How do tourism goal disclosure motivations drive Chinese tourists' goal-directed behaviors? The influences of feedback valence, affective rumination, and emotional engagement

Based on self-determination theory and the broaden-and-build theory of positive emotion, this study investigated the motivations of disclosing tourism goals on social media and its impacts on Chinese tourists’ goal-directed behaviors (GDBs). We proposed and tested a mutual transformation model of tourism goal disclosure motivation under different conditions of feedback valence (positive vs. negative feedback) and examine the mediating role of tourists’ affective rumination and emotional engagement. The results revealed that tourists driven by extrinsic motivations develop a stronger emotional engagement in their tourism goals and exhibit more GDBs after receiving positive feedback on their disclosed tourism goals. However, negative feedback disclosed goals lowers GDBs and leads to affective rumination about tourism goals among those with intrinsic motivations. This study provides theoretical and practical implications for destination marketers to adopt marketing strategies based on the findings.

Disclosures of tourism goals often stay on social media, becoming a means for people to manage their impressions and present themselves online, to refine an ideal self-image in the minds of social network audiences (Lavertu, Marder, Erz, & Angell, 2020). Previous studies have revealed a close correlation between goal disclosure on social media and individual motivation to achieve goals (Beall et al., 2021;Chang et al., 2020;Khan, 2017). For example, previous studies have confirmed individuals proactively participate in social media goal disclosure driven by their extrinsic or intrinsic motivations, such as gaining social recognition or self-expression (Wang & Fesenmaier, 2004). However, most such studies have analyzed the influence of different motivations on goal-directed behaviors (GDBs) from a static perspective (Chang et al., 2020). Rare research has explored the transformation between extrinsic and intrinsic motivations and the subsequent influence on GDBs from a dynamic perspective. Although some studies have examined the transformation of extrinsic motivation to intrinsic motivation under need-supportive environments (Chen et al., 2015;Deci & Ryan, 1985, 2020Van den Broeck, Ferris, Chang, & Rosen, 2016), few studies have explored the bidirectional transformation of tourism goal disclosure motivations and its impacts on tourists' GDBs.
In fact, the transformation between extrinsic and intrinsic motivations would generate a significant impact on individual behaviors (Boone, Vansteenkiste, van der Kaap-Deeder, Soenens, & Verstuyf, 2014;Chen et al., 2015). Organizational behavior and marketing research shows that intrinsic motivation and extrinsic motivation influence individual performance and purchase behavior through different psychological mechanisms (Chang et al., 2020). Intrinsic motivation touches more of the individual's emotional dimension, awakens more individual emotional engagement, and further promotes individual performance (Rich, Lepine, & Crawford, 2010). As a prerequisite for physical engagement, emotional engagement increases the stability, persistence, and pervasiveness of an individual's affective-cognitive state because it strengthens identification with an action, as illustrated in research based in the workplace (Schaufeli, Bakker, & Salanova, 2006). Extrinsic motivation creates a focus on the rewards (such as verbal rewards or material rewards) that are beyond the results of activities or tasks and are often used as a stress experience to act in a specific way , 2020. Because the individual is performing a task that is not necessarily aligned with their core ego, confidence, and attitude, extrinsic motivation will stimulate psychological contradictions and conflicts with other goals, which can trigger a state of rumination (Thomsen, Tønnesvang, Schnieber, & Olesen, 2011).
At the same time, intrinsic and extrinsic motivations are subject to internalization and externalization processes that may alter them significantly (Boone et al., 2014;Chen et al., 2015;Van den Broeck et al., 2016;Vansteenkiste & Ryan, 2013). Self-determination theory (SDT) proposes that external incentives (such as verbal rewards or material rewards, etc.) could impel individuals to absorb behavioral norms and criteria into themselves, which is called internalization (Chen et al., 2015;Ryan & Deci, 2020). On the contrary, basic psychological need frustration (such as autonomy, competence, and relatedness) triggers externalization, in which the reduction of individual self-control leads to considering problems to be someone else's fault (Boone et al., 2014). However, existing studies on the impact of intrinsic and extrinsic motivation focus on internalization and its results, such that many questions remain about the attention on externalization. Additionally, the transformation of intrinsic motivation and extrinsic motivation is not unconditional (Chen et al., 2015). SDT proposes that the autonomous support environment (such as approval, appreciation, encouragement, etc.) could promote the internalization of extrinsic motivation (Deci, Connell, & Ryan, 1989). As a social media platform allowing real-time interaction, it may function as an environment of providing support to individuals (Lavertu et al., 2020). Indeed, social media users increase their relationship with members of their social networks by disclosing tourism goals, and the feedback valence (positive vs. negative) of members affectes individuals' motivation, attitude, and behaviors (Kaplan & Haenlein, 2010). Thus, exploring the boundary conditions of motivation transformation is of great significance for understanding the influence mechanism of goal disclosure motivation on GDBs on social media. However, research has not explored the moderating effect of feedback valence (positive vs. negative) on individual motivation transformation.
Motivation transformation may also affect tourists' subsequent psychological interaction mechanism and change their distal GDBs. SDT proposes that people who perceive intrinsic motivations as the drivers of their behavior will have an intrinsic perceived locus of causality but those who perceive intrinsic motivations as the drivers will have an extrinsic perceived locus of causality . Focusing on intrinsic goal pursuits can increase individuals' emotional engagement and well-being, while focusing on extrinsic goals can trigger anxiety, rumination, and depression (Karatepe & Aleshinloye, 2009;Thomsen et al., 2011). Thus, the dynamic downstream consequences of the dyadic transformation between intrinsic and extrinsic motivations could further induce the change of tourists' psychological mechanisms and GDBs. However, the existing literature has not explored the dynamic downstream consequences generated by the two-way transformation process of motivations.
Being aware of the knowledge gaps above, this study seeks to examine the impact of tourism goal disclosure motivations on tourists' GDBs, and further investigate the underlying psychological mechanism of mutual motivation transformation, feedback valence, and tourists' GDBs. It begins by categorizing tourism goal disclosure motivations into extrinsic and intrinsic types and exploring the differences of tourists' emotional engagement, affective rumination, and GDBs under different tourism goal disclosure motivations. Then, drawing on selfdetermination theory, we analyze and verify the mediation role of emotional engagement and affective rumination between tourism goal disclosure motivations and tourists' GDBs. Finally, by focusing on the role of feedback valence (positive vs. negative) in motivational transformation, this study discusses the interactive effects of tourism goal disclosure motivations and feedback valence on emotional engagement, affective rumination, and tourists' GDBs. In this regard, we provide experimental justification of tourism goal disclosure motivations on social media from a dynamic transformation perspective, which will shed new light on the impact of tourism goal disclosure motivations under different conditions of feedback valence.

Intrinsic and extrinsic motivations
Motivation refers to the internal psychological process and motives that guide and maintain individual behavior and lead it to achieve a certain goal (Deci & Ryan, 1985). An important aspect of motivation involves why people behave in a particular way, which is the reason for their perceived participation in the behavior (Ryan & Deci, 2020). SDT proposed that autonomy, competence, and relatedness, three basic psychological needs, are essential conditions for individual psychological growth, internalization, and mental health (Huertas-Valdivia, Gallego-Burín, & Lloréns-Montes, 2019; Ryan & Deci, 2017, 2020. As a universal motivation theory (Deci & Ryan, 1985;, which aims to systematically explain the dynamic needs, motivation, and well-being of human beings in the social environment, it has been validated in many research fields such as work behavior, customer marketing, social behavior, learning behavior, and information attachment behavior (Chang et al., 2020;Pingel, Fay, & Urbach, 2019;Zhang, Cole, Hirt, & Bilgihan, 2017).
Intrinsic motivation refers to when individuals engage in activities due to their intrinsic pleasure, happiness and satisfaction; it represents a state of high autonomy and self-determination (Ryan & Deci, 2020). Studies have shown that intrinsic motivation significantly increases goal attainment (Liu et a., 2019;Van Hooff & Baas, 2013). In contrast, extrinsic motivation refers when to individuals act under external pressure or in order to obtain external utility, which makes them less likely to act (Zhang et al., 2017). Extrinsic motivation might include the desire to present oneself to others in a particular way; it encourages individuals to keep in touch with others while getting their support, discussing common interests, accepting certain subjective norms, and representing a weak independent state of self-control (Moghimehfar & Halpenny, 2016).
In summary, Toubia and Stephen (2013) have confirmed from the perspective of motivation that individuals often use social networks to meet their diverse social demands. Whether it is intrinsic motivation to pursue goal attainment or extrinsic motivation to pursue successful social interactions, individuals may obtain their own pleasure from tourism goal disclosure. The key difference, however, is that the pleasure of intrinsic motivation is derived from personal interests or values, goals, and aspirations important to an individual and the sense of accomplishment inspired by being witnessed pursuing and achieving goals by many followers (Chang et al., 2020;Liu, Wang, Huang, & Tang, 2019;Zhang et al., 2017). By contrast the pleasure of extrinsic motivation is derived from the satisfaction of social acceptance (e.g., maintaining social relationships, managing impressions, and obtaining social recognition) after the goal is disclosed (Grant & Mayer, 2009;Khan, 2017;Lavertu et al., 2020;Zhang et al., 2017).

The transformation between intrinsic motivation and extrinsic motivation
SDT expounds on the internalization and integration of external values and rules as well as the development and dynamics of extrinsic motivation (Gagné & Deci, 2005;, 2020. SDT indicates that internalization is a psychological process in which an individual absorbs and resets the original external regulation. Once internalization begins, the individual has a certain degree of self-determination . When internalization is incomplete, external regulation and perceived value will remain external or partially internalized in the form of introjection or identification (Ryan & Deci, 2020). When the internalization process reaches the optimum state, the individual will identify the significance of external regulation and perceived value, which will be absorbed and integrated into self-regulation, achieving the integration state (Gagné & Deci, 2005). Deci (2000, 2020) arranged the types of extrinsic motivation on a continuum based on ranging from external regulation to introjected regulation to identified regulation to integrated regulation (see Fig. 1).
SDT indicated that although the internalization of extrinsic motivation is a spontaneously triggered evolutionary process, it does not necessarily occur (Moghimehfar & Halpenny, 2016). The external environment must nourish the internalization of external motivation, and autonomous support is a key factor that determines the degree of internalization of extrinsic motivation (Deci et al., 1989;Gagné et al., 2015;. When individuals meet their independent needs and think about the value of behavior autonomously, they are more likely to recognize and internalize the value of behavior and consequently bring the behavior under autonomous control (Deci & Ryan, 2012;Huertas-Valdivia et al., 2019;Moghimehfar & Halpenny, 2016). Meanwhile, the externalization process occurs when the hindrance of the individual's basic psychological needs causes the individual not to determine their own behavior, which weakens the individual's intrinsic motivation, and then triggers the partial or even complete externalization of intrinsic motivation (Boone et al., 2014;Chen et al., 2015). The individual's attention to obtaining external rewards will prevent them from further exploring the inner interest of activities, hindering them from trying more effective methods to solve problems, thereby reducing individual creativity and personal performance (Gagné & Deci, 2005). Additionally, when the internal or external reasons driving the activity are not clear, individuals tend to be attracted to external rewards and ignore internal reasons, which weakens motivation (Deci & Ryan, 1985). In a similar vein, real-time interaction is a natural attribute of social media (Lavertu et al., 2020). Individuals receive positive or negative feedback from audiences within their social networks, which generate different transformation processes that affect GDB towards tourism goals.

Goal disclosure motivation and goal-directed behaviors
The intrinsic motivation for tourism goal disclosure mainly includes the excitement of setting a new tourism goal, the need for selfexpression, and seeking support for pursuing the goals (Chang et al., 2020). Extrinsic motivations, such as maintaining relationships, obtaining praise and reputation, impression management, establishing new social connections, and gaining social identification, also might drive tourism goal disclosure on social media (Lavertu et al., 2020). Studies of tourists show intrinsic and extrinsic motivations generate significant differences in social interaction behaviors (Munar & Jacobsen, 2014;Toubia & Stephen, 2013). Kang and Schuett (2013) found that tourists who took greater pleasure in their tourism goals were more likely to achieve them. Social media tools facilitate the disclosure of personal tourism goals on a large scale (Tussyadiah & Fesenmaier, 2009), which individuals may use to maintain or enhance their own image, conduct impression management, attract social acceptance and respect, or communicate surplus resources, such as time and energy (Grant & Mayer, 2009), all of which are extrinsic rewards (Locke, 2018).
As SDT suggests, when an intrinsic motivation drives tourism goal disclosure, tourists tend to perform more positive GDBs to satisfy their basic psychological needs (Van Hooff & Baas, 2013). Tourists with intrinsic motivation to disclose their tourism goals on social media usually seek the challenges and pleasure of pursuing their tourism goals rather than seeking external rewards such as praise, prestige, and maintaining social relationships.
Based on the previous studies, the following hypothesis is proposed:

H1.
When the tourism goal is disclosed on social media, intrinsic (vs. extrinsic) motivation will stimulate more tourists' GDBs.

The mediating role of emotional engagement
According to SDT, intrinsic motivation increases task performance, emotional involvement, goal commitment, and subjective well-being . This is because individuals with intrinsic motivation meet the three prerequisites of emotional engagement: meaningfulness, safety, and availability (Kahn, 1990). More specifically, individuals driven by intrinsic motivation (vs. extrinsic motivation) are more inclined to increase their emotional engagement in achieving meaningful goals, can obtain more psychological safety from the happiness and sense of accomplishment from the activities and goals themselves, and focus more on psychological resources on goal achievement, which improves the availability of psychological resources. Karatepe and Aleshinloye (2009) identified intrinsic motivation and active personality as the main antecedent variables of emotional engagement and demonstrated that active individuals create resources through proactive behavior, which makes them more likely to continuously invest in their work. In short, intrinsic motivation (vs. extrinsic motivation) will inspire investment in more emotional resources for goal attainment.
Emotional engagement-a motivational state that reflects the intense and persistent emotional engagement in the individual's role (Kahn, 1990)-is an important psychological resource for establishing a psychological association between the individual and the work and maintaining efficient personal performance (Reina, Rogers, Peterson, Byron, & Hom, 2018). Additionally, emotional engagement is considered a prerequisite for physical engagement and has the characteristics of stability, persistence, and pervasiveness (Schaufeli et al., 2006). Schaufeli et al. (2006) pointed out that people with strong intrinsic motivation will invest more time, money, and emotion in the process of solving problems or goal attainment, make more attempts, and have better perseverance and persistence.
The job engagement model proposed by Britt, Adler, and Bartone (2001) suggests individuals' engagement in their work will increase persistence in GDBs. Harter, Schmidt, and Hayes (2002) found a significant positive correlation between employee emotional engagement and many organizational outcome variables, including overall performance. As Kahn (1992) explained, perceiving and recognizing work as fun produces positive emotions, such as enthusiasm and excitement (i.e., emotional engagement; Rich et al., 2010). Emotional engagement supports full engagement in work and increases their sense of psychological security, which in turn improves retention (Harter et al., 2002). Bakker, Demerouti, and Brummelhuis (2012) confirmed that emotional engagement was positively correlated with task performance, extra work performance, and active learning. Accordingly, we predict emotional engagement has a mediating effect in intrinsic motivation and tourists' GDBs, suggesting: H2. Emotional engagement mediates the relationship between tourism goal disclosure motivation and GDBs for tourists driven by intrinsic motivation (but not extrinsic motivation).

The mediating role of affective rumination
The stress-reactive model of rumination suggests individuals who engage in social interaction with extrinsic motivation are more likely to fall into rumination and suffer social anxiety because they are worried about the outcome of the tourism goal disclosure (Smith & Alloy, 2009). Previous studies have shown that active participation in goal disclosure on social media may reduce mental relaxation from social interaction (Zoccola, Dickerson, & Lam, 2012). Extrinsic motivation increases unpleasant psychological experiences (such as craving, anxiety, and lack of control), leading to psychological problems and interpersonal problems (Chen et al., 2015). It also increases non-adaptive ways of coping with stress, such as escape, social withdrawal, and rumination among individuals participating in social interactions (Pingel et al., 2019). Cropley and Millward's (2009) qualitative study found that external motivation predicts high rumination and a tendency to use external rewards (respect by others, rewards, and recognition from superiors) to prove they work hard. The tourism goal disclosure on social mediato some degree-challenges individual social relationships, social resources, and even social identity and may serve as an additional means of eliciting and/or prolonging rumination (Tran & Joormann, 2015). As a maladaptive coping style, rumination represents an experiential avoidance coping strategy (Pingel et al., 2019). Therefore, extrinsic motivation may be related to the non-adaptive coping style of rumination.
Rumination refers to individuals repetitively and passively thinking about negative situations, causes, and potential consequences (Nolen--Hoeksema, Wisco, & Lyubomirsky, 2008). It is considered a maladaptive response that has negative effects (Smith & Alloy, 2009). The main characteristics of rumination are negative thought content, negative inner experience, abstract construction level, and non-constructive consequences (Donahue et al., 2012). Smith and Alloy (2009) proposed that people who ruminate often respond to the differences between the current state and the goal state in a stable, negative, and extensive way. Thomsen et al. (2011) pointed out that rumination was highly related to extrinsic content of goals as well as to less intrinsic motivation. They further demonstrated that being pushed to pursue disclosed goals under an extrinsic motivation is incompatible with an individual's core self, beliefs, and attitudes and thus stimulates more ambivalence and conflict with other goals. Therefore, based on the extant literature, we hypothesize: H3. For tourists driven by extrinsic motivation (vs. intrinsic motivation), affective rumination mediates the relationship between tourism goal disclosure motivation and GDBs.

The moderating role of feedback value
Social media enables individuals to get positive or negative feedback (e.g., comments, likes, and reposts) from the audiences of their social network circle after their tourism goals are disclosed (Toubia & Stephen, 2013). Kaplan and Haenlein (2010) proposed that the core elements of social media include giving and receiving feedback and forming interactions. Mangold and Faulds (2009) also suggested that feedback (e. g., comments, likes) not only facilitates interaction between users but also lowers the barriers to large-scale real-time interaction. Online feedback extends to the offline world, impacting our behaviors toward real-life goals (Lavertu et al., 2020). Locke, Cartledge, and Koeppel (1968) indicated that giving feedback to those who disclose their goals can effectively modulate individual performance relative to those who do not receive feedback.
In fact, the process of tourism goal disclosure is a process of releasing demand signals and of explicit motivations (Munar & Jacobsen, 2014). Real-time interactive feedback (positive vs. negative) in the context of social media fosters conditions for the bidirectional transformation of extrinsic motivation and intrinsic motivation (Kaplan & Haenlein, 2010). According to the broaden-and-build theory of positive emotion (Fredrickson, 2001), this study proposes that different feedback valence (positive vs. negative) may activate different emotional responses and thus influence individual GDBs (Mangold & Faulds, 2009). Specifically, when an individual received positive feedback, the aroused positive emotions could broaden an individual's momentary thought-action repertoires and quickly widen the array of the thoughts and actions that come to mind (Fredrickson, 2001). Positive feedback satisfies individuals' need to seek immediate recognition and praise and offers them a positive autonomous support environment (Kaplan & Haenlein, 2010). This is the critical condition to facilitate the internalization of extrinsic motivation (Deci et al., 1989;Van den Broeck et al., 2016;Vansteenkiste & Ryan, 2013). Negative feedback immediately triggered negative emotions (Watson, Clark, & Tellegen, 1988). It frustrates individuals' basic psychological needs to be frustrated, promoting rumination and externalization (Boone et al., 2014;Chen et al., 2015). Thus feedback may change the source of an individual's motivation. Hence, we hypothesize: H4. The dominant role of tourism goal disclosure motivation transforms between extrinsic motivation and intrinsic motivation across different feedback valence.
H4a. When intrinsic motivation is dominant, positive feedback strengthens the dominance of intrinsic motivation, but negative feedback undermines the dominance of intrinsic motivation.
H4b. When extrinsic motivation is dominant, positive feedback undermines the dominance of extrinsic motivation, but negative feedback strengthens the dominance of extrinsic motivation.
Previous studies have indicated that feedback valence affected the emotional engagement, affective rumination, and behaviors of users who participate in social media interactions (Barasch & Berger, 2014;Harter et al., 2002). Considering the mutual transformation of intrinsic motivation and extrinsic motivation under different feedback valence situations, the dominance of different motivations will drive tourists' emotional engagement and affective rumination, eventually leading to a change in tourists' GDBs. Specifically, in the positive feedback situation, when intrinsic motivation drove the tourism goal disclosure, the dominant role of intrinsic motivation would be enhanced, which would arouse more emotional engagement, and then have a positive impact on the terminal GDBs (Bakker et al., 2012). However, when extrinsic motivations drive the tourism goal disclosure, positive feedback would facilitate the transformation of extrinsic motivation to intrinsic motivation, and its dominant position will be weakened. Thus, after the internalization of extrinsic motivation, tourists adopted GDBs via emotional engagement.
In the negative feedback situation, when intrinsic motivation drives the tourism goal disclosure, negative feedback triggers psychological and interpersonal pressure, leading to externalization of intrinsic motivation (Gagné & Deci, 2005), and intrinsic motivation will no longer dominate. Nevertheless, when extrinsic motivation drives the tourism goal disclosure, the dominant role of extrinsic motivation would be increased in the presence of negative feedback, which would arouse more affective rumination, and thus a negative impact on the terminal GDBs (Barasch & Berger, 2014). Therefore, after the externalization of intrinsic motivation, tourists will fall into affective rumination and reduce GDBs. Accordingly, we hypothesize: H5. Feedback valence moderates the relationship between emotional engagement, affective rumination, and GDBs.
H5a. When tourists receive positive feedback on their tourism goals, their extrinsic motivations to disclose tourism goals will enhance emotional engagement, which in turn positively influence their GDBs.
H5b. When tourists receive negative feedback on their tourism goals, their intrinsic motivations of disclosing tourism goals will enhance affective rumination, which in turn negatively influence their GDBs.
To test the above hypotheses, we conducted a secondary data study and two experiments. In Study 1, we collected secondary data from the social media platform Sina Weibo to examine the relationship between tourism goal disclosure motivations and tourist's GDBs, testing H1. In Study 2, we conducted a one-factor between-subjects experimental study design to examine the mediating role of tourists' emotional engagement and affective rumination between tourism goal disclosure motivations and tourist's GDBs, testing H2 and H3. In Study 3, using a 2 × 3 factorial between-subjects design, we examine the moderating effect of feedback valence on the relationship between tourism goal disclosure motivation and GDBs, as well as the mutual transformation mechanism between intrinsic motivation and extrinsic motivation under different feedback valence situations (testing H4 and H5). Fig. 2 summarizes the key variables and relationships discussed in this study.

Method
Data collection procedure. The data was collected using a web crawler, which automatically crawled Weibo messages containing "tourism goal," yielding 2866 Weibo messages from the first quarter of 2019. The messages were further screened using the following criteria: a) bloggers must disclose their own tourism goals; b) the content the bloggers' original, c) the messages cannot be a marketing advertisement; d) the messages cannot be experience sharing after travel, and e) the message should be no less than 10 words long. After data cleaning, 876 valid Weibo messages (30.56% of the total) remained for coding.
Coding. Based on the operational definition of tourism goal disclosure motivation (intrinsic motivation vs. extrinsic motivation) and GDBs, two tourism management doctoral students coded the posts for motivation, according to whether the tourism goal disclosure reflects an extrinsic (coded as 1) or extrinsic motivation (coded as 2). In a similar vein, they coded GDBs, as well as longitudinal tracking results. Those who fail to successfully implement the tourism GDBs were coded as 1, and those who succeed in the final implementation of the tourism GDBs were coded as 2. The coding consistency of the two doctoral students was 95% (Perreault & Leigh, 1989). A senior tourism management professor coded the remaining 5%.

Results and discussion
The coding indicated that 325 (37.1%) of the 876 messages reflected Fig. 2. The theoretical model. extrinsic motivation while the remainder reflected intrinsic motivation. Among the extrinsically motivated, 37 bloggers showed evidence of GDBs toward the tourism goal. Among those with intrinsic motivation, 408 undertook GDBs toward their tourism goal. Chi-square test results showed that the impact of tourism goal disclosure motivation on tourists' GDBs is significantly different (χ 2 (1) = 321.158, p < 0.001). Specifically, compared with intrinsic motivation, tourists with extrinsic motivation to participate in tourism goals disclosure tended to adopt fewer GDBs (M GDBs = 11.38%, M no GDBs = 88.62%). Additionally, tourists with intrinsic motivation tended to adopt more GDBs than those with extrinsic motivation (M GDBs = 74.05%, M no GDBs = 25.95%).
The results revealed that motivation predicted GDBs. GDBs intention was stronger among those with intrinsic motivation, and when disclosing tourism goal under an intrinsic motivation, tourists could adopt fewer GDBs. Therefore, H1 was supported (see Fig. 3).

Study 2
To further verify the underlying psychological mechanism behind the findings of Study 1, a one-factor between-subjects (intrinsic motivation vs. extrinsic motivation) experimental design was adopted in Study 2. The purpose of Study 2 is to examine the mediating role of tourists' emotional engagement and affective rumination between tourism goal disclosure motivations and tourist's GDBs, testing H2 and H3.

Participants and procedure
Before the main experiment, a preliminary survey was conducted to test whether participants correctly understood the tourism goals disclosure motivation on social media in the situational experiment materials. We designed two versions of the tourism goal disclosure motivation (intrinsic vs. extrinsic) description (see Appendix A). Additionally, we added a screening item in the survey so that participants who did not disclose their tourism goals on social media were excluded at the beginning of the survey.
A pilot survey was designed and conducted on the Credamo data platform in China (Gai & Puntoni, 2021). After the system automatically rejected eight unqualified participants, the pilot sample consisted of 52 participants (53.8% females, 44.2% were 18-25) who were randomly divided into intrinsic motivation and extrinsic motivation groups. In order to avoid the interference of any real tourism destination brands and product types on the experimental results, we used a fictitious tourism destination, "X," in the study. Participants were randomly assigned to one of the stimuli materials at first and then asked to answer three dichotomous questions (1 = Yes, 0 = No) to verify the scenario's authenticity. Afterward, they separately completed the tourism goal disclosure motivation (intrinsic vs. extrinsic) items on a 7-point scale (1 = Strongly disagree, 7 = Strongly agree; Khan, 2017;Gagné et al., 2015).

Results and discussion
The scenario authenticity test showed that most of the participants (94.2%) reported that the provided scenario was realistic. Moreover, the measurement results showed that all participants could correctly attribute the tourism goal disclosure motivation described in the stimuli (M extrinsic motivation = 5.06, M intrinsic motivation = 5.38, both are greater than the median value of 4). Thus, these findings suggest that participants could distinguish between intrinsic motivation and extrinsic motivation in the context of stimulus materials, which indicates that the pretest successfully manipulates tourism goal disclosure motivation on social media and further shows that the stimulus materials of tourism goal disclosure motivation can be conducted in the experiment below.

Design and measures
The survey and measurement were divided into three parts. Firstly, the items for scenario authenticity test and manipulation check are the same as pretest. All of the participants were shown different versions of the stimulus material adapted from Derfler-Rozin and Pitesa (2020). After that the participants completed manipulation checks: "According to the scenario described by the material, do you think the tourism goal disclosure on social media is based on intrinsic motive or extrinsic motive?" (1 = Strongly disagree, 7 = Strongly agree). The score allowed us to estimate whether participants explicitly distinguished each type of tourism goal disclosure motive. Next, participants were asked to rate their emotional engagement, rumination, and GDBs. Emotional engagement (Rich et al., 2010) and rumination (Donahue et al., 2012) were measured with six and five items, respectively, on a 7-point Likert scale (1 = Strongly disagree, 7 = Strongly agree). The dependent variable, GDBs, was measured by a five-item scale (Perugini & Bagozzi, 2001). Last, we controlled for three variables: goal difficulty (Locke & Fig. 3. The impact of tourism goal disclosure motivation on tourists' GDBs.  (Perugini & Bagozzi, 2001;Su, Lian, & Huang, 2020) and post frequency (Perugini & Bagozzi, 2001) and provided some demographic questions.

Manipulation check and measurement reliability
The scenario authenticity test showed that 93.2% of the participants considered the provided scenario realistic. The measurement results indicated that all distinguished the tourism goal disclosure motives described in the material scenario (M extrinsic motivation = 5.18 vs. M intrinsic motivation = 5.56; both are greater than the median value of 4). Thus, the manipulation of tourism goal disclosure motives was successful. All constructs have a high reliability (emotional engagement α = 0.938; affective rumination α = 0.964; GDBs α = 0.965).

Main effects of tourism goal disclosure motivation on GDBs
An independent-sample t-test was used to test H1. Before hypothesis testing, the sample size estimation test was carried out. G* Power 3.1 was used to calculate the power value of the sample size (Faul, Erdfelder, Buchner, & Lang, 2009). An independent sample t-test was used. When the number of groups was 2, the effect size (f) was 0.5, the significance level was 0.05, and the power value was 0.8, the required sample size was at least 102 (both groups 1 and 2 were greater than 51). Therefore, the sample size of this study had statistical testing power. The result showed that there was a significant difference in GDBs between the groups exposed to the different tourism goal disclosure motives (M extrinsic motives = 4.79 vs. M intrinsic motives = 5.90, SD = 1.17, t = 49.73, p < 0.001, see Fig. 4), thus, H1 was supported again.

The mediating effect of emotional engagement
The mediating role of emotional engagement was tested using SPSS PROCESS macro Model 4 provided by Hayes (2013). We used the tourism goal disclosure motives as the independent variable, emotional engagement as the mediator variable, and GDBs as the dependent variable. Three control variables were added to the model: goal difficulty, tourism experience, and post frequency. There was no significant difference in the control variables of goal difficulty, tourism experience, or post frequency between the two groups (goal difficulty: F(1, 116) = 2.837, p = 0.095; tourism experience: F(1, 116) = 0.424, p = 0.516; post frequency: F(1, 116) = 0.304, p = 0.583). Table 2 shows that the tourism goal disclosure motivation has a significant positive effect on emotional engagement (a 1 = 0.62, p < 0.01). Those individuals assigned to the intrinsic motivation condition had a higher goal emotional engagement than those assigned to the extrinsic motivation condition. Likewise, emotional engagement also has a significant positive effect on GDBs (b 1 = 0.73, p < 0.001). This result suggests that more emotional engagement was aroused in the intrinsic motivation context, the more participants tended to increase their GDBs. Finally, the indirect effect of tourism goal disclosure motivation on GDBs via emotional engagement (a 1 ×b 1 = 0.45) based on 5000 bootstrap samples was significant (Hayes, 2013), with 95% confidence interval (CI) excluding 0 (0.15-0.82).

The mediating effect of affective rumination
The mediating role of affective rumination was also employed SPSS PROCESS macro model 4 by Hayes (2013). A 95% CI of the parameter estimates was obtained by running the samples 5000 times. The tourism goal disclosure motivation was set as the independent variable (coded as intrinsic motivation = 0, extrinsic motivation = 1). Affective rumination was set as the mediator, and GDB was set as the dependent variable. Goal difficulty, tourism experience, and post frequency were set as control variables.
The bootstrapping results show that the tourism goal disclosure motivation has a significant negative effect on affective rumination (a 2 = − 0.70, p < 0.01). Affective rumination also has a significant negative effect on GDBs (b 2 = − 0.12, p < 0.05). The results confirmed that participants driven by extrinsic motivation were more likely to engage in affective rumination, which in turn reduce their GDBs. Finally, as per the discriminant method for the existence of mediating effect proposed by Hayes (2013), 95% CI does not contain 0 (0.01-0.19). Thus the mediating effect of affective rumination was significant (a 2 ×b 2 = 0.08). Therefore, affective rumination plays a mediating role between tourism goal disclosure motive and tourists' GDBs. We thus found support for H3. Table 2 provides more details on the results.

Study 3
The purpose of Study 3 was to examine the moderating effect of feedback valence on tourism goal disclosure motivation and GDBs (testing H4) and to explore the mutual transformation mechanism between intrinsic motivation and extrinsic motivation under different feedback valence situations (testing H5). A 2 (intrinsic vs. extrinsic motivation) × 3 (no feedback vs. negative feedback vs. positive feedback) factorial between-subjects design was adopted.

Pretest of stimuli
A total of 56 undergraduate students from a university in China were recruited in the pre-experiment. Among the participants, 46.7% were male, 53.3% were female, and all were aged 18-35. They were randomly assigned into two scenarios (negative feedback group n = 26 vs. positive feedback group n = 30). After reading the materials, they were asked to complete a scenario authenticity test and manipulation check as well as background information. The feedback was measured with an item on a 7-point Likert scale (1 = Strongly disagree, 7 = Strongly agree).

Results and discussion
The independent sample t-test was used to examine the effectiveness of the manipulation of feedback. The results indicated a significant difference in the feedback of the two groups (t = 4.786, p < 0.01): The positive feedback group (M positive = 5.09, SD = 0.46) had a higher score than the median value 4 (t = 21.81, p < 0.01), while the negative feedback group had a lower score (M negative = 2.93, SD = 0.46) than the median value 4(t = 7.616, p < 0.01). Therefore, the manipulation was successful. Fig. 4. The Influence of tourism goal disclosure motivation on GDBs.

Research design and procedure
This situational experiment is a 2 (extrinsic motivation vs. intrinsic motivation) × 3 (no feedback vs. negative feedback vs. positive feedback) factorial between-subjects design using the mixed materials from the pre-experiment of Study 2 mentioned above. We designed six versions of the tourism goal disclosure motives description and feedback valence as the stimuli materials (see Appendix B). After reading the materials, they were asked to complete the items of scenario authenticity test, manipulation check, emotional engagement, affective rumination, GDBs, and control variables as well as some demographic questions. In order to eliminate the impacts of goal difficulty, travel experience, and post frequency on the feedback valence, we included them as control variables in the model.
In the main experiment of Study 3, we posted recruitment information on the Chinese Credamo data platform (Credamo.com). All 316 respondents participated in the study and 290 complete questionnaires that were collected (extrinsic motivation and no feedback n = 50 vs. extrinsic motivation and negative feedback n = 50 vs. extrinsic motivation and positive feedback n = 46; intrinsic motivation and no feedback = 50 vs. intrinsic motivation and negative feedback n = 47 vs. intrinsic motivation and positive feedback n = 47). Among the 290 participants, 52.4% were male while 47.6% were female; 21.4% were 18-25, 49.3% were 26-35, 20.7% were 36-45, 7.2% were 46-55, 1.4% were 56 or older (the details are shown in Table 3.). Before hypothesis testing, the sample size estimation test was carried out. G* Power 3.1 was used to calculate the power value of the sample size (Faul et al., 2009). Two-way ANOVA was selected. When the number of groups was 6, the effect size (f) was 0.4, the significance level was 0.05, and the numerator df was 1, the power value of 290 samples is greater than 0.99, exceeding the basic level of 0.80, indicating that the effective sample size has statistical testing power.

Manipulation check
Scenario authenticity was determined and 97.6% of the subjects reported that the provided scenario was realistic. The independentsample t-test results indicate that the participants could correctly attribute the tourism goal disclosure motivation (M extrinsic motivation = 5.19, SD = 0.95; M intrinsic motivation = 4.48, SD = 0.94; t = 27.78, p < 0.001, both were significantly higher than the median value of 4). Furthermore, there were significant differences in their judgment of feedback valence (M negative = 2.64 vs. M positive = 5.38; t = − 22.138, p < 0.01). Thus, the manipulation of tourism goal disclosure motivation and feedback valence was successful.

Measurement reliability
The Cronbach's α value of emotional engagement, affective rumination, and GDBs-0.953, 0.884, and 0.942, respectively-were greater than the critical value, 0.700, which indicates that the data in our research have high reliability. Then, the mean score was used in the following analysis.

Mutual transformation of motivation
We examined our proposition that, in line with SDT, the dominant position of tourist intrinsic motivation and extrinsic motivation in tourism goal disclosure could transform under different feedback valence. Using the control group as baseline, we first compared the changes in motivation between the feedback group and the control group, and then revealed the mutual transformation process of intrinsic motivation and extrinsic motivation through independent sample T-test analysis (see Fig. 8). Results of the independent sample T-test analysis showed that under the dominant position of intrinsic motivation, when participants received positive feedback, the dominance of intrinsic motivation was strengthened (M intrinsic motivation = 6.15, SD = 0.56, vs. M control = 6.10, SD = 0.85, t = 0.35, p < 0.05), and extrinsic motivation was also significantly enhanced (M extrinsic motivation = 5.09, SD = 1.28, vs. M control = 4.02, SD = 1.53, t = 3.59, p < 0.001). Although extrinsic motivation was significantly enhanced compared with intrinsic motivation (Δ M intrinsic motivation = 0.05, SD = 0.96, vs. Δ M extrinsic motivation = 1.07, SD = 2.05, t = − 3.08, p < 0.05), the dominant position of intrinsic motivation did not change (M intrinsic motivation = 6.15, SD = 0.56, vs. M extrinsic motivation = 5.09, SD = 1.28, t = 5.59, p < 0.001). Thus, the difference between intrinsic motivation and extrinsic motivation was only narrowed, and there was no significant transformation between intrinsic motivation and extrinsic motivation. On the contrary, when participants received negative feedback, extrinsic motivation was significantly enhanced (M extrinsic motivation = 5.87, SD = 0.69, vs. M control = 4.02, SD = 1.53, t = 7.02, p < 0.001), and in turn intrinsic motivation was significantly weakened (M intrinsic motivation = 3.99, SD = 1.75, vs. M control = 6.10, SD = 0.85, t = − 7.32, p < 0.001). Since extrinsic motivation was significantly enhanced and replaced the dominant position of intrinsic motivation (Δ M intrinsic motivation = − 2.11, SD = 1.98, vs. Δ M extrinsic motivation = 1.85, SD = 1.81, t = − 10.07, p < 0.001), the dominance of intrinsic motivation disappeared (M intrinsic motivation = 3.99, SD = 1.75, vs. M extrinsic motivation = 5.87, SD = 0.69, t = − 6.31, p < 0.001). This suggests intrinsic motivation has been transformed into extrinsic motivation.

Mutual transformation of emotional engagement and affective rumination
Based on the results of the mutual transformation of motivation, we further explored its impact on the mediating path. The independent sample T-test analysis was used to examine the transformation of the mediating path caused by the mutual transformation of motivations (see Fig. 9). Results of independent sample T-test analysis showed when intrinsic motivation was dominant, positive feedback slightly weakened emotional engagement (M emotional engagement = 6.08, SD = 0.64, vs. M control = 6.21, SD = 0.51, t = − 1.10, p > 0.05), and in turn the affective rumination of the participants showed a significant increase (M affective rumination = 4.82, SD = 0.86, vs. M control = 3.30, SD = 1.41, t = 6.04, p < 0.001). Although the increase of affective rumination was greater than emotional engagement in the positive feedback situation (Δ M emotional engagement = − 0.13, SD = 0.84, vs. Δ M affective rumination = 1.52, SD = 1.72, t = − 5.59, p < 0.001), the dominant position of emotional engagement remained unchanged (M emotional engagement = 6.08, SD = 0.64, vs. M affective rumination = 4.82, SD = 0.86, t = 7.46, p < 0.001). Therefore, the difference between emotional engagement and affective rumination was only narrowed, and emotional engagement was not transformed into affective rumination. On the other hand, when participants received negative feedback, intrinsic motivation was gradually externalized, and the participants' emotional engagement towards the tourism goal decreased (M emotional engagement = 3.79, SD = 1.29, vs. M control = 6.21, SD = 0.51, t = − 12.64, p < 0.001), and in turn the rumination was gradually aroused (M affective rumination = 6.06, SD = 0.45, vs. M control = 3.30, SD = 1.41, t = 13.40, p < 0.001). The dominant path of emotional engagement was replaced by affective rumination due to this externalization of intrinsic motivation (Δ M emotional engagement = − 2.42, SD = 1.31, vs. Δ M affective rumination = 2.76, SD = 1.41, t = − 18.18, p < 0.001). Hence, the transformation from intrinsic motivation to extrinsic motivation further triggered the mutual transformation from emotional engagement to affective rumination (M emotional engagement = 3.79, SD = 1.29, vs. M affective rumination = 6.06, SD = 0.45, t = − 11.52, p < 0.001).

Moderated mediation effect
We conducted a moderated mediation analysis using bootstrapping mediation tests with 5000 replications and a 95% CI (Hayes, 2013). In Hayes Model 7, feedback valence served as the moderator for the effect of tourism goal disclosure motivation on tourists' GDBs, and tourists' emotional engagement and affective rumination served as the mediator. Three control variables were added in model: goal difficulty, tourism experience, and post frequency. There was no significant difference in the control variables of goal difficulty, tourism experience, and post frequency between the four groups (goal difficulty: F(1, 186) = 1.576, p = 0.211; tourism experience: F(1, 186) = 2.457, p = 0.119; post frequency: F(1, 186) = 0.118, p = 0.731)).
The results showed that the interaction between goal disclosure motivation and feedback valence significantly impacted GDBs through emotional engagement (β = − 0.21, SE = 0.12, 95% CI: 0.46 to − 0.01). In addition, the interaction between goal disclosure motivation and feedback valence significantly impacted GDBs through the affective rumination (β = − 0.22, SE = 0.11, 95% CI: 0.45 to − 0.02). Therefore, moderated mediation analysis results also supported the moderating role of feedback valence between goal disclosure motivation and GDBs. Specifically, in the positive feedback condition, extrinsic motivation transformed into intrinsic motivation and has a positive impact on GDBs through emotional engagement (E positive = 0.32, SE = 0.08; 95% CI: 0.17 to 0.49). On the contrary, when participants received negative feedback, the conditional indirect effects of goal disclosure motivation on GDBs through emotional engagement was not significant (E negative = 0.11, SE = 0.10; 95% CI: − 0.07 to 0.30). This result confirmed Hypothesis 5a. However, in the negative feedback condition, intrinsic motivation transformed into extrinsic motivation and has a negative impact on GDBs through affective rumination (E negative = − 0.15, SE = 0.06; 95% CI: − 0.26 to − 0.04). While participants received negative feedback, the conditional indirect effects of goal disclosure motivation on GDBs through emotional engagement was not significant (E positive = 0.08, SE = 0.09; 95% CI: − 0.11 to 0.26). Therefore, H5b was confirmed.

Conclusion
Integrating prior work and self-determination theory and the broaden-and-build theory of positive emotion, this study proposed and empirically validated the conceptual model of how the motivations of tourism goal disclosure on social media affect tourists' emotional engagement, affective rumination, and GDBs. Specifically, Study 1 and Study 2 showed that when tourists disclose their tourism goals under the condition of intrinsic motivation, they will be awakened to more emotional engagement related to the tourism goal and activate stronger GDBs. On the contrary, when tourists disclose their tourism goals under the context of extrinsic motivation, they will be more likely to fall into affective rumination related to the tourism goals and weaken the subsequent GDBs. Study 3 explored the moderating effect of feedback valence on the motivation of tourism goals disclosure and tourists' emotional engagement, affective rumination, and GDBs. We found that under the condition of positive feedback, tourists' extrinsic motivation gradually internalizes and awakens more emotional engagement, thereby enhancing the GDBs. Negative feedback weakens and externalizes tourists' intrinsic motivation, enhancing affective rumination, which in turn inhibits the GDBs. Furthermore, we found that emotional engagement and affective rumination play a mediating role in the interaction of tourism goal disclosure motivation and feedback valence on GDBs. Under the case of positive feedback, the impact of this interaction on the GDBs generates an indirect positive effect through emotional engagement, while in the case of negative feedback, the impact of this interaction on the GDBs generates an indirect negative effect through affective rumination. Finally, the significance, theoretical contribution, and managerial implications of this study were discussed.

Theoretical contribution
The goal disclosure process is both a self-disclosure process and a goal pursuit process; it is both a social interaction process and a selfrealization process. Previous studies on goal disclosure motivations examined the intrinsic motivation and the extrinsic motivation separately, based on the belief that intrinsic motivation and extrinsic motivation are opposite to each other, thus neglecting to explore the mutual transformation process between the intrinsic motivation and extrinsic motivation as well as the emotional experience and behavioral results behind the motivations from a dynamic perspective. However, it is critical to understand the boundary conditions of goal disclosure motivations transformation and the influence paths of individual psychological mechanisms under the dominance of different motivations. Therefore, this study validates the effect of feedback valence on tourists' emotional engagement, affective rumination, and GDBs through selfdetermination theory and positive emotion expansion-construction theory. This study enriches the application of self-determination theory in the field of tourism goal disclosure motivation in the context of social media, examines the influence mechanism of goal disclosure motivations on emotional engagement and affective rumination, and expands the theoretical literature regarding tourists' GDBs under the context of tourism goal disclosure motivations.
First, we verified the impact of goal disclosure motivations on tourists' emotional experience and behavioral outcomes in the context of social media, enriched the research on goal disclosure motivations, and expanded the practice of social psychology theory. Lavertu et al. (2020) and Munar and Jacobsen (2014) suggested that the motivations of individual participation in social media goal disclosure significantly affect individuals' online self-expression. However, few studies explore the impact of this feedback on individual GDBs in reality. According to self-determination theory, goal disclosure motivations are considered an important antecedent variable of cognitive, affective, and behavioral consequences related to basic needs (Gagné et al., 2015). Therefore, this study examined the motivation of tourism goals disclosure on social media, and found a significant positive effect on the GDBs of tourists with intrinsic motivation as well as a significant negative effect on the GDBs of tourists with extrinsic motivation. Thus, it further refined and deepened the research on tourist goal disclosure.
The second contribution arises in clarifying the mediating role between goal disclosure motivation and GDBs (Chang et al., 2020). Previous studies have found that work engagement, especially emotional engagement, is an important mediator variable in the work motivation and innovative behavior of employees (Pingel et al., 2019). Studies have also found that employees' work motivation and affective rumination have an effect on employees' absenteeism, emotional exhaustion, turnover intentions (Harter et al., 2002). However, literature discussing how tourist engagement and affective rumination are developed in the context of social media and what factors affect the engagement and rumination of tourists in tourism goals disclosure is minimal (Fang, Zhang, & Li, 2020). This study addressed the gap through examining the mediating role of emotional engagement and affective rumination in tourism goal disclosure motivation and GDBs in the context of social media. Specifically, when tourists disclose their tourism goal with intrinsic motivation, the strong emotional engagement triggered by the sense of accomplishment of goal attainment is awakened, and tourists generate greater GDBs. Whereas tourists disclose their tourism goal with extrinsic motivation, the anxiety about feedback expectations will trigger affective rumination, which will weaken tourists' GDBs.
We also examined the moderating effect of social feedback valence on the transformation of tourists' intrinsic motivation and extrinsic motivation. In the context of social media, tourists with different motivations to disclose their tourism goals will encounter either supportive elements (positive feedback) or thwarting elements (negative feedback) (Kaplan & Haenlein, 2010). A functional significance of self-determination theory is to use a set of rich tools to understand when and how factors such as feedback and evaluation will support and undermine the basic needs of the individual, which in turn triggers the transformation of individual motivation (Ryan & Deci, 2020). Therefore, this study uses the analytical framework of self-determination theory to deeply connect tourist's basic needs satisfaction with their cognitive and emotional underpinnings, and distinguish between two types of feedback valence (positive vs. negative). These results revealed the boundary effect of the transformation mechanism between intrinsic motivation and extrinsic motivation in a more detailed way, as well as the changes in emotional and behavioral outcomes caused by the transformation. It provided some new insights to enrich the feedback literature in the field of tourism goal disclosure motivation.
The study also verified that the mutual transformation between intrinsic motivation and extrinsic motivation is an important reason for the significant existence of moderating effect. We confirmed that, under the condition of positive feedback, the extrinsic motivation is partially internalized into intrinsic motivation, which in turn strengthens the tourists' GDBs; however, under the condition of negative feedback, the intrinsic motivation is partially externalized into the extrinsic motivation, thereby inhibiting the tourists' GDBs. These results not only demonstrated the important assertions of the internalization of extrinsic motivation proposed by self-determination theory, but also creatively put forward the important proposition of the externalization of intrinsic motivation. Therefore, this study effectively enriches and expands the research on self-determination theory in the field of tourism goal disclosure in the context of social media.
In its final contribution to the literature, this study explored the dynamic downstream consequence of the mutual transformation between intrinsic motivation and extrinsic motivation on the change of mediating path, revealing a series of psychological mechanism of the dyadic transformation of intrinsic motivation and extrinsic motivation on tourists' GDBs. The results clarified that the mutual transformation of intrinsic and extrinsic motivation produced dynamic downstream consequences, which further triggered the mutual transformation of tourists' emotional engagement and affective rumination. Combining between-subjects design experiment with within-subjects design experiment was used to cross-verify the dynamic downstream effect generated by the mutual transformation of intrinsic motivation and extrinsic motivation, which effectively remedied the deficiency of existing research on the dynamic downstream consequence generated by the mutual transformation of intrinsic motivation and extrinsic motivation. Therefore, offers better understanding of the psychological interaction mechanism of tourism goal disclosure motivation in the context of social media.

Managerial implications
This study has practical implications for online travel service providers (OTS) and destination marketing organizations (DMOs) seeking to implement precision marketing through social media channels. They are also of great significance for marketers seeking to better understand tourists' goal disclosure motivations and their motivation transformation, so as to adapt the social media marketing strategy according to the appeals of tourists with different motivation types.
The findings of Study 1 suggest marketers should identify the intrinsic and extrinsic motivations of tourism goal disclosure. Segmenting tourists according to goal disclosure motivations can lead to more effective marketing messaging. For example, with the aid of modern information technologies, OTS firms and DMOs could establish a complete set of tourism goals longitudinal tracking management system for tourists with intrinsic motivation to disclose their tourism goal. With the help of official marketing channels (such as official WeChat, Weibo, etc.), social media marketers could regularly push notifications related to tourism goals and destination products. For extrinsic motivation tourists, an online tourism social community might provide a reward mechanism for goal attainment. Social media marketers offer direct active interactions with community members, such as replies to tourists' postings or recommending original boutique tourism strategies to them, thereby reducing tourists' emotional rumination triggered by extrinsic motivations, so as to strengthen the psychological bond and GDBs.
Considering the positive impact of emotional engagement on the GDBs, OTSs could abandon the current one-way social media marketing strategy and cultivate long-term two-way interactive customer goal management. As past research shows, social media marketing of tourism products is a relationship-building process (Li et al., 2021). The active responses of marketers in the two-way social interaction relationship initiated by tourists and their tourism goals are essential to arouse tourists' emotional engagement in tourism goals. In particular, long-standing customer relationships should be the core of social media marketing strategies, because interactions between tourists and OTS firms and tourists' emotional engagement can be developed into valuable relational resources. For example, OTS firms and DMOs can host festival welfare events to provide travel coupons to tourists with tourism goals disclosed, which might effectively arouse the tourists' emotional engagement in tourism goals and create good conditions for the internalization of extrinsic motivation as well as enhance tourists' GDBs.
Finally, by unveiling tourists' externalization tendency for intrinsic motivation under the condition of negative feedback (vs. positive feedback), the present research suggests DMOs should utilize social media interactions and social networks as marketing resources and pay more attention to the comments on tourism goal disclosure, especially the responding to the negative comments quickly, so as to avoid undermining or externalizing tourists' intrinsic motivation by negative feedback, which may lead to the interruption or abandonment of tourists' GDBs. For example, when DMOs find negative comments on tourism goal disclosure, they should take timely remedial measures to eliminate the weakening effect of these negative comments on tourists' GDBs via effective means, such as reparations or apologies, and thus enhance tourists' GDBs. This study suggests feedback valency is effective in the capacity management and marketing strategy planning of destination marketing organizations. During the tourism goal disclosure stage, adopting feedback valency management can potentially be a flexible and effective solution for destination marketing organizations, which are feasible and acceptable social media marketing strategies.

Research limitations and future research directions
This study has some limitations that warrant future research. First, participants were all Chinese. To enhance the generalizability of the research conclusions, future studies might examine participants in other countries and cultures. Second, feedback valence was selected as the moderating variable, and other moderating variables such as selfconstrual and tourism goal craving (Mitev & Irimiás, 2020, p. 103111) could be examined in the future. Finally, GDBs were used as the outcome variable. Future research could explore the impacts of tourism goal disclosure motivations on other behavioral variables such as withdrawal behaviors and tourists' behavioral decision-making.

Impact statements
Although many people disclose their tourism plans on social media every year, not all tourism goals could be fulfilled in individuals' real life. It is still not clear what factors faciliate or hinder individuals to pursue their tourism goals. Based on self-determination theory and the broaden-and-build theory of positive emotion, this study investigated the motivations of disclosing tourism goals on social media and its impacts on tourists' goal-directed behaviors (GDBs). The findings of this study suggest that when tourists receive positive feedbacks on their disclosed tourism goals, extrinsic motivations could enhance individuals' emotional engagement and further promote more GDBs. However, when receiving negative feedback on their disclosed goals, intrinsic motivations could strengthen affective rumination and further hinder GDBs. The findings of this study provide theoretical and practical implications for destination marketers and online tourism service providers to develop matching social media marketing strategies.

Credit author statement
Lujun Su, Ph.D.: Conceptualization, data curation, methodology, project administration, funding acquistion, writing the original draft, review and editing. Xiaojie Yang, Ph.D. candidate: Conceptualization, methodology, data collection, formal analysis, validation, writing the original draft, review and editing. Yinghua Huang, Ph.D.: Conceptualization, methodology, writing the original draft, revising and editing, and responding to reviewers.