits trends of “experientialization” and “emotionalization,” with visitors seeking travel experiences that are emotionally meaningful, memorable, and highly engaging (Liu et al., 2024). Against this backdrop, understanding the formation mechanism of revisit intention is not only practically relevant but also theoretically significant. However, existing studies tend to explain temple tourism behavior from single perspectives, such as religious belief, cultural heritage interest, or place attachment, and lack an integrated framework that incorporates multiple psychological dimensions, including emotional experience (e.g., enjoyment), experiential memory, attitude, and subjective norm (Elshaer et al., 2024). As tourism research increasingly emphasizes leisure experience cognition, emotional responses, and attitude formation processes, incorporating enjoyment, memorability, attitude, and subjective norms into the explanation of revisit intention in temple tourism becomes especially necessary.

Meanwhile, the rise of the digital era has positioned social media as a crucial external factor influencing tourists’ perceptions, experience evaluation, and behavioral decision-making. Prior research has predominantly examined its direct impacts, while relatively few studies have addressed its contextual moderating role in shaping emotional experiences and attitude formation (Hussain et al., 2024). For Mount Wutai, where sacred religious symbolism intersects with modern digital communication, understanding the interaction among emotional experience, cognitive attitude, and social media exposure is particularly important.

Addressing these theoretical and practical gaps, this study develops an integrated model that incorporates emotional experience, cognitive evaluation, social norms, and digital media engagement. It systematically examines how enjoyment, memorability, attitude, subjective norm, and perceived behavioral control influence revisit intention, while introducing social media exposure as a moderating variable. By doing so, the study deepens the understanding of the emotional–cognitive–digital transmission mechanisms in temple tourism and provides more targeted theoretical and practical insights for destination management and communication strategies. That is, this study attempts to explore the following:

(1)

How do the enjoyment, memorability, attitude, subjective norms, and perceived behavioral control influence on tourists’ revisit intentions towards Mount Wutai?

(2)

How does social media exposure moderate the relationships between the psychological factors (including emotional, cognitive, and normative components) and revisit intention in the context of Mount Wutai?

The remainder of this paper is organized as follows: Section 2 reviews the relevant literature, outlining the research progress on revisit intentions in temple tourism; Section 3 presents the research hypotheses of this study; Section 4 describes the research methodology, including variable design and data collection; Section 5 reports the empirical analysis results; and Section 6 concludes the paper by summarizing the findings and highlighting its limitations as well as directions for future research.

2. Theoretical Foundations and Research Hypotheses

2.1. Theoretical Foundations

AT explains how individuals interpret the causes of events and how these interpretations influence emotions and behavior Weiner and Weiner (1985). In tourism, causal attributions can be classified into three dimensions: locus (internal vs. external), stability (stable vs. unstable), and controllability (controllable vs. uncontrollable). Causal attributions influence tourists’ satisfaction, loyalty, and revisit intention (Triantafillidou & Siomkos, 2014). That is, tourists who attribute their positive experiences to internal, stable, and controllable factors, such as spiritual satisfaction and interaction with other tourists, are more likely to have positive attributions and intend to revisit (Chen et al., 2020). Positive attributions positively influence satisfaction and loyalty (L. Su et al., 2017). Recent studies have found that sacred value enhances revisit intention in religious tourism (Y. Su et al., 2025). However, limited studies have examined how attributions explain emotion-based constructs, such as enjoyment and memorability (Wong et al., 2022). Given the locus of control reasoning, it is more appropriate to focus on intention rather than behavior, because tourists intend to revisit when they are constrained by time/resources.

According to TPB, intention is jointly shaped by attitude, subjective norm, and perceived behavioral control, which together predict behavioral outcomes (La Barbera & Ajzen, 2020). Perceived behavior control significantly influences repeat travel (Chou et al., 2025). Regarding religious tourism, subjective norms and cultural values have been shown to have a significant impact (Cai & Jiang, 2024). However, the application of the TPB in East Asian temple tourism remains relatively limited. This research addresses the existing gap by exploring the extent to which attitudes, social influence, and perceived feasibility shape tourists’ intentions to return to Mount Wutai.

ELM explains that individuals interpret persuasive information via two main routes: the central route, which entails a thorough assessment of message content, and the peripheral route, which relies on cues such as the communicator’s credibility or the popularity of the message (Petty et al., 2009). Regardless of the route taken, the final attitudes induced by these routes influence behavioral intentions, including revisit intention (Rehman et al., 2022). ELM is frequently used in marketing and communication research. However, very few studies used this model in religious heritage tourism. Very few studies have examined how both rational elaboration and peripheral cues influence revisit intentions in temple tourism. By using ELM, this study elucidates how tourists’ attitudes towards Wutai Temple influence revisit intention through their perceptions of information quality and exposure to social cues from Wutai Temple, clarifying the psychological process of media exposure-intention.

2.2. Hypothesis Development

Tourists’ enjoyment is a fundamental emotional outcome of tourism experiences that substantially shapes overall evaluations and subsequent behavioral intentions (Juliana et al., 2024). Positive emotional responses—such as pleasure, satisfaction, and relaxation—enhance tourists’ affective attachment to destinations and increase the likelihood of repeat visitation (Yan & Halpenny, 2022). Enjoyment contributes not only to momentary satisfaction but also to the formation of enduring positive impressions that influence future travel choices. In relation to temple tourism, tourists who derive enjoyment from a serene environment, well-organized rituals, or considerate services are more inclined to cultivate affective bonds with the site. These favorable affective associations, in turn, enhance revisit intention. Thus, we advanced that

H1. 

Tourists’ enjoyment promotes their revisit intention to Mount Wutai.

Memorability reflects the emotional depth and uniqueness of a tourism experience and refers to the lasting impressions tourists form after their visit (Rasoolimanesh et al., 2021). Memorable experiences tend to leave a mark on consumers by evoking emotions that trigger related behavior (Karagöz & Ramkissoon, 2023). In other words, memorable activities in experiential tourism significantly enhance tourists’ attachment to destinations and revisit intention. In the case of temple tourism, if tourists have vivid memories of their experiences, such as the impressions they gain from the architecture of temples, meaningful interactions, environment (e.g., peaceful atmosphere), or entertainment, tourists’ revisit intention will be triggered by these recollections when they face similar decision-making in the future (Yin & Jia, 2024). In other words, memory triggers revisit intention. Consequently, the research posits that

H2. 

Tourists’ memorability promotes tourists’ revisit intention to Mount Wutai.

In tourism behavior research, attitude is seen as a strong predictor of behavioral intention, which bridges cognitive evaluations and decision-making in the process of tourism consumption (Rao et al., 2022). Generally, tourists with positive attitudes are interested in, approve of, and are willing to participate in tourism activities. Several studies have shown that positive attitudes greatly enhance revisit intention (Peng et al., 2023). That is, when tourists have positive attitudes, they consider traveling to a destination as meaningful, worthwhile, or enjoyable, and are more likely to revisit. Stronger positive attitudes trigger stronger positive emotions and identify with destinations, which are important drivers of revisit behavior (Isa et al., 2019). Accordingly, we formulate that

H3. 

Tourists’ attitude towards temple tourism positively influences tourists’ revisit intention to Mount Wutai.

Subjective norm is the perceived social pressure to behave in a certain way (Xu et al., 2022). Usually, people consider the opinions of their family, friends, or even people of the same religion when deciding to revisit a destination. Subjective norm has a significant influence on tourists’ attitudes towards destinations and revisit behavior (Zhuang et al., 2020). When tourists believe their families, friends, or even the religious community encourages them to do something, such as visiting the temple, they are more likely to form the intention to revisit. In the context of religious and spiritual tourism, people’s behavior is mainly influenced by cultural or religious norms. Social encouragement, including family traditions, recommendations from religious authorities, and opinions on social media, can enhance tourists’ intention to revisit sacred sites (Hussain & Wang, 2024). Accordingly, we formulate that

H4. 

Subjective norm promotes tourists’ revisit intention to Mount Wutai.

PBC refers to an individual’s perceived ability to perform a behavior, influenced by personal resources and external conditions (Hagger et al., 2022). PBC affects tourists’ ability to revisit the destination by influencing factors such as money, time, and physical access to the site. When tourists feel they have control over these factors and believe that revisiting is possible, they are more likely to have a revisit intention (Baghirov et al., 2023). If individuals think that there are fewer barriers or obstacles to a behavior, for example, visiting a particular destination, they tend to have an intention to perform that behavior (Siegrist et al., 2021). PBC over travel logistics, such as transportation costs or the time required to visit the temple, significantly affects revisit intention to Wutai Temple. Thus, the subsequent hypothesis is formulated:

H5. 

Perceived behavioral control positively influences their revisit intention to Mount Wutai.

Tourists’ enjoyment refers to the pleasure a tourism destination provides, making tourists feel good emotionally and psychologically, as reflected in their evaluations of their experiences as tourists (Pestana et al., 2020). In temple tourism, tourists may feel tranquility, a sense of meaningfulness in Buddhist rituals, or comfort. Enjoyment positively affects tourists’ post-visit evaluations and revisit intention to a tourism destination (González-Rodríguez et al., 2016). That is, tourists will have better evaluations and higher revisit intention for Wutai Temple after experiencing enjoyment in temple tourism. Moreover, the influence of enjoyment on revisit intention may be strengthened by different degrees of media exposure. Suppose tourists are repeatedly exposed to media such as short videos, travel reviews, or symbolic images of Wutai Temple. In that case, these external stimuli will help maintain and enhance tourists’ emotional states by reinforcing the initial emotional significance of Wutai Temple in their minds. Media exposure can reinforce the initial emotional significance of Wutai Temple, which is also spread from private feelings to shared beliefs and culture (Wood & Kenyon, 2018), and help retain the memory of satisfaction with Wutai Temple. Therefore, we formulate that

H6. 

Social media exposure enhances the moderating effect between tourists’ enjoyment and their intention revisit to Mount Wutai.

Memorable experiences leave long-lasting cognitive and emotional traces that are strongly associated with future behavioral intentions (González-Rodríguez et al., 2016). However, the strength of this influence may vary depending on contextual reinforcements. Media exposure refers to external stimuli that can be instrumental in maintaining and amplifying memory-related emotional responses. By interacting with user-generated content, such as revisit content, other tourists can relive past experiences and enhance their psychological attachment (Li & Zhao, 2021). Thus, we posit that

H7. 

Social media exposure strengthens the association between tourists’ memorability and their revisit intention to Mount Wutai.

Positive attitudes are among the strongest predictors of revisit intention (Soliman, 2019). Nevertheless, the magnitude of this predictive link may be shaped by contextual or external factors. Media exposure functions as a symbolic environment that reinforces tourists’ existing attitudes by providing continuous validation and social endorsement of the destination (Arasli et al., 2021). When tourists encounter favorable content—such as testimonials, spiritual reflections, or immersive documentaries—related to temple tourism, their prior positive attitudes may be further strengthened and translated more readily into behavioral intentions (Conner & Norman, 2022). In temple tourism, the impact is heightened by the symbolic significance and affective resonance conveyed through spiritual narratives. Therefore, this study proposes that

H8. 

Social media exposure positively moderates the relationship between tourists’ attitude towards temple tourism and their revisit intention to Mount Wutai.

Social media exposure refers to the extent to which an individual is exposed to and engages with content related to a specific topic through various media platforms, mainly social media (Shahbaznezhad et al., 2022). Media exposure can amplify the visibility and perceived importance of specific destinations through user-generated content, endorsements by opinion leaders, and peer sharing. This external input can reinforce the social expectations or pressures associated with visiting a particular destination, thereby enhancing the role of subjective norms (Lee et al., 2025). Media exposure can strengthen the influence of social factors on behavioral intentions. When individuals see friends, family members, or online community members positively sharing about a destination, the sense of social pressure to visit or revisit that destination becomes stronger. If media content presents spiritual or cultural narratives, it may also reinforce social norms, thereby making individuals more inclined to act in accordance with these perceived social expectations (Singh & Bhuyan, 2025). Consequently, we advance that

H9. 

Social media exposure positively moderates the relationship between subjective norms and tourists’ revisit intention to Mount Wutai.

Social media exposure refers to the extent to which destination-related information, images, and narratives appear on platforms such as WeChat, Weibo, Douyin, or travel blogs (Ouyang & Jiang, 2025). Influence of social media primarily manifests in emotional resonance, experiential imagination, and perceived social identity, rather than in directly altering tourists’ judgments of behavioral feasibility. In other words, social media is more likely to enhance tourists’ emotional responses, strengthen perceptions of social support, or reinforce normative expectations, whereas its impact on perceived behavioral control—such as “whether one has sufficient resources” or “whether the destination is accessible”—is generally limited (Afren, 2024). Although tourists may occasionally encounter travel tips or practical guidance on social media, such fragmented information is often insufficient to alter their overall assessment of a destination’s feasibility significantly. Perceived behavioral control is typically shaped by more stable factors, such as individual resources, prior experiences, and structural constraints, which are not easily influenced by brief media exposure. Consequently, the following hypothesis is formulated in this study:

H10. 

Social media exposure strengthens the linkage between perceived behavioral control and tourists’ intention to revisit Mount Wutai.

2.3. Conceptual Framework

This study integrates AT, TPB, and ELM to construct a comprehensive conceptual framework for explaining tourists’ revisit intentions in temple tourism at Mount Wutai. As illustrated in Figure 1, which presents the theoretical mechanism of this study, the model systematically depicts the logical relationships among emotional attribution, behavioral cognition, and social media–based moderation. Within this framework, AT explains how tourists interpret enjoyment and memorability as emotional attributes of their temple experiences, which, in turn, shape behavioral tendencies. TPB extends this process by incorporating attitude, subjective norms, and perceived behavioral control as cognitive and social determinants that mediate the transition from emotional experience to revisit intention. Meanwhile, the inclusion of ELM clarifies how social media exposure functions as a moderating mechanism, shaping how tourists process information about their experiences and thereby strengthening or weakening the effects of emotional and cognitive antecedents on revisit intention. As shown in Figure 1, enjoyment, memorability, attitude, subjective norms, and perceived behavioral control jointly influence revisit intention (H1–H5), whereas social media exposure moderates these associations through both central and peripheral cue pathways (H6–H10). By integrating emotional attribution, behavioral cognition, and digital persuasion mechanisms, this study proposes a theoretical framework that offers a holistic explanation of how revisit intention is formed in the context of sacred tourism.

3. Research Methodology

3.1. Data Collection

This study used convenience sampling and collected 617 valid responses across two rounds of questionnaires. The respondents were tourists who had visited the famous Buddhist sacred site in China, Wutai Mountain. Given that Wutai Mountain receives over seven million visitors each year, on-site sampling was considered an effective method that could well reflect the features of the overall tourist group. In practice, the researcher conducted randomized intercept surveys at major temple sites (e.g., Xiantong Temple, Tayuan Temple, and Shuxiang Temple), as well as at visitor centers, hotels, and transportation hubs.

This study conducted a questionnaire survey using the Questionnaire Star platform. Given the high mobility, diverse origins, and lack of a complete sampling frame among tourists at Mount Wutai, probability sampling was not feasible. Therefore, a stratified convenience sampling strategy was adopted to maximize sample diversity and representativeness. Data collection followed a two-wave design: the first wave, conducted in mid-June 2025, aimed to obtain preliminary responses and assess the clarity of questionnaire items; the second wave, carried out from late July to early August 2025, coincided with the peak tourist season to ensure that the primary dataset captured a broad and stable visitor population while reducing potential seasonal bias. It should be noted that the two-wave design enhanced data stability and representativeness rather than longitudinal tracking or controlling for common method bias.

Offline data collection was conducted at key locations, including the South Gate entrance, the surrounding areas of Xiantong Temple and Tayuan Temple, and hotels within the central commercial district. QR codes were displayed, and every nth visitor was invited to participate. Online data collection was conducted via WeChat, Xiaohongshu (RED), and collaborating travel agencies, enabling both on-site tourists and recent visitors to respond. To ensure the sample was valid and relevant, screening questions required participants to have visited Mount Wutai within the past 12 months.

A total of 653 questionnaires were collected, and after removing invalid responses, 617 valid questionnaires were retained (effective rate = 94.5%), with offline and online samples accounting for 41.6% and 58.4%, respectively. The sample size is sufficient both numerically and statistically. According to PLS-SEM sampling guidelines (e.g., the 15-times rule and statistical power analysis), models with medium effect sizes typically require fewer than 200 cases. Thus, the 617 responses in this study far exceed minimum requirements and provide strong statistical power and stability for path estimation and moderation analysis.

Furthermore, the measurement of key constructs, enjoyment, memorability, attitude, subjective norms, perceived behavioral control, and revisit intention, as grounded in established theoretical scales. These constructs exhibit apparent structural validity and conceptual alignment, laying a solid foundation for subsequent model testing.

3.2. Instrument

This study designed a structured questionnaire to measure psychological and behavioral variables. The content consists of three parts: (1) the research purpose and confidentiality statement, (2) the basic characteristics of the respondents, and (3) the measurement items of the research variables. The questionnaire items primarily draw on established scales from the existing literature and have been appropriately adjusted for the context of Five-Tai Mountain Temple Tourism. All variables are measured using a five-point Likert scale, with scores ranging from 1 (very disagree) to 5 (very agree).

Tourists’ enjoyment was assessed using three items from Guo et al. (2024) that reflect tourists’ emotional feelings, satisfaction, and enjoyment during temple visits. Tourists’ memorability was assessed using four items from Wang et al. (2021) that reflect tourists’ ability to remember, recall, and retain the temple experience. Attitude toward temple tourism was assessed using five items from Sudarsono et al. (2021) that reflect tourists’ evaluations of cultural value, spirit, and overall temple tourism. Subjective norms are measured using the four items from Al-Swidi et al. (2014), which reflect tourists’ perceptions of social pressure and encouragement from family, friends, and the community. Perceived behavioral control is measured using the six items from Al-Swidi et al. (2014), reflecting tourists’ economic capacity, time availability, and perceived convenience of revisiting Wutai Mountain. The social media exposure measures adopt three items from Michelle and Susilo (2021), reflecting tourists’ opportunities, time, and attention level in accessing information about the temple on social media (WeChat, Douyin, and Xiaohongshu). Finally, the tourists’ willingness to revisit Wutai Mountain is measured using the three items from Oan-Oon and Choibamroong (2025), reflecting tourists’ willingness to revisit, planning, and priority for revisiting. The measurement scales are presented in Appendix A 

Table A1. Measurement.

Variable	Items
Tourists’ enjoyment	TE1. I felt I was engaging in a meaningful and enjoyable experience during my visit to the temple.
	TE2. I found the overall atmosphere of the temple visit to be pleasant and emotionally satisfying.
	TE3. I experienced a sense of joy and peace during my time at the temple.
Tourist Memorability	TM1. You feel very satisfied after visiting this temple.
	TM2. After visiting this temple, you still hope to come again
	TM3. You gain a deeper understanding of culture and spirituality by visiting this temple.
	TM4. By visiting this temple, your interest in temple tourism becomes stronger.
Attitude towards Temple Tourism	ATT1: Based on my personal experience, visiting Mount Wutai provides a spiritually fulfilling and culturally safe environment.
	ATT2: I found my visit to Mount Wutai to be a rewarding and enjoyable spiritual experience.
	ATT3: Mount Wutai helped me gain inner peace and spiritual reflection during my visit.
	ATT4: I tend to choose spiritually and culturally meaningful destinations like Mount Wutai for my travels.
	ATT5: Visiting Mount Wutai for temple tourism was a personally meaningful decision that I would consider making again.
Social Media Exposure	SME1: I frequently come across content related to Mount Wutai temple tourism while using social media platforms such as WeChat, Xiaohongshu, or Douyin.
	SME2: I usually spend a significant amount of time viewing or reading content about Mount Wutai on social media.
	SME3: When I encounter content about Mount Wutai on social media, I pay close attention to it and try to understand what is being shown or discussed.
Perceived Behavioral Control	PBC1: I can make the decision independently to revisit Mount Wutai for temple tourism.
	PBC2: I have the financial ability to afford a revisit to Mount Wutai.
	PBC3: I have sufficient time available to revisit Mount Wutai for temple tourism.
	PBC4: I have complete information and awareness about how to plan a revisit to Mount Wutai.
	PBC5: Travel facilities and access to Mount Wutai are readily available from my location.
	PBC6: I can manage any difficulties (e.g., time, cost, or travel-related issues) that may arise in planning a revisit to Mount Wutai.
Subjective Norms	SN1: The trend of revisiting Mount Wutai among people around me is increasing.
	SN2: People around me generally believe that revisiting Mount Wutai is a meaningful cultural or spiritual activity.
	SN3: My close friends and family members would appreciate it if I revisit Mount Wutai.
	SN4: I would receive full support (e.g., time, encouragement, travel advice) from friends and family if I decide to revisit Mount Wutai.
Tourists’ revisit intention to Wutai temple	TVT1: I would prioritize returning to the temples of Mount Wutai among my future travel destinations.
	TVT2: I intend to revisit the temples of Mount Wutai in the near future.
	TVT3: I will allocate time and budget for revisiting the temples of Mount Wutai.

.

3.3. Data Analysis Tool

This study employed Partial Least Squares Structural Equation Modeling (PLS-SEM) for data analysis. Although the sample size (n = 617) is sufficiently large for covariance-based SEM (CB-SEM), PLS-SEM was selected as the primary analytical approach given the study’s objectives and model characteristics. First, this research is prediction-oriented and aims to maximize the explanatory power of revisit intention rather than emphasizing overall model fit, which aligns more closely with PLS-SEM’s strengths (Legate et al., 2021). Second, the theoretical framework includes multiple latent variables and several moderating paths, making it relatively complex; PLS-SEM offers higher estimation efficiency and robustness when handling such models. Therefore, considering the research goals, model complexity, and data characteristics, PLS-SEM was deemed the more appropriate choice. The analysis proceeded in two stages: the first stage assessed the measurement model, including reliability, convergent validity (AVE), and discriminant validity. The second stage evaluated the structural model by examining path coefficients, coefficient of determination (R²), effect size (f²), and predictive relevance (Q²), thereby testing the model’s explanatory and predictive capabilities.

Before the formal questionnaire survey, a pre-test was conducted on 97 participants to evaluate the clarity of the questionnaire items and the reliability and validity of the measurement tools. Based on the results, the expressions of some questions were slightly adjusted to better align with the theoretical construct. The reliability analysis results showed that Cronbach’s α for all variables exceeded the recommended threshold of 0.70 (Adamson & Prion, 2013).

Table 1. Reliability analysis.

Study Variables	Cronbach’s Alpha
ATT	0.870
PBC	0.902
SME	0.860
SN	0.884
TE	0.850
TM	0.833
TVT	0.804

presents the reliability statistics results of the measurement tools, indicating that each scale has a high degree of consistency. Overall, the measurement model has good reliability and can be used for subsequent structural equation model (SEM) tests.

4. Result

4.1. Descriptive Analysis

Table 2. Sample Information.

Information and Options		Frequency	Percentage
Gender	Male	338	54.8
	Female	279	45.2
Age	18–25	101	16.3
	26–40	251	40.7
	41–60	225	36.5
	61 above	40	6.5
Education level	Undergraduate	173	28.0
	Master	224	36.3
	Doctoral	128	20.7
	Without degree	92	14.9
Region	North China	184	29.8
	East China	190	30.8
	Central China	111	18.0
	South China	132	21.4
Q5	Yes	617	100.0
	No	0	0.0
Q6	Once	219	35.5
	Twice	203	32.9
	Three times	149	24.1
	More	46	7.5
Q7	Yes	617	100.0
	No	0	0.0
Q8	Temple sightseeing	271	43.9
	Temple culture	152	24.6
	Listening to scriptures	141	22.9
	Hiking	53	8.6

Notes: Q5. Have you ever come to travel to Mount Wutai? Q6. How many times have you visited Mount Wutai? Q7. Are you a Buddhist? Q8. The purpose of your visit to Mount Wutai’s Buddhist temple tourism is?

and Figure 2 present the statistical results of the 617 survey respondents, including their demographic characteristics and travel-related features. The gender ratio is generally balanced, with males slightly higher (54.8%) and females accounting for 45.2%. In terms of age structure, the group aged 26–40 has the highest proportion (40.7%), followed by the group aged 41–60 (36.5%), indicating that middle-aged tourists are the leading group for tourism to Wutai Mountain temples. Young tourists under 25 account for 16.4%, while elderly tourists aged 61 and above make up only 6.5%.

Regarding education level, the majority of respondents hold a master’s degree (36.3%), followed by undergraduates (28.0%), doctoral degree holders (20.7%), and those with other educational backgrounds (14.9%). Geographically, the sample is widely distributed across China, with the best representation from East China (30.8%) and North China (29.8%), reflecting Mount Wutai’s strong regional and national appeal. Central China (18.0%) and South China (21.4%) also contribute notable shares of visitors, highlighting the site’s nationwide reach.

All respondents indicated having previously visited Mount Wutai (100%), consistent with the study’s screening criteria. In terms of visit frequency, approximately one-third (35.5%) have visited once, another 32.9% have visited twice, 24.1% have visited three times, and 7.5% have visited more than three times, illustrating a substantial proportion of repeat visits.

Motivations for visiting indicate that temple sightseeing (43.9%) constitutes the primary reason for visitation, followed by engagement with temple culture (24.6%), listening to scriptures (22.9%), and hiking (8.6%). These results highlight that the site’s religious and cultural dimensions are the predominant attractions for tourists.

Overall, the sample composition reflects a diverse yet well-educated group of visitors, with a significant share of middle-aged and repeat tourists, primarily motivated by cultural and spiritual experiences. These characteristics provide important context for interpreting the relationships between enjoyment, memorability, attitudes, social influences, perceived behavioral control, and revisit intention in the subsequent analysis.

Table 3. Descriptive Statistics Results.

	N	Minimum	Maximum	Mean	Standard Deviation
TE1	617	1	5	3.715	0.876
TE2	617	1	5	3.812	0.904
TE3	617	1	5	3.708	0.956
TM1	617	1	5	3.989	1.016
TM2	617	1	5	3.836	1
TM3	617	1	5	3.797	0.916
TM4	617	1	5	3.652	0.9
ATT1	617	1	5	3.901	0.877
ATT2	617	1	5	3.796	0.902
ATT3	617	1	5	3.598	0.955
ATT4	617	1	5	3.771	0.899
ATT5	617	1	5	4.136	0.849
PBC1	617	1	5	3.87	0.85
PBC2	617	1	5	3.763	0.857
PBC3	617	1	5	3.574	0.826
PBC4	617	1	5	3.736	0.929
PBC5	617	1	5	3.643	0.799
PBC6	617	1	5	3.585	0.842
SN1	617	1	5	3.731	0.846
SN2	617	1	5	3.895	0.843
SN3	617	1	5	3.788	0.87
SN4	617	1	5	3.843	0.863
SME1	617	1	5	3.665	0.94
SME2	617	1	5	3.737	1.021
SME3	617	1	5	4.032	0.901
TVT1	617	1	5	3.757	0.823
TVT2	617	1	5	3.784	0.821
TVT3	617	1	5	3.775	0.848

presents descriptive statistics for the research variables, providing an overview of the data’s central tendency, dispersion, and distribution. The Mean values for all items range from 3.574 to 4.136, indicating that respondents generally rated the statements positively across all variables. In particular, Attitude item ATT5 recorded the highest mean score (M = 4.136), suggesting a strong, favorable evaluation of this aspect. In contrast, the Perceived Behavioral Control item PBC3 had the lowest mean score (M = 3.574), indicating relatively lower agreement regarding ease or control over revisiting. The Standard Deviation (S.D.) values range from 0.799 to 1.021, reflecting moderate variability in responses.

Generally, the descriptive results revealed that the respondents have a favorable view of all explained variables, i.e., positive attitudes/experiences towards all attitude/experience-related items and relatively negative attitudes towards all PBC items. The ranges of skewness and kurtosis were acceptable for further CFA and SEM analysis.

4.2. PLS-SEM Analysis

This study employed Partial Least Squares Structural Equation Modeling (PLS-SEM) to examine the causal relationships among variables influencing tourists’ willingness to revisit Wutai Mountain.

4.2.1. Convergent Validity

Convergent validity was assessed using factor loadings, composite reliability (CR), and average variance extracted (AVE) for each latent variable (Cheung et al., 2024). The factor loadings for all observed variables are presented in

Table 4. Convergence Validity.

Latent Variables	Observation Items	Factor Loading	CR	AVE	VIF
Attitude towards Temple Tourism	ATT1	0.859	0.906	0.659	2.468
	ATT2	0.806			2.001
	ATT3	0.809			1.858
	ATT4	0.839			2.235
	ATT5	0.741			1.610
Perceived behavioral control	PBC1	0.851	0.925	0.673	2.638
	PBC2	0.838			2.399
	PBC3	0.764			1.788
	PBC4	0.841			2.474
	PBC5	0.799			2.062
	PBC6	0.826			2.186
Social media exposure	SME1	0.861	0.915	0.781	1.909
	SME2	0.904			2.432
	SME3	0.886			2.391
Subjective norms	SN1	0.895	0.920	0.742	2.689
	SN2	0.853			2.186
	SN3	0.823			1.995
	SN4	0.874			2.502
Tourists’ enjoyment	TE1	0.869	0.909	0.769	2.026
	TE2	0.881			2.259
	TE3	0.880			1.992
Tourists’ memorability	TM1	0.857	0.889	0.668	2.187
	TM2	0.824			1.882
	TM3	0.830			1.906
	TM4	0.754			1.510
Tourists’ revisit intention to Wutai temple	TVT1	0.827	0.885	0.719	1.576
	TVT2	0.860			1.879
	TVT3	0.857			1.860

and range from 0.741 to 0.904, all exceeding the threshold of 0.70. This means that each indicator is highly and significantly correlated with its corresponding latent variable and the indicators sufficiently reflect their corresponding underlying concepts (Legate et al., 2021). CR was calculated as

$$CR={\displaystyle \frac{{\left({\displaystyle \sum {\lambda }_i}\right)}^2}{{\left({\displaystyle \sum {\lambda }_i}\right)}^2+{\displaystyle \sum {\theta }_i}}}$$

where λ_i represents the standardized factor loading, and θ_i indicates the measurement error variance. In addition, the AVE was computed to assess the convergent validity of the constructs as shown below:

$$AVE={\displaystyle \frac{{\displaystyle \sum {\lambda }_i^2}}{{\displaystyle \sum {\lambda }_i^2}+{\displaystyle \sum {\theta }_i}}}.$$

The results show that the AVE values for all constructs exceed the 0.50 threshold. This indicates that the corresponding latent variables account for more than half of the variance in their observed variables. Therefore, the results of the CR and AVE tests prove that each variable has strong internal consistency and good convergent validity.

AVE is used to measure the proportion of the variance in a latent variable that is explained by its observed variables, relative to measurement error. In other words, it reflects the extent to which latent variables account for the variance in their observed indicators. In this study, the AVE values of all constructs exceeded the critical value of 0.50. Thus, the results of CR and AVE indicate that each variable has high internal consistency and good convergent validity.

As shown in Table 4, the variance inflation factor (VIF) values for all measurement indicators range from 1.51 to 2.69, which are well below the commonly accepted threshold of 10. This indicates that multicollinearity is not a concern in this study, that there is no redundancy among the indicators of each latent construct, and that the measurement model is structurally stable for subsequent structural model analysis.

4.2.2. Discriminant Validity

Table 5. Discriminant Validity Test.

	ATT	PBC	SME	SN	TE	TM	TVT
ATT	0.812
PBC	0.328 ***	0.820
SME	0.305 ***	0.237 ***	0.884
SN	0.327 ***	0.363 ***	0.225 ***	0.862
TE	0.459 ***	0.355 ***	0.311 ***	0.312 ***	0.877
TM	0.342 ***	0.377 ***	0.203 ***	0.505 ***	0.358 ***	0.817
TVT	0.388 ***	0.514 ***	0.329 ***	0.488 ***	0.443 ***	0.398 ***	0.848

Note: The diagonal elements represent the square root of the Average Variance Extracted (AVE) for each construct. ATT: Attitude toward Temple Tourism; PBC: Perceived Behavioral Control; SME: Social Media Exposure; SN: Subjective Norm; TE: Tourist Enjoyment; TM: Tourist Memorability; TVT: Tourists’ Revisit Intention to Wutai Temple. *** represents significance at the 1% level.

presents the results of the discriminant validity test, an important step in verifying whether the latent variables in the research can be conceptually distinguished from one another. The condition for the establishment of discriminant validity is that the square root of the average variance extracted for each construct (indicated in bold on the diagonal of the matrix) is greater than the correlation coefficient between that construct and other variables in the model (Rokonuzzaman et al., 2022).

The results show that the square roots of the AVEs for all latent variables are higher than their corresponding cross-conceptual correlation coefficients. For example, the square root of ATT’s AVE is 0.812, which is higher than its correlation values with PBC (0.328), SME (0.305), SN (0.327), TE (0.459), TM (0.342), and TVT (0.388). Other variables also exhibit similar situations, indicating that the shared variance of each concept with its own measurement indicators is greater than its correlation with other concepts. Therefore, this measurement model can effectively distinguish between concepts because no cross-concept correlation coefficient exceeds the square root of the AVE for its corresponding concept. Thus, the measurement model of this study demonstrates good discriminant validity and ensures that each latent variable reflects the expected theoretical dimensions independently.

Table 6. Reliability Statistics.

Study Variables	R2	Q2
TVT	0.722	0.493

Note: TVT: Tourists’ Revisit Intention to Wutai Temple.

presents the test results of the explanatory power and predictive validity of TVT. R² represents the proportion of variance in the dependent variable explained by the independent variables in the model. Generally, when R² exceeds 0.2, the model is considered to have strong explanatory power (Liengaard et al., 2020). In this study, the TVT R² is 0.722, indicating that the model’s predictive variables explain 72.2% of the variance. This indicates that the model is highly effective at explaining tourists’ willingness to revisit Wutai Mountain. Q² is calculated using a blindfold procedure and measures the predictive correlation of the model. When Q² is greater than 0, the model has acceptable predictive ability (Liengaard et al., 2020). The Q² value of TVT in this study is 0.493, which is significantly higher than zero. Therefore, the model shows strong predictive validity for TVT. In summary, the research results show that this model has strong explanatory power and predictive validity, effectively revealing tourists’ willingness to revisit Wutai Mountain and demonstrating its robustness in identifying key influencing factors.

To assess the global fit of the PLS-SEM model, the Goodness-of-Fit (GoF) index proposed by Tenenhaus et al. (2005) was calculated. The GoF is computed as the square root of the product of the average AVE and the average R² of the endogenous constructs. The results show that the GoF value reached 0.719, exceeding the threshold of 0.36 for large effect sizes, indicating that the proposed model demonstrates strong overall fit. This suggests that the model has both satisfactory measurement quality and substantial explanatory power for tourists’ revisit intention.

4.2.3. Path Analysis for Direct Effects

Figure 3 presents the results of the structural equation model generated using SmartPLS, clearly illustrating the hypothesized relationships among the latent variables, including the direct effects and the tested moderating paths. The measurement model is reflected through factor loadings (shown beside the arrows between the yellow rectangles and their corresponding latent variables), and all values exceed the recommended threshold of 0.7 (e.g., TE2 = 0.881, SME2 = 0.904), indicating that the indicators have high reliability and validating the convergent validity as shown in Table 4. Figure 3 also provides intuitive verification of both the measurement and structural models, showing that each construct has good stability and validity and further revealing the direct connections between the research variables, thereby supporting the proposed theoretical framework.

As shown in

Table 7. Structural Model Path Test.

Path	β	S.E.	T	p	Results
ATT → TVT	0.132	0.026	5.040	0.000	Supported
PBC → TVT	0.213	0.027	7.934	0.000	Supported
SME → TVT	0.087	0.042	2.082	0.037	Supported
SN → TVT	0.189	0.027	6.940	0.000	Supported
TE → TVT	0.194	0.027	7.199	0.000	Supported
TM → TVT	0.168	0.029	5.750	0.000	Supported
SME × TE → TVT	0.096	0.031	3.101	0.002	Supported
SME × TM → TVT	0.260	0.049	5.359	0.000	Supported
SME × ATT → TVT	0.155	0.042	3.671	0.000	Supported
SME × SN → TVT	0.128	0.049	2.645	0.008	Supported
SME × PBC → TVT	0.048	0.033	1.468	0.142	Not Supported

, the results of the SEM indicate that most of the proposed hypotheses were statistically supported.

For the direct effects, tourists’ enjoyment (TE → TVT, β = 0.194, p < 0.001) exerted a substantial positive effect on revisit intention, thus confirming H1. Tourists’ memorability (TM → TVT, β = 0.168, p < 0.001) also exhibited a meaningful impact, validating H2. Attitude towards temple tourism (ATT → TVT, β = 0.132, p < 0.001) was a positive predictor of revisit behavior, supporting H3, while subjective norm (SN → TVT, β = 0.189, p < 0.001) proved statistically meaningful, confirming H4. Behavioral control perception (PBC → TVT; β = 0.213, p < 0.001) had the most significant impact among the direct predictors, supporting H5. Moreover, social media exposure (SME → TVT, β = 0.087, p = 0.037) exerted a direct favorable influence on revisit intention, affirming its hypothesized role. Overall, these results provide empirical support for hypotheses H1–H5 and emphasize the critical role of emotional, cognitive, and socio-psychological determinants in shaping revisit intentions.

For the moderating effects, social media exposure significantly strengthened the relationships between tourists’ enjoyment and revisit intention (SME × TE → TVT, β = 0.096, p = 0.002; H6 supported), tourists’ memorability and revisit intention (SME × TM → TVT, β = 0.260, p < 0.001; H7 supported), attitude and revisit intention (SME × ATT → TVT, β = 0.155, p < 0.001; H8 supported), and subjective norm and revisit intention (SME × SN → TVT, β = 0.128, p = 0.008; H9 supported). Notably, the moderating effect is most substantial for memorability. This can be explained by the nature of memorable experiences, which lead to deeper emotional encoding and longer persistence in tourists’ cognitive structures. When such memory traces are later reactivated through destination-related images, narratives, or symbolic cues on social media, they trigger an intensified emotional recall, significantly enhancing tourists’ psychological connection to the site. As a result, social media exposure amplifies memory-based evaluations more strongly than other psychological antecedents, producing the most considerable moderating effect on revisit intention.

However, the moderating effect of social media exposure on the relationship between perceived behavioral control and revisit intention was not significant (SME × PBC → TVT, β = 0.048, p = 0.142), leading to the rejection of H10. As illustrated in Figure 4, the simple slope plots clearly depict the moderating effects of social media exposure, further confirming the significance and direction of the interaction terms in the structural model. For H6–H9, the slopes for high social media exposure are noticeably steeper than those for low exposure, indicating stronger positive effects of enjoyment, memorability, attitude, and subjective norm on revisit intention under conditions of greater exposure. By contrast, the slope difference for H10 is minimal, visually supporting the non-significance of the moderating effect.

The lack of support for H10 may be explained by the fact that perceived behavioral control already exerts a relatively strong and stable direct effect on revisit intention, leaving limited variance to be further amplified through social media exposure. In temple tourism, judgments about feasibility (e.g., time, budget, physical condition, and travel constraints) are often shaped by more structural and enduring factors that are not easily altered by additional online information. At the same time, most digital media content about Mount Wutai tends to emphasize the destination’s attractiveness, atmosphere, and symbolic meaning rather than providing instrumental guidance that could substantially change tourists’ perceptions of feasibility or resource availability. Under these conditions, social media exposure is more likely to reinforce emotional and attitudinal responses than to moderate the PBC–revisit pathway, which may account for the non-significant interaction effect observed for H10.

Overall, the findings support most theoretical assumptions, indicating that social media exposure exerts not only a direct impact on tourists’ revisit intention but also a favorable moderating effect across several critical paths. This underscores the role of psychological dimensions, including attitude, emotional enjoyment, memorability, and normative influences, in shaping tourists’ revisit behavior.

5. Discussion

5.1. Theoretical Implication

Based on the temple tourism context of Mount Wutai, this study establishes an integrated analytical framework combining attribution theory (AT), the theory of planned behavior (TPB), and the elaboration likelihood model (ELM) to examine the determinants of revisit intention. Using PLS-SEM with 617 valid responses, the results show that tourists’ enjoyment, memorability, attitude, subjective norm, and perceived behavioral control all exert significant positive effects on revisit intention, with PBC demonstrating the strongest influence, highlighting the critical role of resource availability, time constraints, and accessibility in revisit decisions. Enjoyment and subjective norm follow in explanatory strength, underscoring the importance of emotional experience and social support, while memorability and attitude also contribute significantly. Regarding moderating effects, social media exposure significantly strengthens the relationships between enjoyment, memorability, attitude, and subjective norm with revisit intention, with the strongest moderation observed for memorability. However, the moderating effect of social media exposure on the relationship between perceived behavioral control and revisit intention is not significant, indicating that digital media is limited in alleviating structural constraints such as cost, time, and transportation.

Previous tourism studies grounded in Attribution Theory have largely explained tourists’ emotions and revisit behavior through functional attributes such as service quality and destination facilities (Tam et al., 2012). However, the present results indicate that, in sacred tourism, enjoyment and memorability are more strongly rooted in internalized spiritual meaning, cultural identification, and symbolic interpretation, suggesting that externally driven attribution pathways may be insufficient in religious heritage contexts. Within the Theory of Planned Behavior, prior leisure tourism research often reports a relatively weak or inconsistent role of subjective norms, whereas the strong normative effect observed here can be explained by the collectivist cultural structure and the socially embedded nature of religious practice in China (Jingyi & Jiangli, 2024). This implies that TPB operates through a more norm-driven mechanism in temple tourism than in conventional leisure settings. Furthermore, this study reveals that perceived behavioral control is the most significant factor in predicting the intention to revisit Wutai Mountain.

Although social media has frequently been treated as a direct predictor of revisit intention (Suo et al., 2024), the results suggest that its primary function lies in amplifying emotional, memorial, and normative pathways rather than directly reshaping feasibility judgments. The non-significant moderation effect on perceived behavioral control is consistent with Carvalho et al. (2025), who argues that evaluations of time, cost, and accessibility are largely governed by structural constraints that are not easily altered through online exposure. Collectively, these differences indicate that the formation mechanism of revisit intention in sacred tourism diverges structurally from that in general leisure tourism, requiring contextualized revisions to AT, TPB, and ELM.

5.2. Practical Implications

From a practical perspective, the findings provide several actionable insights for tourism stakeholders.

For local tourism management authorities, accessibility and basic infrastructure remain important. However, greater emphasis should be placed on developing structured and immersive experience routes, such as guided meditation paths, themed pilgrimage circuits, and integrated cultural interpretation centers. Strengthening the official presence on digital platforms through curated short videos, virtual tours, and real-time visitor service information can enhance destination visibility and maintain visitor engagement beyond their physical visit.

For temple administrations and religious service providers, designing structured experiential programs, such as ritual participation sessions, chanting demonstrations, calligraphy or incense workshops, and novice-for-a-day activities, can deepen visitors’ emotional connection with Buddhist culture. Coordinating these onsite experiences with online dissemination (e.g., livestreamed rituals, digital storytelling, or behind-the-scenes content) can help transform a one-time visit into a sustained spiritual relationship, thereby increasing the likelihood of repeat visits.

For local hospitality businesses and cultural media practitioners, the significant moderating role of social media exposure underscores the need to produce authentic, aesthetically appealing, and story-driven digital content that conveys the symbolic and spiritual essence of Mount Wutai. Hotels, restaurants, craft shops, and media studios can collaborate to develop themed cultural products, such as Buddhist-inspired cuisine, ritual-themed accommodation packages, or festival-based experiential events, that enhance cultural immersion and stimulate revisit intentions through both online influence and offline experience integration.

5.3. Research Limitations and Future Prospects

Despite its contributions, this study has several limitations. First, all variables were measured using self-reported Likert scales, which may introduce social desirability and common-method biases, although anonymity and randomized item ordering were applied; future studies may employ multi-source data, Harman’s single-factor test, or experimental designs to enhance robustness. Second, conceptual proximity may exist among certain constructs, particularly between memorability and attitude, both of which involve affective evaluation; future research could strengthen discriminant validity through finer-dimensional structures or qualitative approaches. Third, this study focuses on the integrated mechanism of emotional, cognitive, and digital factors and does not incorporate demographic characteristics, trip-related variables, or constructs such as destination image, cultural identity, and existential authenticity, which may exert mediating or moderating effects. Fourth, social media exposure is modeled as a static moderator, and the cross-sectional design limits the examination of dynamic digital interactions; longitudinal studies, big-data analytics, or behavioral log analysis are therefore recommended. Finally, the sample was drawn solely from Mount Wutai, which may restrict the generalizability of the findings and calls for further validation through cross-regional or cross-cultural studies.

6. Conclusions

This study demonstrates that enjoyment, memorability, attitude, subjective norm, and perceived behavioral control (PBC) all significantly enhance revisit intention, with PBC emerging as the most powerful predictor. This highlights that, even in spiritually oriented destinations, revisit decisions remain strongly shaped by feasibility-related considerations such as time, cost, and accessibility. At the same time, enjoyment and memorability reflect visitors’ emotionally rich and meaning-laden encounters with sacred space, while the strong effect of subjective norm suggests that revisit intention in Chinese temple tourism is deeply embedded in collective cultural values and socially supported religious practice.

The study contributes to theory in three key ways. First, it extends AT by showing that in sacred tourism, emotional outcomes and memory formation may be less dependent on purely external service or facility attributions and more reliant on internalized spiritual meaning, cultural identity, and symbolic interpretation. Second, it contextualizes TPB by revealing a more norm-driven mechanism than is commonly observed in general leisure settings, reinforcing the importance of culturally situated interpretations of social influence. Third, it clarifies the role of social media within an ELM-informed perspective: rather than directly altering feasibility judgments, social media exposure primarily amplifies emotional, memorial, attitudinal, and normative pathways to revisit intention, with the strongest strengthening effect found for memorability. The non-significant moderation on PBC further implies that digital environments cannot easily offset structural constraints that shape travel feasibility.

Practically, the results suggest that sustainable revisit growth for Mount Wutai requires a dual focus on improving travel feasibility and deepening meaning-centered experiences. Destination authorities should continue enhancing accessibility and visitor services while designing immersive pilgrimage and interpretation routes supported by coherent digital communication. Temple administrations can strengthen repeat visitation by offering structured Buddhist cultural engagement programs and linking onsite rituals with authentic online storytelling. Local businesses and cultural media practitioners should collaborate on aesthetically appealing, spiritually respectful content and themed products that align online inspiration with offline experience quality.