How tourism activity shapes travel experience sharing: Tourist well-being and social context

Why people conduct different sharing about their travel is unclear. Understudied areas include the roles of tourism activity type, tourist well-being, and social context. Under the framework of construal level theory, three studies which combined secondary data and experiments re- vealed that: 1) challenging (relaxing) tourism activities lead to more desirability (feasibility) sharing; 2) eudaimonia (hedonia) occupy the dominant position and mediate the relationship between challenging (relaxing) tourism activity and desirability (feasibility) sharing; and 3) social context induces the transformation of the relationship between eudaimonia and hedonia, and has a signi ﬁ cant moderating impact on the mechanism of travel experience sharing type. Theoretical and managerial implications of travel experience sharing type and mutual transfor- mation between eudaimonia and hedonia are discussed.


Introduction
In the age of social media, people are affected by others' shared experiences, and are willing to share experiences that may affect others (Dedeoğlu et al., 2020). Sharing experiences from tourism activities is very popular on social platforms (Kock et al., 2020;Oliveira et al., 2020). Tourists have a strong motivation to look for suggestions and insights from others' shared experiences to reduce risk and uncertainty (Cheng et al., 2019), because tourism activity cannot be physically displayed or inspected in advance unlike material consumption (Kirilenko et al., 2021); and they also have a strong willingness to share their travel experiences (Wong et al., 2020) as tourism activities enable them to break away from the constraints of daily life (Sarial-Abi et al., 2020) and experience a very different life from their everyday routines (Filep & Laing, 2019;Xu & Zhang, 2021). Travel experience sharing has played a very important role for tourist travel and had a profound impact on the tourism industry (Chen et al., 2020;Kim & Fesenmaier, 2017;Kirilenko et al., 2021;Wong et al., 2020).
Travel experience sharing has also received great attention from academic tourism research (Sotiriadis, 2017;Wu & Pearce, 2016). However, most of these studies are characterized by the assumption that travel experience sharing is a homogenous activity (Ring et al., 2016). The different types of shared content have yet to receive sufficient academic attention (Dedeoğlu et al., 2020). Two types of shared content can be found that are significantly different: 1) desirability sharing which mainly expresses tourists' thinking and aftertaste related to a deeper meaning of tourism activities (Ferguson & Veer, 2015), and 2) feasibility sharing that focuses on the description of the scenery, food, and the process of travel (Brochado et al., 2019). However, why people share different content of travel experiences is still unclear. It has been found that tourists' post-travel behaviors are associated with the types of tourism activity (Mumuni & Mansour, 2014), tourists participating in different types of activities have different intentions to share travel experiences (Su, Cheng, & Swanson, 2020). However, the effect of tourism activity type on travel experience sharing type has yet to be studied.
Previous studies have identified a variety of different activity-based market segments (Mumuni & Mansour, 2014), for example, challenging activity versus relaxing activity (Mehmetoglu, 2007;Su, Cheng, & Swanson, 2020). Accordingly, the type of tourism activities sought after and engaged in by a tourist may be associated with different subsequent behaviors (Chen et al., 2016;Mumuni & Mansour, 2014). For instance, challenging tourism activities have been noted to be particularly emotionally, cognitively, and often physically engaging (Holm et al., 2017). Thus, challenging tourism activities have a stronger positive effect on travel experience sharing than relaxing tourism activities have (Su, Cheng, & Swanson, 2020). However, whether different types of activities lead to different types of sharing, for example, tourists will tend to engage in desirability sharing or feasibility sharing after participating in challenging tourism activities, requires further research.
Tourists may conduct their sharing directly when they participate in tourism activities, but there may also be an intermediate process from participation in tourism activities to experience sharing. This process may be a sense of well-being stimulated by tourism activity (Oliveira et al., 2020;Wu & Pearce, 2016). However, most studies investigating the mediating role of tourist well-being took a hedonic approach which focuses on immediate sensory pleasure, happiness, and enjoyment (Lengieza et al., 2019;Smith & Diekmann, 2017), while neglecting the eudaimonic approach which focuses on the consequences of self-growth and self-actualization (Rahmani et al., 2018;Su, Tang, & Nawijn, 2020). Additionally, previous studies indicated that different types of tourism activity could impact tourist hedonia and eudaimonia differently (Rahmani et al., 2018;Su, Tang, & Nawijn, 2020), and may induce different types of travel experience sharing. Thus, how different types of tourism activity influence travel experience sharing type through tourist eudaimonia and hedonia is valuable to discover.
Consumers are very sensitive in an unfamiliar environment, and their emotions will be affected by the social context (Hofmann et al., 2015;Huang & Wang, 2014;Wenzel & Benkenstein, 2018), thus tourist eudaimonia and hedonia might be influenced by social context in addition to tourism activity. In tourism, it is common to travel with friends (joint social context) or alone (solo social context) no matter what type of tourism activity (Su, Cheng, and Swanson, 2020a). Su, Cheng, and Swanson (2020)) have identified that in a joint context (i.e., traveling with friends), challenging tourism activities will have a significantly larger effect on emotional arousal, relative to a relaxing tourism activity; but in a solo context (i.e., traveling alone), there will be no significant difference in emotional arousal based on activity type (challenging or relaxing). However, they did not consider what kinds of emotions are specifically aroused. Whether the social context effect would be different in different tourism activities, especially in the context of tourist eudaimonia and hedonia, is an intriguing and academically relevant question that has not been explored.
The aim of the paper is thus to fill the above gaps in the literature by exploring the effect of tourism activity type (challenging and relaxing) on travel experience sharing type (desirability and feasibility), using the theory of construal level as the research framework. Three studies that combined secondary data and experiments were conducted to examine: 1) whether the tourism activity type influences travel experience sharing type using secondary data; 2) the impact of tourism activity type on travel experience sharing type and the mediating role of tourist eudaimonia and hedonia using experimental data; and 3) the moderating role of social context on the proposed relationships using experimental data. The findings of this paper contribute to the knowledge on tourist well-being and travel experience sharing, and provide timely suggestions to the tourism managers on how to take advantage of the experiential information shared by tourists via social media.
The type of travel experience sharing: Desirability and feasibility Although tourists have many options to share their travel experiences (Wu & Pearce, 2016), the shared content has remained relatively similar, and can be roughly divided into two types: one reflects travel results that focus on the impact of travel activities on personal growth, goal achievement, and meaning in life, and mainly expresses tourists' thinking and aftertaste of the deeper meaning of travel activities (Oliveira et al., 2020); whereas the other is associated with the travel process that focuses on the description of the scenery, food and the process of tourism, mainly introduces what tourists have seen and heard (Wu et al., 2020).
In general, when people decide to share their travel experiences, they would choose the things that are valuable to them in their travel journey (Wong et al., 2020). Previous studies (e.g., Liberman & Trope, 1998;Liu, 2008) indicated that the value of something could be distinguished between two types associated with the actions being decided: desirability and feasibility. Desirability refers to the value of the end state of an action, whereas feasibility refers to the process of reaching the end state (Liberman & Trope, 1998). Based on these, the sharing reflecting travel results has been found to be associated with desirability value and could be defined as desirability sharing, while the other reflecting travel process has been found to be related to feasibility value and could be defined as feasibility sharing. From construal level theory, desirability sharing means that the shared content is more abstract, because it expresses tourists' thinking and aftertaste of the deeper meaning of travel activities (Oliveira et al., 2020); while feasibility sharing implies that the shared content is more concrete, as it is related to what the tourists have seen and heard (Wu et al., 2020).

Tourism activity type and travel experience sharing type
There are a variety of tourism activities that can meet the needs and motivations of the tourists (Beckman et al., 2017;Sirgy, 2019), and each of them has its dominant features (Mehmetoglu, 2007;Trauer, 2006). Based on these dominant features, scholars have categorized tourism activities into different types, for example, challenging versus relaxing tourism activity (Mehmetoglu, 2007;Su, Cheng, & Swanson, 2020). A challenging tourism activity involves a relatively high level of specialization, high effort, and high risk; whereas a relaxing tourism activity is characterized by a low level of specialization, low effort, and low risk (Mehmetoglu, 2007;Su, Tang, & Nawijn, 2020). Construal level theory suggests that tourists may have a further psychological distance for a challenging tourism activity compared to a relaxing tourism activity, as the challenging tourism activity involves a relatively higher level of specialization, higher effort, and higher risk relative to a relaxing tourism activity (Su, Cheng, & Swanson, 2020).
Since tourists may have different psychological distances for challenging vs. relaxing tourism activities, different construal levels may be activated (Sagristano et al., 2002), which in turn affects their subsequent behavior (Cai & Leung, 2020;Lee & Oh, 2017). Previous studies (e.g., Kim et al., 2008;Liberman & Trope, 1998;Sagristano et al., 2002) have indicated that a higher psychological distance can trigger people's preference for the desirability of an option, whereas individuals show a higher preference for the feasibility of an option when they experience a state with a closer psychological distance. Therefore, a challenging tourism activity may induce more desirability sharing as tourists have a further psychological distance for challenging tourism activity, whereas a relaxing tourism activity may induce more feasibility sharing as tourists have a closer psychological distance for relaxing tourism activity. Based on these theoretical and empirical evidence, the current paper argues that the types of travel experience sharing differ across tourism activity types. The corresponding hypotheses are as follows: H1. Travel experience sharing type differs across tourism activity types.
H1a. Challenging activity induces more desirability sharing than feasibility sharing.
H1b. Relaxing activity induces more feasibility sharing than desirability sharing.

Tourism activity type and tourist well-being
Tourist well-being could be defined as the tourists' affective pleasure and inner feeling of self-growth in the process of satisfying various sensory needs and achieving travel goals (Filep & Laing, 2019;Lengieza et al., 2019;Rahmani et al., 2018;Smith & Diekmann, 2017). Drawing on well-being studies in psychology (Huta & Ryan, 2010;Ryan & Deci, 2001;Waterman, 1993), tourist well-being is increasingly considered to contain two dimensions: eudaimonia and hedonia (Buzinde, 2020;Hao & Xiao, 2021;Lengieza et al., 2019;Scuttari et al., 2021). Tourist eudaimonia reflects tourists' inner feeling of self-growth such as realizing individual potential and self-fulfilment, while tourist hedonia reflects tourists' affective pleasure from satisfaction of sensory needs such as savoring delicious foods or beautiful scenery (Filep & Laing, 2019;Su, Tang, & Nawijn, 2020). From construal level theory, tourist eudaimonia may reflect the abstract side of tourist well-being, because it is related to realizing individual potential and self-fulfilment which represents a higher-order human need; while tourist hedonia may reflect the concrete side of tourist well-being, because it is related to the sensory pleasure which represents a lower-order basic human need (Maslow, 1954).
Previous studies (e.g., Gilbert & Abdullah, 2004;Nawijn et al., 2010) have showed a tight bond between tourism activities and tourist well-being. Additionally, different tourism activity types are reported to influence the two dimensions of well-being differently (Nawijn et al., 2010;Rook & Zijlstra, 2006;Smith & Diekmann, 2017;Sonnentag, 2001). Drawing on construal level theory, tourists may have a different psychological distance for a challenging vs. a relaxing tourism activity which would lead to different construal levels (Sagristano et al., 2002), such construal level might have significant downstream consequences for tourists' wellbeing (Cai & Leung, 2020;Park & Morton, 2015). Specifically, a challenging tourism activity may inspire the more abstract side of tourist well-being (i.e., tourist eudaimonia) as tourists may have a further psychological distance for a challenging than a relaxing tourism activity, whereas a relaxing tourism activity may inspire the more concrete side of tourist well-being (i.e., tourist hedonia) as tourists may have a closer psychological distance for a relaxing vs. a challenging tourism activity. Su, Tang, and Nawijn (2020) indicated that tourist eudaimonia and hedonia have different intensities in certain tourism activities, and they have suggested that tourist eudaimonia is significantly higher than tourism hedonia in a challenging tourism activity. Thus, although both challenging tourism activities and relaxing tourism activities could activate tourist eudaimonia and tourist hedonia, the dominant position between tourist eudaimonia and hedonia may differ across different tourism activity types. The hypotheses are as follows: H2. The dominant dimension of tourist well-being differs across different tourism activity types.
H2a. In challenging activity, eudaimonia occupies the dominant position rather than hedonia.
H2b. In relaxing activity, hedonia occupies the dominant position rather than eudaimonia.

The mediating roles of eudaimonia and hedonia
Existing studies indicated that tourist well-being could result in a series of behavioral outcomes, such as revisit intention (Lin, 2014), willingness to pay (Suess & Mody, 2018), recommending (Dekhili & Hallem, 2020), and travel experience sharing (Oliveira et al., 2020;Wu & Pearce, 2016). Therefore, tourist well-being may play a mediating role between tourism activity and travel experience sharing. Additionally, since tourist well-being includes two dimensions: eudaimonia and hedonia (Rahmani et al., 2018), and the dominant position between tourist eudaimonia and hedonia may differ across different tourism activity types, there are potentially two different mediation mechanisms between tourism activity and travel experience sharing. Specifically, tourist eudaimonia may play a mediating role between a challenging tourism activity and travel experience sharing, whereas tourist hedonia may play a mediating role between relaxing tourism activity and travel experience sharing.
Based on construal level theory, tourist eudaimonia is more abstract than tourist hedonia, it can inspire tourists to a higher level of construal than tourist hedonia, which may lead to a preference of desirability option (Kah et al., 2016). Liberman and Trope (1998) indicated that a low and concrete level of construal would lead people to focus on the feasibility considerations (e.g., how to perform a given action), but a high and abstract level of construal would lead people to emphasize the desirability aspects (e.g., why action would be meaningful). Kah et al. (2016)) also suggested that if a customer was planning a purchase in the distant future (high level of construal), a marketer should emphasize the desirability factors of consumption rather than the feasibility of products. Conversely, if a customer was planning an immediate purchase (low level of construal), a marketer should emphasize feasibility factors of products rather than desirability-related issues. In sum, tourist eudaimonia may lead to more desirability sharing, while tourist hedonia may lead to more feasibility sharing. Hypotheses can be formulated as follows: H3. Tourist well-being mediates the relationship between tourism activity types and travel experience sharing types.
H3a. Eudaimonia mediates the relationship between challenging tourism activity and desirability sharing.
H3b. Hedonia mediates the relationship between relaxing tourism activity and feasibility sharing.

The moderating effect of social context
Social context is defined as social connections with others when people participate in social activities (Bhargave & Montgomery, 2013;Wang et al., 2018). Social context has been established as a key factor to influence consumers' emotions and behaviors through affecting their experience evaluations (Bhargave & Montgomery, 2013;Wang et al., 2018). According to the extent of social connection that people may develop in social activities, Bhargave and Montgomery (2013) divided social context into two types: joint and solo. In joint social contexts individuals feel more connected to others, while they feel less connected to others in solo social contexts (Raghunathan & Corfman, 2006;Ramanathan & McGill, 2007). Social context (from self to other) is an important dimension of psychological distance in construal level theory (Huang et al., 2016), joint context means a close psychological distance that could evoke a low level of construal, while solo context means a far psychological distance which would activate a high level of construal (Bhargave & Montgomery, 2013).
In tourism, it is common to travel with friends (joint context) or alone (solo context) no matter what type of tourism activity (Su, Cheng, & Swanson, 2020). Social context type may also be a key factor in influencing tourist well-being (Huang & Wang, 2014;Wenzel & Benkenstein, 2018). Bhargave and Montgomery (2013) indicated that the evaluation of hedonic experiences is different in joint (vs. solo) social contexts. Su, Cheng, and Swanson (2020)) demonstrated that tourists have a higher level of emotional response and stronger intention of storytelling when traveling with a companion relative to traveling alone. Therefore, we reason that social context may impact tourist eudaimonia and hedonia.
Eudaimonia and hedonia as the two dimensions of well-being are distinguishable but also significantly overlapping (Huta & Ryan, 2010;Rahmani et al., 2018;Ryan & Deci, 2001). The overlap of eudaimonia and hedonia provides a possibility that under proper conditions, there could be mutual transformation between these two (Rahmani et al., 2018). As mentioned earlier, tourist eudaimonia may reflect the abstract side of tourist well-being, while tourist hedonia may reflect the concrete side of tourist well-being. In the joint social context, the close social distance could facilitate the transformation of tourist well-being from a relatively abstract side (eudaimonia) to a relatively concrete side (hedonia). In contrast, in a solo social context, the transformation process of eudaimonia and hedonia may be reversed, the far social distance may promote the transformation process of hedonia to eudaimonia. Due to the different dominance of eudaimonia and hedonia in challenging activities and relaxing activities, when eudaimonia and hedonia change via different social contexts, their relative dominance may also change. Thus, it is proposed that: H4. The dominant dimension of tourist well-being transforms between eudaimonia and hedonia across different social contexts.
H4a. In challenging activities, joint social context makes the dominant position of eudaimonia disappear, but solo social context strengthens eudaimonia's dominance.
H4b. In relaxing activities, joint social context strengthens the dominant position of hedonia, but solo social context makes hedonia's dominance disappear.
From the above analysis, in different social contexts (joint vs. solo), the dominant position of tourist eudaimonia and hedonia may switch in the opposite direction based on the framework of construal level theory (Trope & Liberman, 2010). Such transformation may lead to a change in the relationship between tourism activity types and travel experience sharing types. Previous studies (Bhargave & Montgomery, 2013;Kah et al., 2016;Liberman & Trope, 1998) have shown that people's preference for desirability and feasibility is related to the level of construal that is activated. That is, people who are activated with a high (low) level of construal prefer desirability (feasibility). In challenging activities, when the joint social context makes eudaimonia's dominance disappear, tourists' preference for desirability sharing may also disappear (Kah et al., 2016). Nevertheless, when the solo social context strengthens eudaimonia's dominance, tourists' preference for desirability sharing may increase (Kim et al., 2008). In relaxation activities, the joint social context could enhance the hedonia's dominance which leads to a higher chance of feasibility sharing, while the solo social context might reduce the hedonia's dominance thus leading to less feasibility sharing. As such, we give rise to the following hypotheses: H5. Social context moderates the relationship between tourism activity types and travel experience sharing types.
H5a. In challenging activity, there will be more desirability sharing than feasibility sharing in solo social context, while there will be no significant difference between sharing types in joint social context.
H5b. In relaxing activity, there will be more feasibility sharing than desirability sharing in joint social context, while there will be no significant difference between sharing types in solo social context. Above-mentioned concepts and hypotheses generate an integrated theoretical model (see Fig. 1). To test these hypotheses, we adopt a multi-methods approach: a secondary data analysis and two scenario-based experiments. There are significant benefits to a multi-method approach, as Davis et al. (2011)) pointed out that "using multiple methods to study a phenomenon is proposed to produce results that are more robust and compelling than single method studies" (p. 467).
Study 1: The impact of tourism activity type on travel experience sharing type The purpose of Study 1 is to test hypothesis 1 through secondary data from Sina Weibo (hereafter Weibo). Weibo is the most important platform to share experiences about travel and hospitality products for the Chinese (Su et al., 2019), therefore, we selected Weibo for data collection.

Data collection
In the choice of different types of tourism activity, we took real-life examples as a starting point, that is, the challenging tourism activity identified as skydiving and skiing, while relaxing tourism activity identified as beach resorts and hot springs. The reasons for choosing two types of tourism activity are: 1) According to the definition of challenging tourism activity and relaxing tourism activity (Mehmetoglu, 2007;Su, Tang, & Nawijn, 2020b), participating in skiing or skydiving requires relatively high effort, high challenge, and high risk which meets the definition of challenging tourism activity. Spending time at beach resorts or hot springs, on the other hand, means low effort, low challenge, and low risk which is in line with the definition of relaxing tourism activity; 2) These four tourism activities are very popular in tourism practice (Jin & Sparks, 2017).
The data were collected using a custom web scraping code, which automatically captured hashtags such as "skydiving," "skiing," "beach resort," and "hot spring" in Weibo messages. Considering the impact of the COVID-19 pandemic, Study 1 did not collect data from 2020 but from April to September 2019. As such, 3081 Weibo messages were captured, followed by the process of data cleaning. This data cleaning process complied with the following standards: a) the author of the Weibo messages must be a tourist who participated in a particular tourism activity; b) the words of the messages must not be fewer than 10; and c) the core meaning of messages must be related to travel experiences. After data cleaning, 576 Weibo messages were finally left as samples of Study 1.

Code
With the 576 Weibo messages left, two PhD students in tourism management were invited to code the type of tourism activity and tourist experience sharing type respectively. The authors first explained the definition to them and the characteristics of different types of tourism activity and different types of travel experience sharing, and gave examples. Next, the two PhD students coded the type of tourism activity in strict accordance with the hashtag, challenging tourism activity (skydiving and skiing) was coded as 1, and relaxing tourism activity (beach resort and hot spring) was coded as 2. Furthermore, they coded tourist experience sharing types strictly in line with the definition of tourist experience sharing types (desirability or feasibility). As such, desirability sharing was coded as 1, and feasibility sharing was coded as 2. The coding consistency of the two PhD students reached 95% which shows the reliability of this coding (Perreault & Leigh, 1989). Finally, any remaining inconsistencies related to the coding were assessed by a professor, who was invited from the tourism management major program.

Data analysis and results
The coding results of these 576 Weibo messages showed that 239 of them were coded as challenging tourism activity (175 messages were coded as desirability sharing and 64 messages were coded as feasibility sharing), while 337 messages were coded as relaxing tourism activity (105 messages were coded as desirability sharing and 232 messages were coded as feasibility sharing).
Chi-square tests were employed to examine the impact of tourism activity type on travel experience sharing type. Results showed that the impact of tourism activity type on travel experience sharing type is significantly different (χ 2 [1] = 100.813, p < 0.001). Specifically, compared to relaxing tourism activity, tourists who participate in challenging tourism activity tend to engage in desirability sharing (M desirability = 73.22%, M feasibility = 26.78%); while compared to challenging tourism activity, tourists who participate in relaxing tourism activity tend to engage in feasibility sharing (M desirability = 31.16%, M feasibility = 68.84%). The results illustrated that travel experience sharing type differs across tourism activity type, that is, tourists engage more in desirability sharing than in feasibility sharing for challenging activities but engage more in feasibility sharing than in desirability sharing for relaxing activities. Therefore, hypothesis 1 was supported (see Fig. 2).

Study 2: The dominant position and mediating roles of eudaimonia and hedonia
Study 2 used a scenario-based experiment to replicate the results of Study 1; and to test the hypotheses H2 and H3.

Pretest
Two versions of tourism activity type scenario were designed as the manipulation. 47 MBA students (25 females; average age = 29.85 years) from a university in the south of China were randomly divided across a challenging (G cha, n = 23) and a relaxing tourism activity (G rel, n = 24). After reading the scenario, participants were asked to indicate if the scenario would best be described as either "Challenging" or "Relaxing" (Su, Cheng, & Swanson, 2020). Results of the Chi-square test demonstrated that the manipulation of the tourism activity type was successful (χ 2 [1] = 35.82, p < 0.001).

Main experiment
Using a between-subjects design (challenging vs. relaxing), one hundred participants recruited from an online data platform (Credamo) completed this experiment in exchange for a small cash payment in February 2021. Each participant was randomly assigned to read a tourism-based challenging activity (G cha, n = 50) or a relaxing activity (G rel, n = 50) scenario from the pretest.
After that, participants were asked to complete a four-part questionnaire including the items to test scenario authenticity and manipulation effectiveness, scales of eudaimonia and hedonia, items to measure sharing type intention, and a table of demographic information.
Scenario authenticity was measured by asking participants to indicate Yes or No as to whether the provided scenario "could happen in real life" (Liao, 2007). Manipulation effectiveness was measured in the same way as in the pretest. Eudaimonia and hedonia scales were adapted from Waterman (1993), each of them including six items on a seven-point scale (1 = Strongly Disagree and 7 = Strongly Agree). These scales were subjected to back-translation. When measuring the travel experience sharing type, to make the difference between desirability and feasibility sharing clear to participants, we first described "desirability sharing" and "feasibility sharing" according to their definition: "The travel experience we share can be roughly divided into two types. One is desirability sharing, which focuses on the impact of travel activities on personal growth, goal achievement, and meaning in life, and mainly expresses tourists' thinking and aftertaste of the deeper meaning of travel activities. The other is feasibility sharing, which focuses on the description of the scenery, food and the process of tourism, mainly introducing what we have seen and heard"; and then we used two ways to measure the sharing type. One way was that we directly asked participants to indicate their preference for either "desirability sharing" or "feasibility sharing." Another way to describe in detail the differences between desirability sharing and feasibility sharing, was that we provided two items on a seven-point scale (1 = Strongly Disagree and 7 = Strongly Agree) to measure desirability sharing intention ("If you participated in activities described in the scenario, you would engage in desirability sharing") and feasibility sharing intention ("If you participated in activities described in the scenario, you would engage in feasibility sharing").
Three participants did not complete the experiment, and we finally received 97 (48 in G cha and 49 in G rel ) data sets with a response rate of 97%. Demographic information is displayed in Table 1.

Data analysis and results
Scenario authenticity testing showed that most of the participants (92.8%) reported that the scenario provided was realistic. The manipulation of the tourism activity type was also successful (χ 2 [1] = 71.05, p < 0.001). We then examined the reliability of the eudaimonia and hedonia scales respectively, each of them expressed good reliability (Cronbach's α > 0.8), the details of scales and their reliability are shown in Table 2.
Chi-square tests were employed to test Hypothesis H1. Results showed that the impact of tourism activity type on travel experience sharing type is significantly different (χ 2 [1] = 24.96, p < 0.001). Specifically, compared to G rel , participants in G cha tend to conduct more desirability sharing (M d = 79.16%, M f = 20.84%); while compared to G cha , participants in G rel tend to conduct more feasibility sharing (M f = 71.43%, M d = 28.57%). Second, hypothesis H1 was tested employing Paired-samples t-tests; results indicated that in a challenging tourism activity, desirability sharing intention (M d = 6.06, SD = 0.86) is significantly (t = 9.03, p < 0.001) higher than feasibility sharing intention (M f = 4.54, SD = 0.77), while in a relaxing tourism activity, feasibility sharing intention is higher than desirability sharing intention (M f = 5.94, SD = 1.03, M d = 4.20, SD = 1.32, t = 5.99, p < 0.001). Therefore, consistent with our expectation and the result of Study 1, travel experience sharing type differs across tourism activity type, hypothesis H1 was confirmed again.
To examine the dominant position of eudaimonia and hedonia in G cha and G rel, we compared eudaimonia with hedonia in G cha with the Paired-samples t-tests, and the same comparison was made to G rel . Results showed that eudaimonia (M e = 6.06, SD = 0.60) was significantly (t = 8.67, p < 0.001) higher than hedonia (M h = 5.26, SD = 0.61) in G cha , while hedonia (M h = 6.32, SD = 0.45) was significantly (t = 13.48, p < 0.001) higher than eudaimonia (M e = 5.11, SD = 0.56) in G rel . As expected, eudaimonia occupied a dominant position in challenging tourism activity relative to hedonia (H2a), and hedonia occupied the dominant position in relaxing tourism activity relative to eudaimonia (H2b), consequently hypothesis H2 was confirmed (see Fig. 3).
Based on above analysis, hypothesis H3 was examined by PROCESS, an add-on package of SPSS. According to PROCESS model 4, we used tourism activity type as the independent variable, eudaimonia and hedonia both as the mediating variables, and desirability sharing intention and feasibility sharing intention as dependent variables respectively. Subsequently, a mediation analysis was conducted through bootstrapping tests with 5000 replications and 95% confidence intervals. Results indicated that when the desirability sharing intention was the dependent variable, eudaimonia significantly mediated the relationship between tourism activity type and desirability sharing intention (95% CI: −1.04, −0.27), but hedonia had no significant mediating effect (95% CI: −0.81, 0.07). Therefore, relative to relaxing activities, challenging activities could evoke eudaimonia dominantly and lead to desirability sharing, H3a was confirmed. Meanwhile, when the dependent variable was feasibility sharing intention, hedonia significantly mediated the relationship between tourism activity type and feasibility sharing intention (95% CI: 0.36, 1.21), while eudaimonia had no significant mediating effect (95% CI: −0.04, 0.50). Thus, compared to challenging activities, relaxing activities could evoke hedonia dominantly and lead to feasibility sharing, H3b was consequently supported (see Table 3).

Study 3: The moderating role of social context
The purpose of Study 3 was to examine hypotheses H4 and H5, it was also a scenario-based experiment.

Pretest
Two versions of a social context scenario were designed as the manipulation. 60 MBA students (32 females; average age = 30.67 years) from a university in the south of China were randomly divided across joint (G jo, n = 30) and solo social context (G so, n = 30). After reading the scenario, participants were asked to indicate if the scenario provided would best be described as either "joint" or "solo". Results of the Chi-square test demonstrated that the manipulation of social context (joint vs. solo) was successful (χ 2 [1] = 48.65, p < 0.001). This activity gives me my strongest feelings that this is who I really am This activity gives me my greatest pleasure When I engage in this activity, I feel more intensely involved than I do in most other activities When I engage in this activity, I feel more satisfied than I do when engaged in most other activities When I engage in this activity, I feel that this is what I was meant to do When I engage in this activity, I feel good I feel more complete or fulfilled when engaging in this activity than I do when engaged in most other activities When I engage in this activity, I feel a warm glow I feel a special fit or meshing when engaging in this activity When I engage in this activity, I feel happier than I do when engaged in most other activities

Main experiment
The main experiment utilized a 2 (challenging vs. relaxing tourism activity) × 2 (joint vs. solo social context) betweensubjects design. 200 participants were recruited in the same way as in Study 2, and they were randomly assigned to four groups averagely according to the factorial condition. The tourism activity type manipulation was the same as in Study 2, social context manipulation was the same as in the pretest of Study 3. Participants were asked to imagine themselves in the scenario, and then they completed a four-part questionnaire. Everything in this questionnaire was the same as in Study 2, except for the addition of one item to check the manipulation effectiveness of social context, this item was the same as in the pretest of Study 3.
Next, a 2 × 2 ANOVA was employed to test H5, with tourism activity type and social context as the independent variables, and desirability sharing intention and feasibility sharing intention as the dependent variables respectively. When the dependent variable was desirability sharing intention, results showed a significant interaction between tourism activity type and social context  (F 1, 186 = 4.48, p < 0.05). When the dependent variable was feasibility sharing intention, results showed that the interaction between tourism activity type and social context was also significant (F 1, 186 = 4.18, p < 0.05); therefore, H5 was supported (see Fig. 5). More specifically, when we employed Paired-samples t-tests, results indicated that the difference between feasibility sharing intention (M f = 6.00, SD = 1.00) and desirability sharing intention (M d = 5.76, SD = 0.94) was not significant (t = 1.28, p > 0.05) in G jc , but desirability sharing intention (M d = 5.73, SD = 1.33) was significantly higher than feasibility sharing intention (M f = 4.84, SD = 1.88) in G sc (t = 2.24, p < 0.05). This suggests that in a challenging activity, joint social context makes the difference between sharing types insignificant, but a solo social context induces more desirability sharing than feasibility sharing, so H5a was supported. Moreover, feasibility sharing intention (M f = 5.94, SD = 0.95) was significantly higher than desirability sharing intention (M d = 4.90, SD = 1.25) in G jr (t = 3.75, p < 0.001), but desirability sharing intention (M d = 5.64, SD = 1.11) and feasibility sharing intention (M f = 5.56, SD = 1.13) had no difference (t = 0.33, p > 0.05) in G sr , indicating that in relaxing activity, joint social context induces more feasibility sharing than desirability sharing, but solo social context makes the difference between sharing types insignificant, thus, H5b was supported.

Conclusions and discussion
Using secondary data and two scenario-based experiments, the current paper examined a conceptual model that focused on why people share different content of their travel experiences. The research led us to the following key findings: 1) challenging tourism activities lead to more desirability sharing and relaxing tourism activities lead to more feasibility sharing; 2) eudaimonia (hedonia) occupies the dominant position over hedonia (eudaimonia) in a challenging (relaxing) tourism activity; 3) eudaimonia (hedonia) plays the mediating role in the relationship between a challenging (relaxing) tourism activity and desirability (feasibility) sharing; 4) joint (solo) social context facilitates the transformation from eudaimonia (hedonia) to hedonia (eudaimonia); 5) social context has a significant moderating effect on the relationship between tourism activity type and travel experience sharing type.

Theoretical contributions
Previous studies (e.g., Kang & Schuett, 2013;Munar & Jacobsen, 2014;Oliveira et al., 2020;Wu & Pearce, 2016) only regarded travel experience sharing as a kind of homogeneous activity, leading to a lack of investigation into the heterogeneity of shared contents (Ring et al., 2016). Drawing on construal level theory (Trope & Liberman, 2010), the current study's findings indicate that different types of travel experience sharing and tourist well-being represent different levels of construal, while different types of tourism activities and social contexts activate different psychological distances. Therefore, the relationship between these four concepts has been systematically explored to reveal the reasons why tourists share different contents under different conditions, which achieved a better understanding of the heterogeneity of travel experience sharing and the role of tourism activity type. Currently, as an increasing number of researchers use the construal level theory in tourism research (e.g., Cai & Leung, 2020;Chatterjee & Mandal, 2020;Kim et al., 2020), this article further expands the application of construal level theory in tourism contexts and demonstrates the great potential of this theory in explaining complex tourism phenomena.
Additionally, the current paper incorporates tourist eudaimonia as the mediating role in the relationship between tourism activity and travel experience sharing, which further expands the latest findings of tourist well-being from a positive psychological perspective (e.g., Buzinde, 2020;Filep & Laing, 2019;Hao & Xiao, 2021;Rahmani et al., 2018;Smith & Diekmann, 2017). Moreover, the findings indicate that different types of tourism activity (challenging or relaxing activity) could activate different dominant dimensions of tourist well-being (tourist eudaimonia or hedonia), and further lead to different types of travel experience sharing (desirability or feasibility sharing). Although previous studies (e.g., Su, Cheng, & Swanson, 2020) have pointed out that emotional arousal played the mediating role between tourism activity and experience sharing, they failed to clearly indicate the kind of emotion involved. In addition, while previous studies (e.g., Oliveira et al., 2020;Wu & Pearce, 2016) have examined the role of tourist well-being in travel experience sharing, they did not distinguish the effects of different well-being dimensions. Thus, the underlying mechanism from tourism activity to experience sharing is explained more clearly and deeply in the current research.
Moreover, this paper reveals that the dominant position could transform between eudaimonia and hedonia when taking social context into consideration, which suggests the boundary conditions of the effect from tourism activity type to experience sharing. Thus, this paper elaborates on previous studies (e.g., Oliveira et al., 2020;Wu & Pearce, 2016) which focused on the relationship between tourist experience, tourist well-being, and travel experience sharing. The current study finds that the dominant position could transform between eudaimonia and hedonia, which echoes with a wide range of previous studies, including the ones that argued that eudaimonia and hedonia were significantly overlapping (e.g., Huta & Ryan, 2010;Ryan & Deci, 2001;Waterman, 1993), and that eudaimonia could transform into hedonia (Rahmani et al., 2018). Our conclusion also adds to the study conducted by Su, Tang, and Nawijn (2020)) that compared the relative intensity between eudaimonia and hedonia in different tourism activities but did not consider the transformation of this relative intensity.

Practical implications
In the age of social media, it is more and more convenient to obtain the experience sharing of tourists, but it is not easy to effectively use this information (Lalicic et al., 2021). The current paper provides an approach to classify travel experience sharing into desirability sharing and feasibility sharing and points out that different types of sharing are related to specific types of tourism activities. Therefore, we can more accurately recommend specific tourism activities to tourists according to the type of experience sharing. For example, if a tourist often engages in desirability sharing, he (she) may prefer challenging tourism activities, so the tourism managers could recommend challenging tourism activities to him (her) in a targeted manner.
The current study further suggests that tourist eudaimonia and hedonia could activate different experience sharing, which provides ideas for how to promote tourists to share travel experiences positively. For example, tourism managers could add clues related to life goals and meanings in the design of challenging activities, so that tourists who participate in challenging activities will be activated higher eudaimonia, which in turn leads to more desirability sharing. In the process of replying to tourists sharing content, tourism managers of relaxing tourism activities could provide more clues about sensory experiences (e.g., delicious food or beautiful scenery) to produce more tourist hedonia and more feasibility sharing.
Furthermore, considering that our findings indicate that the social context has the power to change the well-being effect of a challenging activity and a relaxing activity, tourism managers may provide advice to tourists about how to achieve more eudaimonia or hedonia through tourism. For example, they can provide discounts to encourage those who are pursuing hedonia to invite friends to travel together, so that these tourists have a joint social context in the travel process, thereby enhancing their hedonia. Meanwhile, if a person wants to improve her (his) eudaimonia, choosing a challenging tourist destination and setting out alone maybe a good suggestion.

Research limitations and future directions
The current study also has its limitations that should be addressed in future research. First, the samples and participants are from mainland China, which may hamper the external validity of the current paper. Future studies can utilize samples and recruit participants from different countries to enhance the external validity. Additionally, our study illustrates the difference of sharing type after participating in a challenging tourism activity and a relaxing tourism activity, it is subsequently valuable to study what the impact is of the sequence of tourism activity types on travel experience sharing types. Future studies are encouraged to test how mixed tourism activities influence the type of travel experience sharing, for example, framing (which comes first: a challenging or a relaxing tourism activity) effect is recommended to be tested (Chi et al., 2021). Finally, the current paper has identified the social context as a key factor leading to the dominant position being transformed between eudaimonia and hedonia, other factors that further moderate this transformation need to be explored in future studies, for which a potential candidate is mood management strategy (down vs. up regulate; Gao & Kerstetter, 2018).

Declaration of competing interest
None.