The Threshold of Authenticity and the Limits of Cultural Involvement: A Non-Linear Perspective on Heritage Tourist Loyalty

Abstract

As urban heritage tourism faces the challenge of balancing conservation and development, understanding the non-linear drivers of tourist loyalty becomes crucial for sustainable development of urban heritage tourism. This study delves into how perceived authenticity and cultural involvement affect tourist loyalty in an urban heritage district and tests whether these effects vary across different experience levels. Data was collected using a structured survey questionnaire, which was administered through on-site convenience sampling. To ensure the reliability and validity of the data, a face-to-face recruitment approach was employed, with participants invited to complete the survey immediately after their visit. The sample consisted of 469 domestic tourists in the Pingjiang Historic Block. The study applied quadratic modeling within Partial Least Squares Structural Equation Modeling (PLS-SEM) to analyze both linear and non-linear relationships among the key factors. The results indicate that loyalty is driven primarily by experiential and involvement-related processes, with cultural motivation and object-based authenticity influencing loyalty mainly through existential authenticity and cultural involvement, rather than through direct effects. The findings showed that the non-linear relationships were observed in the process of loyalty formation, whereas the relationships among antecedents and cultural involvement remain largely linear. Specifically, both object-based authenticity and existential authenticity have a minimum credibility threshold that must be met before loyalty increases meaningfully, while cultural involvement follows an inverted U-shaped pattern, implying diminishing returns beyond an optimal level. This study contributes to a further understanding of the complex relationship between authenticity and tourist loyalty. The findings provide a novel perspective for policymakers, offering insights into how the non-linear relationships of loyalty can be leveraged to promote the sustainable development of urban heritage tourism.

1. Introduction

Urban heritage tourism has become a significant method through which cities pursue sustainable development while safeguarding traditional heritage [1]. As urban historic districts are increasingly repositioned as experiential cultural spaces, heritage assets are expected to generate not only short-term visitation but also longer-term social and economic value for host communities [2,3]. Therefore, heritage destination competitiveness in urban heritage districts is less dependent on sheer visitor volume and more dependent on whether cultural experiences can be translated into enduring relationships with visitors [4,5]. Accordingly, there is a growing need to clarify how visitors develop stable commitments to heritage districts within urban contexts.

Heritage tourism refers to travel that focuses on the exploration of historical, cultural, and heritage sites [6]. It involves experiencing tangible cultural landmarks, such as monuments, historic buildings, and landscapes, alongside intangible cultural elements like traditions, rituals, and local practices [7]. This form of tourism has gained momentum with the increasing demand for authentic cultural experiences, as tourists seek to connect with local heritage and history in a more immersive and meaningful way [8]. Urban heritage tourism, specifically, is emerging as a key area of growth, where cities use their historical and cultural assets as a vehicle for regeneration, aiming to boost economic development while preserving cultural identities [6,9].

Over the past few decades, urban heritage tourism has become a critical sector for cities worldwide, enabling them to balance the preservation of cultural heritage with urban development needs [1,10]. However, it also faces several challenges, including the risk of over-tourism, commercialization of heritage sites, and the tension between maintaining authenticity and accommodating large numbers of visitors [3]. These challenges underscore the importance of understanding how heritage tourism can contribute to sustainable development by fostering long-term relationships between tourists and heritage sites [11].

Tourist loyalty is a pivotal concept in heritage tourism because tourists’ revisiting intention and positive word-of-mouth determine the sustainability of tourism industry [12]. Many studies have discovered that tourists’ sustained commitment is more likely to develop when they perceive local cultural meanings and narratives, or when they further participate in local cultural activities that deepen their connection to the place [13,14]. Prior research has commonly considered perceived authenticity to serve as a central antecedent of deeper engagement and subsequent loyalty [8,15]. In these studies, authenticity is typically conceptualized as multidimensional, encompassing object-based authenticity that reflects the tangible and intangible heritage attributes, and existential authenticity that captures visitors’ self-related feelings of being true to themselves during the experience [8,16].

Moreover, scholars found these authenticity dimensions foster cultural involvement by strengthening visitors’ cognitive appraisal of heritage and their experiential resonance with the destination and uncovered the relationship between involvement and loyalty [13,17]. Despite the relationships among authenticity and cultural involvement, as well as cultural involvement and tourist loyalty being confirmed, specifying these relationships as strictly linear may be overly restrictive. Recently, Yeh et al. [17] uncovered the complex nature of the relationship between authenticity and involvement, behavior intention as well as nostalgia, explicitly highlighting the existence of non-linear effects. This finding suggests that the marginal effects of authenticity and cultural involvement may vary across their levels.

Although traditional linear models have provided valuable insights into the relationship between authenticity and tourist loyalty as well as that between involvement and tourist loyalty, they often fail to capture the full complexity of how loyalty develops. Specifically, these models overlook the possibility of non-linear effects, where the influence of authenticity and involvement may change at different levels of tourist loyalty [17]. To address this limitation, this study adopts PLS-SEM with quadratic modeling, which allows for the modeling of non-linear relationships between key constructs. Quadratic modeling enables the identification of turning points where the marginal effects of authenticity and cultural involvement shift, providing a more nuanced understanding of how these factors influence tourist loyalty. This approach is particularly useful for exploring the “threshold effects” and diminishing returns that may occur as tourists’ experiences intensify. By incorporating quadratic terms, this study offers a more accurate depiction of loyalty formation, highlighting how the impact of authenticity and involvement evolves across different levels of experience.

Consequently, to bridge these gaps, this study aims to achieve two research objectives within the context of urban heritage tourism. First, the study seeks to empirically verify the mediating role of cultural involvement, delineating how distinct dimensions of authenticity (object-based and existential) translate into loyalty through this psychological state. Second, this study introduces a quadratic setting in PLS-SEM to examine whether the marginal effects change at different levels of constructs, and then more meticulously depicts the formation mechanism of loyalty in urban heritage tourism.

2. Literature Review

2.1. Consumer-Based Model of Authenticity

The consumer-based model of authenticity (C-BA model) was proposed by Kolar and Zabkar [8], which is a seminal research model. The model shifts the heritage tourism focus from the “authenticity of objects” to the “authenticity of the tourist experience” [15]. Unlike traditional approaches that view authenticity solely as an inherent property of artifacts, the C-BA model conceptualizes it as a multi-dimensional construct perceived by the tourists [8]. They propose object-based authenticity (OBA) as tourists’ perception of genuine artifacts and heritage and existential authenticity (EA) as the tourist’s intrapersonal state of breaking free from daily constraints and finding their true self [8]. Another significant contribution of the model is the validation of the relationship between OBA and EA. Kolar and Zabkar [8] were the first to empirically confirm that OBA serves as a critical condition for triggering tourists’ EA.

The C-BA model represents a robust and expandable structural framework that has been widely applied across diverse research contexts [15,18,19]. Drawing upon these studies, this study positions cultural involvement (CINV) as a critical “experiential bridge” facilitating the transmission of authenticity’s effects to loyalty. Furthermore, it advances the model by scrutinizing the non-linear relationships.

2.2. Cultural Motivation

Cultural motivation (CM) serves as the primary antecedent variable in the C-BA model. It is fundamentally defined as the internal psychological drive that compels individuals to travel for the purpose of acquiring new knowledge and experiencing different cultures [8,20]. Distinct from general leisure motivation, cultural motivation is characterized by a specific cognitive orientation towards learning, exploration, and the appreciation of heritage assets. Some studies found that highly motivated tourists are more sensitive to cultural cues and possess a stronger desire to verify the authenticity of their experiences [21].

Tourists driven by strong cultural motivation are not passive observers but active information seekers [6,7]. This is because tourists with a heightened cognitive desire tend to verify the origins and factual accuracy of heritage sites [16]. Consequently, tourists’ CM acts as a psychological mechanism that directs attention towards specific physical cues, such as historic buildings, relics, and craftsmanship, thereby enhancing their sensitivity to OBA. Previous empirical studies have consistently confirmed that tourists who have the stronger CM, have the higher perception of OBA [8,15,21,22]. Therefore, the study proposes the following hypothesis:

H1: 

CM has a positive influence on OBA.

Moreover, CM serves as a potent catalyst for tourists’ EA. According to Wang [16], existential authenticity is activated when people seek to escape the daily life and recover their “true selves” through tourism activities. Tourists with high CM possess a stronger desire for such self-discovery and spiritual enrichment [20,23]. Therefore, they are more likely to shed social constraints and engage in authentic self-expression [8]. Brown [24] supports this point, suggesting that tourists who have stronger CM to experience a different culture they have higher EA. Building on the aforementioned theoretical arguments and empirical findings, the study proposes the following hypothesis:

H2: 

CM has a positive influence on EA.

2.3. Object-Based and Existential Authenticity

In the context of heritage tourism, OBA provides the necessary material cues and credible atmosphere to arouse tourists’ EA [8,15,16]. When tourists believe they are interacting with “real” heritage, they can reduce suspicion and engage in the existential quest for their true selves [16,23]. Empirical studies have repeatedly validated this relationship, confirming that higher perceptions of OBA significantly enhance tourists’ perceptions of EA [8,15,21,25]. Thus, the study proposes the hypothesis:

H3: 

OBA has a positive influence on EA.

2.4. Cultural Involvement

Cultural involvement refers to a psychological state of motivation, arousal, or interest between an individual and a specific cultural destination [26,27]. As a multidimensional construct, it entails the allocation of both cognitive and emotional resources, reflected in heightened attention, mental effort, and information processing [2,17]. Unlike general leisure involvement, cultural involvement entails a higher degree of cognitive effort as tourists actively interpret the symbolic meanings of heritage assets. Crucially, this study focuses on “situational involvement” which refers to a temporary state triggered by immediate environmental stimuli (i.e., tangible and intangible heritage), rather than “enduring involvement” that reflects long-term personal interest. Hence, CINV reflects the intensity of the tourist’s immersion in the cultural experience.

Departing from traditional linear perspectives, cultural involvement is examined herein through a quadratic framework. Drawing on Fotiadis et al. [28] and Gao et al. [2], excessive immersion can lead to cognitive and emotional saturation, potentially hindering positive tourist experiences. Yeh et al. [17] further suggested that involvement may exhibit non-linear patterns, such that the marginal effects may vary across levels of involvement. This exploration offers a fresh theoretical perspective, contributing to a comprehensive understanding of involvement’s complex role in the intricate relationship between perceived authenticity and loyalty.

CM represents the specific desire to acquire knowledge, which compels tourists to dedicate their energy to the tourism [29]. Higher levels of CM typically mobilize greater attention, thereby deepening the state of CINV [30]. On the other hand, drawing on Yeh et al.’s [17] study, this study further explores a potential quadratic relationship, postulating that the positive influence of motivation may not be monotonic. Moreover, if CM reaches very high levels, its association with CINV may not increase proportionally, suggesting a potential curvilinear relationship. Consequently, this study proposes the following hypotheses:

H4 Line: 

CM has a positive influence on CINV.

H4 Quadratic: 

CM is curvilinearly related to CINV.

According to Bryce et al. [15], perceived authenticity functions as a dynamic catalyst that transforms tourists from passive observers into engaged participants. Specifically, OBA, grounded in credible heritage cues, primarily ignites tourists’ cognitive involvement. It compels tourists to expend interpretive effort in information search and meaning-making regarding tangible heritage assets. Conversely, EA fosters behavioral involvement, capturing the enactment of culture through situated practices such as ritual participation, hands-on making, and interactive interpretation.

On the other hand, Kontogeorgopoulos [31] noted that highly authentic settings may introduce inconvenience, implying that the effect of perceived authenticity on involvement may be non-monotonic and thus warrants a curvilinear test. While Yeh et al. [17] further uncovered the identification of a quadratic relationship between perceived authenticity and involvement, this research delves deeper into the specific internal dimensions. Consequently, this study proposes the following hypotheses:

H5 Line: 

OBA has a positive influence on CINV.

H5 Quadratic: 

OBA is curvilinearly related to CINV.

H6 Line: 

EA has a positive influence on CINV.

H6 Quadratic: 

EA is curvilinearly related to CINV.

2.5. Loyalty

Tourist loyalty is widely recognized as the ultimate dependent variable in destination marketing research, serving as a critical indicator of a destination’s long-term competitiveness and economic sustainability. Loyalty refers to a deeply held commitment to rebuy a preferred product or service consistently in the future [32]. In the context of tourism, loyalty is typically operationalized through tourists’ behavioral intentions. Specifically, it encompasses two primary facets, the intention to revisit the destination and the willingness to recommend it to others [33,34]. Cultivating tourists’ loyalty is paramount for developing heritage tourism, as retaining existing visitors is significantly more cost-effective than acquiring new ones, and positive word-of-mouth serves as a credible, organic promotion channel [35].

Many scholars have proven the influence of CM on tourists’ loyalty [8,15]. Yoon and Uysal [34] also found that tourists who are highly motivated to understand a culture are more likely to value the experience and subsequently commit to revisiting or recommending the site. Meanwhile, when tourists perceive the genuineness and originality of heritage, they can differentiate the destination from its competitors. This perception of uniqueness acts as a catalyst, not only reinforcing their intention to revisit but also compelling them to disseminate this exclusive discovery through positive word-of-mouth [12]. Moreover, EA fosters a deep emotional bond between the tourist and the destination. Such profound psychological resonance is a potent driver of loyalty, as tourists are intrinsically motivated to repeat experiences that sustain their authentic self-identity [36,37].

Crucially, CINV serves as a proximate determinant of loyalty, acting as the psychological engine that converts experience into commitment. Prebensen et al. [38] found that high levels of cognitive and emotional involvement significantly enhance the perceived value and memorability of the tourism experience. Empirical studies consistently confirm that such deep engagement is a prerequisite for enduring loyalty, directly fuelling the desire to revisit and the willingness to recommend [30,39].

Curvilinear effects are particularly plausible for involvement-driven loyalty. Drawing on the ‘Too-Much-of-a-Good-Thing’ perspective and diminishing marginal utility, very high levels of involvement may yield diminishing returns due to saturation or fatigue, potentially weakening the incremental contribution of involvement to loyalty [40]. The finding aligns with Sweller’s [41] research that excessive cognitive load can lead to cognitive overload and psychological fatigue, thereby dampening tourists’ overall satisfaction and subsequent desire to revisit.

For CM, OBA, and EA, prior research suggests that the relationships between antecedents and loyalty may deviate from strict linearity. For example, Antón et al. [42] explicitly reported linear and non-linear effects when explaining destination loyalty intentions. In the authenticity literature, Yeh et al. [17] further adopted a quadratic model and reported that authenticity demonstrates a non-linear pattern in its associations with behavioral intention and involvement. Moreover, conceptual work has argued that authenticity effects may involve an “equilibrium point” with other experiential attributes, implying that “more” authenticity is not necessarily linearly beneficial [43]. Therefore, this study tests whether these effects are curvilinear without imposing a directional form a priori. Thus, this study proposes the following hypotheses:

H7 Line: 

CM has a positive influence on tourists’ loyalty.

H7 Quadratic: 

CM is curvilinearly related to tourists’ loyalty.

H8 Line: 

OBA has a positive influence on tourists’ loyalty.

H8 Quadratic: 

OBA is curvilinearly related to tourists’ loyalty.

H9 Line: 

EA has a positive influence on tourists’ loyalty.

H9 Quadratic: 

EA is curvilinearly related to tourists’ loyalty.

H10 Line: 

CINV has a positive influence on tourists’ loyalty.

H10 Quadratic: 

CINV is curvilinearly related to tourists’ loyalty.

Overall, this study proposes a research framework based on the C-BA model (as shown in Figure 1). The framework aims to uncover the line and curvilinear relationships between CM, OBA, EA, CINV and tourists’ loyalty.

3. Research Methodology

3.1. Study Location

The empirical setting for this research is Pingjiang Historic Block, recognized as the epitome of the ancient urban landscape in Suzhou. Pingjiang Historic Block covers a total area of 116.5 hectares and stands as one of the best-preserved historic urban quarters in the Yangtze River Delta region [44]. Pingjiang Historic Block has more than 800-year history, and is a significant component of the Grand Canal, a World Heritage Site [45]. The block is one of the most famous urban heritage blocks, honored with 2005 Asia-Pacific Heritage Conservation Award, Chinese first batch Chinese Historical Culture Famous Street in 2009, the most valuable cultural tourism destination in 2014, and notional-level tourist area in 2022 [46,47]. The block has one World Heritage Site (Garden of Couple’s Retreat), 3 provincial-level heritage, 19 municipally recognized heritage buildings and more than one hundred historic buildings, which show the unique Jiangnan urban life and spatial organization, the archetype of traditional water-town buildings and the spiritual and esthetic essence of traditional Chinese literati culture [48]. Pingjiang Historic Block represents a “living museum” that successfully integrates heritage conservation with a well-developed tourism industry (shown in Figure 2). As a mature tourism destination, the block attracts a substantial volume of visitors; preliminary statistics indicate an annual tourist volume of 10.75 million in 2019. This block offers tourists diverse opportunities for immersive engagement. Specifically, the block hosts numerous intangible cultural heritage workshops (e.g., Su embroidery and Song brocade), interactive tea houses featuring Kunqu Opera and Pingtan performances, and experiential retail spaces such as Hanfu (traditional clothing) photography studios.

This study selected Pingjiang Historic Block primarily due to its synergy between heritage authenticity and the modern tourism industry. The abundance of interactive activities allows visitors to transition from passive sightseeing to active cultural involvement, making it an ideal laboratory for examining the mechanisms of cognitive and behavioral involvement proposed in this study. Furthermore, its sustained popularity and diverse visitor demographic ensure the representativeness and reliability of the data collection for the structural model analysis.

3.2. Data Collection

A three-part questionnaire was developed, comprising background information, socio-demographic characteristics, and construct measurement items. The scale was originally designed in English and translated into Chinese using a standard back-translation method. To ensure face validity and readability, the initial draft was reviewed by three academic experts and four tourists, followed by a pilot study with 75 participants to verify the questionnaire’s feasibility. The data collection was collected in the Pingjiang Historic Block with hardcopy questionnaire, only for confirming the readability of the Chinese questionnaires and the answer time. These questionnaires were excluded from the final survey.

To ensure content validity, the measurement constructs were adapted from established scales in the existing literature. Specifically, the items for CM were adapted from a four-item scale by Nguyen and Cheung [49]. The scales for OBA and EA were derived from Kolar and Zabkar [8] and the related literature (e.g., Bryce et al., 2015 [15]; Kesgin et al., 2021 [18]). The measurement items for CINV were adapted from Yeh et al. [17]. Finally, regarding tourists’ loyalty, this study utilized a composite loyalty scale adapted from Bryce et al. [15]. All survey items were measured using a seven-point Likert scale.

Formal data collection took place during the autumn of 2024 and the spring of 2025. Surveys were distributed in person within the Pingjiang Historic Block, utilizing a mixed-mode approach that combined online Quick-Response (QR) codes (via Power CX) and on-site printed forms. The online survey received a total of 2810 access, resulting in 1291 complete responses. In addition, 200 printed questionnaires were distributed through offline channels, and 134 responses were collected. Invalid responses were excluded based on completion time (<2 min), reverse-coded attention checks, with significant missing data (in paper-based version), and straight-lining behavior, yielding a total of 469 valid responses (378 online and 91 offline). This sample size meets the academic standards required for the statistical analysis employed [50,51].

Then, given the self-reported and single-source nature of the survey, potential common method bias (CMB) was assessed using a full-collinearity approach [52]. All full collinearity Variance Inflation Factor (VIF) values were below the recommended threshold of 3.3, suggesting that CMB is unlikely to meaningfully affect the interpretation of the structural relationships.

3.3. Data Modeling

The empirical data were analyzed using SmartPLS 4 (version 4.1.0.0). The analytical strategy integrates Partial Least Squares Structural Equation Modeling (PLS-SEM). Significance testing relied on a bootstrapping procedure with 5000 subsamples. PLS-SEM was utilized to test the structural relationships, chosen for its robustness in handling non-normal distributions and its capability to model both linear and non-linear (quadratic) relationships [53].

Unlike linear models, quadratic PLS-SEM captures non-linear effects by modeling threshold effects and diminishing returns, which are crucial for understanding how loyalty develops as authenticity and involvement increase. Compared with quadratic PLS-SEM, other common research methods (e.g., multi-group analysis) cannot account for the non-linear dynamics [17]. Quadratic PLS-SEM allows for the identification of turning points, where the influence of authenticity and cultural involvement on loyalty changes. This method provides a more accurate and nuanced understanding of loyalty formation, offering insights into how these factors interact at different levels of intensity and contributing to a more detailed model of tourist loyalty in heritage tourism.

4. Results

4.1. Sample Characteristics

Participants exhibited a balanced gender ratio, with the majority falling into the 25 to 34 (39.4%) and 35 to 44 (30.3%) age brackets. The sample reflects a high level of education (n = 241) and income (n = 312 > 6000 RMB/month), ensuring representativeness of the heritage tourist population. Non-local tourists (n = 240) and repeat visitors (59%) formed the core of the sample. Visits were predominantly social (Family: n = 218; Friends: n = 107) rather than solitary. Cultural immersion (n = 176) and leisure (n = 127) were identified as the leading motivations of tourists, supported by the site’s unique landscape and culinary offerings.

4.2. Assessment of Measurement Model

Strict criteria regarding construct validity and reliability were applied to evaluate the measurement model. All outer loadings fell between 0.715 and 0.878, well above the standard cut-off of 0.708, thus confirming indicator reliability [54]. To assess internal consistency, Cronbach’s alpha (CA) coefficients were examined following the guidelines of Hair et al. [54]. The results indicate that all CA values spanned from 0.770 to 0.867, exceeding the required 0.7 level. Additionally, the Composite Reliability (CR) scores ranged from 0.867 to 0.909, surpassing the 0.7 limit. In terms of convergent validity, the Average Variance Extracted (AVE) values ranged from 0.566 to 0.715, thereby satisfying the 0.5 threshold. In summary, the data from

Table 1. Reliability and validity of the measurement model.

Constructs	Items	Loadings	CA	CR	AVE
Cultural Motivation	CM1	0.775	0.867	0.909	0.715
	CM2	0.872
	CM3	0.854
	CM4	0.876
Object-Based Authenticity	OBA1	0.751	0.805	0.872	0.632
	OBA2	0.820
	OBA3	0.849
	OBA4	0.755
Existential Authenticity	EA1	0.724	0.846	0.887	0.566
	EA2	0.784
	EA3	0.756
	EA4	0.775
	EA5	0.757
	EA6	0.715
Cultural Involvement	CINV1	0.746	0.812	0.876	0.640
	CINV2	0.809
	CINV3	0.821
	CINV4	0.820
Loyalty	LOY1	0.859	0.770	0.867	0.687
	LOY2	0.878
	LOY3	0.742

Note: CM: cultural motivation; OBA: object-based authenticity; EA: existential authenticity; CINV: cultural involvement; LOY: loyalty; CA: Cronbach’s alpha; CR: Composite Reliability (rho_c); AVE: Average Variance Extracted.

confirms that the survey instrument possesses adequate reliability and validity.

Then the study uses the Fornell–Larcker criterion and the Heterotrait–Monotrait (HTMT) ratio to verify the empirical distinctiveness of the constructs. As detailed in

Table 2. Results of discriminative validity (Fornell–Larcker and HTMT).

Fornell–Larcker Criterion
	CINV	CM	EA	LOY	OBA
CINV	0.800
CM	0.538	0.845
EA	0.676	0.603	0.752
LOY	0.590	0.423	0.520	0.829
OBA	0.540	0.658	0.701	0.393	0.795
Heterotrait–Monotrait (HTMT) ratio
	CINV	CM	EA	LOY	OBA
CINV
CM	0.634
EA	0.815	0.695
LOY	0.739	0.503	0.644
OBA	0.668	0.783	0.843	0.489

Note: CM: cultural motivation; OBA: object-based authenticity; EA: existential authenticity; CINV: cultural involvement; LOY: loyalty.

, the analysis confirms that the square root of the AVE for every latent construct exceeds the inter-construct correlations, thereby satisfying the Fornell–Larcker requirement [55]. Additionally, the HTMT ratios presented in Table 2 consistently fall below the rigorous threshold of 0.85. The maximum observed value was 0.843 (between OBA and EA), which aligns with the guidelines established by Henseler et al. [56]. Consequently, the measurement model demonstrates satisfactory discriminant validity.

4.3. Assessment of Structural Model

4.3.1. Collinearity Assessment

Following the guidelines established by Hair et al. [54], lateral collinearity was assessed using the VIF matrix. All VIF values fall below the critical threshold of 5.0, with the highest observed value being 3.572. This result confirms that the structural model is free from critical collinearity issues [54,57].

4.3.2. Explained Variance and Predictive Relevance

Prior to evaluating path coefficients, the model’s explanatory power (R²) and predictive relevance (Q²) were examined. As shown in

Table 3. Results of the coefficients of determination and predictive relevance.

	R2	LR2	Q2
OBA	0.433	0.433	0.265
EA	0.527	0.527	0.291
CINV	0.548	0.484	0.329
LOY	0.419	0.380	0.274

Note: OBA: object-based authenticity; EA: existential authenticity; CINV: cultural involvement; LOY: loyalty; R²: coefficient of determination; LR^2: R-square of Linear Regression; Q²: Stone–Geisser’s predictive relevance.

, the model explains a substantial proportion of the variance in CINV (R² = 0.548) and EA (R² = 0.527), while showing moderate explanatory power for OBA (R² = 0.433) and LOY (R² = 0.419). In addition, the comparison between the full model and the linear-only specification (LR²) indicates that incorporating the quadratic terms yields incremental explanatory gains for CINV (ΔR² = 0.064; from 0.484 to 0.548) and LOY (ΔR² = 0.039; from 0.380 to 0.419). Furthermore, all Q² values range from 0.265 to 0.329, indicating satisfactory predictive relevance for all endogenous constructs [54].

4.3.3. Path Coefficients and Hypothesis Testing

The bootstrapping procedure (two-tailed, α = 0.05, 5000 resamples) was employed to test the hypothesized linear and quadratic relationships. As shown in

Table 4. Path coefficients results.

H	Path	β	STDEV	BCa LL	BCa UL	T-Value	p	Support
H1	CM −> OBA	0.658	0.051	0.544	0.744	12.929	0.000	Yes
H2	CM −> EA	0.251	0.058	0.141	0.367	4.322	0.000	Yes
H3	OBA −> EA	0.536	0.070	0.388	0.658	7.638	0.000	Yes
H4-L	CM −> CINV	0.304	0.063	0.182	0.430	4.832	0.000	Yes
H4-Q	CM −> CINV (Q)	0.066	0.041	−0.031	0.131	1.608	0.108	No
H5-L	OBA −> CINV	0.134	0.068	0.018	0.288	1.975	0.048	Yes
H5-Q	OBA −> CINV (Q)	0.062	0.048	−0.056	0.139	1.294	0.196	No
H6-L	EA −> CINV	0.501	0.054	0.392	0.604	9.227	0.000	Yes
H6-Q	EA −> CINV (Q)	0.031	0.042	−0.054	0.109	0.737	0.461	No
H7-L	CM −> LOY	0.107	0.087	−0.062	0.276	1.221	0.222	No
H7-Q	CM −> LOY (Q)	−0.009	0.045	−0.096	0.076	0.209	0.834	No
H8-L	OBA −> LOY	0.057	0.064	−0.068	0.181	0.887	0.375	No
H8-Q	OBA −> LOY (Q)	0.057	0.028	0.000	0.111	2.051	0.040	Yes
H9-L	EA −> LOY	0.328	0.075	0.168	0.466	4.383	0.000	Yes
H9-Q	EA −> LOY (Q)	0.087	0.030	0.024	0.146	2.870	0.004	Yes
H10-L	CINV −> LOY	0.236	0.112	0.015	0.448	2.114	0.035	Yes
H10-Q	CINV −> LOY (Q)	−0.114	0.042	−0.199	−0.031	2.681	0.007	Yes

Note: -L: line; -Q: quadratic; CM: cultural motivation; OBA: object-based authenticity; EA: existential authenticity; CINV: cultural involvement; LOY: loyalty; β: Path Coefficient; STDEV: Standard Deviation; BCa: Bias-Corrected 95% Confidence Interval.

, the structural results support 11 of the 17 hypothesized paths.

As shown in Table 4, CM exerts significant positive effects on OBA (β = 0.658, p < 0.001), EA (β = 0.251, p < 0.001), and CINV (β = 0.304, p < 0.001), whereas its direct effect on LOY is not significant (β = 0.107, p = 0.222). OBA significantly predicts EA (β = 0.536, p < 0.001) and shows a weak direct effect on CINV (β = 0.134, p = 0.048), but it does not directly predict LOY (β = 0.057, p = 0.375). In contrast, EA and CINV both demonstrate significant direct effects on LOY, with EA showing a comparatively stronger effect (β = 0.328, p < 0.001) than CINV (β = 0.236, p = 0.035). Collectively, these findings indicate that LOY is directly driven by experiential and involvement-related constructs rather than by CM or OBA.

4.4. Non-Linear Effects

Following the identification of significant quadratic terms in the structural model, the functional forms of the non-linear relationships were further examined. As reported in Table 4, three quadratic effects on LOY are statistically significant, whereas none of the quadratic terms predicting CINV reach significance. Therefore, the non-linear analysis focuses on the relationships between EA, OBA, CINV, and LOY.

Based on the standardized coefficients, the linear term (β₁ = 0.328, p < 0.001) and the quadratic term (β₂ = 0.087, p = 0.004) of the quadratic relationship between EA and LOY are positive. This functional form indicates a convex relationship, suggesting that the marginal effect of EA on LOY increases at higher levels of EA. Therefore, the quadratic relationship between EA and LOY can be expressed as

$$\mathrm{L}\mathrm{O}\mathrm{Y}=0.328\mathrm{E}\mathrm{A}+0.087\mathrm{E}\mathrm{A}2$$

The estimated quadratic curve corresponding to this relationship illustrates in Figure 3. Based on the vertex of a quadratic function, the turning point occurs at approximately 2.11.

Similarly, the quadratic term of the relationship between OBA and LOY is positive and significant (β₂ = 0.057, p = 0.040). This pattern reflects that the influence of OBA on LOY becomes more pronounced at higher levels of perceived object-based authenticity. The quadratic relationship between OBA and LOY can be represented as

$$\mathrm{L}\mathrm{O}\mathrm{Y}=0.057\mathrm{O}\mathrm{B}\mathrm{A}+0.057\mathrm{O}\mathrm{B}\mathrm{A}2$$

The estimated quadratic curve corresponding to this relationship shows in Figure 4. Based on the vertex of a quadratic function, the turning point occurs at approximately 3.50.

In contrast, the quadratic relationship between CINV and LOY exhibits a U-shaped functional form, where the linear term is positive (β₁ = 0.236, p = 0.035) and the quadratic term is negative (β₂ = −0.114, p = 0.007). This inverted U-shaped form indicates diminishing marginal returns, suggesting that increases in CINV contribute less to LOY beyond a certain level. The quadratic relationship between CINV and LOY can be represented as

$$\mathrm{L}\mathrm{O}\mathrm{Y}=0.236\mathrm{C}\mathrm{I}\mathrm{N}\mathrm{V}-0.114\mathrm{C}\mathrm{I}\mathrm{N}\mathrm{V}2$$

The estimated quadratic curve corresponding to this relationship shows in Figure 5. Based on the vertex of a quadratic function, the turning point occurs at approximately 5.04.

4.5. Total Effects

Finally, the total effects reported in

Table 5. Total effects results.

Path	β	STDEV	BCa LL	BCa UL	T-Value	p
CINV −> LOY	0.236	0.112	0.015	0.448	2.114	0.035
CM −> CINV	0.694	0.054	0.586	0.796	12.801	0.000
CM −> EA	0.603	0.049	0.491	0.689	12.282	0.000
CM −> LOY	0.505	0.053	0.398	0.603	9.591	0.000
CM −> OBA	0.658	0.051	0.544	0.744	12.929	0.000
EA −> CINV	0.501	0.054	0.392	0.604	9.227	0.000
EA −> LOY	0.446	0.068	0.306	0.574	6.553	0.000
OBA −> CINV	0.402	0.076	0.263	0.559	5.315	0.000
OBA −> EA	0.536	0.070	0.388	0.658	7.638	0.000
OBA −> LOY	0.327	0.073	0.188	0.477	4.468	0.000
CM −> CINV (Q)	0.066	0.041	−0.031	0.131	1.608	0.108
CM −> LOY (Q)	0.006	0.048	−0.094	0.091	0.128	0.898
EA −> CINV (Q)	0.031	0.042	−0.054	0.109	0.737	0.461
EA −> LOY (Q)	0.095	0.032	0.023	0.151	2.919	0.004
CINV −> LOY (Q)	−0.114	0.042	−0.199	−0.031	2.681	0.007
OBA −> CINV (Q)	0.062	0.048	−0.056	0.139	1.294	0.196
OBA −> LOY (Q)	0.072	0.033	0.000	0.136	2.170	0.030

Note: CM: cultural motivation; OBA: object-based authenticity; EA: existential authenticity; CINV: cultural involvement; LOY: loyalty; β: Path Coefficient; STDEV: Standard Deviation; BCa: Bias-Corrected 95% Confidence Interval; Q: quadratic.

further show that CM (β = 0.505, p < 0.001) and OBA (β = 0.327, p < 0.001) have significant overall influences on LOY, despite their non-significant direct effects. These findings indicate significant mediation effects, whereby the influences of CM and OBA on LOY are transmitted through OBA, EA, and CINV rather than through direct pathways.

5. Discussion and Conclusions

5.1. Discussion of the Findings

This study uncovers a staged mechanism through which CM and authenticity shape tourist loyalty in urban heritage areas. The findings reveal that CM primarily functions as an upstream driver, channeling tourists toward stronger perceptions of both OBA and EA, which then enhance CINV. Importantly, the study shows that tourist loyalty is primarily driven by experiential and involvement-related processes rather than direct effects from CM or OBA. The relationship between these variables is mediated by EA and CINV, highlighting the importance of deeper psychological engagement in loyalty formation.

Furthermore, the study demonstrates that non-linearity plays a significant role in the loyalty-formation stage, with quadratic effects observed on loyalty but not on CINV. This suggests that the influence of authenticity and CINV is not uniform across all levels. Specifically, the incremental impact of authenticity on loyalty becomes more pronounced as authenticity perceptions intensify, indicating a threshold effect in loyalty formation. This finding contributes to the refinement of the C-BA model, showing that loyalty is shaped by both mediated effects and curvilinear relationships between the key constructs. These results further highlight that tourist loyalty is contingent on the intensity of authenticity perceptions and cultural involvement, with effects varying as these factors shift from lower to higher levels.

The linear structural results suggest that tourist loyalty is not directly driven by CM or OBA but is formed through the mediating effects of EA and CINV. This finding aligns with C-BA model in heritage tourism, which positions authenticity as a key mechanism translating culturally oriented motives into tourist loyalty rather than as a simple direct path [8]. In line with this view, empirical evidence shows that CM is closely tied to perceived authenticity and subsequent CINV, which then drives behavioral consequences [14,15]. Likewise, many scholars found tourists’ perception of EA and their CINV activate their loyalty [13,30,58]. Moreover, the significant mediating effect of EA and CINV further suggests that authenticity-related experiences are central to building tourist loyalty. This mechanism typically operates through intervening psychological processes rather than purely direct effects [59].

The non-linear relationship results reveal that non-linearity is concentrated specifically in the loyalty-formation stage. The quadratic terms are significant for LOY, whereas none of the quadratic terms predicting CINV reach significance. The divergent functional forms reveal distinct underlying mechanisms. The linear influence of antecedents on CINV suggests a resource accumulation mechanism, where CINV deepens proportionally with the investment of cognitive resources [29,30]. In contrast, the non-linearity concentrated in the loyalty-formation stage implies that authenticity must surpass a credibility threshold to effectively trigger behavioral commitment [60].

More specifically, regarding EA and OBA, the convex functional forms suggest a threshold effect in loyalty formation. Once authenticity perceptions reach a certain intensity, their incremental impact on loyalty becomes more pronounced [58,61]. At lower levels of object-based or existential authenticity, the impact on loyalty is negligible, likely because the cues are insufficient to trigger a profound psychological resonance or “suspend disbelief.” This finding aligns with Lehman et al. [60], who adopt the view that authenticity is “not an all or nothing distinction but a matter of degree” (Cooper et al., 2005 [62], as cited in Lehman et al., 2019 [60]).

Furthermore, the study suggests that once the authenticity experience surpasses a specific threshold, it acts as a catalyst that significantly amplifies tourist loyalty. This finding indicates that authenticity operates as a high-impact attribute, requiring superior delivery to activate the latent potential for deep behavioral commitment. This finding is consistent with previous research indicating that tourists’ perception of authenticity shapes their behavioral intentions through some psychological factors and can increase as authenticity cues become stronger [63].

These results differ in part from Yeh et al. [17] and can be interpreted through two boundary conditions. First, the present study operationalizes authenticity as two dimensions, OBA and EA, which may alter how authenticity translates into CINV compared with studies using a single overall authenticity construct. Second, the destination context plays a pivotal role in shaping the functional form of authenticity. The divergence from Yeh et al.’s [17] inverted U-shaped finding in Macau can be attributed to the distinction between a “tourism enclave” and a “living heritage district.”

Macau is often perceived as a mixed-use destination where heritage consumption competes with modern entertainment, especially gaming industry; thus, tangible heritage might introduce inconvenience for leisure-oriented tourists. In contrast, the Pingjiang Historic Block functions as a “living museum.” Object-based authenticity is seamlessly integrated with functional amenities (e.g., teahouses in historic dwellings), minimizing the trade-offs between authenticity and comfort. Consequently, in such a heritage context, higher authenticity perception does not trigger diminishing returns but rather fosters a “threshold effect,” where loyalty grows exponentially once a high level of credibility is established.

In contrast, the inverted U-shaped relationship between CINV and tourist loyalty indicates diminishing returns beyond the turning point, which is theoretically consistent with the “Too-Much-of-a-Good-Thing” theory [40]. This discovery indicates that when tourists’ CINV reaches a certain level, saturation or fatigue may cause tourist loyalty gains brought by additional cultural activities to decline.

The result might be caused by the psychological costs resulting from high CINV. As cultural involvement rises, tourists typically invest more cognitive and emotional resources in interpretation, information seeking, and meaning making [29]. However, once involvement becomes excessive, the same processes may impose a higher attentional burden and reduce the spontaneity and enjoyment of the visit [41].

Furthermore, in heritage tourism contexts, additional cultural activities often demand sustained attention and continuous appraisal of heritage, which can gradually transform the experience from intrinsically enjoyable to effortful. Consequently, the incremental gains in tourist loyalty generated by further increases in cultural involvement diminish beyond a certain point [40]. Moreover, very high cultural involvement may increase tourists’ scrutiny of heritage cues, making them more likely to notice inconsistencies, commercialization, or signs of staged authenticity that conflict with their prior beliefs or cultural expectations [64]. When discrepancies are detected, tourists may perceive the actual experience as falling short of what they expected, thus weakening their loyalty [65]. This mechanism is particularly relevant in heritage settings. Some research shows that tourists’ authenticity perceptions are closely tied to their loyalty [66].

5.2. Theoretical Contributions

This study advances heritage tourism theory by refining the consumer-based model of authenticity. Rather than assuming direct motivational effects, the findings demonstrate that tourists’ cultural motivation and perception of object-based authenticity primarily influence loyalty via existential authenticity and cultural involvement. This finding highlights the importance of experiential and involvement-related processes as drivers of loyalty formation in urban heritage settings.

The study further contributes to the curvilinear literature by demonstrating that non-linearity is not uniformly distributed across the causal chain. The significant quadratic effects are concentrated at the loyalty-formation stage, while the relationships between antecedents and cultural involvement remain largely linear. This implies that the “Too-Much-of-a-Good-Thing” effect is a state-dependent phenomenon that primarily emerges during the on-site consumption stage, rather than in the preparatory motivational stage. This indicates that the non-linear relationship may emerge selectively when tourists’ experiential states are translated into their loyalty.

Finally, this study contributes to authenticity research by highlighting its contextual dependency. Building on prior work showing that authenticity may exhibit relatively weak explanatory power in destinations where heritage is not the primary buying point. The present findings suggest that both the magnitude and the functional forms of the effects of authenticity depend on the context of consumption. In urban heritage districts such as Pingjiang Historic Block, where conservation quality is high and everyday functions are sustained through preservation and adaptive reuse, heritage is less likely to be accompanied by inconvenience for visitors, therefore translated more effectively into loyalty. This interpretation is consistent with Yin and Dai [12] and offers a contextual explanation for why authenticity does not always translate into loyalty.

5.3. Managerial Implications

The findings imply that management should shift emphasis from heritage conservation alone to experience design that converts authenticity into loyalty. In practice, these findings suggest that management should design more participatory cultural activities to help tourists feel the sense of belonging, rather than merely observing heritage as objects. In the Pingjiang Historic Block, some high-value heritage offerings such as Kunqu performances tend to appeal to a relatively niche audience and require substantial time commitment, which many visitors on short urban trips cannot easily accommodate. By contrast, lighter and time-efficient formats appear to better match younger visitors’ preferences, as suggested by the consistently long queues at souvenir stores in Dazhangjia Alley. From a managerial standpoint, these patterns suggest that cultural involvement should be facilitated through tiered experience offerings that match visitors’ time budgets, combining deep, performance-based cultural activities with convenient, modular touchpoints that can be integrated into fast-paced itineraries.

The findings further suggest that managers should avoid a more is always better assumption. Especially in terms of cultural involvement, the inverted U-shaped pattern implies diminishing returns beyond a threshold. So, programming should prevent cognitive saturation and fatigue by pacing activities, reducing information overload, and building in recovery time.

Moreover, the contrast between destination contexts suggests that authenticity strategies should be calibrated to the broader consumption environment rather than applied as a universal template. In heritage-oriented areas with well-maintained functions, heritage itself is less likely to cause inconvenience to tourists, making tourists’ high-authenticity perceptions more feasible to translate into loyalty. Maintaining usability therefore becomes a strategic complement, including managing crowding, ensuring walkable circulation, preserving basic services, and sustaining living functions of heritage through adaptive reuse.

Finally, the non-linear relationships between authenticity and tourist loyalty imply a minimum credible level of perceived authenticity is needed before loyalty begins to increase meaningfully. The results suggest that when tourists encounter multiple cues that they interpret as inauthentic or staged, their loyalty remains suppressed. This suggests that management should prioritize eliminating salient “inauthenticity signals” and ensuring a baseline threshold of credibility first. The convex relationship also offers a strategic warning against “mediocrity”. Given that authenticity yields increasing marginal returns on loyalty only at high levels, for management, curating a few flagship and high-fidelity heritage activities that ensure high-intensity authenticity experience is pivotal to unlocking the exponential growth of tourist loyalty.

5.4. Conclusions

This study examined how perceived authenticity shapes tourist loyalty in an urban heritage district by integrating linear and non-linear relationships within the C-BA model. The findings indicate a significant mediating effect of existential authenticity and cultural involvement. Importantly, non-linear relationships are concentrated in the loyalty formation stage. The convex patterns for object-based authenticity and existential authenticity imply that stronger authenticity perceptions become increasingly consequential for loyalty once credibility and engagement are established, while the inverted U-shaped pattern for cultural involvement points to diminishing returns beyond an optimal level.

These results refine authenticity theorizing by highlighting contextual dependency, explaining why functional preservation and usability can condition how tourists’ perception of authenticity translates into loyalty. From a practical perspective, the study suggests that management should prioritize a more diversified experience portfolio and provide tiered products that align with visitors’ time constraints. Moreover, destinations should ensure a minimum credibility threshold by reducing salient inauthenticity elements, while also pacing cultural activities to prevent involvement fatigue and declining marginal gains at higher levels of participation.

While this study provides valuable insights into the relationship between authenticity, cultural involvement, and tourist loyalty within the context of the Pingjiang Historic Block, the findings should be interpreted with cultural and contextual boundaries in mind. Specifically, the study’s focus on a Chinese urban heritage site may limit the generalizability of the results to non-Asian contexts, where perceptions of authenticity and cultural involvement might differ significantly. Cultural factors, such as local heritage values and tourist behaviors, are likely to influence how tourists perceive and engage with heritage sites, especially in regions outside of Asia.

Moreover, while the study offers a detailed understanding of loyalty formation for domestic tourists, international tourists may experience the relationship between authenticity and loyalty differently. Factors such as cultural background, familiarity with the heritage, and expectations may shape their engagement and loyalty in ways not fully captured in this research. Therefore, further studies should explore these relationships in diverse geographic contexts to test the cross-cultural applicability of the findings.

The findings of this study provide a foundation for tourism managers to design authentic and culturally engaging experiences that foster loyalty. However, these strategies should be tailored to specific cultural contexts to ensure they resonate with a broader international audience. Future research should include international tourist populations and test the applicability of these findings across different cultural and regional settings.