Key Factors and Configuration Analysis of Improving Tourist Loyalty in Forest Park: Evidence from Yingde National Forest Park, South China

Abstract

Tourist perceived value is an important antecedent to loyalty by enhancing satisfaction, revisiting intentions, and recommendations, thereby promoting sustainable development of forest parks. However, existing research has not sufficiently examined the configurations of perceived value in relation to increasing tourist loyalty specifically in the context of forest parks, representing a notable gap in the existing literature that requires further investigation. To address this gap, both covariance-based structural equation model (CB-SEM) and fuzzy-set qualitative comparative analysis (fsQCA) models were conducted to explore the joint effects of perceived value on tourist loyalty and identify pathways of perceived value dimensions to increase tourist loyalty, based on the Value-Satisfaction-Loyalty Chain model. A total of 404 valid questionnaires were collected from 436 visitors to the Yingde National Forest Park in southern China. Among the respondents, 54.2% were male, nearly 50% were over 36 years old, and 60% held a university degree. The results indicate that perceived value significantly influences tourist loyalty, with satisfaction playing a crucial mediating role between perceived value and loyalty. Notably, the indirect effect mediated by satisfaction was found to be greater than the direct effect of perceived value on loyalty. Five distinct pathways were identified for enhancing tourist loyalty, categorized into three models: the economic value-driven model, the functional value and epistemic value dual-core driven model, and the emotional and social value dual-core driven model. Additionally, four pathways were identified for enhancing tourist satisfaction, which subsequently improves tourist loyalty. These four pathways were grouped into two modes: the economic value-driven model and the functional value plus driven model. This study introduces an innovative perspective on the relationship between tourist perceived value and loyalty in forest parks, identifying key factors and configurations within the five dimensions of perceived value that enhance both tourist loyalty and satisfaction. Moreover, it extends the application of the Value-Satisfaction-Loyalty Chain theory to a forest park context. The findings provide valuable insights for forest park managers, guiding them in enhancing perceived value through targeted pathways to increase tourist revisit intentions and recommendations, ultimately supporting the park’s sustainable development. The influence of individual items on tourist satisfaction and loyalty, along with the identification of optimal item combinations to enhance loyalty, necessitates further investigation. Furthermore, a deeper exploration of the heterogeneity of factors and pathways for improving tourist loyalty is required.

1. Introduction

Forest parks have evolved as significant natural and recreational spaces, contributing to both ecological conservation and human well-being. Historically, the establishment of forest parks was driven by the need to preserve biodiversity and provide controlled public access to natural landscapes [1,2]. Many countries have designated forest parks to balance conservation with recreation, ensuring that visitors can experience nature while minimizing environmental degradation [3]. Over the past few decades, the expansion of protected areas and eco-tourism initiatives has led to a steady increase in forest park visitation worldwide [4].

Forest parks play a critical role in global sustainability by maintaining ecosystem functions, preserving biodiversity, and serving as carbon sinks [5]. They also provide cultural, aesthetic, and psychological benefits, contributing to public health and well-being [6]. Beyond their ecological significance, forest parks serve as important economic assets, supporting local economies through tourism, employment, and community-based conservation initiatives [6,7,8].

Tourism in forest parks has become an essential component of sustainable development strategies, fostering environmental awareness and financial support for conservation efforts [9,10,11]. The tourist potential of forests is determined by various factors, such as the forest’s age, habitat moisture, terrain slope, stand density, the presence of undergrowth and underbrush, soil coverage, and species composition [12]. Nature-based tourism, including activities such as hiking, wildlife observation, and camping, enhances visitors’ appreciation of natural landscapes and promotes responsible environmental behavior [13,14]. Studies have indicated that well-managed tourism in forest parks can contribute significantly to local and national economies while ensuring the protection of natural resources [15,16]. In some forest parks, tourism has surpassed the carrying capacity, leading to environmental protection issues, as observed in HengTou Mountain National Forest Park in northeast China [17]. A similar situation has occurred in Mount Rinjani National Park and Lombok Island and other forest parks [18,19,20]. One potential solution to this issue is the implementation of tourist restrictions, as seen in the Five Mountains of China and Yellowstone National Park in the USA. Some forest parks face the opposite challenge, where the number of tourists is insufficient to support the tourism services offered, leading to concerns about the sustainability of these parks [21]. Zhou et al. found that government economic support alone cannot sustain the development of Qiandao Lake National Forest Park, China, while tourism revenue is an important economic supplement, and the relaxation and health benefits for tourists can increase their willingness to pay more [22]. Zhang et al. indicated that competition among urban forest parks is intensifying, and improving tourist satisfaction, which enhances their likelihood of revisiting, is key to promoting the growth of forest tourism [23]. Therefore, the challenge remains in improving visitor loyalty to increase repeat visits and recommendations for attracting more tourists in most forest parks [24,25,26].

With the increasing demand for recreational opportunities in natural settings, understanding how to enhance tourist loyalty has become a critical issue for park management [27,28,29]. Tourist loyalty, which represents repeat visits and recommendations, is essential in sustaining a stable long-term market and advancing sustainable development of forest parks [27,30].

Extensive research has been conducted on the factors influencing tourist loyalty across various tourism contexts. One of the most critical determinants of tourist loyalty is perceived value, which encompasses multiple dimensions, including emotional, functional, epistemic, social, and monetary value [31,32,33,34,35,36]. Each of these dimensions shapes tourists’ overall experiences and their likelihood of revisiting or recommending the park. Studies have shown that perceived value influences tourist satisfaction and loyalty positively [27,37]. Different dimensions of perceived value, including functional, emotional, and social value, exhibit varying effects on tourist loyalty and satisfaction, depending on the context [38]. Joung & Yang et al., indicated that satisfaction serves as a mediator in the relationship between perceived value and loyalty, reinforcing the importance of visitor contentment in loyalty formation [39,40,41].

Despite the extensive literature on perceived value and tourist loyalty, several critical research gaps persist. While much of the existing research focuses on urban tourism, heritage sites, and hospitality, there is a paucity of studies examining tourist loyalty in nature-based tourism settings, such as forest parks [8,13,28,42,43]. Traditional statistical methods, such as SEM and regression analysis, primarily examine linear relationships but fail to capture synergistic interactions among multiple value dimensions that influence tourist loyalty [44]. The relative importance of different perceived value dimensions remains context-dependent, necessitating further exploration in the specific setting of forest parks [43]. While satisfaction has been recognized as a mediator in tourism research, its specific role in the perceived value-loyalty relationship within forest parks requires further empirical validation [8].

While perceived value has been widely acknowledged in tourism research, the precise mechanisms through which it influences tourist loyalty, particularly in forest park settings, remain underexplored.

To bridge these research gaps, this study will focus on answering the following questions:

(1)

Does tourist perceived value have significant positive effects on tourist loyalty and tourist satisfaction in a National Forest Park context?

(2)

Does tourist satisfaction have a perceived value that has significant positive impacts on tourist loyalty, and does it have a mediating role in the relationship between tourist perceived value and tourist loyalty?

(3)

What are the key factors and combinations to improve tourist loyalty in a National Forest Park context?

By integrating Structural Equation Modeling (SEM) with Fuzzy Set Qualitative Comparative Analysis (FsQCA), this study seeks to advance theoretical understanding while providing evidence-based strategies to enhance tourist loyalty in forest parks. The insights will be valuable to park administrators, policymakers, and tourism stakeholders aiming to foster more engaging and sustainable visitor experiences. These findings will be crucial in the development of targeted strategies to improve visitor satisfaction and loyalty, ultimately contributing to the sustainability of forest parks and enriching the overall visitor experience.

The remainder of our study is organized as follows. In Section 2, the literature on tourist perceived value, satisfaction, and tourist loyalty is explored. Then the hypotheses and the conceptual model are presented. In Section 3, the research site is firstly introduced, and then the measurement instrument and data collection process are demonstrated. The data summary is then presented. In Section 4, the measurement model is firstly tested, and then the hypotheses are examined. The configuration analysis to improve tourist satisfaction and loyalty is then conducted with fuzzy set quantitative comparative analysis (FsQCA). At last, the results are discussed, and the limitations and further research are highlighted in Section 5.

2. Literature Review and Hypothesis

2.1. Tourist Perceived Value

Perceived value has become a central concept in tourism research, as it is crucial in shaping tourists’ attitudes, satisfaction, and loyalty. Perceived value originated in early economic and marketing literature, where it was initially viewed as a trade-off relationship between benefits and costs. Zeithaml further advanced a seminal extension of this concept, defining perceived value as the consumer’s overall assessment of the utility of a product based on perceptions of what is received and what is given [45]. This definition emphasized the subjective nature of value, which varies across individuals and contexts.

Researchers recognized that perceived value is not a unidimensional construct but rather a complex interplay of multiple dimensions. Sheth et al. proposed a theory of consumption values, identifying five dimensions of perceived value, including emotional value (EV), epistemic value (EpV), social value (SV), functional value (FV), and conditional value (CV) [46]. The studies show that emotional value relates to the affective responses tourists experience, such as enjoyment and relaxation, while epistemic value concerns the knowledge or learning gained from the visit. Social value is linked to the social interactions and status gained from the experience, while functional value refers to the practical benefits derived from the visit, such as the quality of services and facilities. Conditional value is the utility derived from specific situational contexts. Holbrook further expanded the conceptualization by introducing a typology of consumer value, which included intrinsic/extrinsic, self-oriented/other-oriented, and active/reactive dimensions [47]. This framework highlighted the experiential and symbolic aspects of value beyond mere functional utility. Moreover, studies advised that the perceived value has been adapted to specific contexts, such as service context, which should emphasize the role of service quality, sacrifice, and perceived risk [48], while environmental and ethical dimensions should be incorporated into perceived value, reflecting growing consumer interest in sustainable consumption [49]. In hospitality service, Vandamme & Leunis proposed that price/value dimensions need to be incorporated into perceived value [50].

Perceived value plays a crucial role in shaping tourists’ loyalty and satisfaction. The scale of perceived value, including dimensions such as emotional value, epistemic value, social value, functional value, and conditional value proposed by Sheth et al. [46], has been widely applied in various contexts within tourism research. Zhao and Weng studied the effects of tourist perceived value on tourist satisfaction in urban forest parks [11]. Lu et al. studied tourist perceived value on destination loyalty in Grand Canal Forest Park [51]. The five-dimensional framework of perceived value has been adapted across different research areas to suit the unique characteristics of each context.

2.2. Perceived Value and Tourist Loyalty

Tourist loyalty refers to both the behavioral and attitudinal commitment of a tourist to a specific destination, service, or brand. It is considered one of the key outcomes of perceived value and satisfaction. It is a central factor influencing long-term visitor retention and the economic sustainability of tourism destinations.

The concept of loyalty originated in marketing and consumer behavior literature, where it was initially viewed as a unidimensional construct focused on repeat purchase behavior [52]. With the development of the tourist loyalty study, researchers recognized that loyalty is more complex than just repeat behavior, including both behavioral and attitudinal dimensions. Behavioral loyalty refers to actual behaviors, such as repeat visitation, frequency of visits, and willingness to recommend [53], while attitudinal loyalty reflects psychological commitment, emotional attachment, and preference for a destination [54].

The Value-Satisfaction-Loyalty Chain is a key component of Value-Satisfaction Theory. The chain indicated that a direct link between perceived value and tourist loyalty has been established, where tourists who perceive high value from a destination are more likely to return and recommend it [55]. Yao et al. found the perceived value has positive effects on tourist loyalty in Tongzhou Grand Canal Forest Park [4]. Zhou et al. found that perceived value has significant positive effects on tourist loyalty in Qiandao Lake National Forest Park [22]. Based on the analysis, we infer that:

H1. 

Perceived value has a significant positive effect on tourist loyalty.

2.3. Perceived Value, Satisfaction, and Tourist Loyalty

Tourist satisfaction refers to the overall evaluation of a travel experience based on the comparison between expectations and actual experiences. It directly influences destination loyalty, repeat visitation, and positive word-of-mouth.

Based on expectancy-disconfirmation theory, satisfaction can be measured with the gap between expectations and perceived performance [56]. Perceived value influences satisfaction by shaping expectations and perceptions of performance. When perceived value exceeds expectations, satisfaction increases. The value-attitude-behavior model suggests that perceived value influences attitudes (e.g., satisfaction), which in turn affects behavioral intentions [57].

Numerous studies have demonstrated a strong positive relationship between perceived value and tourist satisfaction. Chen & Chen proved that emotional experiences, such as joy and relaxation, significantly enhance tourist satisfaction [58], while Zeithaml found that epistemic value, such as learning and discovery during travel, positively influences satisfaction [45]. Sweeney & Soutar indicated that social value, such as social interactions and shared experiences, contributes to higher satisfaction levels [31], while Petric showed that tourists who perceive high functional value, such as quality accommodations and efficient transportation, are more likely to be satisfied with their travel experiences [59]. Schneider & Wagemann showed that tourists who perceive good value for money (Price/Value for Money) are more likely to report high satisfaction [60]. Based on the expectancy-disconfirmation theory and the value-attitude-behavior model, we infer the hypothesis:

H2. 

Perceived value has a significant positive effect on tourist satisfaction.

The Value-Satisfaction-Loyalty Chain illustrates the sequential relationships among perceived value, tourist satisfaction, and tourist loyalty. Typically, satisfaction directly influences loyalty, with satisfied tourists being more likely to engage in repeat visitation, recommend the destination, and develop an emotional attachment.

The Value-Satisfaction-Loyalty Chain also indicated that satisfaction often mediates the relationship between perceived value and loyalty, although perceived value usually influences tourist loyalty directly [54]. The Value-Attitude-Behavior Model proposed that perceived value (e.g., functional, emotional, social) influences tourist attitudes (e.g., satisfaction), which then affect loyalty behaviors. Satisfaction plays a critical role as a mediator in the relationship between perceived value and tourist loyalty. Zhang et al. found that tourist satisfaction has a significant moderating role between tourist perceived value and tourist loyalty in urban forest parks [23], while Yao et al. found the mediating role of tourist satisfaction on the relationship between tourist perceived value and loyalty in Grand Canal Forest Park [4]. It is generally acknowledged that tourist satisfaction is a key antecedent to loyalty [61]. When tourists perceive high value in their experiences (i.e., the benefits outweigh the costs), their satisfaction with the destination increases, leading to a higher probability of their loyalty [37]. Thus, we infer the following hypotheses:

H3. 

Tourist satisfaction has a significant positive effect on tourist loyalty.

H4. 

Tourist satisfaction has a significant mediation effect on the relationship between tourist satisfaction and tourist loyalty.

Therefore, we present the theoretical model (Figure 1), and it then will be examined in our study.

3. Materials and Methods

3.1. Study Site: Yingde National Forest Park

Yingde National Forest Park, situated in Yingde City, northern Guangdong Province, China, is a prime example of a protected natural area that combines ecological conservation with sustainable tourism. Established in 2000, it is one of the earliest and the biggest national forest parks in Guangdong Province. The park, renowned for its rich biodiversity, scenic landscapes, and unique cultural heritage, plays an essential role in both environmental protection and the local economy.

Yingde National Forest Park spans over 50,000 hectares of lush forest, characterized by a variety of plant species and diverse ecosystems. The park is a critical habitat for numerous endemic species, including several that are under threat of extinction. The forest functions as a vital carbon sink, contributing to the region’s efforts in mitigating the impacts of climate change. In addition, the park’s unique geographical features, such as its steep mountains, deep valleys, and crystal-clear streams, contribute to the region’s ecological balance.

In recent years, the tourism industry in Yingde has been promoted as a way to boost local economic development while preserving the region’s natural beauty. The park offers a range of activities designed to provide visitors with a deeper connection to nature, including hiking, wildlife observation, and photography. It also features educational programs aimed at raising awareness about environmental conservation and the importance of protecting natural habitats.

Yingde National Forest Park has become a popular destination for eco-tourism, drawing visitors from the Great Bay Areas. The integration of sustainable tourism practices ensures that tourism activities contribute to the local economy without compromising the park’s ecological integrity.

Despite its success in environmental protection, Yingde National Forest Park faces challenges in competing with other national forest parks and rural tourist spots. Additionally, there is the issue of tourism product renewal, which can lead to the degradation of tourism development over time. One of the primary concerns is how to increase tourist loyalty to ensure higher rates of revisiting and recommendations. This challenge has become a major problem for the park, as sustained visitor engagement is crucial for long-term success. Studying how to improve tourist loyalty is therefore critical for the sustainable tourism development of the park and for contributing to the local economic development.

3.2. Measurement Instrument

The initial phase of the research involved designing a questionnaire, guided by Churchill’s scale development procedures [62]. The questionnaire was structured into two primary sections. The first section focused on collecting demographic data, including age, gender, education level, and economic status. The second section measured key variables, such as tourist perceived value, satisfaction, and loyalty.

Tourists’ perceived value was evaluated using the five-dimension scale developed by Sheth et al. [46] and adapted with the advice of Schneider & Wagemann [60]. The adapted scale encompassed five dimensions, including emotional value (EV), epistemic value (EpV), social value (SV), functional value (FV), and price/value for money (PVM). Tourist satisfaction was assessed through the whole satisfaction between the gap of tourist expectation and perceived performance developed by Oliver [56], comprising four items. Tourist loyalty was measured using the scale developed by Oppermann [53], which included three components: intention to revisit, recommend, and pay more. The scale was shown in

Table 1. Summary of items.

Scales	Mean	Std. Deviation	Skewness	Kurtosis
Emotional value (EV)
EV1: This forest park generates a feeling of wellbeing	3.20	1.14	−0.19	−0.90
EV2: This forest park is exciting	3.99	1.11	−0.82	−0.29
EV3: This forest park is stimulating	3.92	1.12	−0.78	−0.33
EV4: This forest park is fun	3.97	1.08	−0.69	−0.61
EV5: This forest park makes me feel happy	3.95	1.12	−0.84	−0.25
Epistemic value (EpV)
EpV1: This forest park makes me feel adventurous	2.71	1.17	0.11	−0.87
EpV2: This forest park satisfies my curiosity	2.73	1.12	0.24	−0.65
EpV3: This forest park provides authentic experiences	2.79	1.08	0.04	−0.73
EpV4: This forest park makes me know new plant species	2.76	1.12	0.10	−0.71
EpV5: This forest park makes me know more about environmental protection	2.75	1.08	0.07	−0.64
Functional Value (FV)
FV1: This forest park has consistent quality	2.90	1.17	0.14	−0.77
FV2: This forest park has acceptable standard of quality	2.88	1.16	0.08	−0.59
FV3: The forest park is easily accessible	2.89	1.15	0.15	−0.58
FV4: The forest park is smart for navigation and helps need	2.92	1.11	0.07	−0.51
Social Value (SV)
SV1: This forest park improves family harmony	2.92	1.17	0.11	−0.71
SV2: This forest park gives social approval from others	3.14	1.03	−0.40	−0.38
SV3: This forest park helps me to feel acceptable to others	3.10	1.04	−0.30	−0.48
Price/Value for money (PVM)
PVM1: The overall experience offers value for money	3.14	1.00	−0.44	−0.21
PVM2: Most of the local products available are reasonably priced	3.10	1.02	−0.38	−0.54
PVM3: The souvenirs sold are worth buying	2.87	0.97	0.38	−0.50
PVM4: The tickets are worth	2.91	1.00	0.44	−0.61
PVM5: The extra fee activities are worth buying	2.88	0.97	0.41	−0.49
Satisfaction (SAT)
SAT1: I am satisfied with my visit to this forest park	3.40	1.02	−1.78	1.98
SAT2: The experience at this forest park meet my expectations	3.41	1.02	−1.80	2.01
SAT3: This park is one of the best park I have visited	3.44	0.99	−1.90	2.02
SAT4: I felt emotionally fulfilled during my visit to this park	3.43	1.02	−1.82	2.02
Tourist Loyalty (TL)
TL1: I have a strong intention to visit this park again	3.51	0.95	−2.01	2.02
TL 2: I would like to recommend this park to others	3.53	0.92	−2.01	2.03
TL 3: I am willing to spend more money at this park	3.52	0.94	−2.02	2.02

Note: EV means emotional value, EpV means epistemic value, FV means functional value, SV means social value, PVM means price/value for money, TL means tourist loyalty, SAT means tourist satisfaction, and TL means tourist loyalty.

. All rating variables were measured on a 5-point Likert scale, where 1 represented “completely disagree” and 5 indicated “completely agree”.

3.3. Data Collection

The study utilized convenience sampling to collect data from visitors at Yingde National Forest Park. This approach was chosen due to the lack of a defined tourist population or a comprehensive visitor list for random sampling [63,64]. It also proved friendly for both researchers and participants.

Data were collected through a questionnaire-based survey conducted on weekends and public holidays between 1 May and 31 October 2024. To mitigate potential sample selection bias inherent in convenience sampling, data were collected from 11 distinct tourist clusters within the park. Additionally, different questionnaire distributors were assigned to collect data at various locations during each survey session. With approval from park management, four trained graduate students administered the questionnaires on-site. Visitors were invited to participate in the survey about their park experience, with a small gift offered as an incentive. In alignment with related studies on tourist loyalty, no strict distinction was made between first-time and repeat visitors [28,65,66]. Questionnaires were distributed to consenting participants during their rest or visit periods [63].

Of the 436 questionnaires collected, 32 were excluded due to missing values, resulting in 404 valid responses and a validity rate of 91.07%. The sample comprised 54.20% male and 45.80% female respondents. By age, 26.50% were under 18, 25.20% were 19–35 years old, 25.50% were 36–60, and 22.80% were over 60. Educationally, 62.60% held a university degree. In terms of annual income, 59.30% earned between 60,000 and 200,000 Yuan RMB, while 23.80% earned above 200,000 Yuan RMB. A summary of these items is provided in Table 1.

3.4. Research Methods

Initially, the data was assessed for normality with the descriptive analysis of SPSS (v26.0), and the common method bias was evaluated with the single-factor test analysis of SPSS. The research hypotheses were subsequently tested using a two-stage approach, as recommended by Hair et al. [67]. In the first stage, a measurement model was developed using CB-SEM (Covariance-Based SEM) to evaluate validity and reliability. In the second stage, a structural model was applied to test the research hypotheses. At last, the configuration analysis was conducted using the FsQCA model to identify the potential path to increasing tourist loyalty [68].

4. Results

4.1. Normality and Bias Analysis

The normality of the scale items was assessed using skewness and kurtosis analysis. Statistics of skewness and kurtosis tests were found to fall within −2 to 2 for most items, and six of them exceed 2 slightly, indicating that the data normal distribution is acceptable [12,69]. As a result, traditional covariance-based Structural Equation Modeling (CB-SEM) was considered suitable for this research [70,71].

Furthermore, a single-factor test was conducted to check for bias among the items. Four items had eigenvalues greater than 1, with the first factor explaining 36.13% of the total variance, which is below the critical value of 40%. This suggests that common method bias did not significantly affect the data analysis [72].

4.2. Reliability and Validity of the Measures

Reliability was assessed using Cronbach’s alpha (α) and composite reliability (CR). A Cronbach’s alpha value (α) exceeding the critical threshold of 0.7 indicates that the construct demonstrates good reliability [70]. Similarly, a construct is considered to have good reliability if the Composite Reliability (CR) value surpasses the critical threshold of 0.7 [73]. In this study, the α values ranged from 0.784 to 0.972, exceeding the 0.7 threshold, while the CR values ranged from 0.784 to 0.972, also surpassing the 0.7 criterion (

Table 2. Results of measurement model analysis.

	Estimate	S.E.	C.R.	p	Cronbach’s Alpha	CR	AVE
EV ← PV	0.973				0.881	0.881	0.614
EpV ← PV	0.616	0.046	10.859	***
FV ← PV	0.922	0.043	9.956	***
SV ← PV	0.838	0.047	10.995	***
PVM ← PV	0.436	0.045	7.513	***
EV1 ← EV	0.725	0.085	13.028	***	0.849	0.849	0.530
EV2 ← EV	0.736	0.082	13.209	***
EV3 ← EV	0.734	0.079	13.185	***
EV4 ← EV	0.737	0.082	13.232	***
EV5 ← EV	0.706
EpV1 ← EpV	0.816	0.042	22.651	***	0.895	0.895	0.630
EpV2 ← EpV	0.780	0.043	20.723	***
EpV3 ← EpV	0.803	0.042	21.916	***
EpV4 ← EpV	0.768	0.042	20.123	***
EpV5 ← EpV	0.835
FV1 ← FV	0.799	0.038	20.961	***	0.846	0.846	0.578
FV2 ← FV	0.798	0.038	20.871	***
FV3 ← FV	0.794	0.037	20.69	***
FV4 ← FV	0.862
SV1 ← SV	0.745	0.082	12.06	***	0.784	0.784	0.548
SV2 ← SV	0.729	0.083	11.966	***
SV3 ← SV	0.747
PVM1 ← PVM	0.954	0.041	23.21	***	0.919	0.918	0.694
PVM2 ← PVM	0.823
PVM3 ← PVM	0.789	0.037	23.993	***
PVM4 ← PVM	0.830	0.038	20.951	***
PVM5 ← PVM	0.820	0.043	20.561	***
SAT1 ← SAT	0.894	0.042	24.67	***	0.920	0.920	0.741
SAT2 ← SAT	0.847	0.044	22.344	***
SAT3 ← SAT	0.833	0.043	21.673	***
SAT4 ← SAT	0.865
TL1 ← TL	0.955	0.022	45.454	***	0.972	0.972	0.920
TL2 ← TL	0.962	0.021	47.492	***
TL3 ← TL	0.960

Note: S.E. means standard error, C.R. means critical ratio, CR means composite reliability, AVE means average variance extracted; EV means emotional value, EpV means epistemic value, FV means functional value, SV means social value, PVM means price/value for money, TL means tourist loyalty, SAT means tourist satisfaction, TL means tourist loyalty; *** p < 0.001.

). These results collectively indicate robust reliability of the measurement scale.

Convergent validity was evaluated through item factor loadings and the average variance extracted (AVE). A construct is considered to demonstrate satisfactory convergent validity if the factor loadings exceed the critical threshold of 0.7. Furthermore, a construct is regarded as having adequate convergent validity if the Average Variance Extracted (AVE) exceeds 0.50 [74]. In this study, the critical ratios (t-statistics) for factor loadings ranged from 7.513 to 47.792, with all p-values below 0.001, confirming statistical significance. The factor loadings ranged from 0.725 to 0.960, all surpassing the 0.70 threshold, thereby supporting acceptable convergent validity. Additionally, the AVE values ranged from 0.530 to 0.923, all exceeding the 0.50 threshold [74], further substantiating the scale’s convergent validity (Table 2).

Discriminant validity was assessed through the covariance between constructs. The covariance was standardized and expressed as correlations to determine whether the constructs were distinct and separate from one another. According to Hair et al., correlations were considered to indicate distinct constructs if their values were lower than 0.9 [75]. Additionally, discriminant validity was further tested by examining the relationships between the average variance extracted (AVE) and the correlations or shared variance (SV) of the variables (AVE-SV). Discriminant validity is supported when the square root of the AVE is greater than the correlations between a latent variable and others [76]. The HTMT (Heterotrait-Monotrait Ratio) analysis was employed to test the discriminant validity too. The HTMT ratio less than 0.85 is generally considered to be discriminant validity of the two constructs [77]. In this study, the results revealed that the highest correlation coefficient was 0.842, lower than the threshold of 0.9 (

Table 3. Discriminant validity analysis.

Validity Analysis
	CR	AVE	MSV	MaxR(H)	EV	EpV	FV	SV	PVM	SAT	TL
EV	0.849	0.530	0.775	0.849	0.728
EpV	0.895	0.630	0.783	0.896	0.494 ***	0.794
FV	0.846	0.578	0.775	0.846	0.680 ***	0.681 ***	0.760
SV	0.784	0.548	0.288	0.785	0.380 ***	0.122 *	0.311 ***	0.740
PVM	0.918	0.694	0.783	1.007	0.614 ***	0.785 ***	0.752 ***	0.537 ***	0.833
SAT	0.920	0.741	0.761	0.921	0.048	0.294 ***	0.122 *	0.529 ***	0.394 ***	0.861
TL	0.972	0.920	0.761	0.972	0.017	0.358 ***	0.072	0.454 ***	0.423 ***	0.852 ***	0.959
HTMT Analysis
	EV	EpV	FV	SV	PVM	SAT	TL
EV	—
EpV	0.43	—
FV	0.75	0.59	—
SV	0.31	0.10	0.25	—
PVM	0.45	0.84	0.63	0.40	—
SAT	0.05	0.27	0.11	0.45	0.37	—
TL	0.02	0.33	0.07	0.40	0.42	0.82	—

Note: EV means emotional value, EpV means epistemic value, FV means functional value, SV means social value, PVM means price/value for money, TL means tourist loyalty, SAT means tourist satisfaction, TL means tourist loyalty; * p < 0.05, *** p < 0.001.

). Moreover, the square root of the AVE of a latent variable is greater than the correlations between the latent variable and other latent variables, showing discriminant validity (Table 3). The HTMT values range from 0.020 to 0.840, indicating that the constructs are distinct (Table 3). Therefore, based on the correlation analysis, the AVE-SV approach, and the HTMT analysis, the latent variables of the scale were found to be significantly distinct. In comparison to reliability and validity analyses conducted in similar forest park studies [4,22,51,63], the lowest CR and Cronbach’s alpha values in our study were 0.74 and 0.76, respectively, while the factor loading and AVE were 0.70 and 0.43, respectively. The corresponding coefficients in our study exceed the minimum thresholds reported in prior research, demonstrating that the constructs exhibit robust reliability and validity.

4.3. Hypotheses Test

4.3.1. Goodness-of-Fit

A structural model was utilized to test the hypotheses, employing a bootstrapping procedure with 5000 subsamples following the assessment of reliability and validity. The overall fit of the model was evaluated using the chi-square test and several goodness-of-fit indices. A model is considered acceptable if the ratio of the χ² statistic to the degrees of freedom is 3 or lower [78]. For the goodness-of-fit indices, values of the Comparative Fit Index (CFI), Non-Normed Fit Index (NNFI), and Incremental Fit Index (IFI) exceeding 0.90 were used as thresholds for an acceptable model fit [79]. Additionally, the Root Mean Square Error of Approximation (RMSEA) values of 0.08 or below were considered indicative of a good fit. The results showed that χ²/df = 2.97, which was lower than the threshold of 3, indicating moderate fit. The CFI was 0.902, and the NNFI was 0.90, indicating a good fit between the proposed model and the data. The IFI was 0.901, suggesting a good fit between the proposed model and the data. Additionally, the RMSEA was 0.079, suggesting a good fit of the model to the data.

4.3.2. Hypothesis Test

The research hypotheses were tested and analyzed based on the path coefficients and significance (

Table 4. Hypotheses test.

	Estimate	S.E.	C.R.	p	Label
SAT ← PV	0.33	0.05	6.41	***	par_24
TL ← PV	0.10	0.03	3.24	0.001	par_23
TL ← SAT	0.84	0.04	20.98	***	par_13

Note: SAF means satisfaction, TL means tourist loyalty, PV means perceived value; *** p < 0.001.

and Figure 2).

The perceived value (β = 0.10, p < 0.01) exerts a significant positive effect on tourist loyalty, indicating that higher perceived value leads to increased loyalty. Thus, hypothesis H1 is supported by the structural model results.

Similarly, perceived value (β = 0.33, p < 0.001) has a positive influence on tourist satisfaction, confirming that an improved perceived value leads to greater satisfaction. Therefore, hypothesis H2 is also supported.

Furthermore, satisfaction significantly and positively affects tourist loyalty (β = 0.84, p < 0.001), supporting hypothesis H3. The findings suggest that enhanced satisfaction leads to higher levels of tourist loyalty.

4.3.3. Mediating Effect Test

Zhao et al., suggested that indirect effects can be assessed using the bootstrapping method [80]. Following this approach, the indirect effects were evaluated using AMOS 28.0. The 95% confidence intervals were derived from 5000 bootstrapping resamples (

Table 5. Results of the indirect effects.

Parameter	Estimate	Lower	Upper	p
SAT ← PV	0.33	0.23	0.42	0.001
TL ← SAT	0.84	0.80	0.87	0.002
TL ← PV (direct effect) (%)	0.10 (26.51%)	0.05	0.15	0.001
Indirect effect (%)	0.28 (73.49%)	0.19	0.35	0.001
Total effect	0.38	0.29	0.46	0.001

Note: SAF means satisfaction, TL means tourist loyalty, PV means perceived value.

). The findings reveal that tourist satisfaction significantly mediates the relationship between perceived value and loyalty (β = 0.28, p < 0.001), thus confirming hypothesis H4. Furthermore, the total effect of perceived value on tourist loyalty was evaluated by examining both the direct and indirect effects. The direct effect (β = 0.10) accounts for 26.51% of the total effect, while the indirect effect (β = 0.28) accounts for 73.49% of the total effect (Table 5).

4.4. Conditional Analysis of Higher Tourist Loyalty

Fuzzy-set Qualitative Comparative Analysis (FsQCA) is a valuable method for identifying potential solutions that explain the outcome of interest or test variable relationships grounded in theoretical frameworks. Given the significant positive effects of perceived value on both tourist loyalty and satisfaction in forest parks, FsQCA (v3.0) will be employed to analyze the possible solutions for improving tourist loyalty associated with specific perceptions of value.

4.4.1. Data Calibration

The first step in the analysis is to compute the values of latent variables, which will serve as inputs for the FsQCA analysis when the latent variables are measured with multiple items. Several methods are available for calculating the values of latent variables, including item means, standardized item sums, weighted sum scores, regression scores, and Bartlett scores [81]. The weighted sum score, which is calculated by multiplying the standardized factor loadings of each item by its corresponding scaled score, is commonly used. This method is particularly advantageous as it applies different weights (i.e., factor loadings) to each item, ensuring that items with the highest loadings have the greatest impact on the factor score [81]. The factor loadings for the items were computed using confirmatory factor analysis (CFA) in AMOS (v28.0) (Table 2).

In FsQCA, calibration of variables is required to construct fuzzy sets, where the values range from 0 to 1 [68]. A fuzzy set can be conceptualized as a group, and the values within this range define the degree of membership of a case in the group, with 0 indicating no membership and 1 indicating full membership. A membership score of 0.5 represents the midpoint, indicating that a case is equally a member and a non-member of the fuzzy set. Such a case is considered to belong to the intermediate set.

Data calibration can be conducted using either direct or indirect methods, with the direct method being more commonly applied and recommended. In this approach, the researcher assigns three values based on substantive knowledge and theoretical understanding, corresponding to full-set membership, full-set non-membership, and the intermediate set [82]. Typically, the 95%, 50%, and 5% percentiles of the latent variables were calculated to define the thresholds for converting the data into fuzzy sets [44,68]. When the value corresponds exactly to the 50% percentile, it is difficult to determine whether the case belongs to full-set membership or full-set non-membership. To address this ambiguity, a small constant (0.001) is added or subtracted when determining the crossover points [44]. In this study, we added a constant of 0.001 to the 50% percentile value for each latent variable. After completing the data calibration, the five explanatory latent variables—Emotional Value, Epistemic Value, Functional Value, Social Value, and Price/Value for Money—along with the mediating latent variable, satisfaction, and the outcome latent variable, tourist loyalty, were entered into FsQCA for necessity analysis and true table analysis.

4.4.2. Necessity Analysis

Before conducting the truth table analysis, a necessity analysis is performed to assess whether there are necessary explanatory conditions for the outcomes. A consistency level greater than 0.9 for an explanatory variable indicates that it is a necessary condition for the outcome variables [60].

In this study, perceived value has a significant positive effect on tourist loyalty, with satisfaction mediating the relationship. Thus, tourist loyalty and satisfaction are treated as outcome variables, while the five dimensions of perceived value are set as explanatory variables. The necessity analysis results reveal that the highest consistency for the five explanatory variables was 0.80 for tourist loyalty and 0.79 for satisfaction as outcome variables (

Table 6. Results of necessity analysis for tourist loyalty.

Outcome Variable: TLF			Outcome Variable: ~TLF
	Consistency	Coverage		Consistency	Coverage
EV	0.63	0.91	EV	0.55	0.24
~EV	0.48	0.78	~EV	0.80	0.39
EpV	0.52	0.86	EpV	0.63	0.31
~EpV	0.58	0.84	~EpV	0.71	0.31
FV	0.62	0.91	FV	0.52	0.23
~FV	0.48	0.77	~FV	0.80	0.38
SV	0.54	0.86	SV	0.67	0.32
~SV	0.57	0.85	~SV	0.71	0.31
PVM	0.60	0.90	PVM	0.58	0.26
~PVM	0.51	0.80	~PVM	0.78	0.37

Note: TLF means tourist loyalty, EV means emotional value, EpV means epistemic value, FV means functional value, SV means social value, PVM means price/value for money, ~ means logical “not”.

and

Table 7. Results of necessity analysis for tourist satisfaction.

Outcome Variable: SATF			Outcome Variable: ~SATF
	Consistency	Coverage		Consistency	Coverage
EV	0.64	0.77	EV	0.67	0.47
~EV	0.55	0.74	~EV	0.66	0.52
EpV	0.57	0.77	EpV	0.64	0.50
~EpV	0.63	0.75	~EpV	0.70	0.49
FV	0.62	0.76	FV	0.63	0.44
~FV	0.54	0.71	~FV	0.66	0.51
SV	0.58	0.76	SV	0.71	0.54
~SV	0.65	0.80	~SV	0.68	0.48
PVM	0.69	0.85	PVM	0.62	0.44
~PVM	0.55	0.71	~PVM	0.79	0.60

Note: SAF means satisfaction, EV means emotional value, EpV means epistemic value, FV means functional value, SV means social value, PVM means price/value for money, ~ means logical “not”.

), both of which are below the 0.90 threshold. This indicates that no single dimension of perceived value is a decisive factor in tourist loyalty and satisfaction. Consequently, it is important to explore the combinations of explanatory variables that influence tourist loyalty and satisfaction.

4.4.3. True Table Analysis and Solutions

Truth table analysis identifies all potential configurations of explanatory variables that may result in the outcomes, along with their associated frequency, consistency, and Proportional Reduction in Inconsistency (PRI) consistency [83].

Frequency indicates the number of cases in the sample explained by a particular configuration, and a frequency threshold must be established to ensure that each configuration is represented by a sufficient number of cases. For samples larger than 150 cases, a threshold of 3 or higher is typically set, while for smaller samples, a threshold of 2 may be applied [44,68]. The consistency threshold reflects the likelihood that a given combination of explanatory variables leads to the outcome variable. Higher consistency scores suggest a stronger probability that the explanatory combination causes the observed outcome. The minimum recommended consistency threshold is 0.75 [82]. The Proportional Reduction in Inconsistency (PRI) consistency reflects whether the corresponding row in the truth table belongs to a subset of result Y rather than its negation (~Y). A higher PRI value indicates a lower likelihood of simultaneous subsets, suggesting greater precision in the results [43,84]. Generally, a PRI score greater than or equal to 0.75 is considered indicative of reliable and accurate conclusions.

In our study, the frequency threshold is set at 3, given that our sample size is 404, and all configurations with a frequency below this threshold are excluded. The consistency threshold is set at 0.80, which exceeds the recommended minimum value of 0.75, while the PRI consistency is set at 0.75.

After setting the parameters, FsQCA generates a complex solution, a parsimonious solution, and an intermediate solution. The complex solution includes all potential condition combinations, while the parsimonious solution simplifies this by highlighting essential conditions. The intermediate solution, derived from counterfactual analysis, incorporates only plausible alternatives. The final result combines the parsimonious and intermediate solutions. Variables appearing in both are core drivers of outcomes, while those appearing only in the intermediate solution are auxiliary.

After conducting the truth table analysis, five distinct configurations of explanatory variables were identified for positive tourist loyalty (

Table 8. Configuration analysis for Tourist loyalty.

Variabie	Solution A	Solution B	Solution C	Solution D	Solution E
EV	◯	●	⬤	⬤	⬤
EpV	◌	⬤	◌	●	○
FV	◯	⬤	●	●	○
SV	◯	◌	⬤	⬤	⬤
PVM	⬤	◯	○	◌	●
consistency	0.90	0.95	0.94	0.97	0.92
raw coverage	0.31	0.30	0.33	0.42	0.16
unique coverage	0.01	0.01	0.02	0.13	0.01
solution consistency	0.91
solution coverage	0.73

Note: Black circle indicates the presence of a condition, white circle indicates its absence, and dotted circle means do not care about condition. Large (black and white) circles indicate core condition, and small (black and white) circles mean peripheral condition.

). The results show that the subset consistency for each solution is above the threshold value of 0.8, with the overall consistency across all combinations being 0.91. This indicates that these five configurations represent sufficient conditions for explaining positive tourist loyalty. Furthermore, solution coverage represents the explanatory power of the configurations, representing the proportion of cases that can be explained by all configurations [82]. Case coverage reflects the empirical relevance or importance of the configurations, similar to R² in regression analysis [44]. In our study, the five configurations collectively explain 73% of the cases.

Based on the configuration analysis of tourist loyalty as an outcome through FsQCA, the pathways to enhancing tourist loyalty can be categorized into three distinct modes: the cost value-driven model, the dual-core model driven by functional and epistemic value, and the emotional and social value-driven model. The cost value-driven model treats the cost incurred by tourists in the forest park as an auxiliary variable, and it explains more than 30% of the cases with tourist loyalty, demonstrating a strongly reliable predictor of tourist loyalty to the park (consistency = 0.90). The dual-core model, driven by functional and epistemic values, identifies these two variables as the primary drivers, with emotional value serving as an auxiliary factor. This pathway explains over 30% of the cases with tourist loyalty and shows a higher likelihood of boosting loyalty (consistency = 0.95). The emotional and social value-driven model consists of three pathways, where emotional and social values serve as core drivers, while functional value is an auxiliary variable in the first path, both functional and epistemic values serve as auxiliaries in the second, and cost value is auxiliary in the third. The first two paths explain more than 30% of the cases with loyalty, while the third path accounts for 16% of the cases with loyalty, with all three paths exhibiting a strong likelihood of increasing tourist loyalty to the forest park (consistency > 0.90).

Additionally, four distinct configurations of explanatory variables were identified for positive tourist satisfaction (

Table 9. Configuration analysis for satisfaction.

Variabie	Solution A	Solution B	Solution C	Solution D
EV	○	●	●	●
EpV	○	●	◌	●
FV	○	⬤	⬤	⬤
SV	◌	◌	●	●
PVM	●	○	○	◌
consistency	0.88	0.86	0.86	0.89
raw coverage	0.38	0.33	0.36	0.46
unique coverage	0.01	0.01	0.02	0.15
solution consistency	0.81
solution coverage	0.83

Note: Black circle indicates the presence of a condition, white circle indicates its absence, and dotted circle means do not care about condition. Large (black) circle indicates core condition, and small (black and white) circles mean peripheral condition.

). The results indicate that the subset consistency for each solution exceeds the threshold value of 0.8, with the overall consistency for all combinations being 0.81. This suggests that these four configurations represent sufficient conditions for explaining positive tourist satisfaction. Moreover, these configurations collectively account for 83% of the cases, demonstrating strong explanatory power.

According to the configuration analysis of tourist satisfaction as an outcome using FsQCA, two primary pathways to improve tourist satisfaction were identified: the economic value-driven model and the functional value plus driven model. In the economic value-driven model, the cost incurred by tourists in the forest park is treated as an auxiliary variable, explaining more than 38% of the cases and showing a high probability of increasing tourist satisfaction with the park (consistency = 0.88). The functional value plus-driven model consists of three distinct paths, with functional value as the core driving variable. In the first path, emotional value and epistemic value serve as auxiliary drivers; in the second, emotional value and social value are auxiliary drivers; and in the third, emotional value, epistemic value, and social value act as auxiliary drivers. Each of these three paths accounts for over 30% of the cases with tourist satisfaction, with all paths demonstrating a high likelihood of enhancing satisfaction with the forest park (consistency > 0.86).

4.4.4. Robust Test

To ensure the reliability of the results, a robustness test is an essential step in QCA research. We conducted a data robustness test by adjusting the crossover point (with increases and decreases of 0.05), raising the consistency threshold from 0.80 to 0.85, and lowering the frequency threshold to 2 [44,85,86]. The configurations of the outcomes remained consistent with the original analysis, with only slight changes in overall consistency and coverage. These results suggest that the configurations are robust and reliable.

5. Conclusions and Discussion

5.1. Research Conclusions

The covariance-based SEM and fsQCA methods were employed to reveal that perceived emotional value, epistemic value, functional value, social value, and price/value for money jointly influence tourist satisfaction and loyalty in forest parks. The following conclusions are drawn.

First, tourist perceived value in forest parks can be represented with emotional value, epistemic value, functional value, social value, and price/value for money significantly. The coefficients of emotional value, functional value, and social value are higher than that of epistemic value and price/value for money, indicating emotional value, functional value, and social value have greater importance in measuring tourist perceived value in forest parks.

Second, tourist perceived value has a significant positive effect on tourist loyalty and tourist satisfaction. Tourist satisfaction plays a significant positive mediating role between perceived value and tourist loyalty in forest parks. The indirect effect of tourist perceived value on tourist loyalty mediated by tourist satisfaction is higher than the direct effect. Tourist satisfaction is an important mediating variable for improving tourist loyalty in forest parks. Studies have shown that perceived value has a significant impact on tourist loyalty across various tourism sectors, including eco-tourism, forest parks, and rural tourism destinations [4,37,61]. Our study aligns with these findings, confirming that perceived value plays a crucial role in shaping tourist loyalty, with satisfaction acting as a mediator in this relationship. Previous research consistently demonstrates that enhancing perceived value can lead to increased tourist satisfaction, which, in turn, fosters greater loyalty through revisit intentions and recommendations. For destination managers, this underscores the importance of improving perceived value as a key strategy to boost tourist loyalty, ensuring higher rates of repeat visitation and positive word-of-mouth.

Third, five pathways to enhancing tourist loyalty are identified with FsQCA analysis, and the five pathways can be categorized into three distinct modes: the economic value-driven model, the dual-core model driven by functional and epistemic value, and the emotional and social value-driven model. The economic value-driven model treats the cost incurred as an auxiliary variable, and the dual-core model, driven by functional and epistemic values, identifies these two variables as the core drivers, while the emotional and social value-driven model consists of three pathways, where emotional and social values serve as core drivers. The pathways indicate that the dimensions of perceived value play different roles across various pathways, with no single dimension consistently playing the most important role in enhancing tourist loyalty across all pathways.

Fourth, four primary pathways to improve tourist satisfaction were identified, and they can be categorized into two modes: the economic value-driven model and the functional value plus driven model. In the economic value-driven model, the cost is treated as an auxiliary variable, while the functional value plus driven model consists of three distinct paths, with functional value as the core driving variable. Different dimensions of perceived value play distinct roles in various pathways for improving tourist satisfaction.

5.2. Theoretical and Managerial Contribution

Firstly, we combine the five dimensions of tourist perceived value to identify pathways for improving tourist satisfaction and loyalty in forest parks, offering an innovative perspective on tourist loyalty research. Previous studies have shown that perceived value significantly affects tourist loyalty [41,64,87]. However, much of the existing research has focused on the impact of a single dimension [4] or a joint dimension [88] of perceived value on tourist loyalty or has primarily examined the measurement of perceived value dimensions [55], often overlooking the influence of the multi-dimensional interplay on tourist loyalty. To fill this gap, we employed fsQCA to explore the effect mechanism of emotional value, epistemic value, functional value, social value, and price/value for money on tourist loyalty. Through this analysis, we identified key factors and pathways that contribute to higher satisfaction and increased loyalty. The results provide valuable theoretical insights for the diversified management of forest parks and contribute to advancing the research on tourist loyalty by incorporating a multi-dimensional approach.

Secondly, our study demonstrates that price/value for money plays an important role in shaping higher tourist satisfaction and increasing loyalty, thereby enriching the research on perceived value. While previous research on perceived value measurement has highlighted the importance of price in tourism, our study offers new insights into this relationship. Recent studies have used structural equation modeling to confirm that price/value for money is a key factor in measuring perceived value and influencing tourist loyalty [60]. Our findings support this by showing that price, as the only indispensable factor in the pathways, leads to higher satisfaction and increased tourist loyalty. Moreover, we identify the interaction effects of functional and epistemic value as well as emotional and social value on tourist loyalty. The combinations of functional and epistemic value and emotional and social value are key to enhancing tourist loyalty. As core factors, the combinations of emotional and social value can be further extended to three paths for improving tourist loyalty, advancing the field of configuration research on tourist loyalty.

Third, economic value (price/value for money) is an essential factor influencing both tourist satisfaction and loyalty [60]. Therefore, forest park managers should prioritize pricing strategies that enhance economic value for tourists, such as offering family packages, parent-child tickets, group discounts, and promotional offers. These initiatives not only improve the perceived economic value but also contribute to higher tourist satisfaction and loyalty.

Fourth, the combinations of emotional and social value can be extended into three distinct pathways for enhancing tourist loyalty. It is crucial for park managers to strengthen emotional connections between tourists and the forest park, which can significantly improve loyalty. For instance, organizing high-quality, interactive activities that foster emotional engagement among visitors can stimulate emotional resonance and attachment. Additionally, enhancing the construction and quality of basic service facilities, such as hiking trails, digital infrastructure (e.g., free Wi-Fi, navigation maps), and leisure amenities within the forest park, will elevate the perceived functional value. These improvements can further boost tourist loyalty by creating a more convenient and enjoyable experience.

Fifth, both over-visitation and under-visitation contribute to the unsustainable development of national forest parks. Over-visitation leads to environmental degradation, while under-visitation results in economic challenges that undermine the parks’ sustainability. To address these issues, it is crucial to strengthen policies related to forest park management, ensuring they integrate environmental, economic, and social goals. Collaboration among relevant government agencies (e.g., forestry, tourism, and environmental departments) should be promoted to develop integrated strategies that balance park development, conservation, and sustainable tourism. Furthermore, clear metrics should be established to assess the effectiveness of forest park management and evaluate the impact of tourism. Regular policy assessments, informed by data-driven insights, are essential for making necessary adjustments and ensuring long-term sustainability.

5.3. Limitations

Our study has made a little breakthrough in perceived value and tourist loyalty in forest parks, but there are still limitations that need to be improved. First, the perceived value-satisfaction-loyalty chain model was employed to study how tourist perceived value influences loyalty and identify the pathways to improve tourist loyalty based on the combination of five dimensions of perceived value. However, this study may not identify all factors that clarify the action mechanism on higher tourist loyalty. Therefore, to extract all factors that influence tourist loyalty, identify key factors, and find combinations of key factors to improve tourist loyalty is one of the research points in the future. It is important to find the key factors of items and combinations of items to improve tourist loyalty. Second, we took Yingde National Forest Park tourists as the research sample; there are still differences about tourist spots, recreation items, and even service quality in forest parks and regions. Therefore, heterogeneity analysis of factors and pathways remains to be further explored next. Third, it is clear that the tourism experiences of different visitor groups (e.g., first-time vs. repeat visitors, weekday vs. weekend visitors, and family vs. individual visitors) vary significantly. Therefore, future research should focus on comparing the differences among these distinct visitor groups.