Tourists’ Adoption of Extended Reality Technologies: A MetaAnalytical Structural Equation Modeling

TAM and its Applications in Extended Reality Technologies Research in the Context of Tourism

TAM was first proposed by Davis (1989) based on social psychology theories (e.g., innovation diffusion theory) to explain a user’s acceptance of a new technology (Koul & Eydgahi, 2018). TAM is composed of the following four constructs: “usefulness,” "perceived ease of use,” “attitude," and “intention to use.” Usefulness is defined as “the degree to which a person believes that using a particular system would enhance their job performance,” while perceived ease of use refers to “the degree to which an individual believes that using a particular system would be free of physical and mental effort” (Davis, 1985, p. 26), and attitude refers to a person forming favorable or unfavorable feelings toward adopting a certain technology (Kim, 2016). At the core of TAM (Davis, 1989) is the belief that both usefulness and perceived ease of use positively influence an individual’s attitude towards the use of a technology (perceived ease of use has a positive influence on usefulness, while usefulness has a positive influence on a consumer’s attitudes toward a technology). A consumer’s positive attitude relating to a technology leads to the “intention to use” a particular technology. Given the simplicity, understandability, and flexibility of TAM (Wang et al., 2003), the original propositions of TAM have been well tested in a range of contexts, including education (Granić & Marangunić, 2019), health systems (Rahimi et al., 2018), business (Vahdat et al., 2021), and tourism (Go et al., 2020). In addition to the original propositions, various external variables have been adopted by previous studies to extend TAM to achieve a higher explanatory power (Chi, 2018; Zhou et al., 2019). These include, self-efficacy (Eraslan Yalcin & Kutlu, 2019; Zheng & Li, 2020), technology anxiety (Zheng & Li, 2020), social norm (Eraslan Yalcin & Kutlu, 2019), social influence (Patel & Patel, 2018), distraction perception (Yang et al., 2021), and perceived enjoyment (Mathew & Soliman, 2021).

TAM has received considerable attention from tourism researchers in the areas of VR and AR (El-Said & Aziz, 2021; Huang et al., 2013; Lee et al., 2017; Park & Stangl, 2020; Wu et al., 2020). For example, drawing upon TAM, Disztinger et al. (2017), Jung et al. (2018), and Vishwakarma et al. (2020) examined which factors influenced tourists’ intention to use extended reality technology. The predictive power of the core TAM constructs (i.e., perceived usefulness, perceived ease of use and usage attitude) was confirmed. Apart from the core variables, their findings also revealed that some external variables like perceived immersion, perceived innovativeness, and social norms were also the influential factors. With respect to external variables, previous VR and AR research has often adopted perceived enjoyment as the external variable to extend the TAM model in the context of tourism. For example, Jung et al., (2018) and Vishwakarma et al. (2020) found that perceived enjoyment had a positive impact on usefulness and perceived ease of use when tourists employed AR/VR in experiencing destinations. Moreover, a review of TAM and its applications in VR and AR research in the tourism and hospitality context indicated that perceived enjoyment was the most frequently adopted external variable among the sample (in existing studies). Hence, the following hypotheses are framed:

VR and AR in the Tourism Context

VR is generally defined as a computer-generated 3D immersive environment in which users are able to navigate and interact, resulting in the impression of being in a real-world environment (Gutierrez et al., 2008; Guttentag, 2010). The three key elements that characterize VR are visualization, immersion, and interactivity (Yung & Khoo-Lattimore, 2019). Virtual Environments and Virtual Worlds are the most widely used terms in VR research (Yung & Khoo-Lattimore, 2019). The commonly accepted definition for AR is “the enhancement of a real-world environment using layers of computer-generated images through a device” (Guttentag, 2010; Jung et al., 2015). There are three distinctive characteristics of AR: a combination of real and virtual; realtime interaction; and three-dimensional (3D) registration (Azuma, 1997). By seamlessly blending computer simulations with real environments, AR creates an enhanced view that supplements the users’ environment with digital content by facilitating the integration of physical and virtual worlds (Flavián et al., 2019). The similarities and differences between VR and AR have been investigated by previous studies. As both VR and AR have to deal with virtual objects, Guttentag (2010) argues that AR is a type of VR. However, this opinion has been challenged by Milgram and Kishino (1994) who argue that AR and VR should be viewed as being located at opposite ends of a Reality-Virtuality continuum (from solely real word objects to solely synthetic or computer-generated objects). The greatest difference between VR and AR is the level of immersion from the perspective of the user (Yung & Khoo-Lattimore, 2019); with VR, the user is fully immersed into a virtual environment, while with AR, the majority of what the user sees is still the real world. Given these differences, the suitability of VR versus AR depends on the context. Virtual reality can be more entertaining for the consumer as it provides an immersive experience, for example, the use of VR devices to study the wine-making process during a wine-tasting session, or the creation of a 360-degree VR video on the topic of a recently taken trip (Flavián et al., 2019). A drawback for VR, however, is the lack of consumer interaction in comparison with AR (Farah et al., 2019); the use of AR applications has increased, in keeping with the development of mobile devices (Wedel et al., 2020). To counteract this, VR applications have often been used in contexts where increased flexibility for consumers has been required (Moorhouse et al., 2018).

VR and AR technologies are widely used in the tourism industry to enhance tourists’ experiences. It has been confirmed that tourists are more likely to make travel decisions when they are engaged in using VR and AR (Buhalis & Law, 2008; Fotakis & Economides, 2008), and, therefore, an increasing number of destinations and tourist attractions have started to provide fully interactive VR and AR experiences on their websites. VR and AR technologies have been broadly used for tourism destination marketing (An et al., 2021; Marasco et al., 2018) where they have provided tourists with the opportunity to experience a destination prior to their actual visit. Lagiewski and Kesgin (2017) found that exposure to digitally enabled experiences using VR and AR technologies increased the number of real visits. Taking into consideration the importance of VR and AR technologies in destination marketing, recent studies have explored design recommendations for VR and AR, thereby providing deeply immersive experiences for the user in manipulated design settings, such as hotel rooms (Siamionava et al., 2018), and the waiting areas of virtual reality restaurants (Hwang et al., 2012). Another trend that has gained momentum due to the increasing popularity of mobile internet and smart phones, is the use of mobile devices that can incorporate VR and AR technologies to reshape tourists’ experiences (Loureiro et al., 2020; Tussyadiah et al., 2018). Tourists already use their mobile devices to take advantage of VR and AR technologies to obtain information relating to destinations, restaurants, hotels, and cultural monuments (García-Crespo et al., 2016), and mobile devices using VR and AR technologies have been identified as an effective instrument in the development of smart cities and the protection of cultural heritage. The advantages of using VR and AR technologies in the tourism industry may be well recognized, but some researchers have included a note of caution. They warn that one of the greatest risks in the use of these technologies in the tourism industry is the potential for virtual travel to take the place of actual travel (Cheong, 1995; Guttentag, 2010).

VR and AR technologies are user-centered, and approaches to increase tourists’ experiences in the context of tourism, and tourists’ adoption of VR and AR technologies are one of the most studied topics in tourism research (Han et al., 2019; tom Dieck et al., 2018). Loureiro et al. (2020) identify that TAM has been widely employed to explore tourists’ adoption of VR and AR by reviewing 20 years of research on VR and AR in the tourism context. Although technology type has been identified as a powerful moderator of the relations among TAM constructs (Scherer et al., 2019; Šumak et al., 2011), there have been no studies to investigate the influence of extended reality type (VR and AR) when exploring tourists’ intentions to use extended reality technologies. Therefore, the aim of this study was to examine the moderating effect of extended reality type (VR and AR) on tourists’ intention to adopt VR and AR technologies based on TAM, and using a meta-analytic technique. Based on the above literature, the following hypothesis is proposed:

Cultural Differences in Extended Reality Technologies Use

Culture is defined as a set of values, beliefs, and norms shared by a particular group of people (Nakata & Sivakumar, 2001; Pizam & Sussmann, 1995). Extant literature has identified cultural differences as an influential factor in tourist behavior and tourism demand (Choi & Totten, 2012; Liu et al., 2021; Tam & Oliveira, 2019). Specifically, the moderating role of cultural differences in customer technology adoption has been identified in various contexts, such as mobile banking (Goularte & Zilber, 2020; Jadil et al., 2021), mobile commerce (Zhang et al., 2012), and mobile TV (Choi & Totten, 2012). Cultural difference has also been identified as an important moderator for tourists’ adoption of VR and AR technologies (Wei, 2019). Lee et al. (2015) and Jung et al. (2018), in particular, have highlighted cultural differences relating to the acceptance of mobile augmented reality by comparing the intentions of South Korean tourists to use mobile augmented reality with those of Irish tourists.

Previous studies have measured the effect of cultural differences on consumers’ intention to use technologies based on TAM (Alshare et al., 2011; Lin, 2014), but, to date, there have been few studies that have investigated cultural differences when examining tourists’ intentions to use extended reality technologies based on TAM. The previous studies were conducted in a variety of cultural settings, and it is therefore worthwhile to determine whether culture may lead to differences in direction and strength among the TAM relationships. Considering the influential role of culture on tourists’ intentions to adopt technologies, this study further investigated the impact of cultural differences on intention to use extended reality technologies in the tourism context based on TAM, using a meta-analytic technique. Based on the above literature, the following hypothesis is proposed:

Literature Search

A comprehensive search using the EBSCO Hospitality and Tourism Complete database and the Google Scholar search engine was performed to identify relevant literature. Google scholar, in particular, was chosen as a supplementary tool for article retrieval due to the presence of divergent types of publications (e.g., conference papers, book chapters) and, more importantly, to comply with a wish to include works that met the inclusion criteria, but were published in non-tourism/hospitality journals. The key terms included, but were not limited to “TAM,” “Technology Acceptance Model,” “Technology Acceptance” and “Tourism.” “Hospitality” and “Travel” were utilized to identify latent articles.

Inclusion Criteria

To achieve the research objectives, the papers included for final selection were required to meet the following criteria:

Study Selection and Coding

Taking the inclusion criteria into account and following the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA; Moher et al., 2009), the literature search and paper evaluation resulted in the retention of 29 studies with 32 independent samples. Specifically, the initial search generated 6,772 works (see Appendix A). A total of 4,192 articles were removed based on their titles, and duplicates were filtered out. The remaining studies were further screened based on the abstract, which led to the rejection of 441 studies. Finally, the full text of the latent samples was checked to ensure their eligibility for inclusion, yielding the final sample. It is relevant to note that in instances where more than one article applied the exact same search item, only one was retained.

A coding scheme was established to ensure consistency, and the coding of the included studies was undertaken by two independent coders using Microsoft Excel 2019. Any disagreement between the coders was resolved by double checking the original article and through discussion with other project team members. Specifically, for each study, statistical information was first coded (e.g., sample size, effect size, and reliability scores) with respect to the relations among the core TAM constructs. Any external factors that could explain variations in the core TAM variables were also documented. In addition, relevant features of the included studies were recorded for potential categorical moderation analysis, that is, the characteristics of the respondents (such as where the respondents were from) and the main topic (i.e., AR or VR).

Data Analysis

We followed Hunter and Schmidt’s (2004) random-effect meta-analytic procedures to estimate the true magnitude of correlations among the TAM constructs. Pearson correlation coefficients were employed as the effect size metric to aggregate the primary findings. To correct observed correlations of each included study for sampling error, sample size-weighted correlations were initially computed. Where possible, such correlations were further corrected for measurement unreliability using Cronbach’s alpha. However, in line with previous meta-analytic works (e.g., Afshardoost & Eshaghi, 2020), in cases where relations had fewer than three independent studies, the meta-analytic results were not calculated in order to avoid second-order sampling errors (see Hunter & Schmidt, 2004) if, for example, the study chose not to perform meta-analysis on the relations of actual VR/AR usage with other TAM constructs.

The 95% confidence interval (CI) of every estimate was calculated to check for statistical significance. Three heterogeneity examination means (i.e., standard deviation (SD), 80% credibility interval, and Q test) were adopted to detect the variation in the corrected effect size. According to Xu et al. (2020), a large SD and/or 80% credibility interval suggests the presence of heterogeneity. Statistically significant Q statistics also can indicate significant heterogeneity and the need for further moderation analysis. To examine the potential moderating effect of (1) the types of extended reality and (2) cultural differences, the included studies were split into subgroups based on the categorical moderators (AR vs. VR; Eastern culture vs. Western culture). Following the precedent of previous meta-analytic studies (e.g., Lim & Ok, 2021), categorical moderation analysis was conducted using the z test, as recommended by Hunter and Schmidt (2004), to ascertain whether the corrected mean effect sizes for the subgroups of individual works were significantly different from each other.

MASEM was then performed to examine the correlations among the constructs in an integrated model, allowing us to investigate the efficacy of TAM in predicting adoption of VR/AR in the tourism context. Following the two-stage MASEM approach introduced by Cheung (2015), a correlation matrix among the TAM constructs was first made based on the meta-analytic correlations. In the second stage, this was submitted to SEM analysis for the assessment of model fit and the estimation of structural standardized coefficients. In the present review, MASEM analysis was conducted respectively on the TAM model with the core constructs, as well as on the extended TAM model involving external variables.

Sample Description

A total of 29 eligible studies published in both tourism-related and non-tourism-related sources were captured, encompassing a total of 32 independent works. Of these, 23 were articles from 20 differing journals, five were conference papers and one was a book chapter. The studies included a total of 10,630 respondents; specifically, the sample size of the individual studies ranged from 101 to 1,042 (M = 333, SD = 243). The samples were conducted in 11 divergent regions (predominantly Korea K = 7, China K = 5 and the US K = 3). More descriptive information of the included studies can be found in Appendix B.

The earliest study detected on the chosen topic was that of Awang et al., 2009, with the number of relevant publications gradually increasing from 2009 through to 2021. There has been a dramatic rise since 2020, with over 50% of the samples used having been published after 2019. Regarding the research theme, half focused on VR with the remaining half focusing on AR. Of these, only two studies investigated the actual usage of AR/VR. All used cross-sectional research, with SEM being the most popular method of analysis.

Psychometric Meta-Analysis

A result of zero did not occur in the 95% confidence intervals of the meta-analyzed relations, and therefore each corrected correlation among the TAM constructs was deemed to be statistically significant, providing support for Hypotheses 1–7. Specifically, first, as exhibited in Table 1, all possible relations among the core TAM constructs were included in our psychometric meta-analysis, and each of these relations was examined at least nine times. The correlation between perceived ease of use and perceived usefulness, the core variables within TAM, was the most frequently investigated, and was also particularly strong (ρ = 0.65). Across studies, behavioral intention had a strong association with attitude (ρ = 0.56), followed by perceived usefulness (ρ = 0.54) and perceived ease of use (ρ = 0.50). A significantly larger unattenuated effect size was found for the perceived usefulness–attitude relationship (ρ = 0.55) than for that of perceived ease of use–attitude (ρ = 0.52).

OPEN IN VIEWER

Table 1.

Meta-Analysis of Correlations Within TAM.

	K	N	r	ρ	SDρ	95% CI	80% CR	Q
Relations Among the Core Constructs
EU-U	26	8450	0.59	0.65	0.24	[0.65, 0.69]	[0.35, 0.97]	940***
EU-A	11	4527	0.46	0.52	0.32	[0.50, 0.55]	[0.11, 0.93]	591***
EU-BI	18	4225	0.43	0.50	0.23	[0.48, 0.53]	[0.22, 0.79]	254***
U-A	12	4641	0.49	0.55	0.22	[0.52, 0.57]	[0.23, 0.86]	393***
U-BI	21	5526	0.48	0.54	0.22	[0.54, 0.59]	[0.28, 0.84]	366***
A-BI	9	3625	0.49	0.56	0.18	[0.53, 0.58]	[0.30, 0.81]	201***
Other Frequently Examined Relations
PE-EU	17	5464	0.60	0.68	0.14	[0.66, 0.70]	[0.49, 0.88]	257***
PE-U	18	5920	0.62	0.69	0.18	[0.69, 0.72]	[0.48, 0.93]	408***
PE-A	3	2076	0.59	0.67	0.15	[0.64, 0.70]	[0.52, 0.82]	55***
PE-BI	16	4197	0.56	0.62	0.17	[0.61, 0.65]	[0.41, 0.84]	221***
SI-BI	4	863	0.49	0.54	0.04	[0.49, 0.60]	[0.54, 0.54]	2
INN-BI	3	1507	0.56	0.66	0.22	[0.63, 0.72]	[0.38, 0.95]	114***
IMM-BI	3	854	0.35	0.42	0.12	[0.35, 0.49]	[0.29, 0.55]	11*

Note. K = number of studies, N = cumulative sample size, r = mean correlation, ρ = average corrected correlation, SD_ρ = standard deviation of ρ, CI = confidence interval, CR = credibility interval, Q = Q statistic, EU = ease of use, U = perceived usefulness, A = usage attitudes, BI = behavioral intention, PE = perceived enjoyment, SI = social influence, INN = personal innovativeness, IMM = perceived immersion.

*

p < .05, ***p < .001.

Among the external variables, perceived enjoyment, social influence, personal innovativeness and perceived immersion were relatively more frequently assessed in the included samples and were eligible (k ≥ 3) to be included in the meta-analysis. In particular, the relations of perceived enjoyment and the core TAM constructs were the most well examined. Among these, the correlations of perceived enjoyment with perceived ease of use (ρ = 0.68) and perceived usefulness (ρ = 0.69) showed relatively stronger effect sizes. The results also indicated that social influence (ρ = 0.54), personal innovativeness (ρ = 0.66) and perceived immersion (ρ = 0.42) also significantly correlated with behavioral intention. However, because only a limited number of studies (k ≤ 5) were found for these three pairwise relations, the results should be interpreted with caution. A list of all the external variables found in the included studies can be seen in Appendix C.

Moderation Analysis

As illustrated in Table 1, the 80% credibility intervals of most meta-analyzed relations were large and Q statistics for most correlations were significant, indicating the heterogeneity of the corrected effect sizes and the need for subgroup analysis. First, virtual technology type differentiated VR studies from the AR studies was examined as a moderator proposed in the present review. As demonstrated in Table 2, empirical works on AR generated stronger relations among the core TAM constructs compared with research on VR. Findings also indicated that the relations of perceived enjoyment with perceived ease of use, perceived usefulness, and behavioral intention were stronger in the AR sample as well. Statistically, the results of the z test revealed that a marginally significant larger effect size was found for perceived ease of use–perceived usefulness when the study had an AR-related research focus (VR: ρ = 0.58 vs. AR: ρ = 0.71, z = -1.78, p < .10), providing support for Hypothesis 8. Statistically significant z scores for other relations, however, were not found, and therefore the results for H8 (partially supported) should be interpreted with caution.

OPEN IN VIEWER

Table 2.

Results of Moderation Analysis—VR Versus AR.

TAM Relations	Total Number ofRaw Effects	Moderators		Difference(Z Test)
		VR	AR
Relations Among the Core Constructs
EU-U	25	0.58 (11)	0.71 (14)	−1.78˟
EU-BI	17	0.47 (10)	0.57 (7)	−1.24ns
U-A	20	0.56 (11)	0.59 (9)	−0.97ns
Other Frequently Examined Relations
PE-EU	17	0.64 (9)	0.72 (8)	−0.78ns
PE-U	18	0.69 (11)	0.72 (7)	−0.01ns
PE-BI	16	0.61 (10)	0.62 (6)	−0.15ns

Note. To avoid second-order sampling error, subgroup meta-analyses were only conducted on the relations with subgroup k >3. EU = ease of use, U = perceived usefulness, A = usage attitudes, BI = behavioral intention, PE = perceived enjoyment. ˟p < .10, ^nsp > .10.

The moderating role of different cultures was also explored, the results of which are shown in Table 3. Among the core TAM constructs, the perceived ease of use–perceived usefulness and perceived usefulness–behavioral intention relations were stronger in the Western cultural context, while the EU-BI relation was stronger in that of Eastern culture. With regard to the relationships between external variables and the core constructs, the perceived enjoyment–perceived usefulness and perceived enjoyment–behavioral intention relations were stronger in the Western cultural context, while perceived enjoyment–perceived ease of use was stronger in that of Eastern culture. Statistically, the results of the z test revealed that marginally significant larger effect size was found for perceived enjoyment–perceived usefulness when the study was conducted in the Western cultural context (Eastern: ρ = 0.62 vs. Western: ρ = 0.73, z = -1.30, p < .10), providing support for Hypothesis 9. Statistically significant z scores for other relations, however, were not found, and therefore the results for H9 (partially supported) should be interpreted with caution.

OPEN IN VIEWER

Table 3.

Results of Moderation Analysis—Eastern Versus Western.

TAM relations	Total Number ofRaw Effects	Moderators		Difference(Z Test)
		Eastern	Western
Relations Among Core TAM Constructs
EU-U	17	0.50 (13)	0.65 (4)	−0.72ns
EU-BI	12	0.56 (8)	0.42 (4)	0.57ns
U-BI	14	0.58 (9)	0.59 (5)	0.13ns
Other Frequently Examined Relations
PE-EU	13	0.66 (9)	0.63 (4)	0.37ns
PE-U	14	0.62 (8)	0.73 (6)	−1.30˟
PE-BI	13	0.60 (5)	0.64 (8)	−0.82ns

Note. To avoid second-order sampling error, subgroup meta-analyses were only conducted on the relations with subgroup k > 3. EU = ease of use, U = perceived usefulness, BI = behavioral intention, PE = perceived enjoyment. ˟p < .10, ^nsp > .10.

Meta-Analytic Structural Equation Modeling

Meta-analytic structural equation modeling of the core TAM constructs

The efficacy of TAM in predicting user adoption of VR/AR in the tourism context was tested via MASEM, using the meta-analytically derived matrix as input (see Appendix D). Findings showed that the basic TAM model exhibited acceptable goodness-of-fit: χ2(1) = 113.8, p < .001, goodness-of-fit index (GFI) = 0.99, comparative fit index (CFI) = 0.98, normed fit index (NFI) = 0.98, Incremental Fit Index (IFI) = 0.98, root mean square residual (RMR) = 0.03. Altogether, the core TAM constructs explained 35% of the variance in usage attitudes toward VR/AR and 39% of the variance in behavioral intention to use VR/AR, and all the paths were positive and significant (see Figure 1).

OPEN IN VIEWER

Figure 1.

MASEM-The Core TAM Model.

The coefficient results indicated that attitude had a positive association with behavioral intention, with a moderate effect size (β = 0.38, p < .001). Both perceived ease of use and perceived usefulness were positively linked to usage attitudes; however, while the correlation between perceived usefulness and attitude exhibited a moderate effect size (β = 0.37, p < .001), the magnitude of the perceived ease of use–attitude relation was weak (β = 0.28, p < .001). Perceived usefulness was also found to have a direct impact on behavioral intention and the effect size was moderate (β = 0.33, p < .001). In addition, consistent with the zero-order meta-analytic results, perceived ease of use had a strong effect size in its prediction of U (β = 0.65, p < .001).

Meta-analytic structural equation modeling of the extended TAM model

Among the external variables, perceived enjoyment was eligible to be included in the MASEM analysis of the extended TAM model, due to the fact that each of the relations of perceived enjoyment with the core TAM variables were examined at least three times by the individual studies. Accordingly, as shown in Figure 2, to examine the effect of external variables, the core TAM model was extended by one external variable, namely users’ perceived enjoyment with VR/AR.

OPEN IN VIEWER

Figure 2.

MASEM-The Extended TAM Model.

The extended TAM model for adoption of VR/AR in the tourism context also demonstrated a moderate fit to the meta-analytic correlation matrix (see Appendix D): χ2(3) = 407.5, p < .001, goodness-of-fit index (GFI) = 0.91, comparative fit index (CFI) = 0.89, normed fit index (NFI) = 0.89, Incremental Fit Index (IFI) = 0.89, root mean square residual (RMR) = 0.08, and all the paths were positive and significant, providing additional support for Hypotheses 1–7. In particular, and consistent with the zero-order meta-analytic results in the extended model (see Table 1), perceived enjoyment was found to be a stronger antecedent of perceived ease of use and perceived usefulness (perceived enjoyment–perceived ease of use: β = 0.68, p < .001; PEU: β = 0.46, p < .001). The inclusion of perceived enjoyment decreased the path coefficient linking perceived ease of use and perceived usefulness (Figure 1: β = 0.65; Figure 2: β = 0.34), and the impact of perceived enjoyment on perceived usefulness (perceived enjoyment–perceived usefulness: β = 0.46, p < .00) demonstrated larger effect size compared to the influence of perceived ease of use on perceived usefulness (perceived ease of use–perceived usefulness: β = 0.34, p < .00).

Theoretical Implications

This study provides the first meta-analytic review of TAM research in explaining tourists’ intention to use extended reality technologies. It contributes to existing literature in several ways. First, the findings of the present study can be used to resolve the contradictions and inconsistencies evident in previous works based on TAM. It supports the original propositions relating to causal relationships within TAM. In line with previous TAM findings (Kim, 2016; Song et al., 2021), this study also found that perceived usefulness and perceived ease of use played significant roles in the formation of attitude which, in turn, led to behavioral intention. Perceived usefulness also had a directly positive influence on behavioral intention. The positive influence of perceived ease of use on perceived usefulness was also identified and it was confirmed that attitude has a positive influence on behavioral intention. In comparison with MASEM undertaken in other fields (Scherer & Teo, 2019; Scherer et al., 2019; Yousafzai et al., 2007), this study found that when explaining the intention of tourists to use extended reality technologies, “perceived ease of use” had a stronger effect on perceived usefulness.

Second, a series of subgroup analyses revealed that the actual types of extended reality technologies and cultural differences to some extent moderated the relationships within TAM. Previous MASEM studies have found that the moderating effects of technology type on TAM constructs depend on the study context (Scherer et al., 2019; Scherer & Teo, 2019). Given the differences, it is worthwhile to explore the moderating role of extended reality technologies on TAM constructs in explaining tourists’ intention to use extended reality technologies. This study found that the pair-wise relationship of perceived ease of use–perceived usefulness was significantly moderated by the “extended reality” type, and that the effect size of perceived ease of use–perceived usefulness was larger in the VR context than in that of the AR context.

Previous MASEM studies also found that the moderating effect of cultural differences on TAM constructs depend on the study context (Scherer & Teo, 2019; Schepers & Wetzels, 2007). Therefore, it is of value to explore the moderating role of cultural differences on TAM constructs in explaining tourists’ intention to use extended reality technologies. This study found that the effect sizes of perceived ease of use–-perceived usefulness and perceived usefulness–behavioral intention were greater in the Western cultural context than in that of Eastern culture, while the relationship between perceived ease of use and behavioral intention was stronger in the Eastern cultural context than in that of Western culture. More importantly, the z test demonstrated that the relation between perceived enjoyment and perceived usefulness was significantly stronger in the Western context. In comparison with preceding studies exploring the moderating effects of types of technology (e.g., Scherer et al., 2019; Šumak et al., 2011) and culture (e.g., Choi & Totten, 2012; Zhang et al., 2012), this study is the first to use the TAM theory to examine the moderating role of types of technology and cultural differences on tourists’ adoption of extended reality technologies. The potential moderating influences on the relationships among the TAM constructs could be taken into consideration for future empirical research.

Third, this study examined the causal relationships within TAM through the use of MASEM. Although some inconsistent results associated with the relationships among the TAM constructs have been identified by the extant literature, based on robust MASEM findings, this study confirms and further supports the original notion of TAM as proposed by Davis (1985; 1989). Therefore, the results from the study’s MASEM strengthens the applicability of TAM in predicting tourists’ adoption of extended reality technologies. To achieve a higher explanatory power, this TAM model has been extended by external variables. For example, self-efficacy, subjective norms, and facilitating conditions of technology use, were adopted by Scherer et al. (2019) to explain teachers’ adoption of digital technology. Therefore, perceived enjoyment as one of the most frequently used variables to extend TAM in the context of the use of extended reality technologies, was adopted by this study. In the extended version of TAM, perceived enjoyment was found to be a significant antecedent of perceived ease of use and perceived usefulness; MASEM results of this study further revealed that perceived enjoyment played a more important role in explaining tourists’ perceptions of the usefulness of extended reality technologies in comparison with perceived ease of use.

Practical Implications

Extended reality technologies are playing an increasingly important role in the tourism and hospitality industry, and tourism and hospitality practitioners are increasingly concerned with answering questions around how to predict the behavioral intentions of tourists associated with the use of extended reality technologies. This meta-analysis provides a range of practical implications for tourism and hospitality industries. First, findings revealed that the correlation between perceived ease of use and perceived usefulness demonstrated relatively stronger effect size among the paths in TAM. Tourism and hospitality industry practitioners could attempt to provide more user-friendly VR and AR technologies to enhance tourists’ perception of the usefulness of the technologies, and ease-of-use should be the first thing considered by tourism and hospitality practitioners when adopting extended reality technologies. Easy-to-understand documentation could be provided for tourists who already have experience with extended reality technologies, and a more specialized approach could be used to educate tourists with no extended reality experience. Second, given the significant role that tourists’ attitudes have in shaping extended reality technologies use intention, tourism and hospitality industry practitioners would be wise to pay more attention to the antecedents of tourist attitude, particularly to perceived usefulness and perceived ease of use. Specifically, along with making it easier for tourists to use VR and AR technologies, tourism and hospitality practitioners should also consider enhancing the effectiveness of extended reality technologies to improve actual tourism experience, thereby boosting tourists’ favorable attitudes towards using the VR and AR technologies. Tourism and hospitality industry practitioners should test the usefulness of an extended reality technology before actually putting it into practice, and those that increase tourists’ perception of quality of service, should adopted by the industry. Third, the findings of the meta-analysis indicated that the pair-wise relationship of perceived ease of use–perceived usefulness was, to some extent, moderated by both virtual technology type and cultural differences. Destination practitioners could therefore develop divergent marketing strategies for VR and AR users as well as for different cultural groups. For example, as ease of use was found to be more important for VR users than AR users, tourism and hospitality industry practitioners should attach more importance to ease of use when providing extended reality technologies for AR users. Perceived enjoyment was found to be a strong antecedent of perceived ease of use and perceived usefulness; therefore, the importance of regularly evaluating and reviewing the user experience to improve individuals’ level of cognitive playfulness and entertainment should not be overlooked. In particular, practitioners may wish to embed more interactive and immersive elements into extended reality technologies for Western tourists, given that the influence of perceived enjoyment on perceived usefulness was found to be significantly stronger for Western tourists than for Eastern tourists.

Limitations and Future Research

There are some limitations to the present meta-analytic review. First, only studies written in English were included and this may influence the generalizability of the findings. Studies in other languages could be included for future review studies to address this deficit. Second, as most of the studies included in the meta-analysis were cross-sectional works with self-reported data, common method bias may be an issue. Third, perceived enjoyment was the only factor considered in the extended TAM model and, therefore, other external factors were not captured in the current review (see Appendix A). Alternatively, those constructs that were less investigated could be considered by future research to examine how they either improve or impair the predictability of TAM in forecasting tourists’ intentions to use extended reality technologies. Finally, this study only explored the moderating roles of technology type and cultural differences; other potential moderators (e.g., gender and sample size) could be explored by future research.