Introduction

Tourism plays a significant and multifaceted role in the Chinese economy, politics, and society. The Official Tourism Destination Website (OTDW) serves as a crucial government-sponsored platform for promoting tourism. Chinese individuals commonly display explicit behaviors that demonstrate their respect towards authority figures, as evidenced by studies (Chien, 2016). This respect for authority is closely associated with a higher level of popular trust, leading people to have a greater inclination to trust agencies endorsed by high-authority organizations, such as the government (Su et al., 2021). Therefore, under the influence of Chinese CV, consumers believe that the official is synonymous with a high level of competence and service compared to other organizations. OTDW enjoys a notable competitive advantage over other commercial travel websites (Cox et al., 2009). Thanks to this environment, and with the strong support of the government, the OTDWs were established and developed rapidly. As early as the end of July 2006, 95% of Chinese OTDWs had been established, covering 1 national ministry of culture and tourism, 33 local cultural and tourism bureaus/offices, over 300 cities, state and county-level cultural and tourism bureaus, and 302 national 5A-level scenic spots (练红宇, 2007).

As important information output platforms of the official authorities, OTDWs should ideally be more reliable and authoritative compared to tourism e-commerce website, tourism social network platforms, and others. However, studies have shown that the usage of official websites is actually lower than that of other commercial websites (Prideaux et al., 2008, Wang, 2014). News reports showed that more than 68 percent of city residents had never used a government website, and only about 12 percent of individuals obtained government information through government websites, lagging behind the percentage of individuals who obtained such information through commercial websites, social media, and text messages (Daily October 28, 2014). In addition, tourism crises, such as scams and misconduct such as seafood scam in Sanya (CCTV, 2012), the outrageous shrimp scam bill in Qingdao (Baidu.com, 2015), “Snow Village rip-offs” in Mudanjiang (Baidu.com, 2017) and incidents of humiliation and threats of tourists by tour guides in Yunnan (Youku, 2020), have weakened the trust and dependence of Chinese tourists on OTDWs. A large number of negative emotions and voices have gradually shifted from tourism enterprises to the official management departments and the OTDWs (Zheng, 2019, Liu and Ding, 2020, Zhu, 2020), resulting in a crisis of trust in OTDWs (Zhu, 2020).

Although some studies have examined the impact of trust on tourism websites, to the author’s knowledge, there is no study examining the impact of combining CV and PASP on tourism websites, especially OTDWs. Moreover, a lack of understanding of how CV and PASP promotes OTDWs trust has induced calls for:

  • Studies on the mechanisms through which CV and PASP contribute to web trust (Lee, 2010, Chakraborty and Sadachar, 2023).

  • Research focusing on trust in OTDWs services and not on trust in tourism goods (Zelenka et al., 2021, Mior Shariffuddin et al., 2023).

Thus, the present study responds to these calls by examining the direct and indirect effects of CV and PASP on OTDWs trust. The existing research on the inherent trust advantage of OTDWs provides valuable insights into this phenomenon. This study aims to explore the underlying factors that cause tourists to hesitate when choosing to use and trust OTDW. We employ an empirical study of surveys to test the trust model (Lin and Yeh, 2013, Rojo et al., 2020). The subsequent part of the paper thoroughly reviews existing literature and explains this study’s theoretical background and methodologies. In order to examine the research hypotheses and establish the connections between the research constructs, the study utilized Partial Least Square Structural Equation Modeling (PLS-SEM). The results and discussion of the study are presented, and the theoretical contributions, practical implications, limitations and future research of the study are introduced.

Theoretical background

Trust theory is expanded in this study by utilizing CV and PASP perspectives to establish the links between CV, PASP, and trust in OTDWs. There is a consensus on Chinese CV characteristics, such as high collectivism (COL), low power distance (LPD), strong uncertainty avoidance (UA), and matrilineal culture (Hofstede and Bond, 1988, Wong and Lau, 2001, Li and Xu, 2007, Tsang, 2011, Qiu et al., 2015). China, Japan and South Korea are typical representatives of high-context (HC) (Moura et al., 2015). In HC cultures, communication between two parties is not solely based on straightforward language but rather relies on indirect methods, such as public knowledge and shared experiences. Horng and Tsai (2010) emphasizes that food is an essential local distinctiveness (LD) in travel, and it can be very valuable to promote on OTDWs. To effectively leverage COL, LPD, UA, HC, and LD to predict tourists’ online trust, it is vital to ensure that the design of OTDWs is easily understandable and trustworthy for potential tourists.

Trust comes from the long-term accumulation of comprehensive factors, such as learning information through personal experience, which constitute the continuous trust evaluation system of tourism service ability (Schneider, 2007). This study utilizes two indicators from Mayer’s trust theory, namely Perceived Benevolence (PB) and Perceived Integrity (PI) (Mayer et al., 1995, Hall and Page, 2009, Taamneh et al., 2022), to assess the perceived trust of OTDW. The measurement dimension of power needs adjustment as the research field changes (Mayer et al., 1995). Perceived Tourism Management Power (PTMP), tourism information service ability, and local culture conveyed by OTDWs play a role in attracting, infecting and touching users (Hall and Page, 2009). Local management departments’ powers, such as Perceived Economic Management Power (PEMP) (Robin and Stephen, 2013) and resource control (Liu et al., 2019), also impact the perception and continued trust of tourists. This implies that a comprehensive understanding of PASP is crucial in this study, connecting various theories as tourists assimilate, select, and comprehend information through multiple interactions with the OTDW. The perception of a OTDW’s value and reputation, formed by tourists throughout this process, significantly impacts their trust outcome.

Hofstede culture value (CV)

Website is not a culturally neutral medium (Singh et al., 2003), as culture value can influence tourists’ expectations, attitudes, and behavior in various aspects of the tourism experience. Buhalis and Law (2008) analyzed the OTDWs of 56 countries and found that culture value, such as COL, power distance, and masculinity-femininity, impacted the design and content of these websites. They observed that countries with LPD had more egalitarian structures and casual language. Another relevant study by Kim et al., 2019 analyzed the OTDWs of 20 Asian countries and identified that cultural value, such as PD and COL impacted the content and functionality of these websites.

Singh et al. (2003) and Laosethakul and Boulton (2007) both pointed out that China, Indonesia and Venezuela were typical representatives of low individualism whose points were 15–48. As this study focuses on Chinese tourists, the COL characteristic is particularly relevant. Although high power-distance cultures tend to create a greater sense of distance, LPD product ratings are considered more helpful and attractive (Filieri et al., 2018, Filieri et al., 2019). Risk and uncertainty are “inherent” to tourism (Holland, 2017). UA is the way people cope with uncertainty and ambiguous situations (Hofstede et al., 1980). Related to personal safety needs, people find their inner motivations to meet personal needs and reduce uncertainty(Li et al., 2013). Hall (1976) introduced the concept of HC culture, which is prevalent in Eastern countries, including China. Würtz (2005) believed that non-verbal communication was more effective than verbal communication in HC cultures. This type of communication is transmitted indirectly, where the message’s meaning is not conveyed directly but rather subtly implied through mood, gesture, expression, and picture, where meaning and context are inseparable. The distinctive material or intangible cultural resources, such as the unique healing practices of the Caribbean rainforest, ethnic tourism, and food in local restaurant, can be summarized as LDs for tourism (Yang et al., 2008, Lin and Yeh, 2013, Miocevic et al., 2022).

Cultural similarities stand on the opposite side of the tourism experience and can have both positive and negative effects on tourism experiences. While a similar cultural background can create positive views and boost purchase intention, it can also reduce novelty and sensory stimulation, which are important for hedonic motivation in tourism (Goossens, 2000). Baack and Singh (2007) confirmed that a website reflecting similar cultural values would lead users to have positive views, boosting purchase intention, as well as enhancing trust and loyalty. However, too much cultural difference may increase visitor anxiety, as different cultures have different sensitivities and influences on customers and tourists (Bhawuk and Brislin, 1992). In the case of Chinese tourists, their deeply rooted CVs significantly impact their behaviors and decision-making process online even without being noticed (Power, 2009). OTDWs need to find a balance between the sense of security brought by the tourists’ familiar CVs and the mystery and excitement brought by the tourist destinations’ uncertain culture values. The subsequent hypotheses are posited based on the comprehensive interactions of the Chinese CV constructs.

  • H1: COL in OTDW has a positive impact on perceived CV of tourists.

  • H2: LPD in OTDW has a positive impact on perceived CV of tourists.

  • H3: UA in OTDW has a positive impact on perceived CV of tourists.

  • H4: HC in OTDW has a positive impact on perceived CV of tourists.

  • H5: LD in OTDW has a positive impact on perceived CV of tourists.

Perceived administration service power (PASP)

PASP is an index that tourists use to assess the level of government service ability (Nunkoo and Smith, 2013). Solely relying on the invisible hand of the market to organize and regulate itself may result in the tourism industry eventually losing sight of its original intention due to a lack of proper coordination. According to the conceptual model, perceived impacts can be categorized into policy-related impacts and tourism-related impacts, encompassing the economy, social culture, and the environment (Qin et al., 2019). Saptu et al., (2020) examined the impact of perceived government support on attitudes towards Agricultural Entrepreneurship, Perceived Behavioral Control, and the impact of social norms on Agricultural Entrepreneurship intention of Sabahan youth. PASP, such as perceived government support, has a significant impact on Agricultural Entrepreneurship, Perceived Behavioral Control, and industrial economy.

Ability, PB and PI indexes in trust theory have a significant impact on tourists’ trust (Mayer et al., 1995). It is also important to note that this relationship may be complex as the definition and operationalization of Ability may vary across studies, and different aspects of this construct may have other effects on PASP. The institutional theory of political trust is based on hypothetical trust, including three indicators of PEMP (Hetherington et al., 1998). Compared with the other two indicators, the effect of PEMP perceived by the public is more direct, and the cycle is shorter. PEMP can encourage capital accumulation and stimulate economic expansion (Nunkoo and Smith, 2013). Governments can do the same in the tourism economy (Bevir, 2009), for example by increasing sustainable tourism economic construction projects that affect their daily lives (Bramwell, 2011). PTMP is the perceived level of tourism management ability of local governments by citizens, which directly affects their trust in tourism management departments (Wong et al., 2011, Nunkoo and Gursoy, 2012). Residents’ perception of PTMP has been found to positively influence their trust in government in the real world (Oskarsson et al., 2009, PerOla and Torsten, 2010, Nunkoo and Ramkissoon, 2012).

Similar to PB, integrity in others has been shown to foster increased trust (Colquitt et al., 2007). However, there are situations that acting benevolence may seem to require compromising one’s integrity, and vice versa. Conflicts between PB and PI are frequently encountered (Lupoli et al., 2018, Moore et al., 2019). Tourists expect PB to reflect the government’s efforts to serve their interests when developing, which requires the government to prioritize tourists’ needs without being solely driven by profits, regardless of whether it is perceived by the tourists or not (Mcknight et al., 1998, Belanger and Carter, 2008). PI and service quality in turn significantly impact PB and privacy concerns (Janssen et al., 2018). In such challenging situations, PB and PI conveyed by OTDWs require special attention. The subsequent hypotheses are posited based on the comprehensive interactions of the constructs. Building upon these observations, we hypothesize:

  • H6: PEMP in OTDW has a positive impact on PASP of tourists.

  • H7: PTMP in OTDW has a positive impact on PASP of tourists.

  • H8: PB in OTDW has a positive impact on PASP of tourists.

  • H9: PI in OTDW has a positive impact on PASP

Trust

Trust plays a crucial role in travel decision-making, affecting tourist satisfaction, well-being, and loyalty (Harris and Goode, 2004, Kim, 2012, Lee et al., 2019). Trust can be divided into online trust (Kim et al., 2011) and offline trust (Abubakar et al., 2017). When using e-government websites, citizens’ trust is primarily shaped by their evaluation of the officials responsible for developing, maintaining, and monitoring the system, rather than the system itself (Dashti et al., 2009). This indicates that the dimension of trustworthiness, namely PASP, is crucial in a network environment. Particularly in situations where the trustee’s ability, benevolence, and integrity are uncertain or there is limited information available, an individual’s disposition to trust becomes even more pivotal. Belanger and Carter (2008) discovered that the disposition to trust has a significantly impact on both Internet trust and government trust. When government departments provide accurate and comprehensive information, it enhances public trust and subsequently increases residents’ trust in the government (Rousseau et al., 1998).

Culture plays a significant role in shaping consumers’ responses, including tourists, from their decision-making process and purchase intention to loyalty and stickiness (Hofstede et al., 1999). The CV quality of virtual travel community significantly influences the tourist’s trust, which, in turn, affects their attitude towards the website and their intention to transact (Bhawuk and Brislin, 1992, Elliot et al., 2013). Trust does not directly influence website stickiness or intention to transact (Elliot et al., 2013), but trust is crucial in reducing uncertain risk, especially when people have limited cognition in decision-making processes (Grabner-Kraeuter, 2002). Tourists can enhance their culture abilities through sustainable social activities, such as pursuing spiritual experiences, which can in turn influence their perception of the CV (Woodside, 2000, Tung and Ritchie, 2011).

OTDW serves not only as an important platform for disseminating official tourism information but also as a platform for displaying administrative service power. Residents of China, South Korea, and Japan tend to have a positive inclination towards experts and authorities, which makes famous online brands more easily accepted and trusted by consumers (Zeithaml et al., 2002). The government’s service capabilities and corresponding policy performance are the most powerful indicators of citizens’ trust (Robin and Stephen, 2013). Once the trust is established, it will positively influence the brand perceptions which will also positively affect transaction intention. The found hints at the fragility of consumer’ trust, and also reflect the extreme caution of Chinese consumers in online transactions (Chen et al., 2008, Lin and Yeh, 2013). It is essential to understand how PASP affects tourists’ trust in OTDWs.

  • H10: CV in OTDW has a positive impact on PASP of tourists.

  • H11: CV in OTDW has a positive impact on Trust of tourists.

  • H12: PASP in OTDW has a positive impact on Trust of tourists.

Altogether, the theoretical framework of tourists’ trust model for OTDW proposed in this paper is shown in Fig. 1:

Fig. 1
figure 1

Study framework.

Full size image

Methodology

SmartPLS 3.3.4 was used to confirm the reliability of the study framework and test the hypotheses. The Partial Least Squares-Structural Equation Modeling (PLS-SEM) has been commonly used in numerous tourism-based research (Usakli and Kucukergin, 2018, Koç et al., 2022). It enables researchers to analyze both the measurement and structural models. PLS-SEM offers several advantages (Hair et al., 2020), as it effectively handles data with complex hierarchical models, even with a small sample size, and it is suitable for data that does not follow a normal distribution(Wang et al., 2019). In this study, the structural model is complex, and since the objective was to explore the intricate relationship among the dimensions of the constructs; hence, selecting PLS-SEM was considered appropriate (Hair et al., 2020).

Accordingly, PLS-SEM tests the measurement and study hypotheses in two steps: assessing the measurement and structural models. SmartPLS was employed to evaluate reliability and validity in the measurement model assessment. First, the measurement model was assessed by analyzing the convergent validity, discriminant validity, composite reliability (CR), average variance extracted (AVE) and correlation analysis of latent variables of the constructs and measurement items. The structural model was subsequently assessed using path analysis and the coefficient of determination. In addition, path analysis models require a sample size of at least 10 times the number of observation indicators corresponding to each latent variable in the model (Chin and Marcoulides, 1998, Chin et al., 2003). As this trust model had 3–5 observation indicators for each latent variable and an effective sample size of 324, it met SmartPLS’s sample size requirements.

For data analysis, other types of software were used in this research, namely SPSS 27.0.1 and Amos 23. SPSS was utilized to conduct descriptive statistics on the survey data, while Amos was employed to perform confirmatory factor analysis on the same data.

Sample

The target population for this research is individuals with both offline travel experience and online tourism website experience, especially those who use OTDWs for travel and tourism services. As trust is a social construct that involves a willingness to rely on or have confidence in another person, institution, or concept (Rousseau et al., 1998), we employed a data collection approach that combined simple sampling with quota sampling. This approach provides a balanced and effective approach to data collection, utilizing the strengths of both methods to ensure the most realistic and reliable data possible.

Data were collected through the platform named WJX.cn. The following four methods were used to distribute the questionnaires from May 2022 until July 2022: (1) via social applications such as QQ friends, QQ groups, and WeChat; (2) through the purchase of promotion services on WJX.cn, allowing for the targeted distribution of questionnaires via email; (3) via electronic questionnaires sent to tourism company employees via QQ; (4) a small number of paper questionnaires were randomly distributed to tourists in the scenic area. A total of 482 questionnaires were collected in this survey. After excluding questionnaires with identical answers, 2–3 duplicate answers, and too many blank items that were not answered seriously, 324 valid questionnaires remained in this survey, accounting for 67.22% of the total. None of the differences were found to be significant by t-tests, suggesting the non-response was less likely to be a cause of concern in subsequent analysis.

Measurement

The questionnaire of the study consisted of four parts. Part I gathered general data regarding gender, age, educational, and years of use internet. Part II gathered data on CV consisting of 21 items. Part III asked about PASP. Part IV consisted of three parts: the first part is about the impact of CV on PASP, the second part is about the impact of PASP on trust, and the last part is about the impact of CV on trust.

The trust model was examined using a Likert 7-level scale. In addition, we considered the possibility of removing indicators with low factor loading coefficients in the future. As a result, each latent variable was designed with 3–5 observation variables (see Table 1). This satisfies the fundamental requirements of structural equation modeling (SEM) for observation variables.

Table 1 Variables and items.
Full size table

The initial version of the trust model questionnaire included seven demographic questions and 50 observational variable questions. Before conducting the pre-survey, the questionnaire’s content underwent revisions for grammatical and other aspects. The revisions were suggested by experts from tourism enterprises and professors from universities to ensure that the questionnaire’s language and format aligned with Chinese expression and reading habits.

Validity and reliability

Before the formal survey, a pilot study was conducted to minimize errors and assess the accuracy and relevance of the items that measure CV, PASP and trust in OTDWs. For this purpose, the pre-survey was conducted with 50 undergraduates, and the data collected was found to be reliable and valid within acceptable ranges. The pilot study produced Cronbach’s Alpha values that surpassed the minimum threshold of 0.71, indicating strong internal consistency. Based on the undergraduates’ feedback, slight revisions were made to enhance clarity, resulting in minimal item rewording and modifications. The formal questionnaire’s complete set of questions can be found in Table 1.

Results

Demographic analysis

Questionnaires with only one answer and incomplete answers were excluded, hence, data from only 324 individuals were used for the statistical analysis (see Table 2).

Table 2 Descriptive statistics.
Full size table

From the 324 respondents, 49.69% were female and 50.31% were male. In addition, 29.01% of the respondents were between 18 and 25 years old, followed by those who were between 31 and 35 years old (26.85%). Concerning the educational level, 68.52% of the sample were either enrolled or had completed a bachelor’s degree, 19.75% were graduates or postgraduates, 0.93% had completed high school education, and 10.80% had attained a junior college diploma. In terms of previous experience with using Internet for tourism, all respondents had varying lengths of experience. About 33.33% of respondents have been using the Internet for 6-10 years to find travel information, followed by 1–5 years (29.94%), 11–15 years (22.22%), more than 16 years (12.97%), and 1.54% have been using the Internet for less than 1 year.

Measurement model

This study tested the second-order factor structure of PASP and constructed the PASP measurement model by taking PEMP, PTMP, PB, and PI as the first-order factors. Similarly, the CV measurement model was constructed based on the data collected from the questionnaire. Both PASP and CV measurement models were analyzed using AMOS to test the factor structure of latent variables. The results of the analysis are presented in Table 3, and the fit indices indicate an acceptable level of fit.

Table 3 Confirmatory factor analysis fit index.
Full size table

As recommended by Hair et al. (2014a), the final model was also examined by assessing the reliability and validity for each construct of the study. Tables 4 and 5 summarize the different reliability and validity indicators for the measurement model. As shown in Table 4, this model has good convergent validity for the following reasons: (1) the load factors corresponding to all the observation indexes of the latent variables in this research exceed 0.7; (2) its reliability is greater than 0.85; (3) the Cronbach’s Alpha coefficients are all greater than 0.78; (4) and the AVE of all the latent variables exceed 0.58 (Fornell and Larcker, 1981, Hair et al. 2010, Hair et al. (2014a), Hair et al. 2014b). To verify the discriminatory validity of the metric model, the study carried out an analysis of the latent variable correlation factor load coefficient and correlation coefficient. As shown in Table 5, the diagonally blacked numbers represent the square root of the AVE of each latent variable, and the off-diagonal elements are the correlation coefficients of each latent variable. The square root of the AVE of all latent variables has been greater than the correlation coefficient of other latent variables (Chin, 1998).

Table 4 Convergent validity.
Full size table
Table 5 Correlation analysis of latent variables.
Full size table

Furthermore, as shown in Table 6, the factor load coefficients of each observation variable attached to the measured latent variable are higher than the factor load coefficients attached to the other latent variables (Gefen and Straub, 2005). All of these outputs demonstrate that the scale has acceptable levels of discriminatory validity.

Table 6 Cross loadings of latent variables.
Full size table

Structural model

A structural model was developed and tested to examine the relationship among the constructs studied: COL, LPD, UA, LD, HC, PEMP, PTMP, PB, PI, CV, PASP, and Trust. Table 7 and Fig. 2 present the results of the structural model. From the latent variable output of this study (Rojo et al., 2020), three major findings: (1) the explained variance R2 showed that UA, HC, LD, COL, and PD could explain 38.7% of CV of OTDW; (2) PEMP, PTMP, PB and PI could explain 62.3% of the variance of PASP; (3) the explained variance of Trust was 72.2%. It showed that the exogenous latent variables of the trust model could better explain the endogenous latent variables. Additionally, Robustness Tests (Neumayer and Plümper, 2017, Rojo et al., 2020) were performed on split samples of gender and educational to avoid endogeneity issues.

Table 7 Results of testing the structural equation model.
Full size table
Fig. 2
figure 2

PLS-SEM results.

Full size image

The support of path analysis to the trust theoretical model is shown in Table 7. At the 5% significance level, whenever the T value is 1.96, the path analysis shows that (1) COL (Standard Error = 0.055, t = 3.896, p < 0.005), LPD (Standard Error = 0.072, t = 2.258, p < 0.005) and HC (Standard Error = 0.048, t = 8.386, p < 0.005) have significant impacts on OTDW’s CV, and the others have no significant impact, thus providing support for hypotheses 1, 2 and 4; (2) only PB (Standard Error = 0.070, t = 2.395, p < 0.005) has a significant impact on PASP, thus providing support for hypotheses 8; (3) When the T critical value is greater than 1.68, PTMP (Standard Error = 0.067, t = 1.1825, p < 0.1) has a weak impact on PASP (Lin and Yeh, 2013), thus providing support for hypotheses 7; (4) finally, PASP (Standard Error = 0.477, t = 8.237, p < 0.005) has a significant impact on Trust, and CV has a significant impact on PASP (Standard Error = 0.576, t = 8.738, p < 0.005), and Trust (Standard Error = 0.033, t = 21.355, p < 0.005), thus providing support for hypotheses 10, 11, and 12. All the factors of education level, age, gender, net age, and experience of using the internet before traveling have no significant impact on Trust.

Discussion

Our results strengthen and refine prior empirical research that has started to study tourist trust toward a tourism destination (Robin and Stephen, 2013, Liu et al., 2019). In fact, the evaluation of the results obtained from testing H11 and H12 reveals a more complicated and nuanced view of the relationship between CV, PASP and Trust than has been established in prior literature (Kim et al., 2011, Nunkoo and Smith Stephen, 2014), indicating that our finding improves OTDWs trust. Our study provides greater insight into this relationship by considering the dimensions of CV and PASP and by drawing on SEM to study the effects of tourist trust OTDWs.

As mentioned above, the SEM approach seeks to identify effective ways that help achieve OTDW trust, based on the premise that what OTDW uses influencing factors—that is, how to express—is more important than the influencing factors themselves (Hew et al., 2016). Following this premise, our results confirm that the effect of OTDW trust does, in fact, depend on the PASP or CV expression of OTDWs. In other words, a single influencing factor is not strong enough to affect trust. Our results thus reveal a combination of strategies where two different variables work together to enhance tourist trust.

First, destination tourism management departments should pay attention to the promotion of PASP. This result aligns with the research conducted by Liang et al. (2016) and Lee and Koo (2017) on trust in the government’s management of tourism-related issues in China, Taiwan, and Korea. They suggested that tourists are more likely to recommend destinations with high PASP levels to others, leading to positive word-of-mouth and increased tourism revenue. Conversely, negative experiences and memories can lead to a crisis of trust. PTMP (H7) and PB (H8) will influence OTDW trust (i.e., degree to which tourists believe that tourist destination government management cares about them and wants to help them). The strong relationship of service quality with benevolence is also supported by prior research (Tan et al., 2008). Destinations must combine PB with PTMP. Yet merely opting to feel PB is insufficient; that is just detailed information listing and route recommendation on OTDWs, tourists will become esthetically fatigued. In other words, the lack of PTMP changes the expected effect of an ambidextrous strategy. Our study argues that this result occurs because PTMP can fully reflect an effect of tourism management department in managing resources and punishing evil deeds, enabling PTMP to develop PB.

Our second recommendation is to focus on and target CV enhancement on OTDWs. As expected, CV has a significant effect on PASP (H10), further confirming the way government express power or PASP on the basis of prior studies. These previous studies confirmed the relationship between administration service power, trust, life, tourism benefits, and others (Kim and Fesenmaier, 2008, Yang and Khoo, 2015, Chou and Lee, 2018, Han and Hyun, 2019). Residents perceived positive and negative impacts of tourism, knowledge of tourism, perceived power in tourism, and satisfaction with tourism significantly predicted their trust in government actors. Residents perceived positive and negative impacts of tourism were also significantly associated with their quality of life (Tichaawa et al. 2023). Perceived effective local government management of tourism had a strong significant effect on the residents’ trust in government actors. Moreover, residents’ PTMP was a significant determinant of perceived tourism benefits (Rodrigues et al. 2020). Residents’ perceptions of the political and economic performance of government actors significantly predicted trust in government actors. Residents’ support is determined by residents’ trust in government actors and perceived benefits (Nunkoo and Ramkissoon, 2012, Nunkoo and Smith Stephen, 2014). These results confirm that perception is one of the premises, that is, the premise for trust to play a role is that the government’s tourism service capability and benevolence can be perceived by residents. However, these previous studies on PASP, PB and PTMP are usually conducted after tourism events, and how residents perceived and perceived the source mostly relied on the vague summaries of past experience. We argue here that the channel for instant access to government perception (that is, in a certain period of time, tourists use all the information on the OTDW to establish a logical perception ability judgment system) and the way for tourists to quickly feel PASP (that is, whether the PASP information is expressed in a manner of CV similarity or CV opposition) is to adopt similar information on CV dissemination PASP in OTDW. Contrary to previous studies, the absence and presence of PASP can only locally affect perceived trust, and more importantly, the effect of CV on trust. PASP can amplify the effect of CV on perceived trust. In this case, the result is the same whether the CV is directly influenced by itself or through the mediating effect of ability.

Our third and last recommendation is that adopting similar CVs is better for promoting perceived trust. COL(H1), LPD(H2) and HC(H4) exhibited a significant direct effect on CV, confirming that the factors why CV differs across different societies and how those differences can shape people’s attitudes, behaviors, and social norm. This result is in line with the previous findings of Triandis (1995) and Chen and Kim (2013), who found that HC cultures valued social relationships and trust over direct communication and task efficiency and was positively related to values such as group harmony, social obligation, and face-saving in China. COL values prioritize group loyalty, social harmony, and family relationships over individual achievement and independence (Yamamoto et al., 2022). This study confirmed that individualism was a strong predictor of self-confidence in America, but even in this country, COL was a significant predictor of recognizing the need for mental health services. But in Japan, where COL culture is stronger, participants, especially male participants, felt more stigmatized toward mental health professionals. In this case, COL was not a significant predictor in Japan. Our study argues that the social characteristics of CVs are not applicable to all scenes, and it is necessary to conduct CV research in each scene to find out the best form of information transmission. By using appropriate CV expressions to convey information to tourists, tourism government management can maintain and establish trust among clients, even in times of uncertainty and crisis.

Finally, regarding LD, our results show that LD has no significant effect on CV and Trust, which our study had expected to have a positive impact. This contributes to the debate concerning the relationship between LD and CV. The existing literature can be divided into several types: some found a positive relationship between LD and CV, some found a negative relationship between LD and CV, and others found a weak correlation between LD and CV. Yousaf and Xiucheng (2018) found that Japan, South Korea and Thailand attempted to strategize their country’s potential as a preferred Halal tourism destination for Muslim tourists by introducing and promoting Halal cuisines, Halal food culture, Halal food restaurants and general Halal services of interest for Muslims. Horng and Tsai (2010) divided tourists into four categories and found that the survey results would also change if the proportion of tourists in a certain category was too large. Therefore, we have reason to believe that the majority of survey participants in this study belong to the latter two groups as our study found no significant relationship between LD and CV. This result makes perfect sense if we interpret it by expanding the sample size or narrowing the LD to a particular topic, such as food.

Conclusion

Theoretical contributions

Key factors influencing tourists’ trust in OTDWs

In this study, the construct of OTDW trust was developed and validated. To identify the key factors that affect tourists’ trust in OTDWs, the study contributed to the existing literature by highlighting the influence of CV and PASP on tourist behavior and emphasizing the pivotal role of trust in decision-making processes. The construct comprises two dimensions—CV (including COL, LPD, UA, LD and HC) and PASP (including PEMP, PTMP, PB and PI). Throughout the related literature, the scale of trust has come under tremendous discussion, given its importance to the success of businesses. For example, Kim et al. (2011) developed a trust scale in the context of online travel agencies, and Han and Hyun (2015) determined the measurements of trust in travel brands. However, to the best knowledge of the authors, the construct of trust has never been developed in the context of an OTDW. For the current study, OTDWs trust was established and validated with the aid of both a quantitative examination of questionnaire survey. The results confirmed that trust in OTDWs involves two dimensions: cultural value and government service power. The dimensional scale of OTDW trust was consistent with some of the previous studies. For example, Kim (2008), Moura et al. (2015), Kim et al. (2019) and Horng and Tsai (2010) mentioned CV components of OTDW trust, including COL, LPD, UA, LD, and HC. Mayer et al. (1995) measured PASP components of OTDW trust, including PB, PI and ability. Nunkoo and Smith (2013), Yousaf and Xiucheng (2018) measured PASP components of residents’ trust in tourism government, including PEMP, PTMP. Our research then further developed the trust in OTDWs scale from a more comprehensive perspective. The findings of key factors that affect tourists’ trust in OTDWs could fill the research gap regarding the preferences of CV and PASP among OTDWs.

Development and validation of OTDW trust model

To address the uncertainty of OTDWs’ trust information, this study proposed a trust model that includes CV and PASP to solve the dilemma of OTDWs trust. The proposed model reflects the bounded rationality such as PASP influence of tourists and the perceptual dimension such as CV influence when using OTDWs, which not only helps to address the uncertainty of OTDWs trust information but also helps to enhance trust. This study was a bottom-up exploratory study to discover the theory from the existing literature and constructed a conceptual model of tourist trust in OTDWs. Previously, most studies developed measurement trust model items from a literature review or some traditional qualitative methods e.g., (Chen and Tsai, 2007, Kwon and Lee, 2020, Rojo et al., 2020). Different OTDWs with distinct CV and PASP emphasize different aspects of their tourism destinations, leading to variations in website design and content. Therefore, this study contributes to our understanding of tourists’ trust behavioral preferences, with the aim of enriching the research findings of trust models for OTDWs. Additionally, it contributes to the body of knowledge on tourism destination network information services conceptualization.

The power of web information expression means OTDW trust

The current study collected about 324 tourists’ questionnaire on the internet to demonstrate the impact of tourist trust in OTDWs. The results of this study also suggest that in addition to COL, LPD, and HC of CV, PB of PASP has the same impact on trust in OTDWs. Impacting on tourist use and trust of the website has been extensively discussed among academics in tourism. For example, some studies have indicated that incongruent CV is conducive to building a positive image of destination websites (Moura et al., 2015). Singh et al. (2003) reported that CV plays an important role in web content. Robin and Stephen (2013) argued that residents’ perceptions of the economic performance of government actors significantly predicted trust in government actors. Our research validated the direct impact of CV and PASP on tourist trust in OTDWs and provided a new perspective for the establishment of a positive destination through OTDWs. The findings broaden the scope and discussion of trust research in OTDWs and improve our understanding of the needs and preferences of the CV and PASP in OTDWs.

Practical implications

Through the research, we found that trust affects the use of OTDW, and that website design details affect trust. This study provides valuable insights for tourism managers to understand tourists’ preferences and needs in OTDWs trust, selection, and use. By understanding these differences, tourism officials and web designers can tailor OTDWs to enhance the trust and satisfaction of tourists, thereby driving increased tourism economic of the destinations.

For online official tourism destination platforms, there are two aspects of practical significance that can be summarized as follows. On the one hand, this study proposes an OTDWs trust model to help official tourism destination websites produce several optional plans to promote trust. The findings imply that OTDWs should provide distinctive CV and PASP supports to meet the requirements of various types of tourists. To increase the trust of domestic tourists, OTDW managers need to fully consider factors such as collectivism, low power distance, high context, and perceived benevolence of websites. On the other hand, this research supports the platform’s ability to guide potential tourists in considering traveling to the destination. Based on our results, local administrative departments may need to redesign their programmers and focus on the significant role played by collectivism images and high context, the growing significance of low power distance and perceived benevolence in tourists’ satisfaction with services, and specific circumstances that affect tourists’ impression of PASP. For example, our study found that perceived benevolence can influence tourists, so OTDWs can recommend local specialty restaurants for food lovers, provide transportation information for self-service travellers, and offer different travel planning guidance for tourists with varying needs. Therefore, it is imperative to consider similar CV for domestic tourists on OTDWs.

Limitation and future research

While this study makes significant contributions to both theory and practice, there are several limitations that can provide rich avenues for further research. First, although the study highlights the benefits of improving website trust, the findings may not be entirely generalizable due to the uniqueness of OTDWs. In the future, researchers can test the trust theoretical framework on several actual operating OTDWs, and collect adjusted operating data before and after to establish a complete OTDW trust research system. Second, the research sample used in the study may not be entirely representative of tourists as a whole and may not accurately reflect the diversity and variability of tourists in China. Therefore, the extent to which the sample used is similar to the tourists under study is limited. Finally, it is hoped that this study can stimulate the interest of other researchers and encourage tourism website managers to apply the trust research results to OTDWs. By doing so, tourists can increase their trust, alleviate ambivalence, increase the utilization rate of OTDWs, and fully leverage the marketing functions (Zhang et al., 2015).