How to improve tourists' trust in official tourism destination websites in China—an empirical research based on CV and PASP

The study addresses declining trust and usage of Official Tourism Destination Websites (OTDWs) among Chinese tourists despite their official status and presumed authority. Prior reports show lower usage of government sites versus commercial platforms and several tourism crises have eroded public trust, shifting negative sentiment toward official management and OTDWs. The research gap lies in understanding how cultural values (CV) and perceived administration service power (PASP) jointly shape trust in OTDWs, as no prior work has integrated both perspectives for tourism websites. The study proposes and tests a model linking Hofstede-derived cultural dimensions (collectivism, low power distance, uncertainty avoidance, high-context communication, local distinctiveness) and PASP components (perceived economic management power, perceived tourism management power, perceived benevolence, perceived integrity) to OTDW trust. Hypotheses H1–H12 articulate expected effects of CV components on CV, PASP components on PASP, and the roles of CV and PASP in predicting trust.

The theoretical background integrates trust theory with Hofstede’s cultural values and PASP. Literature indicates culture shapes website design, content, and user responses; collectivism (COL), power distance (PD), uncertainty avoidance (UA), high-context (HC) communication, and local distinctiveness (LD) are salient in East Asian contexts. Websites reflecting users’ cultural values can foster positive attitudes, intentions, and trust, though excessive cultural disparity may induce anxiety. Mayer’s trust theory introduces perceived benevolence (PB) and perceived integrity (PI) as trustworthiness facets, while ability/power dimensions must be tailored to context. PASP captures perceived government service capacity through perceived economic management power (PEMP), perceived tourism management power (PTMP), as well as PB and PI, all theorized to influence trust in government actors and, by extension, OTDWs. Prior research shows institutional performance and service quality drive political and website-related trust. The study posits CV influences both PASP and trust directly, and PASP further contributes to trust in OTDWs.

The study employed Partial Least Squares-Structural Equation Modeling (PLS-SEM) using SmartPLS 3.3.4 to assess a complex structural model relating CV, PASP, and trust. Measurement model evaluation included convergent and discriminant validity, composite reliability (CR), average variance extracted (AVE), and factor loadings; structural model evaluation used path analysis and R2 values. Additional software: SPSS 27.0.1 for descriptive statistics and AMOS 23 for confirmatory factor analysis (CFA) of second-order constructs (CV and PASP). Sampling targeted individuals with both offline travel and online tourism website experience, especially OTDW users. Questionnaires were distributed via WJX.cn from May–July 2022 using social apps (QQ, WeChat), WJX paid promotion, emails to tourism employees, and limited on-site paper surveys. Of 482 collected responses, 324 valid cases remained after data cleaning (67.22% valid rate). A 7-point Likert scale measured constructs; each latent variable had 3–5 indicators. A pilot with 50 undergraduates yielded Cronbach’s alpha >0.71 and minor wording revisions. The final instrument covered demographics, CV (21 items), PASP, and paths between CV, PASP, and trust. Sample characteristics: 50.31% male, majority aged 18–35, and 68.52% with bachelor’s degrees. Model adequacy checks showed loadings mostly >0.7, CR >0.85, Cronbach’s alpha >0.78, and AVE >0.58. Robustness tests were performed on split samples (gender, education) to mitigate endogeneity concerns.

- Measurement validity and reliability: All constructs exhibited strong psychometric properties (outer loadings >0.7; Cronbach’s alpha >0.78; CR >0.85; AVE >0.58). Discriminant validity satisfied (Fornell-Larcker and cross-loadings). - Explained variance (R2): CV = 38.7%; PASP = 62.3%; Trust = 72.2%, indicating substantial explanatory power of exogenous constructs. - Structural paths (Table 7): • H1 COL → CV: β = 0.215, SE = 0.055, t = 3.896, p < 0.001 (Supported). • H2 LPD → CV: β = 0.162, SE = 0.072, t = 2.258, p < 0.05 (Supported). • H3 UA → CV: β = 0.029, SE = 0.076, t = 0.385, n.s. (Rejected). • H4 HC → CV: β = 0.403, SE = 0.048, t = 8.386, p < 0.001 (Supported). • H5 LD → CV: β = −0.083, SE = 0.080, t = 1.035, n.s. (Rejected). • H6 PEMP → PASP: β = 0.073, SE = 0.067, t = 1.104, n.s. (Rejected). • H7 PTMP → PASP: β = 0.122, SE = 0.067, t = 1.825, p < 0.1 (Weakly supported). • H8 PB → PASP: β = 0.168, SE = 0.070, t = 2.395, p < 0.01/0.05 (Supported). • H9 PI → PASP: β = 0.007, SE = 0.072, t = 0.096, n.s. (Rejected). • H10 CV → PASP: β = 0.503, SE = 0.058, t = 8.738, p < 0.001 (Supported). • H11 CV → Trust: β = 0.508, SE = 0.033, t = 21.355, p < 0.001 (Supported). • H12 PASP → Trust: β = 0.393, SE = 0.048, t = 8.237, p < 0.001 (Supported). - Controls (age, gender, education, prior travel) had no significant effect on Trust. - Substantive insights: Tourists’ trust in OTDWs is significantly driven by collectivism, low power distance, high-context cues via their effects on CV, and by perceived benevolence and (weakly) perceived tourism management power through PASP. CV strongly influences both PASP and Trust.

The findings demonstrate that OTDW trust emerges from how cultural values and administrative service power are communicated on official sites. Expressions aligned with collectivism, low power distance, and high-context communication significantly shape perceived cultural congruence (CV), which in turn elevates PASP perceptions and directly enhances trust. PASP matters for trust, especially via perceived benevolence and, to a lesser extent, perceived tourism management power; perceived economic management power and integrity did not show significant direct effects on PASP in this setting. The results suggest that trust is best improved through coordinated presentation of culturally congruent content and signals of benevolent, capable tourism governance. Adopting culturally similar cues (COL, LPD, HC) appears more effective for domestic Chinese audiences. Contrary to expectations, local distinctiveness did not significantly influence CV or trust, implying that LD effects may be segment-specific or require more focused thematic presentation. Overall, the model clarifies mechanisms by which CV and PASP interplay to reduce perceived risk and strengthen trust in OTDWs.

This study develops and validates a comprehensive trust model for OTDWs that integrates cultural values (COL, LPD, UA, HC, LD) and perceived administration service power (PEMP, PTMP, PB, PI). Empirical results from 324 Chinese tourists show that collectivism, low power distance, and high-context cues significantly enhance perceived cultural congruence, which strongly influences both PASP and trust. Perceived benevolence and, weakly, perceived tourism management power also contribute to PASP and subsequent trust, while perceived economic management power and perceived integrity were not significant predictors of PASP in this sample. The contributions include: (1) conceptualization and validation of an OTDW trust scale spanning CV and PASP; (2) evidence that the expression of culturally congruent content and benevolent, capable governance signals fosters trust; and (3) practical guidance for tailoring OTDW design. Future research should test the framework across multiple operating OTDWs with pre/post evaluations, broaden samples to enhance generalizability, and explore segment-focused LD themes (e.g., culinary content) to clarify their effects.

- Generalizability: Findings may be context-specific to Chinese OTDWs and not fully transferable to other destinations or platforms. - Sample representativeness: The sample may not reflect the full diversity of Chinese tourists. - Design scope: Results are based on survey perceptions rather than observed behavioral outcomes; longitudinal or field experiments on live OTDWs with pre/post data are suggested. - Construct scope: The non-significant role of local distinctiveness may reflect sample composition or broad operationalization; more granular LD themes could yield different results.