Modeling passionate decision

Article history: Received: July 7, 2020 Received in revised format: August 1


Introduction
High student loss is a severe problem faced by many small and private universities (Davis, 2014). In 2019, Fain reported that in the USA, there are around 22 % of students leaving their university with no credentials. In Indonesia, Tejo (2019) reported that in each yearly batch, there were around 40% of students dropped-out finally from their university. To answer why such phenomena occurred, the author offers a consideration from a decision-making point of view. Student loss is a global phenomenon. In short, the decision to leave a university is majorly caused mostly by lousy enrollment decisions.
To go deeper into that consideration, we need a clear understanding of what is the right decision. Scientists generally view the quality of a decision from two considerations. First, the process approach, in which the decision quality is judged based on how it is made. The approach that is generally used in a business context premises that, with the right decision, a decisionmaker has the highest chance to achieve a goal. They believe that a good process should generate good outcomes, and vice versa (Keren & de Bruin, 2003). However, this expectation cannot always be confirmed. There is no guarantee that a good process will generate good outcomes, and ill-defined processes will generate adverse outcomes. In reality, a good process can produce adverse outcomes, and a bad decision can end with good outcomes (Keren & de Bruin, 2003). Moreover, there is no agreement about decision-making process standards (Keren & de Bruin, 2003). The situation becomes more scattered because decision-making may involve subconscious steps free from the decision-makers' or judges' considerations (Willman-Livarinen, 2017).
Second, in the outcomes approach, the decision-makers' quality of decision is determined by the favorability of the decision outcomes (Keren & de Bruin, 2003), indicated by how satisfied the decision-makers are to their decision (Tybusrki, 2017). This point of view holds that the most satisfying choice is the most proper option according to the situation, not the best possible one to generate behavior outcomes (Keren & de Bruin, 2003;Tyburski, 2017). This approach is most common in a personal consumption context.
The problems with the outcomes approach are, first, many decisions are made under uncertainty. The decision-makers have no clear vision about what will happen with future decision outcomes (Tyburski, 2017;Chernev, Bockenholt, & Goodman, 2015). With this uncertainty, a satisfying decision can lead to adverse outcomes (Spetzler, 2017;Keren & De Bruine, 2003), and vice versa. Second, decision outcomes often occur in the long run (Meller, 2000), and decision satisfaction made at the moment of choice has few things to say about upcoming consumption satisfaction (Heitmann, Lehmann, & Hermann, 2007). Moreover, in participative services, decision outcomes are determined mostly by customer participation in the value creation process (Dong, Sivakumar, Evans, & Zon, 2014).
Some services have high credence quality, a category in which service receivers can conclude the quality of service after experiencing them in a particular term of consumption period (Kotler & Keller, 2016). A university service belongs to this category. Sometimes, the students require several years to conclude whether they are successful or fail to achieve their goals. During that period, as Vargo and Lusch's (2016) service dominant logic theory specified, the students are part of the valuecocreation process. The success of this process depends majorly on the commitment to the choice (Bubic, 2014), which is determined strongly by harmonious passion (Hadden & Knee, 2018).
The question, can we use harmonious passion as the primary indicator of decision quality? To answer this question, the author conceptualizes the concept of a passionate decision and then develops the model in which harmonious passion takes the position as the sole mediating variable that bridges the goal-directed decision and future choice engagement. In detail, the objectives of this study are as follows. First, what are the direct and indirect determinants of harmonious passion? Second, how are the inter-relationships among the determinants? Third, how is the efficacy of harmonious passion in detecting students' loyalty intention and leaving likelihood?
The concept, the model, and the research questions are still missing from scientific discourses so far. Hence, they are the original contributions of this study to the scientific world. The findings are also crucial for practical implementation, especially in explaining students high lost phenomenon described above. With the model, a university can make early detection of students' loyalty and leaving likelihood then take proper measures to increase loyalty and reduce burnout.

Passion and Passionate Decision
People with the same capability can demonstrate different success for the same task (Duckworth, Peterson, Matthews, & Kelly, 2007), because, besides capability, people need another personal property to make its implementation successfully (Duckworth et al., 2007;Howard and Crayne, 2019;Constantin, Holman, and Hojbota, 2011).
Perseverance and persistence are personality traits (Duckworth et al., 2007). Although it can be connected with the specific state (Howard and Crayne, 2019), these traits are inherited and unalterable and describe general tendency to stick with a task no matter how the situations are (Constantin et al., 2011). Goal striving and goal commitment are goal-related persistence. It is defined as the importance level of a goal, determined by effort incurred by an individual to reach it and unwillingness to abandon the goals despite he or she is confronted with setbacks or negative feedbacks of goal-achievement (DeShon and Landis, 1997).
Individuals' engagement to a goal is described well by the passion that describes a strong inclination to an essential activity toward which they are ready to incur their efforts and resources (Vallerand et al., 2003). Passion is a post-decisional experience related to a particular activity. Vallerand et al. (2003) divided passion into two categories, i.e., harmonious passion (HP) and obsession passion (OP). With harmonious passion, people internalize the activity into their identity as they see it as the essential thing in life. They motivate to engage in the activity autonomously or without enforcement to do so. With this approach, people are in a position to oschestrate the passionate activity to function in harmony with other aspects of their life.
Obsessive passion (OP) is also connected with an activity intentionally used to create one's identity. With this activity, an individual focuses more on creating excitement, self-esteem, or social acceptance. It has the potential to control one's life. When an individual becomes under its control, the passionate activity takes the dominant aspect in a person's identity and may becomes detrimental to other aspects of the person's life. For example, excessive passion for a long-distance touring activity may cause a person to abandon other essential activities.
In a university context, harmonious passion has the same meaning as Tinto's (1993) students integration concept. As cited by Hennig-Thurau, Langer, and Hansen (2001), he said that students' integration into their university could occur in two forms. First, academic integration, i.e., active participation university-arranged activities. Second, social integration, i.e., social relationships with fellow students. Tinto (1993) argues that, as Hennig-Thurau et al. (2001) stated, the congruence between students' identities with their institutions is required to create a high degree of students' integration to their university. Therefore, students' value systems, skills, and abilities should fit university requirements and educational service specifications. For this reason, in this study, the author chooses the harmonious passion, not an obsession, as the sole indicator of passionate decision.
A decision or choice (both are perceived as the same and will be used interchangeably) is an option or several options chosen from a group of options or alternatives. Decision-making is the process of choosing the most preferred alternative from several available alternatives (Pennington & Hastie, 1993). It is mainly a cognitive process that involves mental properties, such as perception, attention, and memory. In this process, the decision-makers compare the properties of available alternatives to choose the most justified option (Tyburski, 2017).
Passionate choice or decision is the one toward which the decision-makers have strong inclination because they love, internalized, and see it as an essential part of their self-identity. For some degree, passionate decision shares the same characteristics with involvement. First, the two concepts talk about an essential entity in a person's life because of their relevancy and contributions to his/her self-identity. Second, to find the most satisfying choice or decision, passionate decision, and involvement concepts specified, people are eager to spend time, money, and energy. However, the passionate decision is futureoriented, whereas involvement is a present-oriented concept. The passionate decision aims to explain a choice in which anticipated outcomes are generated mainly by individuals' engagement in the co-creation of value. The concept of involvement limits its horizon to an entity's role to satisfy individuals' needs and wants.

The Model of Passionate Decision
The model of passionate decision developed in this study uses Atkinson's (1957) expectancy-value theory, Pennington and Hastie's (1993) explanation theory, and Bandura's (1977) self-efficacy theory as underlying theories. Atkinson (1957) stated that the people involved in a task depend on the expectancy that their involvement will generate valuable outcomes and the perception that they have the capability of self-efficacy to perform the task. Self-efficacy influences goals-setting and the motivation to achieve them (Bandura, 1977). Pennington and Hastie's (1993) theory of explanation stated that people need an explanation for their decision. In the present developed model, the explanation is manifested in justifiability, decision confidence, and positive affect. Decision-makers' inclination or harmonious passion for their task represents the principle of co-creation value, in which a value is generated through collaboration between the service provider and consumers.
As modeled in Fig. 1, as the indicators of choice engagement, loyalty intention, and the likelihood of leaving are determined by harmonious passion. Self-efficacy, decision confidence, and positive affect influence harmonious passion. Positive affect and decision confidence is influenced by self-efficacy through justifiability and goals. Because of its nature as a contextrelated concept (Tyburski, 2017), as Westaby (2005) suggested, the operationalization of a passionate decision model depends on the context of the decision. In this study, the context of the decision is a university choice. In this context, we deal with goal-directed behavior in which decision is made with picture of the future in mind. This presumption colors the more detailed discussion about the components and the conceptual relationships among them in the following section.

Self-Efficacy
In seting personal goals, people usually evaluate their capabilities to check whether the goals are achievable or not (Atkinson, 1957;Eccless et al., 1983). In goal-directed behavior, in which people set goals and develop scenarios to achieve the goals based on, most of all, their capabilities. Self-efficacy is the individuals' belief that they can handle a task and reach goals. It can be individual or collective in scope (Bandura, 1977).
High self-efficacy people set up a higher goal and have more confidence in performing tasks to reach that goal. They perceived difficult tasks as a challenge instead of an obstacle or threat to be avoided (Bandura, 1977). Conversely, low-efficacy people view difficult tasks as obstacles or threats and focus on their deficiencies and look for reasons to get out instead of finding a way to cope with the difficulties. They have low motivation and a weak commitment to establish such a task (Schunk, 1991).
Perceived-difficulty in performing tasks some time is used as a proxy of self-efficacy. The high perceived difficulty is in line with low self-efficacy belief, and low perceived difficulty task is the reflection of high self-efficacy belief (Kraft, Rise, Sutton, &Roysamb, 2005).

Goals
Initially, goals are understood as results, consequences, or outcomes of a decision or behavior (Woodruff, 1997). Sometimes, the results are positive. If a decision or behavior is directed to pursue them, those results are called promotion goals. The result can also be detrimental. When a decision or behavior is directed to avoid or annihilate them, those results are called avoidance goals (Solomon, 2018). In its development, consumption goals also include experience, interconnected events, and ongoing process (Bagozzi & Dholakia, 1999).
Many behaviors are directed by goals (Bagozzi & Dholakia, 1999). Such behaviors can have many goals. The goals are hierarchical and interconnected. However, there is a focal goal that represents or predict other goals best (Bagozzi & Dholakia, 1999;Woodruff, 1997).
Studying in a university is a goal-directed behavior because it is driven by a willingness to achieve specified goals. Those goals may be defined widely according to above Bagozzi and Dholakia's (1999) concept; however, achievement goals concept is more salient in the education field.
In the early phase of its development, achievement goals are defined as the expected outcomes generated by skill-related factors owned by the high ability or self-efficacy people (Ames, 1992;Nicholls, 1984). Further, it conceived as covering the effort to avoid failure. In this way, Elliot (1999) conceptualized achievement goals as having three goals called trichotomous goals. It consists of mastery goals directed at developing task-related skills or competence, performance-approach goals purposed to achieve relative performance for satisfying ego goals, and performance avoidance-goals focused on avoiding being perceived as an incompetent one. High self-efficacy people conceptually have the first two goals, whereas the third category is generally regarded as belonging to low self-efficacy people. Elliot and McGregor (2001) added the fourth dimension, called mastery avoidance. This goal represents individuals' effort to avoid failure in mastering a skill or competence. With this inclusion, the model has two focuses (mastery and performance) and two valences (valence: approach and avoidance). Hence, it now consists of mastery-approach, mastery-avoidance, performance-approach, and performance-avoidance goals.
Many studies confirmed that the most widely used is the 2X2 model (Huang &Zhang, 2011;Rosas, 2015). However, the mastery-avoidance goal has no clear understanding (Pintrich, 2000). The formulation of hypotheses about the relationship of this element with actual performance is also challenging (Hsieh, Sullivan, & Guerra, 2007). Therefore, Elliot's trichotomous model (without mastery-avoidance goals) is the primary choice of many studies (Vandewalle, Nerstad, & Dysvik, 2019).
In the higher education context, we can expect that new students are more focused on approach motivation. For this reason, in addition to the above arguments, this study adopts the trichotomous model.

Justifiability
Justifiability is the decision-makers' ability to justify their choice (Heitmann et al., 2007). It consists of pros and cons toward a decision, cost, and benefits of the decision, and supporting and constraining factors (Westaby, 2005). To justify means to develop reasons, evidence, logic, or arguments to support a decision or choice (Heitmann et al., 2007;Westaby, 2005). Sufficient reasons and evidence ensure the righteousness of the decision or choice. Justifiably is also required to ensure that the decision-making process has gone through a satisfying process. Otherwise, the decision can be not be justified and will be followed by regret (Reb and Connolly, 2007).
Justification is not supposed to ensure that the decision is the best, but the most satisfied (Tyburski, 2017). This concept tells that the decision is justified according to the decision context. For example, enrolling in a second-class university for a student with limited resources may be justified and not be responded with regret. Reb and Connolly (2007) stated that others could make justification, or an individual can rely on his or her justification on others' arguments. This possibility is discussed in the accountability concept. It concerns whether the decision fulfills significant others' standards for justifying decisions and whether those standards are the same with internal standards.
The fulfillment of social standards creates the decision legitimation achieved when the decision makes sense to the decisionmaker and other stakeholders (Kleindorfer, 2008). Moreover, the anticipation of significant others' legitimation may have a significant effect on the decision, especially in more collective society.

Decision Confidence
Decision confidence is a perceived accuracy of a decision experienced shortly after the decision-makers made a decision (Heitmann et al., 2007;Chernev et al., 2015). As stated before, decision-making is the process of choosing the most preferred alternative from several available options (Pennington & Hastie, 1993). What is meant by accuracy is a perception of how close is the chosen option with an ideal option (Zha, Li, & Yan, 2013). The opposite is a feeling of doubt or anxiety about whether a decision-maker has made a good or bad decision.
Decision confidence determined by the easiness of available options comparison (Pennington & Hastie, 1993), availability of information (Heitmann et al., 2007;Lee & Dry, 2010;Phillips et al., 2016;Di Cagno & Grieco, 2019), the source of information and self-efficacy in information processing (Zha, Li, & Yan, 2013), and decision-making self-efficacy (Reed, Mikels, & Löckenhoff, 2012). When information is incomplete and inaccurate (Lee & Dry, 2010) or overload (Heitmann et al., 2007), the decision-makers tend to be less confident with their decision. Overconfident occurs when decision-makers hold a strong belief that they have made an accurate decision, but, is inaccurate (Phillips et al., 2016). When feeling confident with their decision, people tend to more satisfied with their decision (Wang & Sukhla, 2013) and free from negative emotions, such as regret and anxiety (Heitmann et al., 2007;Zeelenberg et al., 2008).

Positive Affect
The term "affect" functions as an umbrella that covers emotion, moods, and attitude (Bagozzi, Gopinath, & Nyer, 1999;Zeelenberg et al., 2008), including evaluation and preference (Zeelenberg et al., 2008). In short, any experience toward which we can address positive and negative valence (such as pleasantness and unpleasantness, pleasure and pain, favorable and unfavorable, goodness and badness, comfortability, and uncomfortably) is an affect (Zeelenberg et al., 2008). Therefore, in this study, when we talk about emotion, we also talk about affect.
Emotion is a mental state of feelings generated by cognitive appraisals of events or thoughts that can be expressed in many ways, such as physical gestures, posture, and facial features. People may take particular actions to manifest or to cope with their emotions. The choice of the activities depends on nature and meaning the feelings for the person who experiences them (Bagozzi, Gopinath, & Nyer, 1999;Zeelenberg et al., 2008).
Emotion function in decision-making (Bagozzi, Gopinath, & Nyer, 1999;Heitmann et al. 2007) because of the limitations of the cognition, information availability, and time the decision-makers have (Hanoch, 2002). Zeelenberg et al. (2008) asserted that many decision-making processes driven by emotions.
Emotions are basically known as positive (such as happiness, joy, pride, pleasure) and negative emotions (such as sad, disappointed, anxiety). That direction depends on whether a situation or experience is consistent or inconsistent with one's motives, needs, or goals. Consistent situations generate positive emotions, and vice versa (Roseman, 1991).
People can experience positive and negative emotions simultaneously, but only for a short period. Theorists stated that in that co-existence, negative emotions create a state of disequilibrium that negatively contributes to one's wellbeing (Lazarus, 1991). To return to a state of equilibrium, people try to weaken negative emotions (Bagozzi, Gopinath, & Nyer, 1999). In this weakening process, positive emotions are used to affect negative emotions (Frederickson, 2001). The effectiveness of that process depends on the efficacy of positive emotions. Thus, people will strengthen positive emotions (Frederickson, 2001;Langeland, 2014) to compensate for negative emotions (Bagozzi, Gopinath, & Nyer, 1999).

Loyalty Intention
At the beginning of its development, scientists used repeat purchasing as the sole indicator of consumer loyalty. In 1994, Dick and Basu added attitude as another indicator of consumer loyalty. They argued that repurchase behavior does not indicate real loyalty. Habits can drive it. It should be accompanied by a positive attitude to the brand to develop true consumer loyalty. Aaker (1991) stated that commitment is the key indicator of consumer loyalty. With the commitment, consumers should have been satisfied by a brand, like the brand, see the brand as a friend, recommend the brand to potential buyers, and defend the brand from contrary opinions. With the same spirit, Oliver (1999) defined consumer loyalty as a deep commitment of the consumer to a product or brand and willingness to hold it through amid the existence of reasons to leave it and switch to other options (Oliver, 1999). Fournier (1998) viewed brand loyalty as a love-based dyadic relationship. She said that true loyalty called committed partnership should be built voluntarily and socially supported relationships based on love, trust, commitment, intimacy, and commitment despite in the presence of adverse circumstances. Beyond that, many forms of relationship between consumers and their band. However, all of the forms shows that the key indicator of consumer loyalty is brand commitment.
The voluntary nature of consumers' involvement in brand relationships, as found in Fournier's (1998) committed partnerships, is almost a utopia. In a free-choice situation, most of the consumers' involvement in a brand relationship is transactional. Consumers' value generated from the relationship drives their commitment. In other words, people will give their commitment as long as the relationship is beneficial.
Although they acknowledged the efficacy of commitment, most researchers treat consumer loyalty as a multidimensional construct. It consists of attitudinal and behavioral loyalty (Dick and Basu, 1994;Oliver, 1999), including switching behavior (Bansal, Taylor, & James, 2005;Oliver, 1999). Attitudinal loyalty describes a relative attitude to a brand, corporate, or store. Repeat purchases identify behavioral loyalty to a brand or repeat shopping to a store. Switching behavior indicated by shifting from one brand to another brand.
The use of repeat purchases should be with caution. It does not describe the commitment to a brand. Repeat purchases can reflect habits (Aaker, 1991) or spurious loyalty (Dick and Basu, 1994). The most important thing, repeat purchase can only be applied to a product or brand that bough frequently on a regular base. In the case of infrequent buying or bough only once commonly, this indicator is not relevant. Attitudinal loyalty is more profound than mere behavior. It answers why individuals purchase a brand repeatedly. The answer may cover factors, such as liking to a brand as a friend (Aaker, 1991) or partner to be loved (Fournier, 1998), toward which the consumers act as marketers that share advocation and positive word of mouth voluntarily (Aaker, 1991). With those factors, one can predict whether the consumers stick to the brand or leave it in the future (Bansal, Taylor, and James, 2005). Consequently, in this study, attitudinal loyalty and leaving likelihood are two indicators of choice loyalty, because, as Evanschitzky, Iyer, Plassmann, Niessing (2006) stated, they reflect the commitment to a brand.

The Impact of Justifiability
People need reasons to support or defend the option they chose (Heitmann et al., 2007;Pennington & Hastie, 1993;Westaby, 2005). Thus, people feel better when they find that their decision is justifiable, and subsequent behavior is reasonable (Pieters & Zeelenberg, 2005). Wise reasoning generates higher wellbeing, social acceptance, and surpress negative affects (Grossman et al., 2013).
A poor decision creates regret or self-blame feeling (Connollly and Zeelenberg, 2002), and the right decision generates positive emotions or feelings (Heittman, Lehman, & Herman, 2007). Individuals should feel better and more confidence with their decision (Pieters & Zeelenberg, 2005;Heitmann et al., 2007). Westaby (2005) asserted that a reason has instrumental value because it helps people to understand their behavior, behavior of others, and causal relationship of their behavior with internal and external factors. He presumed that individuals could trace their memory to find the most justifiable option. When the most justified option found, people can implement their decision with confidence.
Reasons answer the question about why the goals should be pursued. There are two possible answers to this question. First, the pursuit of goals is fun and enjoyable. Second, the outcomes are essential. The enjoyability of goals pursuit and the importance of the outcomes generate a sense-of-challenge (Sommet and Elliot, 2017). The challenge creates self-efficacy. Up to a particular level, the more challenging is a task, the higher is the self-efficacy (Bandura, 1997). Consequently, a more reasonable decision will create higher self-efficacy. Based on the above arguments, the author proposes the following hypotheses: H1: The more justifiable is a decision, the higher is emotional wellbeing. H2: The more justifiable is a decision, the higher is decision confidence. H3: The more justifiable is a decision, the higher is self-efficacy.

The Impacts of Self-Efficacy
The primary impact of self-efficacy is achievement motivation (Bandura, 1977;Domenech-Betoret, Abellan-Rosello, and Gomez-Artiga, 2017). High self-efficacy people are more confident in performing tasks, set up higher goals, and show a willingness to "mastering a task" (Hsieh, Sullivan, and Guerrera, 2007). They are more receptive to challenging tasks, for they perceive it as a challenge to be mastered instead of a threat to be avoided (Bandura, 1977). They also show greater motivation (Bandura, 1977;Domenech-Betoret et al., 2017;Schunk, 1991). Conversely, low self-efficacy people are less receptive to a difficult task. They see it as a threat rather than a challenge. Their commitment to their goals is low. They tend to focus on their deficiencies and look for reasons to get out instead of finding a way to perform a difficult task (Bandura, 1977) successfully.
In the education field, achievement goals orientation represents achievement motivation (Rosas, 2015). In general, previous studies found that self-efficacy positively affects mastery goals (Hsieh, Sullivan, and Guerra, 2007). However, there are inconsistencies regarding the influence of self-efficacy on performance-approach and performance-avoidance goals.
Previous studies found a negative correlation between self-efficacy and performance-avoidance goals (Elliot, 1999;Middleton and Midgley, 1997;Pajares, Britner, & Valiante, 2000). On the other hand, Hsieh, Sullivan, and Guerra (2007) found that self-efficacy correlate positively with performance-avoidance goals. Surprisingly, low self-efficacy students show a tendency to set up high performance-approach goals.
In this study, the author believes that self-efficacy positively influences mastery goals and performance-approach goals. However, following inconsistent results of previous studies, the author has no strong theoretical reasons to formulate the influence of self-efficacy on performance-avoidance goals. Therefore, in this study, the author proposes the following hypothesis: H4a: Self-efficacy influences mastery goals positively. H4b: Self-efficacy influences performance-approach goals positively.
Besides its influence on achievement motivation, another issue about self-efficacy is its influence on harmonious passion. Previous studies revealed that students with higher self-efficacy show greater persistence (Vallerand, 2012). Theorists of the expectancy-value theory (e.g., Atkinson, 1957;Wigfield & Eccles, 2000) stated that the persistence to pursue goals, besides the value of goals achievement, is also influenced by the perception hold by individuals that they can perform the behavior to achieve goals. Besides, people use the passionate activity as the signature of their strength (Forest, Mageau, Crevier-Braud, & Bergeron, 2012). Therefore, the author expects that self-efficacy influences harmonious passion, as stated in the following hypothesis: H5: Self-efficacy influences harmonious passion positively.

The Effect of Achievement Goals on Positive Affection and Decision Confidence
Several researchers (e.g., Dweck and Elliott, 1983;Nicholls, 1984) suggested that each category of achievement goals associate with different patterns of coping and emotion. Elliot and Dweck (1988) reported that children made two general types of coping patterns in achievement-related situations. Children with challenge avoidance tendencies demonstrated low persistence, express negative affect, and negative self-cognition when facing difficulties or obstacles, whereas children with challenge approaches show the reverse. Kaplan and Maehr (1999) outlined the same argument. They specify that task and ego goals trigger different behavioral, coping, and emotive behavior. Students with ego goals view success in social comparison terms. Concerning self-esteem, this point of view is vulnerable to negative emotions because success is a limited commodity. In this situation, being a loser has a higher possibility than being a winner. Only a few students can achieve as the winners, and most are end up as losers. In contrast, when faced with difficult situations, students who pursue task goals view a problematic situation as a challenge, hold more optimistic orientation, maintain positive affect, and implement problem-solving strategies.
Most recent studies confirmed Kaplan and Maehr's (1999) work. Tuominen-Soini, Salmela-Aro, and Niemivirta (2008) found that mastery approach goals were positively associated with various indices of wellbeing. In contrast, they found that mastery avoidance, performance-approach, performance-avoidance correlate negatively with that variable. Tian, Yu, and Huebner (2017) found that performance-avoidance goal orientations negatively correlate with wellbeing. These arguments are formulated in the following hypothesis: H6a: Mastery goals influence anticipated positive affection positively. H6b: Performance-approach goals influence positive affection negatively. H6c: Performance-avoidance goals influence positive affection negatively. Elliot and McGregor (2001) found that mastery goals have the lowest correlation with test anxiety traits than the other two categories. More specifically, they found that test anxiety has a correlation coefficient of r = 0.08 for a mastery-approach goal, of r = 0.29 for a performance-approach goal, and of r = 0.56 for a performance-avoidance goal. They also found no evidence to reject the null hypothesis for the relationship of fear of failure with a mastery-approach achievement goal. On the other hand, positive relationships fear of failure with performance-approach and performance-avoidance goals are confirmed. Vandewalle (1997) also found that fear of negative evaluation correlates negatively with a learning goal orientation and positively with performance-proven as well as performance-avoidance goal orientations. In sum, decision confidence correlates positively with mastery goals, and negatively performance-approach and performance-avoidance goals, as stated as follows: H7a: Mastery goals influence decision confidence positively. H7b: Performance-approach goals influence decision confidence negatively. H7c: Performance-avoidance goals influence decision confidence negatively.

The Effect of Emotional Wellbeing on Harmonious Passion
In Harmonious passion, people inclination to activity because they love it (Vallerand et al., 2003;Whang et al., 2004), not by coercion to do it. Love to an activity or brand is generated by liking (Aaker, 1991;Fornier, 1998). Liking is constructed by positive emotions to its object, such as satisfied, pleasure, happy, interested, and comfortable. Therefore, positive emotions represented by emotional wellbeing will contribute to harmonious passion, as formulated as follows: H8: Positive affection contributes to harmonious passion positively.

The Effect of Decision Confidence
The internalization of activity generates harmonious passion for one's identity, in which people are free to choose, involve, or retract themselves from an activity. They develop passionate activity and energize the effort intrinsically to make it successful because of believing it as an essential thing in developing their identity (Ryan and Deci, 2000;Vallerand et al. (2003).
In decision-making, the chosen activity with which decision-makers feel passionated is the output of the evaluation of alternatives (Pennington and Hastie, 1993). The author proposes that for being passionated, besides its importance to one's identity, the chosen option should satisfy the decision-makers confidence that they have made accurate a decision accurately. If the decision-makers feel doubt or dissonance with the decision, they will not be passionate about it. For example, when the features are difficult to compare (Wang & Sukhla, 2013), or the information or the number of alternatives is overloaded (Heitmann, Lehman, & Herman, 2007;Wang & Sukhla, 2013), the decision-makers will feel confused, and decision satisfaction decreases. Moreover, uncertainty makes decision-makers feel less satisfied with their decision (Politi, Clark, Ombao, Dizon, & Elwyin, 2011). Therefore, to be passionate about their decision harmoniously, the decision-makers need to feel confident about their decision, as stated in the following hypothesis: H9: Decision confidence influences harmonious passion positively.

The Effect of Harmonious Passion
In its definition mentioned before, strong inclination means the decision-makers love the choice, see it as an essential thing in life, and be ready to invest time and energy to make it successful. With this qualities, we can expect that people will develop a commitment or loyalty to their choice (Heitmann, Lehman, & Herman, 2007), advocate the brand, and recommended the brand to others (Aaker, 1991), including developing a personal relationship (Aaker, 1991;Fournier, 1998) or engagement with and reduce the leaving from the choice (Stoeber, Childs, Hayward, & Feast, 2011), as stated in the following hypothesis: H10: Harmonious passion influences loyalty intention positively. H11: Harmonious passion influence leaving likelihood negatively.

Research Site, Sample and Data Collection Method
The research of this kind ideally should be conducted in a university or college toward which the choice is generally made under self-consideration, as Ryan and Deci's (2000) self-determination theory expects. That consideration should be free from the halo effect generated by brand or university popularity. The decision should be based mostly on the evaluation of the pros and cons of the university. Thus, the research took place at a private business college in North Jakarta, Indonesia's capital city. This college accepted 521 new students in 2019, and as many of 350 students accepted the invitation to participate voluntarily (response rate was 67.18%). They consisted of 198 males (56.6%) and 152 (43.4%) females with an average age of 18.29 years, and the median was 18 years.
The author distributed the questionnaires online. The author sent the link to all new students via Whattsapp and demanded they fill the questionnaire any time before the class began in the following week. With this approach, the respondents are free from subjective bias caused by the researcher's presence (Malhotra, 2020). The author ensures their freedom by intentionally designing questionnaires free from personal identification questions. In other words, the author treated all of the participants unanimously. There is no reward given for each participation. The voluntarily and unanimously nature of the research hopefully increases the quality of data.
The order of the questions was intentionally randomized to reduce position bias (Malhotra, 2020). The author limited the number of questions to the point that the participant feels it as acceptable.

Measurements
The measurements were from previous studies. Self-efficacy measurement was from Pintrich, Smith, Garcia, & McKeachie (1991). Elliot and Murayama (2008) was the source of achievement goals measurements, i.e., mastery, performance-approach, and performance-avoidance goals. The author adapted justifiability and decision confidence measurements from Heitmann, Herman, and Lehman (2007). The 5-item World Health Organization Well-Being Index (WHO-5) Version 1988 was adopted to measure positive affection. The measurement of harmonious passion followed the work of Zhao, St-Louis, Vallerand (2015). Loyalty intention and leaving likelihood were from Kumari and Patyal (2015). Due to its sensitivity, the author develops a single item measurement of leaving likelihood for this study so that the respondent should not frequently respond that may disturbing their feeling of privacy.
Three research specialists from Kwik Kian Gie Research Center helped evaluate the first version of the measurements to match the local context. The evaluation ended with the modification of two question items. In the decision confidence measurement, the original question, "It was impossible to be certain which product fits my preferences best," is replaced by: "My decision to choose this university was the best decision possible for me personally." In the decision justifiability measurement, the original question, "In order to choose this university, it was not necessary to make any difficult trade-offs," was found unworkable. It replaced by, "My decision to choose this university was a wise one." Because some respondents have were not fluent in English, the original questions translated into the Indonesian language. Then, to ensure that the translation did not change each question's meaning, the translated version was retranslated into English. Two English lectures evaluate the original and retranslated version of the measurements. When they concluded that both have the same meaning, the author used the Indonesian version of the measurement.
The author utilized five levels of Likert-type scale in all measurements to record the response. It was ranging from 1 (strongly disagree) to 5 (strongly agree). Exceptionally, the likelihood of leaving measurement used adversely scoring, ranging from 5 (strongly disagree) to 1 (strongly disagree).

Validity and Reliability Analysis
Confirmatory factor analysis (CFA) with LISREL 8.8 analyzes the validity of the instruments. The results are displayed in Appendix 1. We can see that all constructs are valid (loading factor>0.5, average variance extracted≥0.5, and composite reliability >0.6, as suggested by Hair et al. (2016) and Malhotra (2020). To achieve these requirements, the author made two adaptations. In harmonious passion measurement, the original statement, "My study in this university will allow me to live a variety of experiences," was skipped because, with this item, the measurement is non-valid (AVE=0.49). For the same reason, the author skipped the original statement, "I will have special attachment or emotional feelings towards this institution" from loyalty intention measurement.
All of the measurements are reliable, as indicated by Cronbach Alpha value that surpasses the minimum threshold of 0.70, as suggested by Hair et al. (2016) and Malhotra (2020). It means that if the instrument is used to repeatedly measure the specified variables, it will proceed with consistent results.

Mean and Correlation
As shown in Table 1, the average of each construct, except the leaving likelihood (below the score of 3), is far above the score of 3 and approaches the score of 4 or above that score (mastery goals). It means that respondents give positive respondents to each construct, except leaving likelihood.
Despite leaving likelihood has the lowest mean, special attention should be paid on it. As many of 106 respondents somewhat agree, 72 respondents agree, and 16 respondents strongly agree or 194 respondents (55.43%) in total to the statements: "In the future, you may leave this university." This result supports the report delivered publicly by Tejo (2019), in which around 40% of the student in Indonesia for various reasons finally leave their university with no credentials. The correlation between constructs that have direct relationships with each other ranges from very low to very high. The correlation between harmonious passion (pas) with loyalty intention is the highest one (R=0.99), and the lowest one is its correlation with leaving likelihood (R=-0.16). Interestingly, the correlations with emotional wellbeing (R=0.83), decision confidence (R=0.80), and self-efficacy (R=0.91) are also high. High correlations also occur between self-efficacy with mastery goals (R=0.97), performance-approach goals (R=0.87), and performance-approach goals (R=0.79), as well as between justifiability and decision confidence (R=0.94).

Structural Model
The As exhibited in Fig. 2, the path between justifiability and emotional wellbeing is significant (γ11=0.55, t=7.97, α<0.0001); therefore, H1: "The more justifiable is a decision, the higher is emotional wellbeing," is supported. The path between justifiability and decision confidence is also significant (γ61=0.91, t=11.92, α<0.0001), and there is enough evidence for H2: "The more justifiable is a decision, the higher is decision confidence." As expected, the path between justifiability and selfefficacy is also significant (γ21=0.68, t=10.61, α<0.0001); therefore, the hypothesis: "The more justifiable is a decision, the higher is self-efficacy" (H3) is confirmed.
Mastery goals (η4) show its efficacy in determining emotional wellbeing (η1) positively and significantly (β41=0.46, t=3.28, α<0.001), as specified in H6a: "Mastery goals influence anticipated positive affection positively." The same result is also apparent in its influence on decision confidence (β63=0.47, t=3.48, α<0.001) that confirms H7a: "Mastery goals influence decision confidence positively." The study fails to confirm the direction and influence of performance-approach goals on emotional wellbeing because the path between the two constructs is positive and non-significant (β14=0.045, t=0.43, α>0.05), therefore H7b, "Performance-approach goals influence decision confidence negatively," is not confirmed. However, the study is successfully to confirm H7b: "Performance-approach goals influence decision confidence negatively" (β64=-0.21, t=-2.08, α<0.05). The path between performance-avoidance goals with emotional wellbeing is negative and significant (β15=-0.22, t=-2.59, α<0.01). Therefore, H7c: "Performance-avoidance goals influence decision confidence negatively" is confirmed. Its negative influence on decision confidence stated in H7c: "Performance-avoidance goals influence decision confidence negatively" is also confirmed by the study (β65=-0.28, t=-3.45, α<0.01). The contribution of emotional wellbeing to harmonious passion specified in H8: "Positive affection contributes to harmonious passion positively" is confirmed by the study (β71=0.27, t=4.95, α<0.001). The same result also was shown by decision confidence, in which its fluence on harmonious passion specified H9: "Decision confidence influence harmonious passion positively" is supported by the study. Interestingly, both constructs, together with self-efficacy, can explain 93% harmonious passion. Therefore, the variance of harmonious passion explained by other variables is only 7%. The influence of harmonious passion on loyalty intention (β87=0.99, t=11.70, α<0.0001), specified in H10: "Harmonious passion influence loyalty intention positively," is confirmed by the study. This construct explains as much of 99% (R 2 =0.99) loyalty intention's variance. As expected in H11: "Harmonious passion influence leaving likelihood negatively" is also confirmed by the study (β87=-0.17, t=-2.90, α<0.01). However, harmonious passion can only explain this construct as much of 2.6%.

Total Influence on Harmonious Passion
What makes a decision being passionated or not passionated? To answer this question, the author investigates the total effect of its direct and direct determinants on the variable under investigation represented by t-value, as shown in Table 2. We can see those positive determinants of harmonious passion are justifiability, emotional wellbeing, decision confidence, self-efficacy, and mastery goals. Performance-avoidance goals demonstrate adverse determination on harmonious passion. Basedon t-value, justifiability and self-efficacy are the two strongest positive determinants of harmonious passion.

The Impact of Justifiability
As expected, justifiability influences decision-makers' wellbeing, decision confidence, and self-efficacy. Its effect on wellbeing repeated the result found by Grossman et al. (2013). They found that wise reasoning is associated with higher life satisfaction and less negative affect. However, they said that we should distinguish abstract reasoning from wise satisfaction. They said wise reasoning based on pragmatic considerations is more reliable than abstract reasoning. It is mainly generated by learning and experience. Although the participants mainly have no experience in choosing a university, their reasoning can fulfill the properties of wise experience. Choosing a university is a high involvement behavior for which individuals collect sufficient information before making a decision. Therefore, the keyword "easy to justify," "good choice," and "wise decision," despite impressing abstract measures, are based on cognitive considerations. Therefore, the author believes that the justifiability in this study is wise reasoning. As stated before, justifiability is the process of compiling reasons for making a decision. When the reasons are there, self-efficacy will follow, as shown by the significant positive influence of justifiability on self-efficacy. However, in that role, the determination of justifiability is limited (R 2 =0.47). It means that in this decisionmaking, self-efficacy is more influenced by other variables than by the justifiability. Decision confidence is the primary impact of justifiability in this study, as shown by the highest influence on this variable (R 2 =0.92) than on wellbeing (R 2 =0.66) and self-efficacy (R 2 =0.47). It means that the perceived accuracy of the decision reflected in decision confidence is based mainly on cognitive considerations.

The Impact of Self-Efficacy
This study reveals that self-efficacy influences achievement goals and harmonious passion positively. Its influence on mastery goals follows the same track with Hsieh, Sullivan, and Guerrera (2007) with high determination (R 2 =0.94). The effect on performance-approach goals is positive and the same with Middleton and Midgley (1997) and Pajares, Britner, and Valiante (2000). However, with the determination coefficient of R 2 =76, the effect on this goal orientation is lower than that of masterygoals. This result is intriguing, because, as stated before, performance-approach is a goal orientation in which people see the competence or talent as the primary factor of success. Success in achieving mastery goals, on the other hand, depend mainly on effort. As specified before, self-efficacy influences harmonious passion positively and significantly. This result asserts the notion proposed by Vallerand (2012) that self-efficacy contributes to persistence positively. Interestingly, among the three direct determinants of harmonious passion, based on t-value, self-efficacy has the highest contribution (Table 2). This result strengthens the premise proposed by expectancy-value theory (Atkinson, 1957;Wigfield & Eccless, 2000) that motivation is influenced partly by a belief that one can perform a task under consideration.

The Impact of Achievement Goals
This study strengthens the previous finding found by Tuominen-Soini, Salmela-Aro, and Niemivirta (2008) in which mastery approach goals were positively associated with various indices of wellbeing. In contrast, in this study, goals that reflect the concern to avoid being incompetent (performance-avoidance goals) influence wellbeing negatively as also found by Tian, Yu, and Huebner (2017). The influence of performance-approach goals on wellbeing is problematic. In this study, although positive, the path between variables are not significant. This result differs from Linnenbrink (2004), who found that performance-approach goals function detrimentally on wellbeing. It seems that we need to think about the context where the self-efficacy work. In Linnenbrik's work, as mentioned earlier, the research was conducted in a math class, and the participants are intentionally stimulated to only focus on performance-approach goals. In this study, the element was measured in the freefrom stimulation situation, and its existence was also stood together with the mastery and performance-approach goals. The impact on decision confidence follows the same pattern, in which mastery goals influence decision confidence positively, whereas performance-approach and performance-avoidance goals influence decision confidence negatively. These results strengthen the notion that students with mastery goals see that success depends mainly on their effort (Kaplan & Maehr, 1999).
On the other hand, students with performance-approach goals know that the opportunity to be a winner is limited. Besides, they still have no idea who their rivals are and how good they are in this business. With all of this, they feel less confident in their decision. The result of this study confirms the notion proposed by Elliot and Moller (2003) that performance-avoidance is deleterious. It creates anxiety and fear of failure (Elliot & McGregor, 1999;Vandewalle, 1997). In a short-term longitudinal setting, Shim and Ryan (2005) found that students with performance-avoidance goals tend to experience diminishing motivation in their study. Therefore, it makes sense if the decision-making, this goal orientation correlates with decision confidence negatively.

The Impact of Emotional Wellbeing Harmonious Passion
The positive influence of positive affect found positive and significant on harmonious passion are in the same way as Curran, Hill, Appleton, Vallerand, Standage's (2015) meta-analysis, in which 24 studies they studied, found the same result. Then, they conclude that positive affect is part of harmonious passion. It means that a strong inclination to an object caused by the fact that passionate people have a comfortable feeling on the object.

The Impact of Decision Confidence on Harmonious Passion
This study reveals that decision confidence influences harmonious passion positively. In rational and high-involvement decision-making, as stated before, a decision is the outcome of evaluating various options (Pennington and Hastie, 1993). The perception that they have made an accurate decision is the most fundamental prerequisite for being passionate about a decision. Curran et al. (2005) stated that decision confidence is the cognitive element of harmonious passion. However, they said that such influence only matters when the decision-making occurs according to the self-determination theory principle. According to Ryan and Deci (2000), motivation depends entirely on the decision-makers' autonomy and competence.

The Impact of Harmonious Passion
As expected, harmonious passion positively influences loyalty intention. The determination is high (R2=98%). It means that in this model, harmonious passion almost being the sole determinant of loyalty intention. This result confirms that harmonious passion identic with persistence to, love for, liking, and positive word of mouth about the choice (Curran et al. 2015). This study also shows that harmonious passion influence leaving likelihood negatively as also found by previous studies. In a study among modern China's government employees, Gong, Zhang, Ma, Liu, and Zhao (2018) found that harmonious passion correlates negatively with employees' burnout. Stoeber, Childs, Hayward, and Feast (2011) find that harmonious passion correlates positively with student engagement (vigor, dedication, and absorption) and negatively with (exhaustion, cynicism, and inefficacy). However, in this study, harmonious passion only explains 2.6% of leaving likelihood. This weak determination may be caused by complex considerations in a leaving decision that includes push, pull, and mooring factors (Bansal Taylor, and James, 2005). Those factors can only be concluded after the decision is implemented. Therefore, harmonious passion for the choice, at the time the educational service has not been experienced, has weak determination in explaining leaving likelihood.

General Discussion
This study finds that passionate decision factors are justifiability, positive affect, decision confidence, self-efficacy, and mastery-goals. Performance-avoidance goals contribute negatively to passionate decision. These factors intersect partly with Heitman, Lehman, and Herman (2007) work which found that the factors that influence decision satisfaction, among others, are decision confidence, justifiability, and positive emotions. However, both concepts are different. Decision satisfaction is a present-oriented concept, whereas harmonious passion is a present and future-oriented concept. ,In their cognitive model of decision making, Small and Venkatesh (2001), decision confidence is the indicator of decision quality. In the emotional-based decision-making, Zeelenberg, Nelissen, Breugelmans, and Pieters (2008) use positive affection as the main factor. Moreover, for Dooley and Fryxell (2017), commitment is the indicator of the decision quality. Harmonious passion used as the sole indicator of passionate decision reflects the persistence or commitment to the decision. Therefore, this model takes the role of commitment proposed by Dooley and Fryxell (2017) as the indicator decision quality. This model also covers Small and Venkatesh's (2001) model through decision confidence and Zeelenberg et al. (2008) model through positive affect. Besides, the passionate model also represented self-efficacy and expected outcomes or goals specified in expectancy-value belief theory (Atkinson, 1957;Eccless et al., 1983). This model confirms the high adaptive nature of mastery-goals shown by the positive influence of this variable on positive affect and decision satisfaction directly and harmonious passion indirectly. The detrimental effect of performance-avoidance goals (Elliot and Moller, 2003;Elliot and McGreagor, 1999;Vandewalle, 1997;Shim and Ryan, 2005) is also confirmed the model shown by its negative effect on positive affect and decision confidence directly and harmonious passion indirectly. Interestingly, this model also covers the inconsistent effect of performanceapproach goals, as Elliot and Moller (2003) stated.

Conclusion
Direct determinants of harmonious passion are positive affect, decision confidence, and self-efficacy, whereas the indirect determinants are justifiability and mastery goals. Performance-avoidance goals act as negative indirect determinants of harmonious passion. Harmonious passion has a high capability in explaining loyalty intention and limited capability in explaining leaving likelihood. As a whole, the model is a good fit and has good nomological validity.

Limitations and Direction for Future Research
This study intentionally focused on harmonious passion, exclude obsession passion, as the representation of passion and.
Although harmonious passion is more adaptive than obsession passion (Vallerand et al. 2003: Vallerand, 2012, further research can investigate whether harmonious passion is better than the obsessive passion for that role.