Nguyen Xuan Quynh
| Hoang Ha* | Quynh Mai Pham | Hoa Thi Thanh Pham | Nguyen Son Tung
© 2026 The authors. This article is published by IIETA and is licensed under the CC BY 4.0 license (http://creativecommons.org/licenses/by/4.0/).
OPEN ACCESS
This study examines how cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience shape tourists' revisit intention through perceived value and whether co-creation participation (CCP) conditions these relationships in northern mountainous Vietnam. Survey data from 350 tourists were analyzed using covariance-based structural equation modeling (CB-SEM) and multi-group SEM. The six-factor measurement model showed good fit (CFI = 0.975, TLI = 0.970, RMSEA = 0.045, SRMR = 0.033), and metric invariance was supported across non-participants (n = 158) and participants (n = 192). Structural relations differed significantly across groups (Wald χ²(9) = 61.745, p < 0.001). CCP significantly conditioned the community interaction-to-perceived value path (β = 0.519 vs. 0.079) and the cultural authenticity-to-revisit intention path (β = −0.123 vs. 0.253). Bootstrap analysis with 5,000 resamples also showed significant between-group differences in the indirect effects of cultural authenticity and community interaction through perceived value. The findings indicate that CCP is a selective boundary condition rather than a uniformly positive driver. The study extends differentiated Stimulus-Organism-Response (S-O-R) explanations of cultural tourism behavior and provides segment-specific implications for community-based destination management.
co-creation participation, ethnic cultural experience, moderated mediation, multi-group structural equation modeling, perceived value, revisit intention, Vietnam
Ethnic cultural tourism has become an important development pathway for many remote mountainous destinations because cultural resources, community life, and indigenous landscapes can be effectively transformed into distinctive tourism experiences. This is particularly relevant in northern mountainous Vietnam, where destinations such as Sa Pa, Bac Ha, Ha Giang, Mu Cang Chai, Mai Chau, Moc Chau, Dien Bien, and Cao Bang are characterized by diverse ethnic communities, traditional architecture, costumes, cuisine, handicrafts, festivals, local markets, and cultural landscapes. Tourism in these areas extends well beyond sightseeing: visitors may interact with residents, observe and learn local customs, and participate directly in variouscommunity-based cultural activities. These characteristics make northern mountainous Vietnam an appropriate setting for examining ethnic cultural experience as a multidimensional phenomenon involving authenticity, interaction, learning, and aesthetic appreciation.
This experiential perspective is consistent with the experience economy theory, which emphasizes that tourism value derives not only from functional service delivery but also from memorable, meaningful, educational, and aesthetically engaging experiences [1]. It also aligns with the Stimulus-Organism-Response (S-O-R) framework, whereby experiential stimuli influence internal evaluations and subsequently shape behavioral responses [2]. In the present study, cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience represent distinct experiential stimuli; perceived value reflects tourists’ internal evaluation of these experiences; and revisit intention represents the behavioral response.
Previous studies have established important relationships among authenticity, local culture, perceived value, memorable tourism experiences, destination image, satisfaction, loyalty, and revisit intention in heritage, rural, festival, and cultural tourism contexts [3-9]. Nevertheless, two related issues remain empirically insufficiently clarified. First, ethnic cultural experience is often approached through selected experiential attributes or aggregate experience measures, leaving less understanding of whether distinct dimensions operate through different pathways to future behavioral intention. Authenticity, community interaction, cultural learning, and aesthetic experience capture cultural credibility, social encounter, cognitive enrichment, and sensory-symbolic appreciation, respectively; their effects therefore need not be homogeneous.
Second, the conversion of cultural experience into revisit intention may depend on how tourists engage with the destination. Tourism co-creation research increasingly recognizes tourists as active participants in experience formation rather than passive recipients of tourism services [10-13]. A tourist who observes cultural life and one who directly participates in cooking, brocade weaving, farming, traditional crafts, folk performances, or community activities may use different experiential cues when evaluating the value of a destination. Co-creation participation may therefore constitute a boundary condition that changes the strength or configuration of experience-value-behavior relationships rather than simply exerting an additional positive effect.
Perceived value provides a theoretically relevant mechanism for explaining these relationships because it reflects tourists’ overall evaluation of benefits relative to financial cost, time, effort, and psychological return [14]. Tourism research consistently identifies perceived value as an important antecedent of subsequent behavioral intentions [6, 15-17]. Accordingly, authentic, interactive, educational, or aesthetically engaging experiences may encourage revisiting partly because tourists perceive them as worthwhile and meaningful. At the same time, certain culturally salient experiences may also influence revisit intention more directly through social, emotional, or symbolic mechanisms. Examining both direct and value-mediated pathways therefore provides a more differentiated account of how ethnic cultural experiences generate future behavioral intention.
Against this background, the present study examines how four dimensions of ethnic cultural experience-cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience-shape revisit intention directly and indirectly through perceived value. It further investigates whether these structural and indirect relationships differ according to co-creation participation (CCP). CCP is operationalized as a binary behavioral condition distinguishing tourists who did not participate in cultural activities with local communities from those who did. Its conditioning role is examined through multi-group structural equation modeling (SEM) rather than assuming that participation uniformly strengthens all relationships.
This study makes three contributions. First, it advances a differentiated S-O-R perspective by examining four theoretically distinct cultural-experience stimuli and allowing their pathways to perceived value and revisit intention to vary. Second, it refines the role of perceived value by examining it as a selective value-conversion mechanism rather than assuming a homogeneous mediation process across all cultural experiences. Third, it introduces co-creation participation as a boundary condition of the experience-value-revisit mechanism, thereby examining whether tourists’ actual participation changes how specific cultural experiences are translated into perceived value and behavioral intention. Practically, this perspective can help destination managers, tourism enterprises, homestay operators, and local communities design cultural experiences that are better aligned with different modes of tourist engagement while preserving cultural authenticity and strengthening destination value.
2.1 Theoretical framing: Differentiated Stimulus-Organism-Response pathways
The S-O-R framework proposes that environmental stimuli influence individuals’ internal states, which subsequently shape behavioral responses [2]. In tourism, destination experiences can therefore be conceptualized as stimuli that affect tourists’ cognitive and affective evaluations before influencing subsequent behavioral intentions. The experience economy perspective complements this logic by emphasizing that consumers derive value from memorable, educational, aesthetic, and engaging experiences rather than from functional service delivery alone [1].
In ethnic cultural tourism, however, experiential stimuli are unlikely to be interchangeable. Cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience represent different forms of engagement with the destination, involving cultural credibility, social encounter, cognitive enrichment, and sensory-symbolic appreciation, respectively. Recent applications of S-O-R in cultural heritage tourism similarly suggest that different experiential cues may activate distinct psychological mechanisms [18].
Accordingly, this study adopts a differentiated S-O-R pathway perspective. Cultural authenticity (CA), community interaction (CI), cultural learning (CL), and cultural aesthetic experience (CE) are specified as distinct experiential stimuli; perceived value (PV) represents the principal organismic evaluation; and revisit intention (RI) represents the behavioral response. The S-O-R logic primarily motivates the mediated paths through PV, while direct paths from the experiential dimensions to RI are retained to examine whether some cultural experiences also exert behavioral effects that are not fully transmitted through perceived value.
2.2 Ethnic cultural experience dimensions and perceived value
CA refers to tourists’ perceptions that local customs, ethnic identity, and cultural practices are presented genuinely and remain connected with the everyday life of the host community. Authenticity is particularly salient in ethnic and heritage tourism because excessive commercialization or staged presentation can weaken cultural credibility and destination distinctiveness. Prior studies associate perceived authenticity with perceived value, memorable tourism experience, satisfaction, and revisit intention [3, 8, 19]. When tourists perceive cultural practices as rooted in local life rather than primarily produced for tourist consumption, they are more likely to regard the experience as meaningful and worthwhile.
H1a. Cultural authenticity positively influences perceived value.
CI captures tourists’ evaluations of friendliness, openness, communication, assistance, and opportunities for direct contact with local residents. Importantly, CI is conceptually distinct from CCP. CI concerns the quality and opportunity of social interaction, whereas CCP concerns whether tourists actually participate in cultural activities with the community. Positive host-guest interaction can reduce social distance, facilitate cultural interpretation, and create relational value. Research on tourism co-creation likewise identifies interaction as an important process through which tourists develop experiential meaning and destination value [10-12].
H1b. Community interaction positively influences perceived value.
CL reflects the extent to which tourism experiences improve visitors’ understanding of local customs, lifestyles, and cultural values. Cultural tourists may seek not only to observe visible cultural elements but also to understand their meanings and social significance. Interpretation and learning can consequently increase the informational, psychological, and experiential worth of a trip [4, 20, 21].
H1c. Cultural learning positively influences perceived value.
CE refers to tourists’ sensory, visual, emotional, and symbolic responses to traditional architecture, costumes, handicrafts, village environments, cultural symbols, and the overall cultural atmosphere. Such aesthetic cues can make a destination distinctive and memorable and can generate emotional responses that extend beyond functional evaluations of tourism services. This reasoning is consistent with the experience economy perspective and empirical research in cultural heritage settings [5, 19].
H1d. Cultural aesthetic experience positively influences perceived value.
2.3 Perceived value, direct effects, and mediation
PV represents tourists’ overall evaluation of the benefits received relative to the financial cost, time, effort, and psychological return associated with an experience [14]. Tourism research consistently identifies PV as an important antecedent of satisfaction, loyalty, and revisit intention [6, 16, 17]. In ethnic cultural tourism, this evaluative process is particularly relevant because similar cultural activities may be regarded as meaningful and worthwhile by some visitors but superficial or overly commercialized by others.
H2. Perceived value positively influences revisit intention.
The four experiential dimensions may also affect RI independently of PV. Authenticity may strengthen destination distinctiveness and credibility; community interaction may create social connection; cultural learning may generate curiosity and motivation for further exploration; and aesthetic experience may produce memorable emotional and visual impressions. These direct paths are therefore specified alongside the value-mediated paths, allowing the model to distinguish experiences that operate predominantly through tourists’ value evaluations from those that may also have more immediate social, cognitive, or affective associations with revisit intention.
H3a–H3d. Cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience each positively influence revisit intention.
PV is further expected to mediate the relationships between the four experiential dimensions and RI. From an S-O-R perspective, tourists first encounter destination-related experiential stimuli, evaluate their worth and meaning, and subsequently form behavioral intentions. Nevertheless, the strength of this conversion mechanism may differ across experiential dimensions: cognitively oriented experiences may depend more strongly on value evaluation, whereas socially or aesthetically salient experiences may retain additional direct associations with RI.
H4a–H4d. Perceived value mediates the effects of cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience on revisit intention.
2.4 Co-creation participation as a boundary condition
Value co-creation research conceptualizes tourists as active contributors to tourism experiences rather than passive recipients of destination offerings [12, 13]. In ethnic cultural tourism, such participation may involve cooking, brocade weaving, farming, traditional crafts, folk performances, culturally appropriate rituals, or aspects of community and family life. These activities differ conceptually from ordinary host-guest interaction because they involve tourists’ direct investment of attention, effort, time, and embodied participation in the cultural experience.
Prior research suggests that co-creation can influence how experiential value is formed and subsequently translated into behavioral outcomes [22-24]. Importantly, participation need not uniformly strengthen every relationship. Direct involvement may make some experiential cues more salient, reduce tourists’ reliance on others, or alter whether an experience influences RI primarily through PV or through a more direct behavioral pathway. Thus, CCP is conceptualized here not as an additional cultural-experience dimension or as a uniformly positive antecedent, but as a boundary condition of the experience-value-revisit mechanism.
Because the study distinguishes tourists according to whether they actually participated in co-created cultural activities, the conditioning role of CCP is examined through multi-group SEM. A directional moderation hypothesis is not imposed because existing theory does not provide sufficiently consistent grounds for predicting that participation will strengthen or weaken every individual structural relationship. Instead, the study examines whether the structural configuration and value-mediated mechanisms differ between participating and non-participating tourists.
RQ1. Do the structural relationships in the proposed experience-value-revisit model differ between tourists who do not participate and those who participate in co-created cultural activities?
RQ2. Do the indirect effects of cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience on revisit intention through perceived value differ between the two CCP groups?
Figure 1 summarizes the proposed experience-value-revisit framework and the grouping role of CCP.
3.1 Research design and context
This study employed a cross-sectional quantitative design to examine the relationships among ethnic cultural experience, perceived value, and revisit intention, and to assess whether these relationships differ according to tourists’ co-creation participation. The empirical context was ethnic cultural tourism in northern mountainous Vietnam, where tourism experiences are closely associated with ethnic minority communities, community-based tourism, homestays, traditional crafts, local cuisine and markets, village life, and direct cultural encounters between visitors and host communities. Data were collected across multiple destinations rather than at a single site to capture variation in tourists’ cultural experiences within the regional context.
The target population comprised tourists aged 18 years or older who had directly experienced at least one ethnic cultural tourism activity during their most recent visit to a surveyed destination. Eligibility was determined on the basis of respondents’ exposure to ethnic cultural content at the destination; individuals who had not experienced such activities were excluded. Respondents participated voluntarily and were informed of the study purpose, their right to decline participation, and the confidentiality of the information provided.
Given the study’s focus on tourists’ post-experience evaluations, the unit of analysis was the individual tourist. The design enabled the estimation of relationships among the latent constructs and subsequent comparison of these relationships between tourists who participated and those who did not participate in co-created cultural activities.
3.2 Sampling and data collection
A multistage on-site sampling strategy was used to obtain respondents across six northern mountainous provinces of Vietnam, with the main survey sites concentrated in the Northwest region, including Lao Cai, Lai Chau, Dien Bien, and Son La. Data collection was conducted at cultural attractions, community-based tourism villages, homestays, local markets, and spaces where ethnic cultural activities were organized.
Data collection was conducted between May and July 2026. During the fieldwork periods, visitor groups were approached sequentially at tourist-frequented locations, and one or two eligible tourists from each approached group were invited to participate. Eligibility was confirmed before questionnaire administration according to the criteria described in Section 3.1. This procedure reduced excessive concentration of responses within the same visitor group while broadening the diversity of individual respondents.
More than 500 questionnaires were distributed and more than 400 were returned. Completed questionnaires were screened before analysis. Responses were excluded when they were incomplete, failed to satisfy the eligibility criteria, or showed clear indications of invalid or inattentive responding. After screening, 350 valid questionnaires were retained for analysis.
For the multi-group analysis, respondents were classified according to their actual participation in cultural activities with local communities. Of the 350 respondents, 158 (45.14%) were classified as non-participants (CCP = 0), whereas 192 (54.86%) were classified as participants (CCP = 1). The resulting group sizes were sufficiently comparable to permit estimation and comparison of the proposed structural relationships across the two participation conditions. Detailed operationalization of CCP is presented in Section 3.3.
3.3 Measures
The measurement instrument comprised six reflective latent constructs and one binary categorical moderator, CCP. CA, CI, CL, and CE were each measured using four items, whereas PV and RI were each measured using three items. All latent-construct items were assessed on a five-point Likert-type scale ranging from 1 = strongly disagree to 5 = strongly agree.
The measurement items were adapted from established concepts and previously used measures in tourism research and contextualized to ethnic cultural tourism in northern mountainous Vietnam. The adaptation retained the conceptual meaning of each construct while aligning item wording with tourists’ experiences of ethnic culture, community interaction, cultural learning, and destination aesthetics. CA captures tourists’ perceptions of the genuineness and community embeddedness of ethnic cultural practices; CI reflects the perceived quality and opportunity for interaction with local residents; CL captures tourists’ perceived cultural learning; CE reflects the sensory, visual, emotional, and symbolic qualities of the cultural environment; PV represents tourists’ evaluation of the worth of the experience relative to financial cost, time, and psychological benefits; and RI captures their intention and likelihood of revisiting the destination.
CCP was specified as a binary categorical moderator and operationalized as an observed behavioral indicator. Respondents were coded 0 when they did not participate in cultural activities with local communities and 1 when they did. Participation included direct involvement in activities such as cooking, brocade weaving, farming, traditional crafts, folk performances, or other culturally embedded activities undertaken with community members. CCP therefore captures actual participation status rather than the perceived quality, intensity, or value of co-creation. Its moderating role was subsequently examined through multi-group SEM by comparing structural relationships and conditional indirect effects between the non-participation and participation groups.
CCP is conceptually distinct from CI. CI measures tourists’ perceptions of the quality and opportunity for host-guest interaction, whereas CCP identifies whether tourists actually participated in co-created cultural activities. This distinction allows the study to examine whether participation changes the structural mechanisms through which perceived cultural experiences are translated into perceived value and revisit intention.
The measurement items and their conceptual bases are summarized in Table 1.
Table 1 summarizes the item wording, construct codes, and conceptual bases used to operationalize the six reflective constructs and CCP.
Table 1. Measurement items and operationalization
|
Construct/Variable |
Code |
Measurement Item |
Measurement/ Conceptual Basis |
|
Cultural Authenticity (CA) |
CA1 |
I feel that the cultural activities at the destination are authentically presented. |
[3, 4, 8, 19] |
|
CA2 |
I feel that local customs are naturally maintained throughout the tourism experience. |
[3, 4, 8, 19] |
|
|
CA3 |
I notice that the local ethnic identity is clearly reflected in tourism activities. |
[3, 4, 8, 19] |
|
|
CA4 |
I perceive that the cultural experience at the destination is closely linked to the life of the local community. |
[3, 4, 8, 19] |
|
|
Community Interaction (CI) |
CI1 |
I feel that the local residents communicate with me in a friendly manner. |
[10, 11] |
|
CI2 |
I feel that the local people share cultural information openly. |
[10, 11] |
|
|
CI3 |
I had the opportunity to interact directly with the local community during the trip. |
[10, 11] |
|
|
CI4 |
I perceive that the assistance from local people during cultural activities is effective. |
[10, 11] |
|
|
Cultural Learning (CL) |
CL1 |
I gain a better understanding of the local ethnic customs after the trip. |
[4, 5, 20, 21] |
|
CL2 |
I feel that the cultural activities are explained in an easily understandable manner. |
[4, 5, 20, 21] |
|
|
CL3 |
I learned useful knowledge about the lifestyle of the local community. |
[4, 5, 20, 21] |
|
|
CL4 |
I have a deeper awareness of the local cultural values after experiencing the destination. |
[4, 5, 20, 21] |
|
|
Cultural Aesthetic Experience (CE) |
CE1 |
I feel that traditional costumes create a striking visual impression. |
[1, 5, 19, 21] |
|
CE2 |
I perceive that traditional architecture makes the cultural space more attractive. |
[1, 5, 19, 21] |
|
|
CE3 |
I notice that local cultural symbols are vividly represented at the destination. |
[1, 5, 19, 21] |
|
|
CE4 |
I feel that the cultural atmosphere at the destination leaves a memorable visual impression. |
[1, 5, 19, 21] |
|
|
Perceived Value (PV) |
PV1 |
I feel that the cultural experience at the destination is worth the financial cost. |
[6, 14, 17] |
|
PV2 |
I feel that the cultural experience at the destination is worth the time I spent on the trip. |
[6, 14, 17] |
|
|
PV3 |
I perceive that the cultural experience at the destination brings positive psychological value to me. |
[6, 14, 17] |
|
|
Revisit Intention (RI) |
RI1 |
I intend to return to this ethnic cultural destination in the future. |
[3, 6, 7, 9] |
|
RI2 |
I am likely to choose this destination again for my next trip. |
[3, 6, 7, 9] |
|
|
RI3 |
I will prioritize this destination when returning to the Northern mountainous region of Vietnam. |
[3, 6, 7, 9] |
|
|
Co-creation Participation |
CCP |
0 = non-participation; 1 = participation in cultural activities with the community. |
[10, 11, 13] |
3.4 Data analysis
Data analysis was conducted using covariance-based SEM (CB-SEM) in R with the lavaan package (version 0.6-20; [25]) and proceeded in four stages.
Because all latent constructs were measured using self-reports, common-method bias was addressed both procedurally and statistically. Procedurally, respondents were informed of the study purpose, participation was voluntary, they could decline participation, and the confidentiality of their responses was assured. Statistically, an unrotated single-factor diagnostic was examined, and a one-factor confirmatory factor analysis (CFA) was estimated and compared with the hypothesized six-factor measurement model [26].
First, the data were screened and descriptive statistics were examined before assessing the six-factor measurement model through CFA. Internal consistency was evaluated using Cronbach’s alpha and composite reliability (CR), whereas convergent validity was assessed using standardized factor loadings and average variance extracted (AVE). Discriminant validity was evaluated using the heterotrait-monotrait ratio (HTMT; [27]). Measurement quality was assessed in accordance with current SEM reporting recommendations [28]. Overall model fit was evaluated using χ², χ²/df, the comparative fit index (CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR), with the indices interpreted jointly rather than on the basis of a single cutoff criterion [29].
Second, multi-group CFA was conducted to establish measurement comparability between the non-participation (CCP = 0) and participation (CCP = 1) groups before testing the moderating role of CCP. Configural invariance was assessed by estimating the same factor structure across both groups, followed by metric invariance in which factor loadings were constrained to equality. Because the focal analysis concerns cross-group differences in structural relationships rather than latent means, metric invariance was treated as the relevant measurement requirement for the subsequent structural comparisons. Changes in CFI, RMSEA, and SRMR, together with the χ² difference test, were considered jointly when evaluating invariance [30, 31].
Third, the moderating role of CCP was examined using multi-group SEM. The metric-invariant measurement structure was retained while the nine structural paths (CA, CI, CL, and CE to PV and RI, together with PV → RI) were freely estimated across the two CCP groups. A global Wald test was first used to examine whether the set of structural relationships differed jointly between groups. Path-specific Wald tests were then conducted to identify the particular structural coefficients that differed significantly across participation conditions. Equality tests were applied to the corresponding structural parameters, whereas standardized coefficients (β) were reported to facilitate substantive interpretation of the group-specific effects. A statistically significant difference between the corresponding paths, rather than significance in one group but not the other, was taken as evidence of moderation.
Fourth, conditional indirect effects of CA, CI, CL, and CE on RI through PV were estimated separately for CCP = 0 and CCP = 1 using 5,000 bootstrap resamples. The difference between each pair of group-specific indirect effects was then estimated directly. Moderated mediation was considered supported when the 95% bootstrap confidence interval for the between-group difference in the indirect effect did not include zero. This procedure follows moderated-path logic whereby a categorical moderator may alter one or more constituent paths and consequently change the magnitude of an indirect effect [32].
CCP was self-selected rather than experimentally assigned. Therefore, the multi-group comparisons are interpreted as conditional associations rather than causal effects of participation. The analysis is intended to identify whether the experience-value-revisit relationships differ across observed participation conditions, while recognizing that unobserved demographic or trip-related heterogeneity may also contribute to between-group differences.
4.1 Descriptive statistics and measurement quality
Descriptive statistics and correlations are reported in Table 2.
Table 2. Descriptive statistics and correlations
|
Construct |
N |
Mean |
SD |
CA |
CI |
CL |
CE |
PV |
RI |
CCP |
|
CA |
350 |
3.289 |
0.798 |
1.000 |
0.476*** |
0.251*** |
0.528*** |
0.507*** |
0.589*** |
0.133* |
|
CI |
350 |
3.540 |
0.660 |
0.476*** |
1.000 |
0.432*** |
0.545*** |
0.555*** |
0.623*** |
0.177*** |
|
CL |
350 |
3.471 |
0.682 |
0.251*** |
0.432*** |
1.000 |
0.383*** |
0.426*** |
0.448*** |
0.142** |
|
CE |
350 |
3.361 |
0.778 |
0.528*** |
0.545*** |
0.383*** |
1.000 |
0.541*** |
0.696*** |
0.185*** |
|
PV |
350 |
3.545 |
0.565 |
0.507*** |
0.555*** |
0.426*** |
0.541*** |
1.000 |
0.695*** |
0.459*** |
|
RI |
350 |
3.447 |
0.670 |
0.589*** |
0.623*** |
0.448*** |
0.696*** |
0.695*** |
1.000 |
0.263*** |
|
CCP |
350 |
0.549 |
0.498 |
0.133* |
0.177*** |
0.142** |
0.185*** |
0.459*** |
0.263*** |
1.000 |
As shown in Table 2, mean scores for the six latent constructs ranged from 3.289 to 3.545, with adequate variation across respondents. All correlations among the focal constructs were positive and statistically significant. The strongest associations were CE-RI (r = 0.696) and PV-RI (r = 0.695), followed by CI-RI (r = 0.623). Correlations between the four experience dimensions and PV ranged from 0.426 to 0.555. None of the correlations among latent constructs exceeded 0.70, suggesting that the constructs were related but not excessively overlapping. CCP was most strongly associated with PV (r = 0.459) and more moderately with RI (r = 0.263). These descriptive relationships provide preliminary support for the proposed model but do not constitute evidence of moderation.
Common-method-bias diagnostics did not indicate that a single common factor dominated the data. In the unrotated single-factor analysis of the 22 latent-construct items, the first factor accounted for 45.27% of the total variance, below the commonly used 50% heuristic. More importantly, the one-factor CFA showed poor fit (χ² = 2,012.734, df = 209, CFI = 0.662, TLI = 0.626, RMSEA = 0.157, SRMR = 0.108), substantially worse than the hypothesized six-factor model (CFI = 0.975, TLI = 0.970, RMSEA = 0.045, SRMR = 0.033). These results suggest that common-method variance was unlikely to be the dominant source of covariance among the study constructs, although common-method bias cannot be ruled out completely.
Reliability, convergent validity, and standardized loadings are reported in Table 3.
As reported in Table 3 (Reliability, convergent validity, and standardized factor loadings), standardized factor loadings ranged from 0.657 to 0.924, with most exceeding 0.70. Cronbach’s alpha values ranged from 0.847 to 0.920 and CR values from 0.847 to 0.922, indicating strong internal consistency. AVE ranged from 0.582 to 0.797, exceeding the 0.50 criterion for all constructs and supporting convergent validity. HTMT values were below 0.85, with the highest value (0.779) observed between CE and RI. Overall, the six reflective constructs demonstrated satisfactory reliability, convergent validity, and discriminant validity.
Table 3. Reliability, convergent validity, and standardized factor loadings
|
Construct |
Item |
Loading |
Cronbach’s α |
CR |
AVE |
|
CA |
CA1 |
0.823 |
0.897 |
0.898 |
0.688 |
|
CA2 |
0.792 |
||||
|
CA3 |
0.864 |
||||
|
CA4 |
0.837 |
||||
|
CI |
CI1 |
0.663 |
0.860 |
0.863 |
0.614 |
|
CI2 |
0.834 |
||||
|
CI3 |
0.814 |
||||
|
CI4 |
0.811 |
||||
|
CL |
CL1 |
0.770 |
0.847 |
0.847 |
0.582 |
|
CL2 |
0.707 |
||||
|
CL3 |
0.744 |
||||
|
CL4 |
0.826 |
||||
|
CE |
CE1 |
0.657 |
0.856 |
0.863 |
0.615 |
|
CE2 |
0.863 |
||||
|
CE3 |
0.779 |
||||
|
CE4 |
0.822 |
||||
|
PV |
PV1 |
0.867 |
0.920 |
0.922 |
0.797 |
|
PV2 |
0.891 |
||||
|
PV3 |
0.919 |
||||
|
RI |
RI1 |
0.924 |
0.913 |
0.917 |
0.787 |
|
RI2 |
0.810 |
||||
|
RI3 |
0.922 |
The full-sample CFA fit indices are reported in Table 4.
Table 4. Full-sample confirmatory factor analysis (CFA) fit
|
Fit Index |
Value |
|
χ² |
329.161 |
|
df |
194 |
|
χ²/df |
1.697 |
|
CFI |
0.975 |
|
TLI |
0.970 |
|
RMSEA (90% CI) |
0.045 [0.036, 0.053] |
|
SRMR |
0.033 |
As shown in Table 4, the six-factor measurement model provided a satisfactory overall fit and a sound basis for subsequent group comparisons. Before comparing structural relations, the measurement model was estimated separately for each CCP group and then tested jointly. The CCP = 1 group showed good standalone fit (CFI = 0.975, TLI = 0.970, RMSEA = 0.042, SRMR = 0.043). The CCP = 0 standalone model showed weaker global fit (CFI = 0.902, TLI = 0.883, RMSEA = 0.096, SRMR = 0.059), which warrants caution. However, group-specific diagnostics did not indicate a general breakdown in measurement quality. In the CCP = 0 group, standardized loadings ranged from 0.686 to 0.978, CR ranged from 0.852 to 0.938, and AVE ranged from 0.592 to 0.835. Inspection of residual correlations indicated localized residual associations, particularly around CL2, but no absolute residual correlation exceeded 0.20. Thus, the weaker global fit appears to reflect localized model misfit rather than broadly weak indicators. The joint configural model remained acceptable and metric invariance was supported.
4.2 Measurement invariance across co-creation participation groups
Measurement-invariance results are summarized in Table 5.
Table 5. Measurement invariance across co-creation participation (CCP) groups
|
Model |
χ² |
df |
CFI |
TLI |
RMSEA |
SRMR |
ΔCFI |
ΔRMSEA |
ΔSRMR |
|
Configural |
737.642 |
388 |
0.9361 |
0.9240 |
0.0718 |
0.0482 |
- |
- |
- |
|
Metric (equal loadings) |
760.420 |
404 |
0.9349 |
0.9255 |
0.0710 |
0.0551 |
−0.0012 |
−0.0008 |
+0.0069 |
As shown in Table 5, the configural model showed acceptable fit across the two CCP groups (CFI = 0.9361, TLI = 0.9240, RMSEA = 0.0718, SRMR = 0.0482). Constraining factor loadings to equality produced only minor changes in model fit, and the χ² difference test was non-significant, Δχ²(16) = 22.778, p = 0.120. These results support metric invariance, indicating that the factor loadings were sufficiently comparable across non-participants and participants. Structural coefficients can therefore be meaningfully compared between the two CCP groups.
4.3 Structural differences across co-creation participation groups
The multi-group structural model retained equal factor loadings and allowed the nine focal structural paths to vary across CCP groups. Model fit was acceptable (χ² = 760.420, df = 404, χ²/df = 1.882, CFI = 0.935, TLI = 0.926, RMSEA = 0.071, SRMR = 0.055). A global Wald test rejected equality of the structural paths, Wald χ²(9) = 61.745, p < 0.001, indicating that the experience-value-revisit mechanism differs materially between non-participating and participating tourists.
Group-specific structural paths and moderation tests are presented in Table 6.
As reported in Table 6, the multi-group structural model showed acceptable fit (χ²/df = 1.882, CFI = 0.935, TLI = 0.926, RMSEA = 0.071, SRMR = 0.055). The global Wald test was significant, Wald χ²(9) = 61.745, p < 0.001, indicating that the overall structural pattern differed across CCP groups.
Two path-specific differences were statistically significant. The CI → PV effect was substantially stronger among non-participants (β = 0.519, p < 0.001) than participants (β = 0.079, p = 0.437), with a significant between-group difference (p = 0.0001). In contrast, CA → RI was non-significant among non-participants (β = −0.123, p = 0.104) but positive and significant among participants (β = 0.253, p < 0.001), with a significant cross-group difference (p = 0.0002).
PV positively predicted RI in both groups (β = 0.617 vs. 0.217), but the difference did not reach statistical significance (p = 0.0565). The remaining path differences were also non-significant. Thus, CCP operates as a selective moderator, primarily affecting the CI → PV and CA → RI relationships.
Table 6. Multi-group structural paths and path-specific moderation tests
|
Path |
β CCP = 0 |
p |
β CCP = 1 |
p |
Wald χ² |
p Difference |
Moderation |
|
CA → PV |
0.310 |
<0.001 |
0.178 |
0.065 |
3.270 |
0.0705 |
No |
|
CI → PV |
0.519 |
<0.001 |
0.079 |
0.437 |
16.228 |
0.0001 |
Yes |
|
CL → PV |
0.215 |
0.001 |
0.095 |
0.329 |
1.157 |
0.2821 |
No |
|
CE → PV |
0.049 |
0.594 |
0.360 |
0.001 |
2.033 |
0.1539 |
No |
|
PV → RI |
0.617 |
<0.001 |
0.217 |
0.001 |
3.637 |
0.0565 |
Not supported |
|
CA → RI |
−0.123 |
0.104 |
0.253 |
<0.001 |
14.270 |
0.0002 |
Yes |
|
CI → RI |
−0.020 |
0.846 |
0.178 |
0.012 |
2.680 |
0.1016 |
No |
|
CL → RI |
−0.030 |
0.625 |
0.044 |
0.510 |
0.666 |
0.4145 |
No |
|
CE → RI |
0.508 |
<0.001 |
0.365 |
<0.001 |
2.671 |
0.1022 |
No |
The model explained 76.8% of PV and 86.0% of RI among non-participants, compared with 35.6% and 72.3%, respectively, among participants. These R² differences are interpreted descriptively because no formal test of their cross-group differences was conducted.
4.4 Conditional indirect effects and moderated mediation
Conditional indirect effects were estimated with 5,000 bootstrap resamples. The results demonstrate that the role of PV is not homogeneous across participation groups.
Conditional indirect effects and between-group differences are reported in Table 7.
Table 7. Conditional indirect effects by co-creation participation (CCP) group and differences between indirect effects
|
Indirect Pathway |
Effect CCP = 0 |
BCa 95% CI |
Effect CCP = 1 |
BCa 95% CI |
Δ Indirect (1−0) |
BCa 95% CI |
Moderated Mediation |
|
CA → PV → RI |
0.148 |
[0.059, 0.279] |
0.034 |
[0.001, 0.093] |
−0.114 |
[−0.251, −0.010] |
Supported |
|
CI → PV → RI |
0.358 |
[0.210, 0.573] |
0.020 |
[−0.038, 0.109] |
−0.338 |
[−0.566, −0.174] |
Supported |
|
CL → PV → RI |
0.107 |
[0.030, 0.210] |
0.025 |
[−0.029, 0.119] |
−0.082 |
[−0.198, 0.028] |
Not supported |
|
CE → PV → RI |
0.032 |
[−0.117, 0.225] |
0.074 |
[0.017, 0.160] |
0.042 |
[−0.149, 0.218] |
Not supported |
As shown in Table 7, the conditional indirect effects show that CCP selectively changes the mediating role of PV. For CA, the indirect effect on RI through PV was stronger among non-participants (0.148) than participants (0.034), and the between-group difference was significant (Δindirect = −0.114, 95% CI [−0.251, −0.010]). A stronger moderated-mediation pattern emerged for CI: its indirect effect was 0.358 among non-participants but only 0.020 among participants, with a significant difference of −0.338, 95% CI [−0.566, −0.174].
In contrast, the between-group differences in the indirect effects of CL (Δ = −0.082, 95% CI [−0.198, 0.028]) and CE (Δ = 0.042, 95% CI [−0.149, 0.218]) were not significant. Therefore, moderated mediation was supported for CA and CI, but not for CL or CE. These results reinforce the conclusion that CCP changes selected value-mediated pathways rather than uniformly moderating the entire model.
4.5 Discussion
The findings provide a differentiated explanation of how ethnic cultural experiences are translated into revisit intention and show that this process is not structurally identical for tourists with different modes of engagement. The significant global Wald test indicates that the overall experience-value-revisit mechanism differs between tourists who participate and those who do not participate in co-created cultural activities. However, only selected structural paths and indirect effects differ significantly across groups. CCP therefore operates as a selective boundary condition rather than as a uniformly positive influence on the entire model.
4.5.1 Co-creation participation and value formation
The clearest moderation effect concerns CI. Among non-participants, CI was the strongest predictor of PV, whereas the same relationship was weak and non-significant among participants. The cross-group difference was statistically significant, and the CI → PV → RI indirect effect was also substantially stronger among non-participants. Community interaction therefore appears particularly important for tourists who remain primarily observers of ethnic cultural activities.
For non-participants, friendly communication, opportunities to ask questions, assistance from residents, and access to local cultural information may help transform an observational visit into a meaningful experience. Interaction can reduce social distance and facilitate interpretation of the cultural environment, thereby strengthening tourists’ evaluations of whether the trip was worthwhile. This interpretation is consistent with previous research emphasizing host-guest interaction and relational processes in tourism value co-creation [10-12].
Among participants, direct involvement in activities such as cooking, weaving, farming, traditional crafts, or folk performance may provide experiential information beyond ordinary interaction. Participation allows tourists to learn through doing and encounter cultural practices more closely. Consequently, the marginal role of ordinary interaction in explaining perceived value may decline. This does not imply that interaction becomes unimportant; rather, participation appears to change the cues tourists rely on when evaluating value.
The finding therefore extends tourism co-creation research by showing that interaction and participation are related but not interchangeable mechanisms. Participation may substitute for some value-forming functions of social interaction instead of simply amplifying them. This interpretation is compatible with recent research portraying co-creation as multidimensional and context-dependent [13, 22, 23].
4.5.2 Participation changes how authenticity influences revisit intention
The second important moderation effect concerns CA. The direct CA → RI path was not significant among non-participants but became positive and significant among participants, and the difference between the coefficients was statistically significant. At the same time, the indirect effect of CA on RI through PV was significantly stronger among non-participants. Taken together, these findings suggest that participation changes the route through which authenticity becomes behaviorally consequential.
For non-participants, authenticity appears to operate primarily through value evaluation. Tourists who observe local practices may assess whether customs, ethnic identity, and cultural activities are genuinely maintained and connected to community life and then incorporate these perceptions into an overall evaluation of whether the experience was worthwhile. Revisit intention consequently emerges partly through the CA → PV → RI pathway.
For participants, authenticity appears to have a more direct association with revisit intention. Direct engagement may make cultural practices personally salient because tourists encounter local knowledge, materials, people, and cultural meanings in a more immediate setting. Authenticity may therefore influence behavioral intention without being fully translated through an explicit value assessment. Because the underlying psychological process was not directly measured, this interpretation should remain cautious; nevertheless, the combination of significant direct-path moderation and moderated mediation is consistent with a shift from a predominantly value-mediated route toward a more direct behavioral route.
This finding adds nuance to previous studies reporting positive relationships between authenticity, perceived value, memorable experience, satisfaction, and revisit intention [3, 8, 19]. Rather than contradicting this literature, the present study indicates that authenticity may operate through different mechanisms for different modes of tourist engagement.
4.5.3 Perceived value as a selective conversion mechanism
PV positively predicted RI in both CCP groups, confirming its importance in tourism behavior and supporting previous evidence that tourists are more likely to revisit when they consider an experience worthwhile relative to its financial cost, time, effort, and psychological benefits [6, 16, 17]. Although the PV → RI coefficient was descriptively larger among non-participants, the cross-group difference did not reach statistical significance. The results therefore do not support moderation of the PV → RI relationship itself.
The more informative differences occur in the pathways leading into PV. Among non-participants, authenticity, community interaction, and cultural learning significantly contributed to PV, whereas the pattern among participants differed. The model also explained substantially more variance in PV among non-participants. Because the R² difference was not formally tested, it should be interpreted descriptively, but it suggests that the four modeled cultural-experience dimensions provide a more complete explanation of value formation for non-participants.
For participating tourists, additional value sources may become relevant, including immersion, personal involvement, agency, social bonding, novelty, or emotional engagement. These mechanisms were not measured and therefore remain issues for future investigation. Nevertheless, the findings refine the S-O-R interpretation by suggesting that PV is not a universal conduit through which every experiential stimulus must pass. Instead, it functions as a selective value-conversion mechanism whose role depends on both the type of cultural experience and tourists’ mode of engagement.
4.5.4 Cultural learning and aesthetic experience
CL significantly predicted PV among non-participants but not among participants, and its indirect effect through PV was significant only in the non-participation group. However, neither the structural-path difference nor the difference between the two indirect effects was statistically significant. Thus, the evidence does not establish moderation or moderated mediation for CL. The results nevertheless remain consistent with previous research indicating that cultural learning and interpretation can enhance the meaning and value of heritage and cultural tourism experiences [4, 20, 21].
CE showed a different pattern. CE had a strong positive direct relationship with RI in both groups, while the cross-group difference was not significant. Aesthetic experience therefore appears to be a comparatively stable driver of revisit intention regardless of participation status. Traditional architecture, costumes, cultural symbols, craft spaces, village environments, and the broader cultural atmosphere may create memorable visual and emotional impressions for both observers and participants. This result is consistent with the experience economy perspective and previous research emphasizing aesthetic experience in cultural heritage tourism [1, 19].
The indirect CE → PV → RI effect was significant only among participants, but its cross-group difference was not significant. Accordingly, this pattern should not be interpreted as moderated mediation. The more defensible conclusion is that CE maintains a robust direct association with RI across participation conditions.
Overall, the findings reveal a differentiated structure: CI → PV and CA → RI are participation-sensitive; the indirect effects involving CA and CI are conditional on participation; whereas CE → RI and PV → RI are comparatively stable across groups. This pattern demonstrates why a pooled experience-value-revisit model may conceal meaningful heterogeneity and extends S-O-R applications by showing that both the type of experiential stimulus and tourists’ mode of engagement shape behavioral mechanisms.
4.6 Theoretical implications
The study contributes theoretically in three respects. First, it advances a differentiated S-O-R perspective by demonstrating that cultural authenticity, community interaction, cultural learning, and cultural aesthetic experience should not be treated as interchangeable stimuli. Their direct and value-mediated relationships differ, while selected relationships vary further according to participation status. S-O-R applications in cultural tourism should therefore consider both the composition of experiential stimuli and the boundary conditions under which particular stimuli are converted into organismic evaluations and behavioral responses.
Second, the findings refine the role of perceived value. PV is not simply a universal mediator between cultural experience and revisit intention. The moderated-mediation results for CA and CI indicate that the value-conversion mechanism depends partly on tourists’ participation status. This shifts attention from whether perceived value mediates experience-behavior relationships toward when and for whom particular mediation pathways operate.
Third, the study contributes to co-creation research by distinguishing actual participation from perceived community interaction and treating CCP as a categorical moderator. Participation does not uniformly strengthen positive relationships; instead, it selectively reorganizes existing experience-value pathways. This provides a more conditional interpretation of tourism co-creation and complements research emphasizing its multidimensional and context-dependent nature.
4.7 Managerial and policy implications
The findings suggest that ethnic cultural tourism should not rely on a single experience-design strategy. Within the binary participation classification used in this study, community interaction was particularly important for perceived value among non-participants. Destinations may therefore provide respectful opportunities for contact with residents through local guides, homestay hosts, craft producers, market vendors, and community interpreters, while recognizing that these recommendations do not distinguish among different intensities or forms of participation.
Among respondents classified as participants, authenticity showed a stronger direct relationship with revisit intention. Participatory activities such as cooking, weaving, farming, crafts, or performances should therefore remain credibly connected to community practices and avoid excessive theatricalization. Local communities should retain meaningful influence over what is presented, how cultural knowledge is interpreted, and which practices are appropriate for tourist participation. This recommendation should be interpreted cautiously because CCP was measured only as participation versus non-participation and did not capture duration, intensity, type, voluntariness, cultural depth, or perceived quality of participation.
Cultural learning should be strengthened through storytelling, interpretation, demonstrations, and accessible explanations of local customs and values, particularly for less participatory visitors. Aesthetic experience should also be protected across both groups because CE directly predicts RI under both participation conditions. Traditional architecture, textiles, cultural landscapes, craft spaces, costumes, and visual symbols should therefore be preserved without excessive standardization.
At the policy level, local authorities and destination management organizations may strengthen community capacity in cultural interpretation, visitor communication, participatory product design, and benefit sharing. Co-creation initiatives should emphasize cultural legitimacy, voluntary participation, community ownership, and mutual benefit rather than simply increasing the number of participatory activities. The present results should not be read as implying that all forms or intensities of co-creation generate equivalent effects.
5.1 Conclusion
This study examined how four dimensions of ethnic cultural experience influence revisit intention directly and indirectly through perceived value and whether these mechanisms differ according to co-creation participation in northern mountainous Vietnam. The overall structural pattern differed significantly across participation groups, but moderation was selective. CCP significantly conditioned the CI → PV and CA → RI relationships, while the remaining structural-path differences were not statistically supported. PV positively predicted RI in both groups.
The conditional indirect-effect analysis further showed that CCP significantly conditioned the value-mediated effects of CA and CI, whereas moderated mediation was not supported for CL or CE. CCP should therefore be understood as a selective categorical boundary condition that changes how particular cultural experiences are translated into perceived value and revisit intention rather than uniformly strengthening the entire model.
5.2 Limitations and future research
Several limitations should be acknowledged. First, the cross-sectional design does not establish temporal causality; longitudinal research should examine whether revisit intentions persist and translate into actual repeat visitation. Second, although multiple destinations were included, a complete population-based sampling frame was unavailable. CCP was self-selected rather than randomly assigned, so observed cross-group differences may partly reflect unobserved demographic or trip-characteristic heterogeneity. Accordingly, the findings should be interpreted as conditional associations rather than causal effects of participation. Future studies should retain respondent-level profile variables within a linkable analytical structure and apply control-adjusted, matching-based, longitudinal, or quasi-experimental designs to better isolate the role of participation.
Third, CCP was measured as a binary participation indicator and therefore does not capture participation intensity, duration, type, voluntariness, cultural depth, or perceived co-creation quality. Future studies should combine behavioral participation indicators with validated multi-item co-creation measures and test whether different forms or intensities of participation produce distinct or nonlinear effects.
Fourth, the latent constructs were measured through self-reported responses. Although the procedural safeguards, the single-factor diagnostic, and the poor fit of the one-factor CFA provide no indication that a single common method factor dominated the covariance structure, common-method and socially desirable responding cannot be excluded completely. Future studies should combine surveys with behavioral records, actual repeat-visit data, experience diaries, or longitudinal follow-up.
Finally, the weaker standalone measurement-model fit for the non-participation group (particularly TLI and RMSEA) warrants replication with larger samples. Although group-specific loadings, reliability, AVE, residual diagnostics, and metric invariance supported retention of the measurement structure, the CCP = 0 estimates should be interpreted cautiously. Future research could also examine mechanisms particularly relevant to participating tourists, including immersion, personal involvement, social bonding, memorability, place attachment, agency, and emotional engagement. Replication across different ethnic groups, destination-development stages, tourist markets, and forms of cultural participation would further clarify the generalizability of the selective boundary-condition mechanism.
[1] Pine, B.J., Gilmore, J.H. (1999). The Experience Economy: Work is Theatre & Every Business a Stage. Harvard Business School Press.
[2] Mehrabian, A., Russell, J.A. (1974). An Approach to Environmental Psychology. The MIT Press.
[3] Zhou, G.F., Chen, W.K., Wu, Y.T. (2022). Research on the effect of authenticity on revisit intention in heritage tourism. Frontiers in Psychology, 13: 883380. https://doi.org/10.3389/fpsyg.2022.883380
[4] Zhou, Q.X., Pu, Y.J. (2022). Impact of cultural heritage rejuvenation experience quality on perceived value, destination affective attachment, and revisiting intention: Evidence from China. Asia Pacific Journal of Tourism Research, 27(2): 192-205. https://doi.org/10.1080/10941665.2022.2046619
[5] Darvishmotevali, M., Tajeddini, K., Altinay, L. (2023). Experiential festival attributes, perceived value, cultural exploration, and behavioral intentions to visit a food festival. Journal of Convention & Event Tourism, 24(1): 57-86. https://doi.org/10.1080/15470148.2022.2131668
[6] Rasoolimanesh, S.M., Iranmanesh, M., Seyfi, S., Ragavan, N.A., Jaafar, M. (2023). Effects of perceived value on satisfaction and revisit intention: Domestic vs. international tourists. Journal of Vacation Marketing, 29(2): 222-241. https://doi.org/10.1177/13567667221086326
[7] Peng, J.M., Yang, X.Y., Fu, S.H., Huan, T.C. (2023). Exploring the influence of tourists’ happiness on revisit intention in the context of Traditional Chinese Medicine cultural tourism. Tourism Management, 94: 104647. https://doi.org/10.1016/j.tourman.2022.104647
[8] Satrya, I.D.G., Adityaji, R., Susilo, C., Karya, D.F., Hendra. (2024). The role of authenticity, local culture, tourism perceived value, and memorable tourism experience for heritage tourism revisit intention. International Journal of Sustainable Development and Planning, 19(3): 1163-1170. https://doi.org/10.18280/ijsdp.190334
[9] Seow, A.N., Foroughi, B., Choong, Y.O. (2024). Tourists’ satisfaction, experience, and revisit intention for wellness tourism: E word-of-mouth as the mediator. SAGE Open, 14(3): 1-16. https://doi.org/10.1177/21582440241274049
[10] Xu, F., Tan, J., Lu, L.J., Li, S.S., Qin, L.G. (2021). How does value co-creation behavior affect destination loyalty? A role switching perspective. Journal of Theoretical and Applied Electronic Commerce Research, 16(5): 1805-1826. https://doi.org/10.3390/jtaer16050101
[11] Monteiro, C., Franco, M., Meneses, R., Castanho, R.A. (2023). Customer co-creation on revisiting intentions: A focus on the tourism sector. Sustainability, 15(21): 15261. https://doi.org/10.3390/su152115261
[12] John, S.P., Supramaniam, S. (2024). Value co-creation research in tourism and hospitality management: A systematic literature review. Journal of Hospitality and Tourism Management, 58: 96-114. https://doi.org/10.1016/j.jhtm.2023.11.008
[13] Deng, Y.T., Lee, H.M., Lee, T.J., Hyun, S.S. (2024). Co-creation of the tourist experience: A systematic assessment scale. Tourism Management Perspectives, 51: 101212. https://doi.org/10.1016/j.tmp.2023.101212
[14] Zeithaml, V.A. (1988). Consumer perceptions of price, quality, and value: A means-end model and synthesis of evidence. Journal of Marketing, 52(3): 2-22. https://doi.org/10.1177/002224298805200302
[15] Sodawan, A., Hsu, R.L.W. (2022). Halal-friendly attributes and Muslims’ visit intention: Exploring the roles of perceived value and destination trust. Sustainability, 14(19): 12002. https://doi.org/10.3390/su141912002
[16] Ko, J.H., Kang, H.S., Seo, J.W. (2023). Experiential value, overall image, revisit intention, and willingness to pay a premium in the context of environmentally certified hotel restaurants: The moderating effect of environmental consciousness. Sustainability, 15(17): 12913. https://doi.org/10.3390/su151712913
[17] Qian, J.W., Li, X. (2024). Perceived value, place identity, and behavioral intention: An investigation on the influence mechanism of sustainable development in rural tourism. Sustainability, 16(4): 1583. https://doi.org/10.3390/su16041583
[18] Liu, H.Y., Park, K.S., Wei, Y.Q. (2024). An extended stimulus-organism-response model of Hanfu experience in cultural heritage tourism. Journal of Vacation Marketing, 30(2): 288-310. https://doi.org/10.1177/13567667221135197
[19] Genc, V., Gulertekin Genc, S. (2023). The effect of perceived authenticity in cultural heritage sites on tourist satisfaction: The moderating role of aesthetic experience. Journal of Hospitality and Tourism Insights, 6(2): 530-548. https://doi.org/10.1108/JHTI-08-2021-0218
[20] Zhang, S.N., Ruan, W.Q., Yang, T.T. (2021). National identity construction in cultural and creative tourism: The double mediators of implicit cultural memory and explicit cultural learning. SAGE Open, 11(3): 1-16. https://doi.org/10.1177/21582440211040789
[21] Zhou, Q.X., Pu, Y.J., Su, C. (2023). The mediating roles of memorable tourism experiences and destination image in the correlation between cultural heritage rejuvenation experience quality and revisiting intention. Asia Pacific Journal of Marketing and Logistics, 35(6): 1313-1329. https://doi.org/10.1108/APJML-11-2021-0829
[22] Cui, M.X., Cheng, L., Shang, Y.F. (2024). The influence of experiencescape of home-based accommodation on tourists’ subjective well-being at cultural heritage sites: The role of value co-creation. Journal of Destination Marketing & Management, 31: 100845. https://doi.org/10.1016/j.jdmm.2023.100845
[23] Wang, L.F., Mao, Z.Q. (2025). How the co-creation of cultural experiences between Hanfu tourists and the destination experiencescape influences revisit intention: A moderated mediation model based on memorability, authenticity, and place attachment. SAGE Open, 15(3): 1-19. https://doi.org/10.1177/21582440251378163
[24] Kumar, N., Adhikari, K. (2026). Role of co-creative experience toward driving tourist satisfaction and revisit intention: Insights from a spooky destination. Consumer Behavior in Tourism and Hospitality, 21(2): 274-288. https://doi.org/10.1108/CBTH-07-2024-0242
[25] Rosseel, Y. (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2): 1-36. https://doi.org/10.18637/jss.v048.i02
[26] Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., Podsakoff, N.P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5): 879-903. https://doi.org/10.1037/0021-9010.88.5.879
[27] Henseler, J., Ringle, C.M., Sarstedt, M. (2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43: 115-135. https://doi.org/10.1007/s11747-014-0403-8
[28] Cheung, G.W., Cooper-Thomas, H.D., Lau, R.S., Wang, L.C. (2024). Reporting reliability, convergent and discriminant validity with structural equation modeling: A review and best-practice recommendations. Asia Pacific Journal of Management, 41: 745-783. https://doi.org/10.1007/s10490-023-09871-y
[29] Hu, L.T., Bentler, P.M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1): 1-55. https://doi.org/10.1080/10705519909540118
[30] Cheung, G.W., Rensvold, R.B. (2002). Evaluating goodness-of-fit indexes for testing measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 9(2): 233-255. https://doi.org/10.1207/S15328007SEM0902_5
[31] Chen, F.F. (2007). Sensitivity of goodness of fit indexes to lack of measurement invariance. Structural Equation Modeling: A Multidisciplinary Journal, 14(3): 464-504. https://doi.org/10.1080/10705510701301834
[32] Edwards, J.R., Lambert, L.S. (2007). Methods for integrating moderation and mediation: A general analytical framework using moderated path analysis. Psychological Methods, 12(1): 1-22. https://doi.org/10.1037/1082-989X.12.1.1