Tourism, especially international tourism, has been recognised to contribute to the economy in various ways. Statistics show that tourism and its associated sectors account for 10% of the global GDP. For example, the tourism industry is directly linked with travel agents, airlines, tour operators, local economies, and maintenance of the travel destinations [19]. The outbreak of coronavirus disease (Hereafter, COVID-19) has led to a severe hit on the tourism industry due to the lockdowns in different countries and social distancing restrictions, causing a 6% negative impact on the global GDP [20]. The tourism industry has not only amplified the impacts of tourism on the world’s economy, but it also indicates how important the tourism industry is to many sectors [21], [22]. For example, COVID-19 has led to a monetary loss of between 60% and 80% in revenue from international tourism (USD 910 billion to USD 1.2 trillion). Monetary loss. And around 100 to 120 million tourism-associated jobs at risk [23]. Based on the economic impacts of the tourism industry, travel destinations need to reconsider their marketing strategies to enhance tourists’ visit intentions in order to contribute to the recovery of the world’s economy. Zheng et al. [4] indicate that the promotion of a travel attraction can largely stimulate consumption and promote regional economic development.

The adoption of digital technologies in the tourism industry has been suggested by the previous research [24]. Advanced digital technologies contribute to the tourism industry in various ways. Examples include using robots or chatbots in the hotel or airport to handle parts of the service process, using VA to provide useful travel information to tourists, and using virtual reality to help tourists see and experience an attraction without travelling to the site [25]. While there has been some research on service delivery, promoting virtual or actual travel using advanced digital technologies [26], there is a lack research focussed on how advanced digital technologies can be used to promote and attract tourists to search for more attraction information (Showing an interest in the travel attraction that is been promoted) and to encourage the intention to visit the travel destination (Actually wanting to visit that attraction) more effectively and efficiently [26]. Such an issue is important, as it can indirectly facilitate the recovery of the world’s economy by arousing tourists' interests in visiting the attraction [4].

A VA is a digital assistant that uses voice recognition, language processing algorithms, and voice synthesis to listen to specific voice commands and return relevant information or perform specific functions as requested by the user. VA is one of the digital technologies being popularly used in the tourism industry, particularly in Western countries. It is easily accessible for digital-gadget users, and it provides opportunities for tourists/practitioners to search for/promote attraction information or purchase/sell tickets to an attraction online [3]. VA offers tourism managers a new channel to promote and communicate their travel destination more effectively with more than 4.2 billion VA users across the globe [1]. Previous research has recognised that VA can be involved in tourists’ travel planning processes [27], and, from here, it can engage tourists in order to motivate them to search for more information and make a travel decision [28]. Such a planning process can also happen during tourists’ travel. For example, some tourists may need to decide whether they would go to a restaurant, museum, attraction, or other recreational activity after they have landed at their destination. In this case, VA can be used to provide other tourists’ reviews and background information about the attraction [29].

Providing more information to tourists can be crucial to influencing tourists’ decision-making because tourists’ visit intentions can be affected by tourists’ subjective knowledge about the travel destination [30], which could be formed by the information provided by the VA. With the popular use of social media and other user-generated content (Hereafter, UGC) by tourists to obtain information [29], VA can provide more relevant and detailed information about a travel destination. Such relevance, which ultimately correlates to the persuasiveness of the information [31], depends on different sources of UGC [32].

Social identity theory argues that information provided by people within the social network (i.e., in-group) would be more persuasive than that provided by people outside of the social network (i.e., out-group) because the identity salience fosters processing identity-relevant information, i.e., schematic or abstract processing, by the individual [7], [28]. That is, individuals tend to be more receptive to information provided by the in-group compared to that offered by the out-group. The social self-schema explains that when an individual is exposed to in-group information, there is increased processing of the information because it is more relevant to their self-identity. In this scenario, the individual tends to be more favourable to and pay more attention to information that is relevant to the group [7].

On the other hand, previous research shows that individuals are less likely to adopt suggestions from listening to information provided by out-groups [8]. Individuals have less tendency to adopt recommendations from listening because the out-group information communicated by the VA is less relevant to the individuals’ social self-schemas, and they are less likely to process the information explicitly. Consequently, tourists may be more likely to be inspired by the in-group information communicated by the VA during their travel compared to out-group information. Thus, when the VA refers to the review or information from the tourists’ in-group VA users (As opposed to out-group VA users), it may result in the tourist being more willing to listen to the information.

Although in-group recommendations seem to be a more convincing VA information source, social identity theory provides a foundation for this view, and its underlying social self-schema suggests an opportunity for VA to make a more persuasive message to tourists, recent studies [9] show that the influence of in-group and out-group information sources on consumer behavioural intention can vary due to individual perceptions. Some consumers are found to be more willing to take out-group suggestions instead of proposals made by in-group members. The section below explains how the sources of information provided by the VA may become an antecedent to tourist inspiration and how the inspiration may influence tourists’ behavioural outcomes.

Inspiration is a feeling state that can motivate tourists to put their received information into action [10]. Inspiration can be explained by the Cognitive Appraisal Theory of emotions (CAT), which involves a process of eliciting inspiration and behavioural outcomes after an individual experiences inspiration [33]. In other words, inspiration can be illustrated by cognition of an incoming stimulus, emotional response towards the stimulus, and motivation of action [34]. With the use of CAT, a customer inspiration model has been developed by Bottger et al. [9].

To further explain the model with reference to CAT, inspiration has been conceptualised by three components or stages: Evocation, transcendence, and motivation [35]. These explain the cognition-emotion-action process of the CAT. First, evocation explains how inspiration can be evoked [36]. Evocation may involve an external incoming stimulus being exposed by tourists, for example, an attraction or other information that tourists are exposed to during their travel. Second, transcendence illustrates the emotional response component of inspiration, which involves feelings of positivity, clarity, and self-enhancement [9]. In the context of tourism, transcendence involves mainly positive emotional responses associated with inspiration, such as the joy over receiving useful information from the VA or the excitement the travel experience generates towards an attraction [37]. Finally, motivation describes the action state of the tourist after being inspired by the cognition of the incoming stimulus and its associated motivational elements of the emotional response [38].

Looking further into the concept of inspiration, it is also comprised of two sub-components that explain the mechanism of inspiration using the concepts of evocation, transcendence, and motivation. These two components illustrate that inspiration is an intrinsic motivational state [38]. ‘Inspired by’ explains the activation component and ‘inspired to’ delineates the behaviour and intention component. These two components are considered as second-order constructs in the model, as they can illustrate the full episode of inspiration while distinct from the aforementioned cognition-emotion-action process [9].

Cohen and Garcia [44] propose the concept of collective threats, which is about the risks related to a group of people when making decisions. People make decisions differently when they face a threat as a group [4]. Some research suggests that people act more selfishly when they face a threat collectively as a group [45], whereas other research argues that people may be more cooperative with other people by sharing resources and advice when they are facing collective threats [46], [47].

In the context of tourism, threats can be commonly seen when tourists are traveling both internationally and domestically. For example, tourists may not know the area where they stay, be familiar with the transportation system, or the language etc. The current research focuses on collective threats that are related to a group of tourists. The focus is on collective threats because many tourists travel as a group, regardless of the purpose of their travel. For instance, a group of employees travel abroad for a business trip, family members travel together for a family trip or foreign tourists travel together for a packaged trip. The concept of severity of collective threats has been applied to the current study to refer to the degree of urgency or danger that the tourists face during their travel [4].

When a group of tourists is under more severe threats and in a more intense situation during their travel, they may be more likely to resolve the threats as a collective group [12]. According to the social identity theory, when a group of tourists is under severe collective threats, they are more likely to listen to the advice provided by in-group members, as they trust the advice communicated by the people close to and similar to themselves [13]. Similarly, the construal-level theory also indicates that people adopt a more abstract information processing style when they can refer to information that they can trust [48]. For example, when a group of tourists have a very limited time and are not able to visit all the planned travel destinations during the trip on the final (Versus first) travel day, they tend to seek advice available from their in-group (Out-group) members, such as people in the same online community or sharing similar travel preferences.

Collective threats are expected to change the source-characteristics aspect of the inspiration model in relation to how tourists may be inspired by in-group (Versus out-group) information provided by the VA. In particular, tourists in a more severe collective-threats situation during their travel are more likely to be inspired by social self-schema-consistent information. Consequently, VA in-group travel information may be more effective and persuasive when tourists are under more severe collective threats through the elicitation of an inspiration state. On the other hand, when the severity of collective threats is low (Compared to high), individuals are more likely to be affected by out-group peers. Therefore, we propose:

H3a: When tourists are under more severe collective threats, they are more likely to be inspired by the in-group travel information provided by the VA.

H3b: When tourists are under less severe collective threats, they are more likely to be inspired by the out-group travel information provided by the VA.

The inspiration model and CAT explain that the elicitation of inspiration depends not only on the incoming stimulus, but on an individual’s subjective information processing of the incoming stimulus [9], [49]. The implicit beliefs explain the differential interpretation, prediction, and control of different people in the social world. The beliefs of different people are kept deeply in people’s minds [50], affecting their decision-making and information decoding processes. Previous research indicates that implicit beliefs can influence how individuals understand and interpret things based on their human nature [51], [52]. Some studies also suggest that implicit beliefs are associated with an individual’s malleability towards different things in the world, including products and services [53], [54].

There are two types of people based on which of the belief systems they hold: incremental-oriented and entity-oriented beliefs. The two show different perspectives on personal goals, values, and characteristics. Individuals who possess an incremental belief system believe that their abilities and characteristics are malleable through learning [15]. They focus on improving themselves through self-development and gaining experience. Incremental tourists perceive that making mistakes is part of their learning process. Therefore, they are willing to take risks even if the mistake may make them look unsmart [16].

Previous research [55] indicates that individuals with incremental beliefs usually emphasise the consistency of the schema less because they adopt a more concrete information processing style. Based on this learning pattern, incremental tourists are likely to adopt a concrete processing style on the out-group travel information provided by the VA, which allows them to learn more than the in-group or social self-schema consistent information. This links with the inspiration model in that the incremental tourists are more likely to be inspired by out-group information than by in-group information, enhancing the likelihood to be inspired by out-group travel information.

The inspiration model is indicated considering their concrete information processing and their favourability of learning from schema-inconsistent information. Based on the characteristics of incremental belief, individuals with this belief system are more likely to be receptive towards out-group travel information provided by the VA and perceive such information as a way for them to learn and improve their knowledge. H4a: Tourists with incremental beliefs are more likely to be inspired by the out-group than in-group travel information provided by the VA.

Entity individuals believe that their abilities are stable and fixed [15]. Individuals with entity belief focus on signalling and proving their positive abilities to others but hiding their negative abilities in order to exhibit their ‘good sides’ to others [17]. Hence, these individuals are less likely to take risks when making decisions. In terms of perceiving travel information, individuals with entity beliefs are likely to rely on their initial trait information when making causal attributions or subsequent judgements [53].

The research conducted by Flaherty and Pappas [55] confirms that individuals with entity beliefs are more likely to adopt an abstract processing style in which an overall picture is evaluated, looking at schema-consistent information, which allows them to prove their performance. Based on this, tourists with entity beliefs are more likely to be inspired by the in-group travel information provided by the VA because it is social-self-schema consistent, which fits into their abstract processing style. Thus, this proposes:

H4b: Tourists with entity beliefs are more likely to be inspired by the in-group than out-group travel information provided by the VA.

Furthermore, an individual’s implicit belief is also expected to affect the moderation effect of perceived severity of collective threats on the direct impact of source of the VA recommendation message on tourist inspiration. Tourists who have an incremental belief system learn from the process of trial and error. When the collective threats that they face are perceived to be more likely to influence the results, they are more likely to take a risk and be open to both in-group and out-group suggestions. After all, incremental tourists learn more through these intense experiences that can help them improve. Thus, they are equally inspired by both in-group and out-group messages from their VA devices. When they perceive a low collective threat, out-group suggestions can widen their thoughts and help them think outside the box.

On the other hand, entity tourists prefer making a reasonable decision without taking significant risks. Therefore, when they perceive a high collective threat, they tend to be conservative, and their inspiration is likely to be evoked by in-group recommendations. However, when they perceive the collective threats are not that serious, the possibility of making a regrettable decision is low, and they are more willing to see how people outside their circle are making choices.

For tourists with incremental beliefs:

I. H5a: When they perceive a less severe collective threat, they are more likely to be inspired by the out-group than in-group travel information provided by the VA.

II. H5b: When they perceive a more severe collective threat, there is no difference between their inspiration from VA messages provided by out-group or in-group sources.

For tourists with entity belief:

I. H5c: When they perceive a more severe collective threat, they are more likely to be inspired by the in-group than out-group travel information provided by the VA.

II. H5d: When they perceive a less severe collective threat, there is no difference between their inspiration from VA messages provided by out-group or in-group sources.

The theoretical framework is presented in Fig. 1.

The purpose of Study 1 is to examine the effect of VA social identity prime recommendation message sources on tourist inspiration and subsequent behavioural and attitudinal outcomes in the context of different levels of tourists’ perceived collective threats. A two (VA social identity prime recommendation message source: In-group vs. out-group peers) by two (Severity of collective threats: More vs. less) between-subject scenario-based experiment was designed. The experimental website was built with the aforementioned combined scenarios for the study. The between-subject experimental design removes the possibility of fatigue among the respondents when answering the scaled questions, as well as any biased opinions, by allowing them to view more than one scenario.

In Study 2, an individual’s implicit belief is expected to moderate the two-way interaction effect of the VA recommendation source and the severity of collective threats on tourist inspiration. A two (VA social identity prime recommendation message source: in-group vs. out-group peers) by two (Perceived Severity of collective threats: More vs. less) by two (Implicit belief: Incremental vs entity) scenario-based experiments (with the same structure but different treatments for implicit beliefs) were conducted.

An independent samples t-test revealed evidence to support H1. In-group VA information sources (M=5.43, SD=1.44) exhibited a higher Tourist Inspiration compared to out-group VA information sources (M=4.21, SD=1.21; t(178)= 14.37, p < .01). After conducting a median split on Tourist Inspiration (Mdn = 4.29), a Chi-squared analysis unveiled a significant association between a heightened level of tourist inspiration and an increased likelihood of seeking additional information (X2 = 14.92, p < .01; Table 1). Therefore, H2a was supported.

Searched for more Did not Search for more Attraction Information Attraction Information High inspiration 68 (Exp. count: 55.3) 25 (Exp. count: 37.7) Low inspiration 39 (Exp. count: 51.7) 48 (Exp. count: 35.3)

Supporting H2b, a linear regression analysis indicated that tourist Inspiration had a positive influence on an individual’s willingness to visit the recommended attraction (β = .59, SE = .04, p < .01; R2 = .51).

On the other hand, after conducting a median split on severe collective threats (Mdn=4.12), a Two-way ANCOVA (Gender, age, education level were controlled as covariates) demonstrated a significant two-way interaction effect between collective threats and information sources (F(1, 173) = 21.41, p < .01, Table 2, Fig. 2). Particularly, when tourists faced more severe collective threats, they were more likely to be inspired by the in-group travel information provided by the VA (M more collective threats + in-group = 5.92, SD = 1.10 vs. M less collectivethreats + in-group= 4.21, SD = .84). Contrariwise, when tourists faced less severe collective threats, they were more likely to be inspired by the out-group travel information provided by the VA (Fig. 2). The findings provided support for H3a and H3b.

Supporting H4a and H4b, a two-way ANCOVA revealed that different implicit beliefs among travelers significantly interfered with the impact of VA recommendation sources on inspiration (F(1, 353) = 15.21, p < .01, Table 3). Travelers with incremental beliefs were more prone to finding inspiration in out-group travel information provided by the VA (M = 5.72, SD = 1.02) as opposed to in-group information (M = 4.66, SD = .83). Conversely, individuals with entity beliefs were found more inclined to be inspired by in-group travel information recommended by the VA (M = 5.51, SD = 1.14) rather than out-group information (M = 4.02, SD = 1.21). See Fig. 3.

In addition, for individuals with incremental beliefs, when they perceived a lower level of collective threat, they were more likely to be inspired by the external group’s travel information provided by the VA (M = 5.57, SD = .98), rather than the internal group’s information (M = 4.45, SD = 1.23; t =21.33, p < .01). However, when they perceived a higher level of collective threat, there was no difference found in their inspiration from the external (M = 4.71, SD = 1.43) or internal (M=4.52, SD = 1.20; t = 2.34, p = .53) group’s information provided by the VA (Fig. 4). Thus, H5a and H5b were both supported.

For individuals with entity beliefs, when they experienced a higher level of collective threat, they were more likely to be inspired by the internal group’s travel information provided by the VA (M = 5.53, SD = 1.42; t =28.12, p < .01), rather than the external group’s information (M = 4.02, SD = 1.10). When they perceived a lower level of collective threat, no difference was found in their inspiration from the external (M = 4.31, SD = .91) or internal (M = 4.14, SD = 1.21; t = 1.77, p = .65) group’s information provided by the VA (Fig. 5). H5c and H5d were both supported.