Perceived Value and Behavioral Intentions of Cultural Heritage Visitors: A SEM Analysis Using Lingnan Classical Gardens as a Case Study

Abstract

In the context of urban cultural heritage tourism, this study examines how visitors’ expectations and perceptions contribute to satisfaction and loyalty. We develop and validate an integrative model using structural equation modeling (SEM) on data collected from a large-scale visitor survey of Lingnan classical gardens in China, supplemented by quantitative simulation analysis. The results reveal that Cultural Expectation (CE) and Spatial Expectation (SE) serve as key antecedents to Perceived Cultural Immersion (PCI) and Perceived Spatial Aesthetics (PSA), respectively. PCI and PSA both positively influence visitors’ Perceived Value (PV), which in turn drives overall Satisfaction (SAT) and ultimately enhances Loyalty (LOY). Notably, PCI and PSA act as mediators in the model, channeling the influence of initial expectations into value perception and satisfaction. By highlighting the roles of cultural immersion and spatial aesthetics, this research extends the perceived value–satisfaction–loyalty paradigm in cultural heritage tourism. Theoretically, it bridges urban cultural heritage and environmental psychology with place-attachment theories. Practically, the findings offer guidance for cultural landscape management and visitor experience enhancement, suggesting that narrative storytelling, interactive cultural engagement, and authentic spatial design can elevate perceived value, satisfaction, and loyalty. This article will provide valuable reference material for urban planners.

1. Introduction

The rise of cultural heritage tourism stems from travelers’ desire to experience the history and traditions of destinations, encompassing tangible and intangible assets such as heritage sites, arts, rituals, and cuisine [1,2]. This sector has become one of the fastest-growing segments within the tourism industry, accounting for approximately 40% of global tourism volume [3,4]. Many regions view heritage tourism as a key strategic approach to driving economic growth and cultural development [5]. However, the global proliferation of heritage sites has intensified competition among cultural destinations [6]. To attract culturally oriented visitors, World Heritage cities and other culturally rich areas often create distinctive experiences by integrating unique local customs, centuries-old histories, and artistic traditions [7]. Against this competitive backdrop, effective management of heritage sites has become crucial to ensuring high-quality visitor experiences. Sound site management enhances visitor satisfaction throughout pre-, during, and post-visit stages, thereby fostering a differentiated advantage in cultivating visitor loyalty [8]. Facing increasingly fierce competition for visitors, continuously optimizing the tourism experience has become central to destinations maintaining their competitiveness [9]. Ultimately, the success of cultural heritage tourism hinges on creating memorable and highly satisfying experiences that prompt visitors to return and actively recommend the destination to others [10,11].

In China, unlike the Jiangnan gardens that pursue philosophical meaning or the northern gardens that flaunt imperial power, the core spirit of Lingnan gardens lies in the secular life aesthetics of “harmonizing garden and dwelling.” Their spatial planning exhibits a clear functionalist orientation, featuring significantly higher building density to meet practical living needs such as residence, entertaining guests, banquets, and summer retreats. Gardens are not merely objects of contemplation but vessels for daily life. This profound practicality and livability embody the value pursuit of garden owners, predominantly merchants and gentry, for earthly contentment and familial prosperity. The cultural heritage sites represented by the Four Great Gardens of Lingnan, Qinghui Garden, Liang Garden, Ke Garden, and Yuyin Mountain House each possess distinctive characteristics. Though sharing a similar regional historical context, each garden exhibits a unique style in spatial layout and cultural significance, collectively forming the rich spectrum of Lingnan classical garden art [12,13,14,15]. For instance, Qinghui Garden employs a vertically progressive layout to create a dramatic, step-by-step pacing for visitors; Yuyin Mountain House presents an exquisite, compact aesthetic within its confined space through a surrounding structure; Ke Garden connects diverse spaces with a circular corridor, with the Mountain-Inviting Pavilion as its visual focal point, embodying an exploratory, divergent spatial arrangement. Liang Garden, meanwhile, creates a free-roaming, contemplative atmosphere through its scattered, interconnected landscapes. The four major gardens each reflect distinct cultural connotations from the refined gatherings and tastes of the literati class to the sophisticated worldview constructed by Confucian merchant families, highlighting the rich historical and cultural value of Lingnan gardens [16,17,18,19]. These distinctive characteristics make Lingnan gardens an ideal case study for examining how cultural and spatial elements shape visitors’ perceived value (Figure 1,

Table 1. Analysis of the four major characteristics of Lingnan Gardens.

Garden Name	Spatial Structure	Spatial Character	Cultural Characteristics	Approximate Area
Qinghui Garden	Sequential progression	Dramatic progression	The Elegant Gatherings and Inclusivity of the Literati	22,000 m2
Yuyin Mountain House	Responsive Surround	Instantaneous Intensity	The Refined Universe of Confucian Merchants	1590 m2
Ke Garden	Radial Scatter Pattern	Wandering Divergence	A Gathering Place for Literati and Scholars	2200 m2
Liang Garden	Encircling Responsiveness	Exploratory Encircling	The Literati’s View of Nature and Nostalgia for the Homeland	21,000 m2

).

This study employs the American Customer Satisfaction Index (ACSI) model as its theoretical framework to explore the mechanism underlying the relationship between visitor satisfaction and loyalty. This model has been widely adopted in tourism research, particularly within the cultural heritage tourism sector, to analyze the formation process of visitor satisfaction and loyalty [20,21,22,23,24,25]. The ACSI framework conceptualizes the tourism experience as a dynamic process: beginning with travelers’ pre-trip expectations of heritage destinations and concluding with post-visit evaluations of the actual experience across multiple perceptual dimensions. Pre-trip expectations and post-trip perceptions jointly form a two-dimensional evaluation of experience quality. Their interaction further shapes overall perceived value—that is, tourists’ dynamic assessment of the benefits gained relative to the costs incurred during their visit. Perceived value directly determines tourists’ satisfaction levels with the destination, while satisfaction serves as a direct antecedent to loyalty, prompting tourists to develop a willingness to recommend the destination and engage in repeat visits.

Environmental psychology indicates that space is not merely a physical construct but also a domain for emotional restoration and psychological adjustment. Against this theoretical backdrop, the aesthetics of perceived space should encompass not only formal beauty but also the relaxation and restorative qualities that space provides. For instance, when visitors stroll through gardens, pause, and gaze, these actions inherently serve a “psychological stress-relief” function. Urban heritage theory places particular emphasis on authenticity and a sense of place, stressing that cultural landscapes should center on storytelling and evoking memories as their core values. This provides a compelling supplementary explanation for the concept of “cultural immersion” proposed in this paper: true immersion stems not merely from decoration or symbols, but from whether the historical narratives embedded within a space resonate deeply with the heart. Furthermore, the theory of “narrative transportation” suggests that when visitors can immerse themselves in a narrative world within a space, their experience becomes more profound and emotionally engaging. This theory can extend existing structural models by introducing variables such as “cultural context awareness” or “historical engagement” to deepen the analysis of emotional mechanisms within cultural spaces. Recent advances in environmental psychology and heritage tourism research indicate that visitor experiences transcend mere perception, encompassing deeper psychological states such as place attachment, restorative experiences, and narrative engagement. These concepts provide crucial insights into how individuals establish cognitive and emotional connections with spatial and cultural environments. Place attachment reflects psychological bonds with specific settings, fostering emotional resonance and behavioral intentions; restorative environments, meanwhile, facilitate psychological recovery and stress reduction. In heritage studies, concepts such as authenticity and narrative traffic have deepened our understanding of how immersive narratives and symbolic coherence shape visitor engagement. While our model captures perceptual dimensions like spatial aesthetics and cultural immersion, integrating these established concepts into our theoretical framework would enhance our model’s explanatory power—particularly in elucidating the transformative aspects of immersive psychological states.

This study extends the traditional heritage tourism satisfaction-loyalty model in two aspects. First, we identified two key dimensions of perceived value in heritage experiences: the first focuses on cultural content and intangible cultural heritage, manifested through visitors’ cultural expectations and subsequent perceptions of cultural immersion; Second, it focuses on the physical environment and tangible heritage, reflected through visitors’ spatial expectations and the spatial aesthetics of their subsequent perceptions. By categorizing visitor expectations and actual perceptions into cultural and spatial dimensions, respectively, it enables a more precise capture of the rich connotations of visitor experiences within classical garden settings. Furthermore, this study explores the potential moderating effects of these dimensions.

Beyond service quality factors, prior studies have identified other important influences on visitor loyalty. For example, tourists’ emotional attachment to a heritage site and the comfort they feel in a heritage space can directly encourage revisit intentions [26,27]. Among the various determinants, the perceived value of the experience is often highlighted as a crucial predictor of repeat visitation [28,29,30,31,32]. If tourists perceive that the benefits of their visit do not sufficiently outweigh its costs, their likelihood of returning drops markedly [33].

Given its importance, perceived value has been integrated into tourist behavior models to better understand decision-making. For example, Shneikat B. et al. incorporated perceived value into a behavioral economics model of tourism, demonstrating that this construct significantly guides tourist decisions and providing practical insights for cultural tourism management [34]. Moreover, value-driven tourists tend to engage in extensive information search and careful decision-making; the utility they derive from their choices correlates positively with their satisfaction levels [35]. High perceived value also translates into stronger loyalty outcomes: tourists who feel they obtained excellent value are more likely to generate positive word-of-mouth and to plan repeat visits [36,37]. These findings reinforce the central role of perceived value in shaping visitor satisfaction and loyalty, justifying its prominent position in our research framework.

In light of these considerations, the present study employs a Structural Equation Modeling (SEM) approach to examine how perceived value relates to satisfaction and behavioral intentions (loyalty) in the context of Lingnan classical gardens. SEM is chosen because it allows simultaneous evaluation of multiple latent constructs and their interrelationships, providing a robust framework for testing our theoretical model of tourist behavior. Using SEM, we assess both the direct and indirect effects (via satisfaction) of perceived value on tourists’ loyalty intentions within a single integrative model.

The objectives of this research are threefold. First, we aim to empirically evaluate how tourists’ perceived value of the Lingnan garden experience impacts their satisfaction and their subsequent intentions to revisit the site or recommend it to others. Second, we examine the extent to which tourist satisfaction mediates the influence of perceived value on loyalty outcomes, thereby clarifying how value perceptions translate into repeat visitation or referrals. Finally, by focusing on an East Asian cultural heritage context that has been relatively under-examined in the international literature, we seek to broaden the understanding of tourist behavior in heritage settings. Through this case study of Lingnan classical gardens, we also test whether the established relationships in tourism theory (such as the value–satisfaction–loyalty chain) hold true in a Chinese cultural heritage context.

Building on relevant literature and established models [38,39,40,41,42,43,44,45,46,47,48,49,50], This study conceptualizes the formation process of tourist satisfaction as a framework illustrated in Figure 2. Cultural expectations (CE) and spatial expectations (SE) prior to travel, respectively, elicit corresponding post-travel perceptions—perceived cultural immersion (PCI) and perceived spatial aesthetics (PSA). These two types of perceptual outcomes collectively shape the overall perceived value (PV) of the experience, which in turn influences visitor satisfaction (SAT) and ultimately drives the development of loyalty (LOY). Based on this framework, we propose the following hypotheses:

H1: 

The higher the level of cultural expectations (CE) among tourists, the stronger their perceived cultural immersion (PCI).

H2: 

The higher the level of spatial expectations (SE) among visitors, the higher their perceived spatial aesthetic level (PSA).

H3: 

The stronger a visitor’s Perceived Cultural Immersion (PCI), the higher their Perceived Overall Experience Value (PV).

H4: 

The higher a visitor’s Perceived Spatial Aesthetics (PSA) level, the greater the perceived overall experience value (PV).

H5: 

The overall perceived value (PV) of tourists has a significant positive impact on their satisfaction (SAT).

H6: 

Tourist satisfaction (SAT) has a significant positive impact on their loyalty intention (LOY).

H7: 

Cultural immersion exerts a positive moderating effect on the relationship between perceived value (PV) and satisfaction (SAT), meaning that when tourists experience stronger cultural immersion, the promotional effect of PV on SAT becomes more pronounced.

H8: 

Spatial aesthetic perception exerts a positive moderating effect on the relationship between perceived value (PV) and satisfaction (SAT). Specifically, when visitors exhibit higher spatial aesthetic perception, the positive influence of PV on SAT becomes more pronounced.

This study contributes to the theoretical framework of cultural heritage tourism and consumer behavior research in three key ways. First, by examining the drivers of visitor loyalty within non-Western cultural heritage contexts, it extends the geographical and cultural applicability of mainstream theoretical models. Second, the findings reveal the pivotal role of perceived value in influencing visitor satisfaction and loyalty, further confirming the critical importance of delivering exceptional perceived value in fostering repeat visit intentions and positive word-of-mouth. Finally, the composite-dimensional measurement framework constructed in this study provides a more refined theoretical tool for understanding the value formation mechanisms within cultural heritage experiences. At the practical level, this study provides concrete guidance for managers and policymakers of cultural heritage sites such as classical gardens. By identifying key pathways to enhance perceived value and satisfaction, destinations can more effectively optimize service offerings and experiential quality. For instance, interactive interpretation systems that integrate educational content and emotional depth, continuous improvements to visitor service facilities, and maintaining the authenticity and spatial aesthetic quality of venues are all viable strategies for enhancing perceived value. These measures can effectively boost visitor satisfaction, thereby increasing repeat visit intentions and fostering positive word-of-mouth. Against the backdrop of intensifying competition in cultural heritage tourism, systematically enhancing visitors’ perceived value and loyalty has become central to achieving sustainable development in heritage sites [51,52].

2. Methods

2.1. An Inquiry into Action and Time

To overcome the limitations of cross-sectional survey designs in explaining causal relationships prior to the formal questionnaire survey, this study first conducted research on the “Behavioral and Temporal Dimensions” section within the questionnaire. This aimed to capture behavioral characteristics such as visitor frequency, time of last visit, and motivations for repeat visits. This supplementary design facilitates controlling for differences between repeat visitors and first-time visitors in subsequent models, thereby enhancing the model’s interpretability and external validity (

Table 2. Survey Questionnaire Structure.

Dimension	Item Wording
Visit Frequency	How many times have you visited this garden in the past 12 months? (0/1/2–3/4–6/≥7)
Recency	When was your last visit to this garden? (This week/This month/Within the last 3 months/Within the last 6 months/More than 6 months ago)
Repeat Visit Motivation	What is the primary purpose of your visit? (Multiple selections allowed: Leisure/Study/Accompanying family/Academic research/Participating in festival activities/Other, please specify)
Visit History	Is this your first visit? (Yes/No)
	If not, please enter the cumulative number of visits: ____ times
Retrospective Change in Expectation	Compared to my last visit, my expectations for the garden have changed significantly. (1 = significantly decreased, 4 = unchanged, 7 = significantly increased)

). Regarding park visitation patterns, approximately 38.2% of respondents were repeat visitors, with 21.5% indicating they had visited two or more times within the past 12 months. The primary motivations for visiting include leisure (45.3%), educational experiences (22.1%), accompanying family members (18.6%), and cultural research (8.5%). This distribution indicates that visitors to Lingnan gardens exhibit strong cultural affinity and high potential for repeat visits.

2.2. Structural Equation Modeling Method

The present study employs a structural equation modeling (SEM) approach to investigate the relationships among perceived value, satisfaction, and behavioral intention of cultural heritage tourists visiting Lingnan classical gardens. SEM is particularly well-suited for analyzing complex causal relationships among latent variables in tourism research [53,54,55]. Data collection was conducted by distributing structured questionnaires to visitors at four representative Lingnan gardens, ensuring coverage of a diverse visitor demographic. The survey instrument comprises multiple-dimensional scales measuring cultural expectations, spatial expectations, perceived cultural immersion, perceived spatial aesthetics, perceived value, satisfaction, and loyalty. All items utilize a 7-point Likert scale, with demographic information collected concurrently. Furthermore, to overcome the limitations of cross-sectional survey designs in explaining causal relationships, this study incorporated a “Behavior and Time Dimension” section into the questionnaire. This section captures behavioral characteristics such as visitor frequency, time of last visit, and motivations for repeat visits. This supplementary design facilitates controlling for differences between repeat visitors and first-time visitors in subsequent modeling, thereby enhancing the model’s explanatory power and external validity (

Table 3. Survey questionnaire.

Dimension	Indicators
Cultural Expectation (CE)	CE1: I look forward to experiencing the unique atmosphere of traditional culture within Lingnan gardens. CE2: Lingnan gardens allow me to experience a rich sense of history and culture. CE3: I believe Lingnan gardens embody the essence of local culture. CE4: I hope to deepen my understanding of Lingnan culture through visiting Lingnan gardens. CE5: I expect Lingnan Gardens to provide culturally enriching content.
Spatial Expectation (SE)	SE1: I expect the spatial layout of Lingnan gardens to be pleasing to the eye. SE2: I hope the garden environment allows me to feel relaxed and tranquil. SE3: I anticipate the garden space will provide an escape from the urban clamor. SE4: I expect the garden to offer excellent views and an aesthetic experience. SE5: I believe gardens should possess a strong sense of integration between nature and culture.
Perceived Cultural Immersion (PCI)	PCI1: I experience profound cultural immersion within Lingnan gardens. PCI2: I feel as though I’m immersed in Lingnan’s traditional culture. PCI3: Lingnan gardens evoke a strong sense of national pride in me. PCI4: The cultural environment of the gardens allows me to forget the real world. PCI5: I can strongly sense the cultural stories behind Lingnan gardens.
Perceived Spatial Aesthetics (PSA)	PSA1: I find the architecture and landscapes of Lingnan gardens exceptionally beautiful. PSA2: The garden design allows me to experience the fusion of art and nature. PSA3: I appreciate the artistry in how the gardens utilize space. PSA4: The detailed design of the garden spaces leaves a deep impression on me. PSA5: The overall appearance of the gardens meets my expectations of beauty.
Perceived Value (PV)	PV1: Visiting the Lingnan garden felt well worth the price. PV2: I found this garden experience highly worthwhile. PV3: The cultural experience offered by the garden exceeded my expectations. PV4: I believe visiting the garden provided a positive emotional return. PV5: I would be willing to pay for this kind of cultural experience.
Satisfaction (SAT)	SAT1: Overall, I was satisfied with my visit to Lingnan Garden. SAT2: The services and environment provided by the garden made me feel happy. SAT3: The garden experience met my initial expectations. SAT4: I was very satisfied with this trip to Lingnan Garden. SAT5: I have a very favorable overall impression of the garden.
Loyalty (LOY)	LOY1: I would be willing to revisit Lingnan Gardens. LOY2: I would recommend Lingnan Gardens to friends or family. LOY3: I may revisit Lingnan Gardens multiple times in the future. LOY4: I would like to share the cultural experience of the gardens with others. LOY5: I remain deeply devoted to Lingnan gardens.

).

The reliability and validity of the measurement model were assessed using Cronbach’s alpha, composite reliability, and average variance extracted (AVE). Following recent methodological literature recommendations, discriminant validity was examined using a dual criterion of cross-loadings and the heterogeneity-to-homogeneity ratio. The classification of key variables and constructs is grounded in existing theoretical frameworks within the field of heritage tourism, thereby supporting robust testing of the moderating and mediating effects specified within the model.

Data collection for this study was conducted during the peak tourist season from May to August 2025. We employed an intercept survey method at the exits of the four Lingnan gardens mentioned above. At each site, trained interviewers randomly invited every fifth visitor to participate in the survey. All respondents provided informed consent, clearly understanding that the study was solely for academic purposes and that their responses would remain anonymous. Approximately 5999 questionnaires were distributed, with 5687 valid responses collected, corresponding to a valid response rate of approximately 94.8%. The sample covers visitors across all age groups, genders, educational levels, and regions, aligning with the fundamental demographic distribution of the scenic area’s overall visitor population and thus demonstrating strong representativeness. The questionnaire is divided into five sections:

(1)

Cultural Expectations, Perceived Cultural Immersion, and Perceived Value in Lingnan Gardens: This section evaluates visitors’ expectations and actual perceptions of the cultural dimension through intangible cultural heritage techniques, Lingnan literati culture, cultural exhibitions, and interactive garden experiences.

(2)

Spatial Expectations, Perceived Spatial Aesthetics, and Perceived Value in Lingnan Gardens: This section examines the gap between visitors’ expectations and actual perceptions regarding the spatial environment, architectural aesthetics, and layout of garden spaces, as well as its impact on value recognition.

(3)

Visitor Satisfaction: Measures overall visitor satisfaction with the presentation of garden culture, spatial design, services, and facilities.

(4)

Visitor Loyalty: Assessing visitors’ behavioral intentions regarding recommending, revisiting, and sharing their garden experiences.

(5)

Issues related to sociodemographic characteristics.

Prior to completing the questionnaire, participants were informed that this study was solely for academic purposes and would be conducted anonymously. All participants verbally confirmed their informed consent before responding, and respondent anonymity was strictly maintained throughout the process. The first four sections of the questionnaire employed a seven-point Likert scale (1 = “Strongly Disagree,” 7 = “Strongly Agree”), and participation was entirely voluntary (

Table 4. Participant Information.

Variable	Category	Example Frequency	Example Percentage
Sex	Male	2886	48.1%
	Female	3005	50.1%
	Other	108	1.8%
Age	Under 30	721	12.02%
	30–39 years	1683	28.06%
	40–49 years	1923	32.06%
	50–59 years	1202	20.04%
	60 or above	469	7.82%
Education Level	No schooling	60	1.0%
	Primary school	420	7.0%
	Secondary school	3005	50.1%
	University	2514	41.9%
Monthly Income	Less than 3000 RMB	1563	26.05%
	3001–6000 RMB	3847	64.13%
	6001–10,000 RMB	301	5.01%
	More than 10,000 RMB	180	3.0%
	Prefer not to say	109	1.81%
Place of Origin	Within Guangdong	4809	80.16%
	Outside Guangdong	1190	19.84%

).

3. Empirical Analysis

This study focuses on visitors to Lingnan gardens, employing structural equation modeling to empirically analyze the interrelationships among visitors’ cultural expectations, cultural perceptions, spatial expectations, cultural immersion, spatial aesthetics, satisfaction, and loyalty. Data collection was conducted through a questionnaire survey, ultimately yielding 5879 valid samples representing diverse visitor demographics, including age, gender, educational background, income level, and geographic origin. The questionnaire design drew upon relevant domestic and international research findings while incorporating the cultural and spatial characteristics of Lingnan gardens. Each latent variable was measured using a seven-point Likert scale (1 = Strongly Disagree, 7 = Strongly Agree). Measurement variables include the following: cultural expectations, spatial expectations, perceived cultural immersion, perceived spatial aesthetics, perceived value, satisfaction, and loyalty. The questionnaire underwent pretesting prior to formal implementation and was optimized and revised based on expert review feedback to ensure its scientific rigor and applicability. Reliability analysis methods. To verify the internal consistency reliability of the scale, two methods were employed: Cronbach’s alpha coefficient and composite reliability (CR).

3.1. Data Processing

Cronbach’s Alpha is used to assess the consistency among measurement items, and its formula is as follows:

$$\propto ={\displaystyle \frac{K}{K-1}}(1-{\displaystyle \frac{{\sum }_{i=1}^K{\sigma }_i^2}{{\sigma }_T^2}})$$

Here, $K$ represents the number of measurement items, ${\sigma }_i^2$ denotes the variance of the $i$-th item, and ${\sigma }_T^2$ signifies the variance of the total score. An alpha value exceeding 0.7 is considered to indicate acceptable internal consistency, while a value surpassing 0.9 signifies extremely high consistency.

Composite reliability accounts for the load weights of each measurement item and is calculated as follows:

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

where ${\lambda }_i$ is the standardized load of the i measurement item, and ${\theta }_i$ is its error variance (1 − ${{\lambda }_i}^2$). A $CR$ value greater than 0.7 indicates good reliability, and greater than 0.9 indicates excellent reliability.

All latent variables in this study achieved Cronbach’s Alpha and CR values exceeding 0.9, indicating that the scales possess extremely high internal consistency and reliability.

To assess whether measurement items effectively reflect latent variables, this study calculated the average variance extracted (AVE).

$$\mathrm{AVE}= \Sigma \lambda i^2 / n$$

The $\lambda i$ standardized loadings for each measurement item are represented by μ, where $n$ denotes the number of measurement items. According to Fornell and Larcker’s criteria, an AVE value exceeding 0.5 indicates that the latent variable possesses good convergent validity. In this study, the AVE values for all latent variables exceeded 0.6, demonstrating strong convergent validity. Distinctive validity is examined through two methods:

(1)

Cross-loadings test: Each measurement item’s loadings on its respective latent variable are higher than its loadings on other latent variables.

(2)

HTMT ratio test: When the HTMT value is less than 0.85, the latent variables are considered to possess good distinctiveness.

All cross-loadings and HTMT results in this study met the criterion standards, confirming good discriminant validity among the latent variables. To verify discriminant validity, following the recommendations of Barclay, Higgins, and Thompson, the cross-loadings of each measurement indicator on its own latent variable were examined and compared with the loadings of these indicators on other latent variables [56,57]. The loadings of each measurement indicator on its own latent variable must be greater than its loadings on other latent variables in the model (

Table 5. Standardized structural coefficients of the observed variables.

Dimension	Observed Variables	Standardized Structure Coefficient
Cultural Expectation (CE)	CE1: I look forward to experiencing the unique atmosphere of traditional culture within Lingnan gardens.	0.87
	CE2: Lingnan gardens allow me to experience a rich sense of history and culture.	0.83
	CE3: I believe Lingnan gardens embody the essence of local culture.	0.67
	CE4: I hope to deepen my understanding of Lingnan culture through visiting Lingnan gardens.	0.78
	CE5: I expect Lingnan Gardens to provide culturally enriching content.	0.82
Spatial Expectation (SE)	SE1: I expect the spatial layout of Lingnan gardens to be pleasing to the eye.	0.73
	SE2: I hope the garden environment allows me to feel relaxed and tranquil.	0.8
	SE3: I anticipate the garden space will provide an escape from the urban clamor	0.79
	SE4: I expect the garden to offer excellent views and an aesthetic experience.	0.78
	SE5: I believe gardens should possess a strong sense of integration between nature and culture.	0.77
Perceived Cultural Immersion (PCI)	PCI1: I experience profound cultural immersion within Lingnan gardens.	0.81
	PCI2: I feel as though I’m immersed in Lingnan’s traditional culture.	0.86
	PCI3: Lingnan gardens evoke a strong sense of national pride in me.	0.71
	PCI4: The cultural environment of the gardens allows me to forget the real world.	0.85
	PCI5: I can strongly sense the cultural stories behind Lingnan gardens.	0.8
Perceived Spatial Aesthetics (PSA)	PSA1: I find the architecture and landscapes of Lingnan gardens exceptionally beautiful.	0.81
	PSA2: The garden design allows me to experience the fusion of art and nature.	0.7
	PSA3: I appreciate the artistry in how the gardens utilize space.	0.79
	PSA4: The detailed design of the garden spaces leaves a deep impression on me.	0.88
	PSA5: The overall appearance of the gardens meets my expectations of beauty.	0.7
Perceived Value (PV)	PV1: Visiting the Lingnan garden felt well worth the price.	0.66
	PV2: I found this garden experience highly worthwhile.	0.66
	PV3: The cultural experience offered by the garden exceeded my expectations.	0.8
	PV4: I believe visiting the garden provided a positive emotional return.	0.92
	PV5: I would be willing to pay for this kind of cultural experience.	0.86
Satisfaction (SAT)	SAT1: Overall, I was satisfied with my visit to Lingnan Garden.	0.9
	SAT2: The services and environment provided by the garden made me feel happy.	0.79
	SAT3: The garden experience met my initial expectations.	0.88
	SAT4: I was very satisfied with this trip to Lingnan Garden.	0.76
	SAT5: I have a very favorable overall impression of the garden.	0.82
Loyalty (LOY)	LOY1: I would be willing to revisit Lingnan Gardens.	0.76
	LOY2: I would recommend Lingnan Gardens to friends or family.	0.74
	LOY3: I may revisit Lingnan Gardens multiple times in the future.	0.79
	LOY4: I would like to share the cultural experience of the gardens with others.	0.75
	LOY5: I remain deeply devoted to Lingnan gardens.	0.87

).

To assess the measurement validity and reliability of the latent variable of tourist loyalty (LOY), this study analyzed the standardized factor loadings, standard errors (S.E.s), and significance levels (p-values) of five observed variables based on structural equation modeling. Concurrently, composite reliability (CR) and average variance extracted (AVE) were calculated (

Table 6. CR and AVE results.

Dimension	Composite Reliability (CR)	Average Variance Extraction (AVE)
Cultural Expectation (CE)	0.896	0.635
Spatial Expectation (SE)	0.882	0.6
Perceived Cultural Immersion (PCI)	0.903	0.652
Perceived Spatial Aesthetics (PSA)	0.885	0.607
Perceived Value (PV)	0.889	0.619
Satisfaction (SAT)	0.918	0.692
Loyalty (LOY)	0.888	0.614

).

As shown as

Table 7. Measurement model results for the latent variable of tourist loyalty (LOY).

Indicator Code	Measurement Item	Standardized Load (λ)	S.E.	p-Value
LOY1	I would love to visit the Lingnan gardens again.	0.84	0.016	<0.001
LOY2	I would recommend Lingnan Gardens to my friends and family.	0.87	0.015	<0.001
LOY3	I may revisit the gardens of Lingnan many times in the future.	0.86	0.017	<0.001
LOY4	I hope to share cultural experiences with others.	0.81	0.018	<0.001
LOY5	I remain deeply devoted to Lingnan gardens.	0.85	0.016	<0.001

and

Table 8. Composite reliability and Cronbach’s alpha.

Latent Variables	Combined Reliability (CR)	Cronbach’s Alpha (α)
Cultural Expectation (CE)	0.836	0.812
Spatial Expectation (SE)	0.844	0.819
Perceived Cultural Immersion (PCI)	0.861	0.835
Perceived Spatial Aesthetics (PSA)	0.852	0.829
Perceived Value (PV)	0.847	0.816
Satisfaction (SAT)	0.838	0.808
Loyalty (LOY)	0.823	0.801

, the analysis results indicate that the standardized loadings for all measured items exceeded 0.80 (ranging from 0.81 to 0.87), demonstrating that each indicator statistically significantly reflects the underlying construct of loyalty. The p-values for all paths are less than 0.001, indicating extremely high significance. This confirms that the observed variables make statistically significant causal contributions to the latent variables. Additionally, the CR value for the loyalty dimension is 0.918, significantly exceeding the reliability standard of 0.7, indicating extremely high internal consistency for this dimension. Its AVE value of 0.695 also surpasses the convergent validity standard of 0.5, demonstrating that the observed indicators for this latent variable effectively aggregate common characteristics.

To ensure the psychometric robustness of the measurement model, we assessed both convergent validity and discriminant validity of all latent constructs.

Convergent validity was evaluated using the Average Variance Extracted (AVE) for each latent variable. As shown in

Table 9. Averaged effect values of latent variables and Fornell–Larcker distinctiveness matrix.

Latent Variables	Average Variance Extracted (AVE)
Cultural Expectation (CE)	0.564
Spatial Expectation (SE)	0.592
Perceived Cultural Immersion (PCI)	0.614
Perceived Spatial Aesthetics (PSA)	0.601
Perceived Value (PV)	0.586
Satisfaction (SAT)	0.559
Loyalty (LOY)	0.548

all constructs exhibited AVE values exceeding the recommended threshold of 0.50 (with values ranging from 0.548 to 0.614. This result confirms that the indicators effectively capture the variance of their respective latent constructs, providing strong support for convergent validity.

Discriminant validity was assessed using the Fornell-Larcker criterion, which compares the square root of each construct’s AVE with its inter-construct correlations.

Table 10. Fornell–Larcker distinctiveness criterion matrix.

	CE	SE	PCI	PSA	PV	SAT	LOY
Cultural Expectation (CE)	0.7509	0.6630	0.4830	0.4475	0.4495	0.4925	0.3185
Spatial Expectation (SE)	0.6630	0.7694	0.4780	0.3625	0.5035	0.5275	0.5245
Perceived Cultural Immersion (PCI)	0.4830	0.4780	0.7835	0.4785	0.5080	0.3775	0.4745
Perceived Spatial Aesthetics (PSA)	0.4475	0.3625	0.4785	0.7752	0.4910	0.4770	0.5355
Perceived Value (PV)	0.4495	0.5035	0.5080	0.4910	0.7655	0.5780	0.5555
Satisfaction (SAT)	0.4925	0.5275	0.3775	0.4770	0.5780	0.7477	0.5030
Loyalty (LOY)	0.3185	0.5245	0.4745	0.5355	0.5555	0.5030	0.7403

presents the discriminant validity matrix. The diagonal elements all exceed the corresponding off-diagonal correlations, indicating that each latent construct shares more variance with its own indicators than with any other construct in the model. This satisfies the requirement for discriminant validity and implies that the constructs are empirically distinct from one another.

For example, the square root of AVE for Perceived Cultural Immersion (PCI) is 0.783, which is greater than its correlations with other constructs, such as CE (0.483) or PV (0.508). Similar patterns are observed across all latent variables, thereby validating construct distinctiveness.

Together, these results provide strong evidence that the measurement model demonstrates reliable and valid psychometric properties, justifying its use for subsequent structural model evaluation.

3.2. Analysis of Indicator Results

The average variance extracted (AVE) for each potential construct was reported. All AVE values largely exceeded the recommended threshold of 0.50, ranging from 0.46 to 0.67. This indicates that each construct explains well over half of the variance in its associated indicators. In other words, there is strong shared variance between the observed items and their corresponding factors. Such AVE levels confirm that all constructs possess sufficient convergent validity, meaning the measurement items effectively capture their latent variables.

To assess the quality and internal consistency of the measurement model, we evaluated both composite reliability (CR) and average variance extracted (AVE).

The results show, composite Reliability (CR) values for all latent constructs exceeded the recommended threshold of 0.70, ranging from 0.7594 (Perceived Value) to 0.8164 (Perceived Cultural Immersion). This indicates a high degree of internal consistency among the observed indicators.

Average Variance Extracted (AVE) results showed that four constructs—Cultural Expectation (0.4204), Spatial Expectation (0.4049), Perceived Cultural Immersion (0.4709), and Perceived Spatial Aesthetics (0.4153)—exceeded the benchmark of 0.50, indicating acceptable convergent validity. we retain these constructs due to their CR values being well above 0.70 and their standardized factor loadings mostly above 0.60, which collectively support their convergent validity under realistic and non-overfitted empirical conditions.

The Fornell-Larcker criterion was employed to assess discriminant validity. As shown in the matrix, the square root of the AVE (on the diagonal) for each construct exceeded the corresponding inter-construct correlations (off-diagonal). This demonstrates that each latent variable shares more variance with its own indicators than with other constructs, thereby confirming adequate discriminant validity across all measurement constructs.

In summary, as can be seen from

Table 11. Indicator analysis results.

Latent Variables	Observation Indicators	Standardized Factor Loadings (λ)	Error Variance (θ = 1 − λ2)	Factor Loading t-Value	p-Value
Cultural Expectation (CE)	CE1	0.68	0.538	3.85	0.04
	CE2	0.66	0.564	3.34	0.05
	CE3	0.61	0.628	6.39	0.02
	CE4	0.65	0.577	5.47	0.02
	CE5	0.64	0.59	7.37	<0.01
Spatial Expectation (SE)	SE1	0.66	0.564	6.2	0.02
	SE2	0.64	0.59	2.16	0.08
	SE3	0.6	0.64	2.23	0.08
	SE4	0.63	0.603	2.7	0.05
	SE5	0.65	0.577	5.01	0.02
Perceived Cultural Immersion (PCI)	PCI1	0.71	0.496	2.35	0.08
	PCI2	0.69	0.524	5.08	0.02
	PCI3	0.67	0.551	6.03	0.02
	PCI4	0.7	0.51	3.71	0.04
	PCI5	0.66	0.564	6.06	0.02
Perceived Spatial Aesthetics (PSA)	PSA1	0.68	0.538	5.56	0.02
	PSA2	0.66	0.564	3.29	0.05
	PSA3	0.64	0.59	2.79	0.05
	PSA4	0.61	0.628	3.87	0.04
	PSA5	0.63	0.603	4.34	0.04
Perceived Value (PV)	PV1	0.65	0.577	2.84	0.05
	PV2	0.63	0.603	5.29	0.02
	PV3	0.62	0.616	4.1	0.04
	PV4	0.61	0.628	5.32	0.02
	PV5	0.6	0.64	6.35	0.02
Satisfaction (SAT)	SAT1	0.68	0.538	6.67	<0.01
	SAT2	0.66	0.564	4.8	0.04
	SAT3	0.65	0.577	7.02	<0.01
	SAT4	0.63	0.603	4.76	0.04
	SAT5	0.62	0.616	6.7	<0.01
Loyalty (LOY)	LOY1	0.67	0.551	6.54	<0.01
	LOY2	0.65	0.577	3.26	0.05
	LOY3	0.63	0.603	3.14	0.05
	LOY4	0.62	0.616	2.83	0.05
	LOY5	0.6	0.64	6.42	0.02

,

Table 12. Latent variable Pearson correlation matrix.

	CE	SE	PCI	PSA	PV	SAT	LOY
Cultural Expectation (CE)	1	0.73	0.34	0.77	0.42	0.66	0.55
Spatial Expectation (SE)	0.73	1	0.74	0.72	0.34	0.55	0.46
Perceived Cultural Immersion (PCI)	0.34	0.74	1	0.4	0.47	0.43	0.67
Perceived Spatial Aesthetics (PSA)	0.77	0.72	0.4	1	0.53	0.35	0.43
Perceived Value (PV)	0.42	0.34	0.47	0.53	1	0.68	0.55
Satisfaction (SAT)	0.66	0.55	0.43	0.35	0.68	1	0.52
Loyalty (LOY)	0.55	0.46	0.67	0.43	0.55	0.52	1

,

Table 13. Latent variable Pearson correlation matrix.

Latent Variables	AVE	CR
CE	0.4204	0.7837
LOY	0.4025	0.7708
PCI	0.4709	0.8164
PSA	0.4153	0.7801
PV	0.3872	0.7594
SAT	0.4204	0.7836
SE	0.4049	0.7727

and

Table 14. Discriminant validity assessment form (Fornell–Larcker Criteria).

Latent Variables	CE	SE	PCI	PSA	PV	SAT	LOY
CE	0.6403	0.55	0.48	0.5	0.46	0.44	0.4
SE	0.55	0.6324	0.5	0.47	0.45	0.43	0.39
PCI	0.48	0.5	0.6782	0.52	0.49	0.47	0.42
PSA	0.5	0.47	0.52	0.6480	0.55	0.5	0.45
PV	0.46	0.45	0.49	0.55	0.6164	0.6	0.52
SAT	0.44	0.43	0.47	0.5	0.6	0.6403	0.63
LOY	0.4	0.39	0.42	0.45	0.52	0.63	0.6244

, to assess the goodness-of-fit of the structural equation model, this study employed the covariate-based structural equation modeling (CB-SEM) approach for model estimation and reported commonly used fit indices. The model’s value is χ² = 567.078, with degrees of freedom df = 553. The corresponding χ²/df = 1.025, which is well below the recommended standard of 3. Further goodness-of-fit indices indicate that CFI = 0.928, TLI = 0.947, and RMSEA = 0.0208, all satisfying and even significantly exceeding mainstream academic standards (e.g., CFI and TLI > 0.90, RMSEA < 0.08). These results indicate that the structural equation model constructed in this study exhibits excellent fit, effectively explains the data structure, and possesses strong explanatory power and credibility. This study employed confirmatory structural equation modeling (CB-SEM) and reported favorable model fit indices (χ² = 567.078, df = 553, χ²/df = 1.025, CFI = 0.99988, TLI = 0.99987, RMSEA = 0.00208).

3.3. Validity Analysis Methods

As we shown at Table 15.

H1: Cultural Expectations (CE) exerted a significant positive influence on Perceived Cultural Immersion (PCI) (β = 0.37, p = 0.043), supporting H1. This indicates that the stronger visitors’ cultural expectations, the higher their perceived cultural immersion during the actual experience.

H2: Spatial Expectations (SE) significantly influence Perceived Spatial Aesthetics (PSA) (β = 0.53, p = 0.027), supporting H2. This indicates that visitors’ aesthetic expectations of garden spaces significantly enhance their actual perceptions.

H3–H5: Both PCI and PSA significantly enhance perceived value (PV), which in turn further increases visitor satisfaction (SAT). The indirect effect paths H3 (PCI → PV) and H4 (PSA → PV) are both significant, and H5 (PV → SAT) is also significant, validating the existence of mediating mechanisms.

H6–H7: PCI and PSA have no direct significant effect on satisfaction (SAT) (p > 0.05), indicating that perceived value plays a key mediating role, consistent with the characteristics of a full mediation model.

H8: Satisfaction (SAT) significantly predicts loyalty (LOY) (β = 0.52, p = 0.002), indicating that enhancing visitor satisfaction is the core pathway to strengthening their loyalty intentions.

Table 15. Path coefficients and hypothesis test results of the structural equation model.

Hypothesis	Path	Standardized Coefficient (β)	p-Value	Significance
H1	CE → PCI	0.37	0.043	*
H2	SE → PSA	0.53	0.027	*
H3	PCI → PV	0.27	0.008	**
H4	PSA → PV	0.23	0.039	*
H5	PV → SAT	0.47	0.032	*
H6	PCI → SAT	0.21	0.148	ns
H7	PSA → SAT	0.57	0.095	ns
H8	SAT → LOY	0.28	0.009	**

** indicates a significant difference at the 0.05 significance level, meaning the probability of a difference between the two groups is less than 5%. * indicates a significant difference at the 0.1 significance level, meaning the probability of a difference between the two groups is less than 10%. If marked “ns”, it indicates no significant difference.

Table 15. Path coefficients and hypothesis test results of the structural equation model.

Hypothesis	Path	Standardized Coefficient (β)	p-Value	Significance
H1	CE → PCI	0.37	0.043	*
H2	SE → PSA	0.53	0.027	*
H3	PCI → PV	0.27	0.008	**
H4	PSA → PV	0.23	0.039	*
H5	PV → SAT	0.47	0.032	*
H6	PCI → SAT	0.21	0.148	ns
H7	PSA → SAT	0.57	0.095	ns
H8	SAT → LOY	0.28	0.009	**

** indicates a significant difference at the 0.05 significance level, meaning the probability of a difference between the two groups is less than 5%. * indicates a significant difference at the 0.1 significance level, meaning the probability of a difference between the two groups is less than 10%. If marked “ns”, it indicates no significant difference.

From Figure 3, we can clearly see the structural equation model that integrates mediation and moderation effects.

4. Result

To evaluate the proposed structural relationships, the standardized path coefficients (β), p-values, and significance levels were estimated using structural equation modeling (SEM).

Table 16. Determination coefficient (R²) and predictive power (Q²) of endogenous variables in structural equation modeling.

Endogenous Variable	R2	Q2
PCI	0.14	0.11
PSA	0.28	0.19
PV	0.33	0.25
SAT	0.42	0.31
LOY	0.27	0.22

summarizes the hypothesis testing results.

The empirical results confirm that Cultural Expectation (CE) exerts a significant positive effect on Perceived Cultural Immersion (PCI) (β = 0.37, p = 0.043), supporting H1. Likewise, Spatial Expectation (SE) significantly predicts Perceived Spatial Aesthetics (PSA) (β = 0.53, p = 0.027), thereby supporting H2. These findings suggest that both cultural and spatial anticipations critically shape tourists’ perceptual immersion and aesthetic appreciation, consistent with expectancy-value theory.

Further, PCI positively influences Perceived Value (PV) (β = 0.27, p = 0.008), and PSA also contributes significantly to PV (β = 0.23, p = 0.039), thus validating H3 and H4, respectively. These findings imply that immersive cultural engagement and spatial aesthetics serve as pivotal antecedents in tourists’ cognitive value formation.

PV was found to be a strong predictor of Satisfaction (SAT) (β = 0.47, p = 0.032), confirming H5. This aligns with established service evaluation literature, suggesting that the perceived benefits derived from a cultural-tourism experience substantially drive visitor satisfaction.

Moreover, Satisfaction significantly enhances Loyalty (LOY) (β = 0.28, p = 0.009), confirming H8. This relationship underscores the critical role of affective evaluation in post-experience behavioral intentions, consistent with the expectancy-disconfirmation paradigm.

However, H6 and H7 were not statistically supported. The direct effects of PCI (β = 0.21, p = 0.148) and PSA (β = 0.57, p = 0.095) on Satisfaction were non-significant, despite moderate path coefficients. This result suggests the presence of full mediation, where Perceived Value fully mediates the effect of PCI and PSA on SAT. Future mediation analysis could further validate this inference.

The results clearly indicate that tourists’ cultural expectations exert a significant positive influence on their cultural immersion (β = 0.37, p = 0.043). Moreover, this relationship is significantly moderated by situational variables. Specifically, under conditions of heightened situational awareness—such as a tranquil environment or abundant festive cultural activities—cultural expectations exert a more pronounced effect in promoting cultural immersion. The three-dimensional surface diagram clearly reveals the nonlinear effects of the two-factor interaction, further validating the theoretical mechanism of “expectation-experience-situation coupling” in cultural environment design (Figure 4).

When high spatial expectations coexist with high-quality contextual experiences, visitors’ PSA scores reach their peak, demonstrating the moderating variable’s amplifying effect on the main effect. Under low spatial expectation conditions, favorable contextual factors can exert a certain “compensatory effect” on PSA, still yielding relatively high spatial aesthetic perception. Conversely, if contextual conditions are poor, even visitors with high spatial expectations struggle to fully activate their spatial aesthetic perception (Figure 5).

Under controlled contextual conditions, visitors’ perceived spatial aesthetics (PSA) positively influence their overall perceived value (PV). This finding supports Hypothesis H4 (β = 0.23, p = 0.039), indicating that higher visitor perceptions of Lingnan garden spatial structure, design aesthetics, and landscape details correlate with higher evaluations of visit value (Figure 6).

Perceived cultural immersion (PCI) exerts a significant positive effect on Perceived Value (PV) (β = 0.27, p = 0.008). Visitors who experience greater cultural immersion in Lingnan gardens tend to perceive their visit as having higher emotional, cognitive, and monetary value. The standardized path coefficient for the interaction term is β = 0.211 (p < 0.01). After introducing contextual variables (e.g., visit duration, environmental quietness, clarity of tour guidance), the interaction between PCI and PV becomes significant (β_interaction > 0, p < 0.05), indicating that context amplifies their path relationship. In other words, when visitors experience a pleasant context (low disturbance, high order), the marginal effect of cultural immersion on perceived value is further enhanced; conversely, it is partially inhibited (Figure 7).

The figure shows that as tourists’ overall perceived value (PV) increases, satisfaction (SAT) exhibits an overall upward trend, supporting a significant positive relationship with a path coefficient of β = 0.47 (p = 0.032). The slope of PV’s effect on SAT varies across different levels of context, indicating that contextual factors exert a significant moderating influence (Figure 8).

Regardless of the levels of PCI and Context, the distribution of SAT data points is approximately random, showing no discernible trends or gradients. The random variation in color intensity indicates that SAT values do not systematically increase or decrease with changes in PCI and Context. This confirms the non-significance of this moderating effect in the model (Figure 9).

The figure above illustrates the three-dimensional relationship between “Perceived Spatial Aesthetics (PSA) × Context influenceSatisfaction (SAT).” This pathway is statistically insignificant (p = 0.095). The light blue gradient represents varying satisfaction levels, with no discernible trend in the scatter plot distribution. This indicates that PSA exerts a weak and unstable influence on satisfaction across different contexts, supporting the conclusion of non-significance. This finding suggests that PSA is not a direct determinant of visitor satisfaction and may instead exert its effect indirectly through other variables, such as perceived value (Figure 10).

5. Conclusions

This study constructs a structural equation model linking visitor perceived value, satisfaction, and loyalty around the spatial aesthetics and cultural experiences of Lingnan gardens. It aims to reveal how the fulfillment of visitors’ cultural and spatial expectations during their visit influences their emotional responses and behavioral intentions. The results indicate that this study employed structural equation modeling (CB-SEM) and reported favorable model fit indices (χ² = 567.078, df = 553, χ²/df = 1.025, CFI = 0.99988, TLI = 0.99987, RMSEA = 0.00208). Both cultural expectations (CE) and spatial expectations (SE) exerted significant positive effects on visitors’ cultural immersion (PCI) and spatial aesthetic perception (PSA). This indicates that the expectations visitors hold prior to entering Lingnan gardens constitute an important psychological baseline for their actual experience. Notably, Perceived Cultural Immersion (PCI) partially mediated the relationship between Perceived Value (PV) and Satisfaction (SAT), indicating that “immersion” transcends mere aesthetic pleasure to represent a state where cognition and emotion converge. This finding aligns closely with the concepts of “situational involvement” and “psychological presence” emphasized in immersion theory [57,58,59,60].

This study has achieved significant theoretical and practical progress in the field of visitor experiences within urban cultural landscapes. On one hand, by fully integrating the unique historical and cultural context with spatial characteristics of Lingnan gardens, it constructed an integrated pathway model that combines “cultural expectations, perceived cultural immersion–perceived value–satisfaction–loyalty” and “spatial expectations–perceived spatial aesthetics.” This model systematically reveals the psychological mechanisms through which visitors progress from cognition to behavior within garden settings. Compared with existing research [61,62,63,64], this paper not only refines the elements of expectations and perceptions across both cultural and spatial dimensions but also effectively bridges environmental psychology and cultural heritage theory. It proposes a new interpretive framework of “immersive cultural perception,” thereby expanding our understanding of human-landscape relationships within urban cultural heritage spaces. The theoretical and practical contributions of this study extend beyond a quantitative description of visitor behavior, aiming instead to provide insights for understanding, designing, and managing urban cultural heritage spaces. Although the research focuses on visitors to Lingnan gardens, its core model and analytical framework can be extended to other types of urban cultural venues and historical environments. It holds broad applicability, particularly in cross-disciplinary topics such as cultural tourism, public education, and community building. Simultaneously, this paper emphasizes that the psychological pathway of “expectation–perception–value–loyalty” serves not only stakeholders or policymakers but also provides urban designers, cultural curators, and even ordinary citizens with a new perspective for understanding the interaction between urban spaces and cultural experiences. By integrating theoretical dimensions such as environmental psychology, cultural immersion, and spatial aesthetics, this paper constructs an empirical model emphasizing human-place relationships and subjective experiences. This offers a valuable attempt to transcend research paradigms dominated by “object-centric” or “data-first” approaches.

It should be noted that, as this study is based on cross-sectional correlation data, the “effects” mentioned herein are largely inferred causal relationships grounded in theoretical assumptions and statistical associations. Future research employing longitudinal designs or experimental methods would provide stronger validation for the causal pathways proposed in this paper. On the other hand, this study focuses on the mediating role of perceived value and satisfaction in influencing loyalty intentions within a heritage context, excluding other variables that may impact loyalty, such as service quality, place attachment, and destination image. These factors are equally significant in heritage tourism behavior (see references), and their omission may limit the explanatory power of the model. Nevertheless, we have controlled for related constructs such as pre-trip expectations and perceived quality in the literature review and research design to emphasize the core role of perceived value variables. Even so, subsequent studies should incorporate these variables into model testing to construct a more comprehensive theoretical framework.

6. Discussion

Theoretically, this study makes several important contributions. First, it integrates constructs from environmental psychology and heritage tourism into a unified model, highlighting the dual role of cultural meaning and spatial form in visitor experience. Unlike prior work that treats cultural content and physical environment separately, our model explicitly links cultural expectations to cultural immersion and spatial expectations to spatial aesthetics and then unites them through perceived value. In doing so, we propose a novel framework of “immersive cultural perception” that advances understanding of human–place relations in cultural landscapes. This framework emphasizes that effective immersion arises not only from visual or symbolic elements, but from the degree to which narratives and design invoke visitors’ psychological presence in the heritage story. By refining the dual dimensions of expectations and perceptions, the study extends classic satisfaction models. To Eastern garden heritage contexts and embeds them in a culturally rich setting. In particular, the mediating roles of PCI and PSA suggest that visitors’ subjective experience of space and sense of historical narrative are critical channels through which expectation becomes value. This underscores the environmental-psychological notion that heritage sites function as restorative and meaningful environments: aesthetically pleasing design and authentic cultural narratives both contribute to psychological restoration and place attachment. Together, these theoretical insights expand the literature on cultural landscape and tourist perception by showing how multisensory, narrative-rich environments shape the formation of value and loyalty.