How Passion in Entrepreneurship Develops Over Time: A Self-Regulation Perspective

Abstract

We adopt a self-regulation perspective to present a model of the development of passion in entrepreneurship. We argue that entrepreneurial self-efficacy and performance influence the two components of passion—positive feelings and identity centrality—over shorter and longer time horizons, respectively. Furthermore, we argue for the recursive effects of passion on entrepreneurial self-efficacy and performance. Three longitudinal studies with measurements over three weeks (n = 65) and three months (n = 150 and n = 180) support our hypotheses. We contribute to a theory of passion that integrates the different time horizons over which the components of passion change.

Keywords

venture performance time/temporal aspects start-up Bottom of the Pyramid (BOP)emotions

Passion is a critical driving force in entrepreneurship (Breugst et al., 2012; Cardon et al., 2009; Huber et al., in press). Scholars often describe passion as “a core characteristic of great wealth creators” (Baum & Locke, 2004, p. 588) or as the “fire in the belly” (Smilor, 1997, p. 342) that characterizes successful entrepreneurs. To investigate passion, researchers have predominantly built on either Vallerand et al.’s (2003) or Cardon et al.’s (2009) theories of passion. Vallerand et al. (2003) conceptualize passion as a “strong inclination toward an activity that people like, that they find important, and in which they invest time and energy” (p. 757). Cardon et al. (2009) conceptualize passion as “consciously accessible, intense positive feelings experienced by engagement in entrepreneurial activities associated with roles that are meaningful and salient to the self-identity of the entrepreneur” (p. 517). Common to both conceptualizations is that passion involves two components: positive feelings toward an activity and identity centrality, which refers to the importance of the activity to the person’s identity (Murnieks et al., 2014). Research building on this conceptualization has shown that passion drives entrepreneurial behavior and success (Cardon & Kirk, 2015; Cardon et al., 2013; Drnovsek et al., 2016; Mueller et al., 2017). In summary, current research suggests that the two underlying components of passion are positive feelings and identity centrality, and that passion drives success in entrepreneurship.

In this study, we go beyond current research to provide novel theoretical insights into the development of passion and the role of time in this process. Recent research has begun to address the question of how passion develops, examining antecedents of passion in terms of entrepreneurial self-efficacy and performance (Cardon & Kirk, 2015; Collewaert et al., 2016; Dalborg & Wincent, 2015; Gielnik, Spitzmuller et al., 2015). Extending the research on the antecedents, we argue that the conceptualization of passion as consisting of the components of positive feelings and identity centrality requires theorizing about time to better understand the development of passion. Specifically, we argue that the two components of passion develop in distinct fashion over time. Positive feelings and identity centrality operate on different levels of self-regulation and, in extension, develop over different time horizons—or, in terms of self-regulation theory, have different cycle times (Lord et al., 2010). Specifically, identity centrality is shaped at a higher level of self-regulation, whereas positive feelings are shaped at a lower level of self-regulation. Self-regulatory processes at higher levels of self-regulation have longer cycle times than do processes at lower levels of self-regulation. Accordingly, we argue that time is crucial to the theoretical understanding of the development of passion in entrepreneurship.

In this study, we draw from self-regulation theory (Baumeister & Vohs, 2004; Lord et al., 2010) to cover new ground and present a theoretical model of the development of passion that takes into account the different cycle times of the components of passion. Self-regulation deals with the cognitive and emotional processes of performance and goal accomplishment (Lord et al., 2010) and has been discussed as an overarching theoretical framework for understanding passion (Cardon et al., 2009). Three basic assumptions of self-regulation theory are critical to our study. First, self-regulation comprises a feedback loop that involves setting a standard, monitoring goal progress, and evaluating one’s performance and oneself. Any such evaluation is capable of triggering strong emotional reactions (Ilies & Judge, 2005). Second, self-regulation occurs at different hierarchical levels of abstraction, each level being associated with its own feedback loop and cycle time to run through the loop. Higher, more abstract levels of regulation are associated with longer cycle times (Lord et al., 2010). Third, cognitions and emotions are interrelated, influencing each other dynamically over time as part of the feedback loop (Baumeister et al., 2007; Lord et al., 2010).

Building on self-regulation theory, we posit that passion develops in a feedback loop and that, because they operate on different levels of abstraction in the self-regulatory hierarchical structure, the components of passion have different cycle times in this loop. Identity centrality operates on a high level of abstraction and therefore has a longer cycle time (Ibarra, 1999; Lord et al., 2010). By contrast, positive feelings related to the performance of tasks and activities operate on an intermediate level of abstraction with relatively shorter cycle times (Lord et al., 2010). Finally, we posit that passion is both an outcome and an antecedent of entrepreneurial self-efficacy, reflecting the dynamic interrelationship between cognitions and emotions over time (Lord et al., 2010). Self-efficacy is cognition that plays a key role in self-regulation as it is the cognitive representation of one’s past performance, as well as a cognitive driver of future performance (Bandura, 1989). As such, the reciprocal relationships between passion and entrepreneurial self-efficacy, as well as between entrepreneurial self-efficacy and performance, suggest a recursive cycle in which passion is both an outcome and a driver of entrepreneurial performance through entrepreneurial self-efficacy (Figure 1).

Figure 1

Model of the development of passion with its components of positive feelings and identity centrality, and recursive effects on entrepreneurial self-efficacy and performance.

Our study contributes to the literature in at least two ways. First, our theoretical model describes the process through which passion develops and considers the different self-regulatory levels on which the two components of passion—identity centrality and positive feelings—operate. The model thus specifies the time horizons over which the two components develop. Addressing the important role that cycle times play in the two components of passion provides a richer description of the development of passion. Second, we posit a recursive model in which passion is the pivotal point of a cycle. In this cycle, entrepreneurial performance shapes passion via entrepreneurial self-efficacy; passion, in turn, influences entrepreneurial performance through entrepreneurial self-efficacy. By proposing recursive relationships, we integrate hitherto fragmented perspectives on the role of entrepreneurial self-efficacy and performance for passion (e.g., Cardon & Kirk, 2015; Murnieks et al., 2014).

A Self-Regulation Perspective on the Reciprocal Relationships of Passion, Self-Efficacy, and Performance in Entrepreneurship

Self-regulation theory is a grand theory describing cognitive and emotional processes relevant for goal achievement and performance (Lord et al., 2010). Importantly, a self-regulation lens emphasizes that performance requires the integration of self-regulatory processes on different levels of abstraction that have different time horizons. On high and more abstract regulatory levels, people regulate their possible selves and identities. On intermediate and more concrete levels, people regulate their performance on specific tasks. On low and micro levels of self-regulation, people regulate integrated and automatic task behaviors based on scripts and flexible action patterns (Lord et al., 2010). Furthermore, each level involves a feedback loop that comprises setting a standard, comparing the progress to the standard, and evaluating one’s performance. The evaluation then triggers cognitive and emotional responses, which in turn can start a new cycle (Ilies & Judge, 2005).

Important to our study is the relationship between levels of self-regulation and the cycle times to go through the feedback loop. Specifically, “the level of abstraction inversely relates to the cycle times of feedback loops [...]. Longer cycle times tend to be associated with higher-level, more abstract constructs” (Lord et al., 2010, p. 547). In our study, we focus on the high and intermediate level on which identity centrality and the consciously accessible positive feelings of passion operate. Self-regulation on a high level has cycle times that take months or years. Self-regulation on an intermediate level involves cycle times that can last several days, as people work on a task and receive feedback. The low and micro levels of self-regulation play out in much shorter time frames, from seconds to milliseconds (Lord et al., 2010).

We use the notion of different cycle times on the levels of self-regulation as a starting point for our investigation of the different time horizons over which the components of passion develop. Passion in entrepreneurship is the composite of two components: positive feelings and identity centrality (Cardon et al., 2009). Positive feelings associated with passion in entrepreneurship are emotional experiences—such as excitement and enjoyment—when engaging in entrepreneurial tasks and are thus different from positive affect in general (Cardon et al., 2013). Identity centrality refers to the importance of entrepreneurship to one’s self-concept. Specifically, people with high identity centrality consider entrepreneurship to be an important part of who they are (Murnieks et al., 2014). We argue that it is conceptually justified and theoretically important to examine the development of the two components separately. It is justified because positive feelings and identity centrality are “conceptually and empirically distinct from each other” (Cardon et al., 2013, p. 374). Furthermore, it is theoretically important because disentangling positive feelings and identity centrality allows theorizing about the two components on different regulatory levels, resulting in different predictions regarding the time horizons over which the two components develop. Consistent with the idea of cycle times on intermediate and high regulatory levels, we predict a shorter cycle time for positive feelings and longer cycle time for identity centrality.

Self-regulation theory further suggests that cognitions and emotions influence each other over time (Lord et al., 2010). Specifically, emotions are often the result of cognitive evaluation of one’s performance or oneself. Moreover, emotions influence cognitions in terms of expectations about future performance and outcomes (Baumeister et al., 2007). To capture the cognitive evaluations of one’s past performance and expectations of one’s future performance, we rely on entrepreneurial self-efficacy. In contrast to general self-efficacy, which captures people’s beliefs in their overall capabilities, entrepreneurial self-efficacy reflects people’s beliefs in their capabilities to successfully perform future entrepreneurial tasks (Chen et al., 1998). Entrepreneurial self-efficacy also captures people’s past performance as it varies over time as a result of mastery experiences (Gielnik et al., 2020; Zhao et al., 2005). We posit that entrepreneurial self-efficacy is a mechanism underlying the recursive relationship between passion and entrepreneurial performance because it captures people’s cognitions about their past and future performance. In the following, we explain the theoretical rationale for a self-regulation perspective underlying the reciprocal relationships between passion, entrepreneurial self-efficacy, and performance on different levels of self-regulation and with different cycle times.

Shorter Cycle Time of Positive Feelings

We argue that positive feelings develop as part of a feedback loop of achieving entrepreneurial performance and then cognitively evaluating the performance. Importantly, we posit that the evaluation of entrepreneurial tasks and performance takes place on an intermediate regulatory level with shorter cycle times (Lord et al., 2010). Accordingly, we hypothesize that entrepreneurial performance is positively related to positive feelings, and this relationship unfolds over shorter time horizons. Self-regulation theory suggests that performance and making progress reduce the discrepancy between the current state and a desired goal, resulting in positive feelings like excitement and enjoyment (Baumeister et al., 2007). Indeed, research in entrepreneurship supports this reasoning by showing that performance predicts positive feelings of passion (Gielnik, Spitzmuller, et al., 2015).

We further theorize that the effect of entrepreneurial performance on positive feelings is indirect through entrepreneurial self-efficacy. According to the self-regulation theory, it is the cognitive evaluation of one’s performance that elicits an emotional response (Lord et al., 2010). Specifically, people evaluate their own performance and derive conclusions about their task-related capabilities. Believing themselves to be capable of performing a task and expecting that future efforts will lead to desired outcomes enhances people’s positive feelings toward this task (Bandura, 1989). Research provides empirical support for the links in the indirect relationship between entrepreneurial performance and positive feelings through entrepreneurial self-efficacy. Mastery experiences and past performance are determinants of self-efficacy (Sitzmann & Yeo, 2013). Moreover, self-efficacy influences positive feelings during task engagement (Schutte & Malouff, 2016). In entrepreneurship, research shows that entrepreneurial self-efficacy is associated with positive feelings of passion (Cardon & Kirk, 2015; Dalborg & Wincent, 2015; Gielnik et al., 2017).

Hypothesis 1: (a) Entrepreneurial performance has a positive effect on positive feelings, (b) which is mediated by entrepreneurial self-efficacy. The effect crystallizes over shorter time horizons.

Longer Cycle Time of Identity Centrality

We argue that identity centrality develops in a feedback loop as a result of entrepreneurial performance and the gradual development of entrepreneurial self-efficacy. Moreover, identity centrality is regulated on a high level of abstraction with longer cycle times. Therefore, the relationship between entrepreneurial performance and identity centrality through growth in entrepreneurial self-efficacy unfolds over longer time horizons.

Building on concepts from self-regulation theory, research on identity suggests that the relationship between people’s identities and their behavior follows the principle of a feedback loop (Burke, 2006). That is, people’s identity influences their behavior and their behavior feeds back into their identities. This feedback loop constitutes a verification process by which people seek to establish congruence by confirming their identities through their behavior. For example, showing high performance in a domain validates and reinforces people’s identities within this domain (Burke, 2006; Burke & Stets, 2009). Research supports this reasoning, showing that professionals gradually develop an identity by evaluating their behavior based on internal standards and external feedback (Ibarra, 1999).

Furthermore, we argue that developing entrepreneurial self-efficacy is a mechanism underlying the relationship between entrepreneurial performance and identity centrality. People’s identities are linked to a belief in mastery and control (Burke, 2006). Identities become more central when people develop the belief that they are competent and can fulfill the expectations of the identity (Stets & Burke, 2000). Competence-confirming feedback in a domain increases beliefs of efficacy and, consequently, enhances the importance of this domain to people’s identity (McCall & Simmons, 1978). Developing a belief of being competent therefore constitutes a key mechanism strengthening the importance of an identity and enhancing its centrality (Ashforth & Schinoff, 2016). We note that research on identity emphasizes the cumulative pattern of changes in one’s competence beliefs for identities to change overall (Burke & Stets, 2009). That is, identities change when people experience an increase or decrease in self-efficacy over time. This argumentation is in line with research demonstrating that identity construction happens over time and is intertwined with the development of skills and associated growth in self-efficacy (Donnellon et al., 2014; Ewing & Ewing, 2017; Miscenko et al., 2017). Accordingly, we hypothesize that growth in entrepreneurial self-efficacy leads to changes in identity centrality.

Hypothesis 2: (a) Entrepreneurial performance has a positive effect on identity centrality, (b) which is mediated by the growth in entrepreneurial self-efficacy. The effect crystallizes over longer time horizons.

The Reverse Route: Effects of Passion on Entrepreneurial Self-Efficacy and Performance

Self-regulation theory suggests that positive activating emotions influence the effort that people invest in a task, and this effect is likely to be mediated by performance expectancies (Lord et al., 2010). Furthermore, self-regulation theory explicitly refers to a feedback loop in this process, arguing that positive emotions generated by previous performance may influence subsequent performance when the emotions spill over into the next self-regulatory cycle (Ilies & Judge, 2005). Passion is an activating positive emotional experience, and it is the composite of positive feelings and identity centrality. Accordingly, we argue that the components of passion complement each other and their interplay influences entrepreneurial performance. Passion energizes entrepreneurs to devote great effort to their entrepreneurial task, thus increasing the likelihood of succeeding in entrepreneurship (Cardon et al., 2009). Indeed, empirical research provides evidence of a positive effect of passion on performance in entrepreneurship (Cardon & Kirk, 2015; Drnovsek et al., 2016; Mueller et al., 2017).

Furthermore, we argue that the effect of passion on entrepreneurial performance is mediated by people’s beliefs in their capabilities. Passion, which involves positive feelings and identity centrality, influences people’s beliefs in their capabilities. Specifically, negative feelings are interpreted as signs of low capability, whereas positive feelings are interpreted as signs of high capability (Bandura, 1989). Furthermore, believing oneself to be competent is an outcome of the verification process of people’s identities (Burke & Stets, 2009). When people show behavior that is consistent with the standards in a domain, they verify their identity. The verification leads to a heightened sense of efficacy because people become confident about their ability to fulfill the expectations in this domain. Entrepreneurial self-efficacy in turn promotes entrepreneurial performance. A strong belief in one’s capabilities positively impacts self-regulatory processes that are conducive to entrepreneurial performance (Chen et al., 1998). People with strong confidence in their abilities set more challenging goals, are more committed to these goals, invest more effort, and are more likely to persist. Consistent with theoretical arguments, research provides evidence that entrepreneurial self-efficacy is positively linked to entrepreneurial performance (Rauch & Frese, 2007).

Hypothesis 3: (a) Passion has a positive effect on entrepreneurial performance, (b) which is mediated by entrepreneurial self-efficacy.

Study I: Methods

Sample

We sampled entrepreneurs across the three phases of the entrepreneurial process (prelaunch, launch, and postlaunch phase). The final sample comprised 65 entrepreneurs from Tanzania, selected using the following criteria: being active on the business during the three weeks of our study and speaking English. We recruited participants through lists provided by a university and an entrepreneurship hub. Additionally, we asked participants to introduce other entrepreneurs to us. This approach resulted in 369 potential participants who were contacted via email or phone calls. In total, 68 entrepreneurs were eligible for participation and agreed to take part in the study (18% response rate). Out of the 68 participants, two participants dropped out over the course of the study and one participant was excluded for being unable to read the questionnaire, resulting in a final sample of 65 participants.

Of the final sample, 34 (52%) participants were male. Participants ranged in age from 22 to 65 years with an average age of 38 years (standard deviation [SD] = 11.5). Of the participants, 58% held a university degree, 29% held a diploma, 5% had a certificate, 6% had completed secondary school education, and 2% had completed primary school. Most participants (52%) had started another business prior to their current business, and 89% had received entrepreneurship education. We note that they had participated in entrepreneurship programs other than the program of Study II and Study III. On average, our participants had started 1.9 businesses (SD = 1.3) and had been managing their current business for 5.5 years (SD = 5.6). The businesses ranged in size from 0 to 40 full-time employees (M = 4.0; SD = 6.4). On average, these businesses generated 19,809,973 Tanzania Shilling sales per month (approximately 9364 USD). Furthermore, 51% were in the service sector; 25% were in the trade sector; 11% represented agriculture, forestry, and fishing; 8% represented manufacturing; 5% represented transportation and public utilities; and 1% represented finance, insurance, and real estate.

Study Design and Procedure

We employed a repeated measures design with three measurements (T1–T3) and a time lag of one week to capture cycle times on an intermediate regulatory level. At each wave, we visited the participants to collect the data. At T1, we collected data on demographic and control variables using structured face-to-face interviews. After the interview, participants completed a questionnaire on positive feelings, identity centrality, entrepreneurial self-efficacy, and performance. At T2 and T3, the 65 participants completed a questionnaire to once again measure positive feelings, entrepreneurial self-efficacy, and performance. This approach resulted in three observations per participant and a total of 195 observations.

Measures

Components of Passion

We measured passion by its two components of positive feelings and identity centrality. Positive feelings are variable over time and we therefore ascertained positive feelings weekly at T1, T2, and T3. We used four items from Cardon et al. (2013) to capture entrepreneurs’ positive feelings for inventing. One item, for example, was “It is exciting to figure out new ways to solve unmet market needs that can be commercialized”. We asked the participants to respond to the four items with regard to their current feelings, using a seven-point answer scale anchoring from 1 (not at all) to 7 (absolutely). We computed the mean across the four items. Cronbach’s α over the three measurements was α = 0.81. According to Cardon et al. (2017), passion can result from multiple sources. Indeed, recent research suggested that entrepreneurs can be passionate about different objects, such as entrepreneurship, their venture, or a particular business opportunity (Cardon et al., 2017; Warnick et al., 2018). In Study I, we focused on positive feelings about inventing for two reasons: first, as suggested by Cardon et al. (2009), inventing is one of the three key roles in entrepreneurship (the other two being founding and developing). Second, the items for inventing refer to the identification of new business opportunities. The items therefore applied to all participants of our sample. In contrast, the other two roles involve nurturing a business and dealing with employees, which did not apply to some of our participants who were still in the first phase of the entrepreneurial process.

We measured identity centrality once, at T1, as it is more stable and less likely to change over the course of the three weeks. We used four items developed by Callero (1985) and adapted to the entrepreneurship context (Farmer et al., 2011; Murnieks et al., 2014). For example, one item was “Being an entrepreneur is an important part of who I am”. We used a seven-point response scale from 1 (not at all) to 7 (absolutely). We computed the mean of the four items (Cronbach’s α = 0.68). We measured identity in entrepreneurship broadly, which is in line with previous research (Murnieks et al., 2014, 2016). This approach takes into account that the specific roles share conceptual similarities that reflect individuals’ general experience when engaging in entrepreneurship. We followed Cardon et al.’s (2013) conceptualization and modeled passion as the interaction between positive feelings and identity centrality when predicting outcomes of passion. We thus included the main effects of positive feelings and identity centrality, as well as the interaction term of the two mean-centered components in our statistical analyses.

Entrepreneurial Self-Efficacy

We assessed entrepreneurial self-efficacy weekly at T1, T2, and T3, using 10 items based on previous research (Frese et al., 2007). The items referred to tasks and activities that are specifically related to entrepreneurship. All items started with the stem “How confident are you that you can” followed by specific entrepreneurial activities, such as “perceive business opportunities well” and “do the marketing of your products well”. Participants answered all items on a seven-point answer scale ranging from 1 (20%) to 7 (100%). We computed the mean over the 10 items to create a scale of entrepreneurial self-efficacy (Cronbach’s α = 0.86).

Entrepreneurial Performance

We measured entrepreneurial performance weekly at T1, T2, and T3. We used three items adapted from previous research to capture entrepreneurs’ subjective performance as the owner of the business (Liden et al., 1993). For example, one item was “In the last week, my performance as a business owner was high”. Participants provided their answers on a seven-point response scale from 1 (not at all) to 7 (absolutely). We computed the mean of the three items (Cronbach’s α = 0.88). To validate the self-report measure of entrepreneurial performance, we ascertained the total profit that the participants made with their business during our study period. Analyses revealed a significant correlation (r = 0.30, p = .013) between the mean of the self-report measure over the three measurements and the total profit gained during the three weeks, thus providing evidence for the validity of the self-report measure (Bosco et al., 2015).

Control Variables

We used entrepreneurs’ gender, age, entrepreneurial experience, and entrepreneurship education as control variables. Prior research shows that these constructs influence entrepreneurial self-efficacy and performance (Davidsson & Honig, 2003; Klapper & Parker, 2011). We assessed all control variables at T1. We assessed entrepreneurial experience by asking participants for the number of businesses that they had ever started. To ascertain entrepreneurship education, we asked participants whether they had ever received entrepreneurship or business training (0 = no, 1 = yes). We further included the measurement wave as control variable in our statistical analyses to control for learning or trend effects.

Method of Analysis

To account for the dependency in our data (multiple observations nested in individuals), we used random coefficient modeling (Bliese & Ployhart, 2002). We created lagged versions of each repeated measures variable, that is, variables representing each respective variable one week later. This data structure allowed us to test the lagged effects of the predictor in one week on the dependent variable in the next week. We included the dependent variable at the prior measurement to control for autoregression and model change in the dependent variable. We grand mean centered the predictor variables. The variables therefore contained both between-person and within-person variance (Enders & Tofighi, 2007). Our approach was in line with self-regulation theory, which holds that both between-person and within-person variance components are relevant in predicting self-regulatory outcomes (Lord et al., 2010). For example, people need to exceed a certain threshold (or absolute between-person level) in self-efficacy to experience positive self-regulatory effects (Vancouver et al., 2008). Similarly, within-person effects of self-efficacy vary with people’s absolute (between-person) levels of self-efficacy (Beck & Schmidt, 2012). We report conditional and marginal R² as statistics of explained variance (Nakagawa & Schielzeth, 2013). To test indirect effects, we applied the Monte Carlo method (Selig & Preacher, 2008). The Monte Carlo method provides a confidence interval around the indirect effect based on 20,000 replications. The indirect effect is significant if the 95% confidence interval around the indirect effect excludes zero.

Study I: Results

Table 1 shows the descriptive statistics of the variables. We tested for multicollinearity by calculating variance inflation factor (VIF) scores in all models. All VIFs were below 2.5, indicating that multicollinearity was not a concern in the analyses (O’Brien, 2007). Our repeated measures variables displayed substantial within-person variance over time with 40% of the total variance in positive feelings, 26% of the total variance in entrepreneurial self-efficacy, and 58% of the total variance in entrepreneurial performance.

Table 1

Study I: Descriptive Statistics of Variables on Level 1 (Level of Observations) and Level 2 (Level of Participants).

Variables	M	SD	1	2	3	4	5	6	7	8
Level 2
1. Gender^a	0.52	0.50
2. Age	38.03	11.46	−.16
3. Entrepreneurial experience	1.91	1.27	.08	−.09
4. Entrepreneurship education^b	0.89	0.31	.07	.04	−.06
5. Identity centrality	6.54	0.51	−.03	−.10	.09	−.07
6. Entrepreneurial self-efficacy^c	5.65	0.59	−.08	.03	−.17	−.19	.29*
7. Positive feelings^c	6.15	0.64	.04	−.04	.06	−.07	.40**	.50**
8. Entrepreneurial performance^c	4.93	1.06	−.23^†	.18	.10	−.07	.18	.42**	.38**
Level 1
1. Entrepreneurial self-efficacy T1	5.69	0.59
2. Entrepreneurial self-efficacy T2	5.61	0.64	.77**
3. Entrepreneurial self-efficacy T3	5.63	0.71	.71**	.76**
4. Positive feelings T1	6.23	0.64	.37**	.34**	.28*
5. Positive feelings T2	6.15	0.78	.41**	.36**	.29*	.63**
6. Positive feelings T3	6.08	0.79	.51**	.46**	.54**	.55**	.64**
7. Entrepreneurial performance T1	4.80	1.49	.20	.32**	.16	.21^†	.24^†	.01
8. Entrepreneurial performance T2	5.06	1.14	.35**	.34**	.35**	.32**	.27*	.31**	.33**
9. Entrepreneurial performance T3	4.94	1.39	.24^†	.33**	.46**	.33**	.30*	.38**	.42**	.54**

Note. Number of participants = 65; number of observations = 195; ^a0 = female, 1 = male; ^b0 = no, 1 = yes; ^caggregated across the three measurement waves; SD = standard deviation; ^†p < .10, *p < .05, **p < .01.

Hypotheses Testing

Hypothesis 1 states that entrepreneurial self-efficacy mediates the effect of entrepreneurial performance on positive feelings. We tested the hypothesis in a weekly time interval to show that the effect crystallizes over a time frame of days. Table 2 displays the results. Entrepreneurial performance had a positive effect on entrepreneurial self-efficacy (Model 2: b = 0.08, p = .009) and positive feelings (Model 4: b = 0.09, p = .038). Furthermore, entrepreneurial self-efficacy had a positive effect on positive feelings (Model 5: b = 0.33, p = .001). The Monte Carlo method showed a significant indirect effect of entrepreneurial performance on positive feelings through entrepreneurial self-efficacy (indirect effect = 0.03, p = .009). The results thus supported Hypothesis 1.

Table 2

Study I: Random Coefficient Models Testing the Effect of Entrepreneurial Performance on Positive Feelings Through Entrepreneurial Self-Efficacy.

	Entrepreneurial self-efficacy(lagged, weekly)				Positive feelings(lagged, weekly)
	Model 1		Model 2		Model 3		Model 4		Model 5
	b	SE	b	SE	b	SE	b	SE	b	SE
Measurement wave	0.08	(0.08)	0.06	(0.08)	0.00	(0.10)	−0.03	(0.11)	−0.01	(0.10)
Gender^a	0.09	(0.08)	0.12	(0.08)	0.04	(0.10)	0.08	(0.11)	0.11	(0.11)
Age	0.00	(0.01)	0.00	(0.01)	0.00	(0.01)	0.00	(0.01)	0.00	(0.01)
Entrepreneurial experience	−0.01	(0.03)	−0.02	(0.03)	0.00	(0.04)	−0.01	(0.04)	0.03	(0.04)
Entrepreneurship education^b	−0.03	(0.13)	−0.03	(0.13)	0.16	(0.15)	0.17	(0.17)	0.27	(0.17)
Entrepreneurial self-efficacy	0.84**	(0.07)	0.79**	(0.07)					0.33**	(0.10)
Positive feelings					0.70**	(0.07)	0.67**	(0.08)	0.57**	(0.08)
Entrepreneurial performance			0.08**	(0.03)			0.09*	(0.04)	0.06	(0.04)
Deviance (−2LogLikelihood)	182.08		180.23		263.49		263.49		255.19
R² marginal/conditional	0.57/0.57		0.60/0.60		0.40/0.40		0.42/0.42		0.47/0.47

Note. Number of participants = 65; number of lagged observations = 130; unstandardized regression coefficients (b’s) are shown; ^a0 = female, 1 = male; ^b0 = no, 1 = yes; SE = standard error; *p < .05, **p < .01.

Hypothesis 3 states that passion affects entrepreneurial performance via entrepreneurial self-efficacy. Table 3 shows the results. Passion had a positive effect on entrepreneurial self-efficacy (Model 1: b = 0.30, p = .013) and entrepreneurial performance (Model 3: b = 0.75, p = .041). As passion is reflected by the interaction term between positive feelings and identity centrality, we conducted simple slope analyses. The analyses revealed that positive feelings had a positive effect on entrepreneurial self-efficacy (b = 0.21, p = .027) and entrepreneurial performance (b = 0.80, p = .004) in the case of high identity centrality (+1 SD above the mean). In the case of low identity centrality (−1 SD below the mean), the effects of positive feelings on entrepreneurial self-efficacy (b = −0.10, p = .225) and entrepreneurial performance (b = 0.05, p = .839) were not significant. Entrepreneurial self-efficacy had a positive effect on entrepreneurial performance (Model 4: b = 0.53, p = .009). The Monte Carlo method showed that the indirect effect of passion on entrepreneurial performance through entrepreneurial self-efficacy was significant (indirect effect = 0.16, p = .018), supporting Hypothesis 3.

Table 3

Study I: Random Coefficient Models Testing the Effect of Passion on Entrepreneurial Performance Through Entrepreneurial Self-Efficacy.

	Entrepreneurial self-efficacy (lagged, weekly)		Entrepreneurial performance(lagged, weekly)
	Model 1		Model 2		Model 3		Model 4
	b	SE	b	SE	b	SE	b	SE
Measurement wave	0.07	(0.08)	−0.16	(0.17)	−0.13	(0.17)	−0.08	(0.16)
Gender^a	0.06	(0.08)	−0.36	(0.25)	−0.42^†	(0.23)	−0.42^†	(0.23)
Age	0.00	(0.01)	0.01	(0.01)	0.01	(0.01)	0.00	(0.01)
Entrepreneurial experience	−0.01	(0.03)	0.09	(0.10)	0.05	(0.09)	0.12	(0.09)
Entrepreneurship education^b	0.01	(0.13)	−0.06	(0.39)	−0.02	(0.37)	0.11	(0.38)
Entrepreneurial self-efficacy	0.81**	(0.07)					0.53**	(0.20)
Positive feelings	0.05	(0.06)			0.43*	(0.18)	0.28	(0.19)
Entrepreneurial performance			0.19*	(0.08)	0.16*	(0.08)	0.10	(0.08)
Identity centrality	−0.01	(0.08)			0.10	(0.23)	−0.01	(0.24)
Passion (positive feelings × identity centrality)	0.30*	(0.12)			0.75*	(0.36)	0.73*	(0.36)
Deviance (−2LogLikelihood)	184.65		418.91		407.13		401.77
R² marginal/conditional	0.59/0.59		0.10/0.39		0.20/0.44		0.24/0.51

Study I: Discussion

Study I provided evidence for the effect of entrepreneurial performance on positive feelings through entrepreneurial self-efficacy in a weekly research design. This finding supported our notion of a shorter cycle time for changes in positive feelings as a result of entrepreneurial performance and self-efficacy. Study I also showed an effect of passion as the composite of positive feelings and identity centrality on entrepreneurial performance through entrepreneurial self-efficacy, providing evidence for the reverse route of effects. However, Study I did not examine antecedents of identity centrality, which constitutes the component of passion operating on a high regulatory level. We therefore conducted Study II to extend the findings of Study I by examining the effects of entrepreneurial performance and growth in entrepreneurial self-efficacy on identity centrality over a longer horizon time of three months.

Study II: Methods

Study Design and Procedure

To examine the different cycle times of the components of passion, we conducted a study that included repeated measures over two different time periods. First, we used a repeated measures design with biweekly data collections over 12 weeks (W2–W12). We used this design to replicate the findings from Study I regarding the relationship between entrepreneurial performance and positive feelings over shorter cycle times. The 12 weeks were part of an entrepreneurship training program, during which participants started and operated real micro-businesses (Frese et al., 2016). Second, we conducted two measurements with a time lag of three months (T1–T2). The two measurements were taken the week before and the week after the training program. We used the second time period to examine changes in participants’ identity over a longer cycle time of three months.

The training program took place at a university in Tanzania and was designed for undergraduate students to engage in entrepreneurship. We collected data using questionnaires that were distributed at the end of the weekly training session. In the first session, participants formed 35 teams of between four and seven members. The participants then identified a business opportunity to launch a business with the goal of making profit within the 12-week training period. The participants went through the whole entrepreneurial process. They developed a business opportunity, introduced a product or service into the market, and managed a venture under real business conditions. The teams started different types of businesses, including selling t-shirts, providing a supply of phone accessories, or opening a bakery. The participants thus performed all major activities required by entrepreneurs.

Sample

Our sample comprised 150 participants of the entrepreneurship training program, recruited through student mailing lists, leaflets, and personal communication. Students from all faculties and years of study were eligible for participation. The program was taught in English. The program was voluntary and not part of the regular curriculum. Participants received certificates of successful participation, but were not graded. In total, 420 students applied for the training program by completing an application form and a baseline questionnaire. Due to limited training capacities, we randomly assigned 220 students to the program who were then divided into four classes with approximately 55 students. Seventy participants (32%) did not attend the training regularly (i.e., attended less than eight out of 12 sessions). We excluded these participants from our analyses to assure that all study participants were working on their businesses, resulting in a final sample of 150 participants. The t-tests showed that participants who dropped out did not differ significantly in terms of gender, identity centrality, and entrepreneurial self-efficacy. Participants who dropped out were significantly earlier in their studies (M = 2.06 vs. M = 2.43) and were more likely to have prior entrepreneurial experience (M = 0.56 vs. M = 0.39). However, interviews with participants who dropped out showed that the main reasons were lack of time and incompatibility of the training schedule with the university schedule.

In the sample, 74% of the participants were male, 39% had started a business before, and 49% were in the third year of their studies. They studied at different faculties, including the Business School (37%), the College of Social Sciences (32%), and the College of Natural and Applied Sciences (25%). During the training program, we measured positive feelings and entrepreneurial performance every second week from the second week onwards. On average, participants completed 4.96 questionnaires leading to a total number of 744 observations. We measured entrepreneurial self-efficacy and identity centrality before and after the training program (T1–T2). We collected data from all 150 participants at the two measurements. Furthermore, we collected data at T2 about the profits generated by the participants during the training program as an alternative measure of entrepreneurial performance.

Measures

Components of Passion

We measured the two components of passion—positive feelings and identity centrality. We measured positive feelings during the training program (W2–W12), using two items. We used a shortened scale, which is common practice in a repeated measurement study to reduce participants’ burden (Foo et al., 2009; Uy et al., 2010). Given that our intention was to replicate the findings from Study I, we used two items capturing positive feelings for inventing (Cardon et al., 2013) that were identical to two of the four items of Study I. We employed the full scale at T1 and selected the two items that yielded the highest item total correlation (0.62 and 0.60 vs. 0.39 and 0.55). The two items were “Searching for new ideas for products/services to offer is enjoyable to me” and “Scanning the environment for new opportunities really excites me”. We used item total correlation as a selection criterion, which is in line with the approach by Cardon et al. (2013) to develop the scale. The correlation between the original scale with all four items and the subscale with the two selected items was high and significant (r = 0.87, p < .001), suggesting that there was substantial overlap between the original and the shortened scale. We used a seven-point response scale ranging from 1 (not at all) to 7 (absolutely) and computed the mean across the two items. Cronbach’s α over the six measurement waves was α = 0.79.

We measured identity centrality before (T1) and after the training program (T2). To do so, we used the two items that had the highest item total correlation in Study I (0.73 and 0.71 vs. 0.56 and 0.41). The items were “Being an entrepreneur is an important part of who I am” and “For me, being an entrepreneur means more than just running my business”. Participants responded by using a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). Cronbach’s α was α = 0.67.

Entrepreneurial Self-Efficacy

We measured entrepreneurial self-efficacy before (T1) and after the training program (T2). We used 12 items developed by Gielnik, Frese et al. (2015). The general stem was “How confident are you that you can” followed by 12 entrepreneurial activities (e.g., “perceive business opportunities well”). We used a five-point response scale ranging from 1 (20%) to 5 (100%). Cronbach’s α was α = 0.93.

Entrepreneurial Performance

We assessed entrepreneurial performance during the training program (W2–W12) with two items based on previous research (Pearce & Sims, 2002). The items were: “In the last two weeks, we made good progress in starting and running our team business” and “In the last two weeks, we were very effective at getting things done in our business”. The items reflected perceived progress in new venture creation (Uy et al., 2015). We used a seven-point response scale from 1 (not at all) to 7 (absolutely). We computed the mean of the two items (Cronbach’s α = 0.91). To validate the measure, we asked participants at the end of the training program how much profit they had made in the course of the 12 weeks. We found a significant positive correlation between the mean of the self-report measure and the profit reported at the end of the training program (r = 0.34, p < .001), providing evidence for the validity of the self-report measure (Bosco et al., 2015).

Control Variables

We measured the control variables at T1 before the training program. We included gender (0 = female, 1 = male), year of studies, entrepreneurial experience (whether participants were business owners at the time of the study or had ever started a business in the past), and entrepreneurial team size as control variables (Klotz et al., 2014). We also controlled for the measurement wave in all analyses.

Method of Analysis

We used the biweekly data to test the relationship between entrepreneurial performance and positive feelings. The data consisted of multiple observations nested in individuals and teams. We therefore used the same statistical approach as in Study 1 and performed random coefficient modeling with grand mean centered predictor variables. Furthermore, we tested lagged effects, controlling for the dependent variable at the previous measurement.

We used condition-based regression analyses to test Hypothesis 2, which states that growth in entrepreneurial self-efficacy mediates the effect of entrepreneurial performance on identity centrality. We operationalized growth in entrepreneurial self-efficacy as being the difference in entrepreneurial self-efficacy from T1 to T2. Condition-based regression analysis calculates the magnitude of the difference effect of a variable on an outcome variable (Humberg et al., 2018). However, condition-based regression analysis does not combine the two variables into a single difference score, which implies that it avoids the disadvantages of reduced reliability of difference scores. Condition-based regression analysis tests whether the difference between two predictors has any effect on an outcome variable beyond the main effects of the predictors. In our case, we regressed identity centrality on entrepreneurial self-efficacy at T1 and T2, and extracted the respective coefficients (c_T1 and c_T2). The conditions to provide evidence for a significant effect of growth in entrepreneurial self-efficacy from T1 to T2 on identity centrality are that the coefficients are not zero and have opposite signs. Specifically, we need to show that (1) the effect of entrepreneurial self-efficacy at T1 (c_T1) is negative and (2) the effect of entrepreneurial self-efficacy at T2 (c_T2) is positive (Humberg et al., 2018). To avoid testing a hypothesis with two tests, Humberg et al. (2018) developed a theorem that combines the two conditions (both coefficients are not zero) into a single condition. They showed that testing the two conditions is mathematically equivalent to testing the condition abs: = |c_T2 – c_T1| – |c_T2 + c_T1 |> 0. The condition abs can be reformulated according to a lemma. In our case, the reformulated condition is: if c_T2 – c_T1 ≥ 0 and c_T2 + c_T1 ≥ 0, then abs = –2c_T1 (cf., Humberg et al., 2018). If the test shows that abs is significantly positive, then the data provide support that higher values in the difference between the predictor variables are related to higher values in the outcome. Condition-based regression analysis can be applied to any context in which researchers seek to examine the relationship between the difference of two variables and an outcome (Humberg et al., 2018).

Study II: Results

Table 4 presents the descriptive statistics of the variables. Based on the weekly data, we computed a set of null models to estimate the within-person variance. The null models showed that 47% of the total variance in positive feelings and 60% of the total variance in entrepreneurial performance was within-person variance. Multicollinearity was not a problem in this study, as all VIF scores were below 2.5 (O’Brien, 2007).

Table 4

Study II: Descriptive Statistics of Variables on Level 1 (Level of Observations) and Level 2 (Level of Participants).

Variables	M	SD	1	2	3	4	5	6	7	8	9	10	11
Level 2
1. Gender^a	0.74	0.44
2. Year of studies	2.43	0.92	−.05
3. Entrepreneurial experience^b	0.39	0.49	.13	−.08
4. Entrepreneurial team size	5.00	1.22	.04	.04	−.07
5. Identity centrality T1	4.45	0.72	−.09	−.30**	.14^†	.00
6. Identity centrality T2	4.40	0.61	.01	−.12	.14^†	−.05	.21*
7. Entrepreneurial self-efficacy T1	4.15	0.63	−.02	−.04	.14^†	−.05	.29**	.06
8. Entrepreneurial self-efficacy T2	4.28	0.53	.00	.00	.00	−.07	.11	.38**	.32**
9. Mean positive feelings (W2-W12)^c	6.01	0.72	.02	−.03	.04	.17*	.06	.30**	.16^†	.39**
10. Mean entrepreneurial performance (W2-W12)^c	5.73	0.79	.00	.01	−.01	.14^†	.07	.26**	.32**	.30**	.59**
Level 1
1. POF W2	6.20	0.78
2. POF W4	6.03	0.91	.59**
3. POF W6	6.07	0.79	.48**	.57**
4. POF W8	5.93	0.99	.60**	.60**	.66**
5. POF W10	6.05	0.93	.55**	.51**	.56**	.59**
6. POF W12	5.93	0.99	.46**	.39**	.66**	.59**	.49**
7. PER W2	5.83	1.11	.27**	.21*	.28**	.07	.24*	.32**
8. PER W4	5.83	0.97	.27*	.30**	.36**	.25*	.16	.25*	.34**
9. PER W6	5.69	1.05	.25*	.30**	.58**	.49**	.24*	.45**	.22*	.63**
10. PER W8	5.71	1.12	.28**	.20*	.51**	.56**	.40**	.38**	.13	.38**	.63**
11. PER W10	5.84	1.06	.35**	.23*	.37**	.43**	.43**	.38**	.23*	.35**	.47**	.53**
12. PER W12	5.69	1.14	.28**	.23*	.39**	.44**	.40**	.61**	.16^†	.36**	.59**	.58**	.63**

Note. Level 2: N = 150; ^a 0 = female, 1 = male; ^b 0 = no, 1 = yes; ^c aggregated across the weekly measurement waves of training program; Level 1: 84 < N < 147; POF = positive feelings; PER = entrepreneurial performance; SD = standard deviation; ^†p < .10, *p < .05, **p < .01.

Hypotheses Testing

Table 5 shows that entrepreneurial performance predicted positive feelings (Model 2: b = 0.08, p = .025), and positive feelings predicted entrepreneurial performance (Model 4: b = 0.14, p = .009). We thus replicated the recursive relationships between entrepreneurial performance and positive feelings over shorter cycle times from Study I.

Table 5

Study II: Random Coefficient Models Testing the Reciprocal Relationship Between Entrepreneurial Performance and Positive Feelings.

	Positive feelings(lagged, biweekly)				Entrepreneurial performance(lagged, biweekly)
	Model 1		Model 2		Model 3		Model 4
	b	SE	b	SE	b	SE	b	SE
Measurement wave	0.01	(0.02)	0.01	(0.02)	0.02	(0.02)	0.02	(0.02)
Gender^a	−0.02	(0.10)	−0.01	(0.10)	0.11	(0.11)	0.10	(0.11)
Year of studies	−0.06	(0.05)	−0.06	(0.05)	−0.07	(0.06)	−0.07	(0.06)
Entrepreneurial experience	0.07	(0.10)	0.07	(0.10)	0.03	(0.10)	0.02	(0.10)
Entrepreneurial team size	0.02	(0.05)	0.02	(0.05)	0.09	(0.07)	0.08	(0.07)
Positive feelings	0.29**	(0.04)	0.26**	(0.05)			0.14**	(0.05)
Entrepreneurial performance			0.08*	(0.04)	0.32**	(0.04)	0.29**	(0.04)
Deviance (−2LogLikelihood)	1067.78		1067.63		1225.70		1223.08
R² marginal/conditional	0.10/0.41		0.12/0.41		0.14/0.43		0.16/0.43

Note. Number of teams = 35; number of participants = 139; number of lagged observations = 490; unstandardized regression coefficients (b’s) are shown; ^a 0 = female, 1 = male; *p < .05, **p < .01; SE = standard error.

We tested Hypothesis 2, which states that growth in entrepreneurial self-efficacy mediates the effect of entrepreneurial performance on identity centrality. Table 6 displays the results. We used entrepreneurial performance during the training program (W2–W12) as the independent variable, growth in entrepreneurial self-efficacy from before to after the training program (T1–T2) as the mediator, and identity centrality after the training program (T2) as the dependent variable. We aggregated the measures of entrepreneurial performance during the training program (W2–W12) and computed the mean to obtain a variable for the total entrepreneurial performance across the 12 weeks. The results showed that the mean of entrepreneurial performance during the training program (W2–W12) had a significant effect on entrepreneurial self-efficacy (Model 2: b = 0.15, p = .009) and on identity centrality (Model 3: b = 0.20, p = .002). In Model 4, we included entrepreneurial self-efficacy at T1 and T2 to conduct condition-based regression analysis. Entrepreneurial self-efficacy at T1 was negatively (b = −0.18, p = .022) and entrepreneurial self-efficacy at T2 was positively related to identity centrality (b = 0.41, p < .001). The lemma was significant (lemma = 0.37, p = .021), suggesting that growth in entrepreneurial self-efficacy predicted identity centrality beyond the main effect of entrepreneurial self-efficacy at T2. Furthermore, we calculated the indirect effect of entrepreneurial performance on identity centrality through growth in entrepreneurial self-efficacy based on the lemma. The Monte Carlo method showed a significant indirect effect (indirect effect = 0.05, p = .028), thus supporting Hypothesis 2.

Table 6

Study II: Testing the Effect of Entrepreneurial Performance and Growth in Entrepreneurial Self-Efficacy on Changes in Identity Centrality Over Three Months (T1–T2).

	Entrepreneurial self-efficacy (T2)						Identity centrality (T2)
	Model 1			Model 2			Model 3			Model 4
	b	SE	β	b	SE	β	b	SE	β	b	SE	β
Gender^a	0.01	(0.10)	0.01	0.01	(0.09)	0.01	0.00	(0.11)	0.00	0.00	(0.10)	0.00
Year of studies	0.01	(0.05)	0.01	0.00	(0.04)	0.01	−0.04	(0.05)	−0.06	−0.04	(0.05)	−0.06
Entrepreneurial experience^b	−0.05	(0.09)	−0.05	−0.04	(0.09)	−0.03	0.14	(0.10)	0.11	0.17^†	(0.09)	0.14
Identity centrality (T1)							0.14*	(0.07)	0.16	0.15*	(0.07)	0.18
Entrepreneurial self-efficacy (T1)	0.27**	(0.07)	0.32	0.21**	(0.07)	0.25				−0.18*	(0.08)	−0.19
Entrepreneurial self-efficacy (T2)										0.41**	(0.09)	0.36
Mean entrepreneurial performance(W2–W12)				0.15**	(0.05)	0.22	0.20**	(0.06)	0.25	0.16**	(0.06)	0.21
F	4.19			4.89			4.08			6.50
R ²	0.10			0.15			0.12			0.24

Note. Number of participants = 150; ^a0 = female, 1 = male; ^b0 = no, 1 = yes; ^†p < .10, *p < .05, **p < .01; SE = standard error.

As a robustness check, we used the profit generated by the participants during the training program as an alternative measure of entrepreneurial performance. We used the same control variables in these models. The results were consistent: Profit predicted entrepreneurial self-efficacy (b = 0.01, p = .041) and the indirect effect on identity centrality through growth in entrepreneurial self-efficacy was significant (indirect effect = 0.001, p = .036), replicating the results based on an objective measure of entrepreneurial performance.

Study II: Discussion

The results of Study II provided evidence for the effect of entrepreneurial performance on identity centrality through entrepreneurial self-efficacy over a longer cycle time of three months. Furthermore, Study II replicated the findings from Study I regarding the recursive effects of entrepreneurial performance and positive feelings over a shorter cycle time. However, Study II did not investigate entrepreneurial self-efficacy as a mediator of the short-term relationship between entrepreneurial performance and positive feelings. We therefore conducted Study III to replicate and extend the findings from Study II. Furthermore, we conducted Study III to provide evidence for the generalizability of the theoretical model by demonstrating that the findings hold when using other objects of passion in entrepreneurship.

Study III: Method

The setting and procedure was similar to Study II. We used the entrepreneurship training program and a research design with two time periods, and collected data during the 12 weeks of the training program (W1–W12) as well as before and after the training program (T1–T2). The program took place at the same university but with a different cohort of students. The participants formed 34 teams of four to seven members and engaged in the entrepreneurial process. The sample consisted of 180 students who were recruited using the same procedure as in Study II. In total, 406 students applied for the training. From these applicants, we randomly selected 224 students. We then excluded 44 participants (20%) who attended fewer than eight out of 12 sessions. The t-tests showed that the participants who dropped out did not significantly differ from the remaining participants. In the sample, 78% of the participants were male and 46% had previously started a business. Most participants were in the third year of their studies (69%), and studied at different faculties, such as the Business School (70%) or the College of Social Sciences (12%).

During the training program, we measured positive feelings and entrepreneurial self-efficacy four times every three weeks. We measured entrepreneurial performance every week from the second week onwards. On average, participants completed 10.5 questionnaires leading to a total number of 1,889 observations. We measured entrepreneurial self-efficacy and identity centrality both before and after the training program (T1–T2), and measured the profit generated by the participants after the training program (T2). We collected data from 180 participants at T1 and from 179 participants at T2.

Measures

Components of Passion

We used different objects for the components of passion than those used in Study II. We assessed positive feelings at W2, W5, W8, and W11 with two items that referred to the business idea the students were working on during the training program. This approach is consistent with prior research that used the business idea as the object of passion (Li et al., 2017). The business idea was an important element of the participants’ entrepreneurial activity; they identified their business idea in the first week, and then implemented it in the following weeks. We developed two items by rephrasing those used by Cardon et al. (2013): “I am very excited about the business idea of our team” and “I feel energized when I think about putting our business idea into practice”. Participants provided their answers on a 7-point response scale ranging from 1 (not at all) to 7 (absolutely). We combined the two items by computing the mean (α = .63).

We measured identity centrality before (T1) and after the training program (T2). We measured identity centrality aspiration, which refers to the relative importance of entrepreneurship to one’s future self-concept (Farmer et al., 2011). Specifically, identity aspiration captures the relative importance of becoming an entrepreneur and is a strong predictor of people’s actions (Farmer et al., 2011). Aspirations serve as a standard that people seek to confirm and validate, fulfilling the function of identity standards in the feedback loop (Burke & Stets, 2009; Farmer et al., 2011). We used three items developed by Farmer et al. (2011). An example item was “Becoming an entrepreneur would be an important part of who I am”. Participants responded by using a 5-point response scale ranging from 1 (strongly disagree) to 5 (strongly agree). Cronbach’s α was α = 0.86.

Entrepreneurial Self-Efficacy

We measured entrepreneurial self-efficacy during the training program with weekly measurements. We assessed entrepreneurial self-efficacy four times, at W1, W4, W7, and W10. We used four items developed by Gielnik, Frese et al. (2015) that referred to activities relevant in the beginning of the entrepreneurial process. All items started with the general stem “How confident are you that you can” followed by “start a business”, “become self-employed”, “overcome problems when starting a business”, and “manage a business well”. We used a 7-point response scale anchoring from 1 (20%) to 7 (100%). We computed the mean of the four items (Cronbach’s α = 0.90). Furthermore, we measured entrepreneurial self-efficacy before (T1) and after the training program (T2). We used the same 12 items as in Study II (Cronbach’s Alpha α = 0.93).

Entrepreneurial Performance

Using the same two items as in Study II, as well as two additional items (“In the last week, the team did a very good job” and “In the last week, we achieved the goals we have set for the team business”), we assessed entrepreneurial performance during the training program with weekly measurements. We computed the mean of the four items (Cronbach’s α = 0.93). Similar to Study II, we measured the profit generated by the participants at the end of the training, which correlated significantly with the mean of the self-report measure across the 12 weeks (r = 0.24, p < .01).

Control Variables

We used the same control variables as in Study II.

Study III: Results

Table 7 presents descriptive statistics. Null models showed that 72% of the total variance in positive feelings, 57% in entrepreneurial self-efficacy, and 57% in entrepreneurial performance was within-person variance. All VIF scores were below 2.5 (O’Brien, 2007).

Table 7

Study III: Descriptive Statistics of Variables on Level 1 (Level of Observations) and Level 2 (Level of Participants).

Variables	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18
Level 2
1. Gender^a	0.78	0.41
2. Year of studies	2.63	0.71	−.12
3. Entrepreneurial experience^b	0.46	0.50	.16*	.09
4. Entrepreneurial team size	5.89	1.00	.05	.03	.07
5. Identity centrality T1	4.66	0.49	.02	.09	.10	−.02
6. Identity centrality T2	4.82	0.37	−.03	.11	.09	−.04	.16**
7. Entrepreneurial self-efficacy T1	4.17	0.60	−.05	.04	.08	−.01	.12	.03
8. Entrepreneurial self-efficacy T2	4.39	0.51	.00	.06	.07	.01	.13^†	.42**	.33**
9. Mean entrepreneurial self-efficacy (W2–W12)^c	5.90	0.69	−.09	.09	.11	−.06	.20**	.24**	.52**	.51**
10. Mean positive feelings (W2–W12)^c	5.93	0.69	−.08	.16*	.16*	.13^†	.15*	.27**	.21**	.33**	.40**
11. Mean entrepreneurial performance (W2–W12)^c	5.72	0.71	−.26**	.10	−.05	.07	.17*	.24**	.25**	.30**	.57**	.51**
Level 1
1. ESE W1	5.58	0.99
2. ESE W4	5.86	0.85	.50**
3. ESE W7	6.04	0.82	.46**	.50**
4. ESE W10	6.13	0.76	.42**	.42**	.59**
5. POF W2	5.98	0.95	.11	.09	.14	−.02
6. POF W5	5.96	0.98	.25**	.25**	.31**	.22*	.25**
7. POF W8	5.98	0.92	.25**	.20*	.46**	.29**	.27**	.19*
8. POF W11	5.87	1.00	.18*	.20**	.26**	.47**	.23**	.31**	.38**
9. PER W2	5.62	1.10	.33**	.24**	.16^†	.19*	.26**	.16	.18*	.22**
10. PER W3	5.61	1.00	.27**	.24**	.32**	.21**	.12	.39**	.26**	.25**	.27**
11. PER W4	5.50	1.06	.17*	.35**	.23**	.17*	−.07	.24**	.20*	.22**	.22**	.65**
12. PER W5	5.80	0.80	.34**	.31**	.31**	.33**	.03	.27**	.31**	.25**	.48**	.39**	.46**
13. PER W6	5.81	0.80	.30**	.29**	.47**	.36**	.08	.17^†	.51**	.37**	.36**	.39**	.36**	.57**
14. PER W7	5.77	0.90	.46**	.23**	.49**	.38**	.06	.25**	.33**	.16^†	.26**	.51**	.38**	.47**	.54**
15. PER W8	5.78	0.90	.39**	.20*	.45**	.34**	−.05	.10	.44**	.23**	.30**	.38**	.28**	.42**	.60**	.63**
16. PER W9	5.74	0.95	.19*	.26**	.34**	.52**	.00	.13	.40**	.51**	.18*	.29**	.44**	.36**	.51**	.38**	.59**
17. PER W10	5.67	1.10	.23**	.25**	.34**	.50**	.03	.10	.35**	.53**	.18*	.22**	.31**	.39**	.46**	.44**	.42**	.70**
18. PER W11	5.73	1.16	.34**	.30**	.35**	.50**	−.03	.20*	.44**	.60**	.29**	.38**	.40**	.47**	.63**	.51**	.57**	.66**	.79**
19. PER W12	5.98	1.08	.29**	.26**	.23**	.38**	.03	.17*	.33**	.47**	.25**	.27**	.23**	.37**	.45**	.41**	.43**	.35**	.45**	.69**

Note. Level 2: 178 < N < 180; ^a0 = female, 1 = male; ^b0 = no, 1 = yes; ^caggregated across the weekly measurement waves of training program. Level 1: 105 < N < 169; ESE = entrepreneurial self-efficacy; POF = positive feelings; PER = entrepreneurial performance; SD = standard deviation; ^†p < .10, *p < .05, **p < .01.

Hypotheses Testing

Table 8 shows the results based on the weekly data obtained during the training program. Entrepreneurial performance had a positive effect on both entrepreneurial self-efficacy (Model 2: b = 0.24, p < .001) and positive feelings (Model 4: b = 0.32, p < .001). Moreover, entrepreneurial self-efficacy predicted positive feelings (Model 5: b = 0.30, p < .001). The indirect effect of entrepreneurial performance on positive feelings through entrepreneurial self-efficacy was significant (indirect effect = 0.07, p < .001), providing support for Hypothesis 1. Furthermore, we found evidence for the reverse effect of positive feelings on entrepreneurial self-efficacy (Table 9, Model 1: b = 0.10, p = .009) and on entrepreneurial performance (Table 9, Model 3: b = 0.10, p = .011). Entrepreneurial self-efficacy had a positive effect on entrepreneurial performance (Table 9, Model 4: b = 0.29, p < .001), and it mediated the effect of positive feelings on entrepreneurial performance (indirect effect = 0.03, p = .008), replicating the findings from Study I.

Table 8

Study III: Testing the Direct and Indirect Effect of Entrepreneurial Performance on Positive Feelings Through Entrepreneurial Self-Efficacy in a Weekly Research Design.

	Entrepreneurial self-efficacy(lagged, weekly)				Positive feelings(lagged, weekly)
	Model 1		Model 2		Model 3		Model 4		Model 5
	b	SE	b	SE	b	SE	b	SE	b	SE
Measurement wave	0.03*	(0.01)	0.03*	(0.01)	−0.02	(0.02)	−0.03	(0.02)	−0.05**	(0.02)
Gender^a	−0.23*	(0.09)	−0.11	(0.09)	−0.13	(0.14)	0.06	(0.13)	−0.06	(0.13)
Year of studies	−0.04	(0.05)	−0.05	(0.05)	0.15^†	(0.08)	0.14^†	(0.08)	0.10	(0.08)
Entrepreneurial experience^b	0.14^†	(0.07)	0.15*	(0.07)	0.14	(0.11)	0.14	(0.10)	0.11	(0.10)
Entrepreneurial team size	−0.02	(0.04)	−0.06^†	(0.04)	0.05	(0.07)	0.03	(0.06)	0.04	(0.06)
Entrepreneurial self-efficacy	0.46**	(0.04)	0.38**	(0.04)					0.30**	(0.07)
Positive feeling					0.16**	(0.05)	0.14**	(0.05)	0.05	(0.05)
Entrepreneurial performance			0.24**	(0.04)			0.32**	(0.05)	0.29**	(0.06)
Deviance (−2LogLikelihood)	767.34		736.74		831.05		799.60		670.91
R² marginal/conditional	0.31/0.31		0.37/0.37		0.07/0.23		0.19/0.26		0.26/0.35

Note. Model 1 and Model 2: number of teams = 34, number of participants = 169, number of observations = 362; Model 3 and Model 4: number of teams = 34, number of participants = 159, number of observations = 316; Model 5: number of teams = 34, number of participants = 154, number of observations = 281; unstandardized regression coefficients (b’s) are shown; ^a 0 = female, 1 = male; ^b 0 = no, 1 = yes; ^†p < .10, *p < .05, **p < .01. SE = standard error.

Table 9

Study III: Testing the Direct and Indirect Effect of Positive Feelings on Entrepreneurial Performance Through Entrepreneurial Self-Efficacy in a Weekly Research Design.

	Entrepreneurial self-efficacy (lagged, weekly)		Entrepreneurial performance(lagged, weekly)
	Model 1		Model 2		Model 3		Model 4
	b	SE	b	SE	b	SE	b	SE
Measurement wave	0.02	(0.01)	0.03**	(0.01)	0.03**	(0.01)	−0.03	(0.02)
Gender^a	−0.16^†	(0.09)	−0.18^†	(0.09)	−0.17^†	(0.09)	−0.21	(0.13)
Year of studies	−0.08	(0.05)	0.04	(0.05)	0.03	(0.05)	0.04	(0.08)
Entrepreneurial experience^b	0.09	(0.07)	0.07	(0.07)	0.04	(0.07)	−0.16	(0.10)
Entrepreneurial team size	0.00	(0.04)	0.11	(0.05)	0.10*	(0.04)	0.02	(0.07)
Entrepreneurial self-efficacy	0.44**	(0.04)					0.29**	(0.07)
Positive feelings	0.10**	(0.04)			0.10*	(0.04)	0.07	(0.05)
Entrepreneurial performance			0.49**	(0.04)	0.45**	(0.04)	0.37**	(0.06)
Deviance (−2LogLikelihood)	648.40		1,304.92		1,302.58		787.95
R² marginal/conditional	0.33/0.33		0.30/0.34		0.31/0.35		0.27/0.38

Note. Model 1: number of teams = 33, number of participants = 165, number of observations = 325; Model 2 and Model 3: number of teams = 34, number of participants = 180, number of observations = 544; Model 4: number of teams = 34, number of participants = 159, number of observations = 316; unstandardized regression coefficients (b’s) are shown; ^a0 = female, 1 = male; ^b0 = no, 1 = yes; SE = standard error; ^†p < .10, *p < .05, **p < .01.

As in Study II, we used condition-based regression analysis to test Hypothesis 2. Table 10 shows that entrepreneurial performance during the training program predicted entrepreneurial self-efficacy at T2 (Model 2: b = 0.18, p < .001) and identity centrality (Model 3: b = 0.12, p = .003). Entrepreneurial self-efficacy at T1 had a negative (b = −0.10, p = .033) and entrepreneurial self-efficacy at T2 had a positive effect on identity centrality (b = 0.29, p < .001). The lemma was significant (lemma = 0.19, p = .032), providing evidence that growth in entrepreneurial self-efficacy predicted identity centrality beyond the main effect of entrepreneurial self-efficacy at T2. We calculated the indirect effect of entrepreneurial performance on identity centrality through growth in entrepreneurial self-efficacy based on the lemma. The indirect effect was significant (indirect effect = 0.04, p = .030), supporting Hypothesis 2 and replicating the findings from Study II.

Table 10

Study III: Testing the Effect of Entrepreneurial Performance and Growth in Entrepreneurial Self-Efficacy on Changes in Identity Centrality Over Three Months (T1–T2).

	Entrepreneurial self-efficacy (T2)						Identity centrality (T2)
	Model 1			Model 2			Model 3			Model 4
	b	SE	β	b	SE	β	b	SE	β	b	SE	β
Gender^a	0.01	(0.09)	0.01	0.08	(0.09)	0.07	0.03	(0.07)	0.03	0.00	(0.06)	0.00
Year of studies	0.03	(0.05)	0.04	0.02	(0.04)	0.03	0.04	(0.04)	0.08	0.03	(0.04)	0.06
Entrepreneurial experience^b	0.04	(0.07)	0.04	0.05	(0.09)	0.05	0.06	(0.05)	0.08	0.05	(0.05)	.007
Identity centrality (T1)							0.08	(0.06)	0.10	0.07	(0.05)	0.09
Entrepreneurial self-efficacy (T1)	0.28**	(0.06)	0.33	0.23**	(0.07)	0.27				−0.10*	(0.04)	−0.16
Entrepreneurial self-efficacy (T2)										0.29**	(0.05)	0.40
Mean entrepreneurial performance(W2–W12)				0.18**	(0.05)	0.25	0.12**	(0.04)	0.23	0.08^†	(0.04)	0.15
F	5.64			7.07			3.33			7.00
R ²	0.12			0.17			0.09			0.22

Note. Number of participants = 178; ^a0 = female, 1 = male; ^b0 = no, 1 = yes; ^†p < .10, *p < .05, **p < .01. SE = standard error.

Analogous to Study II, we used the profit generated by the participants during the training program as an alternative measure of entrepreneurial performance. Again, the results were consistent. Profit predicted entrepreneurial self-efficacy (b = 0.01, p = .003) and had an indirect effect on identity centrality through growth in entrepreneurial self-efficacy (indirect effect = 0.001, p = .034), providing evidence for the robustness of our results.

Study III: Discussion

Study III provided support for the hypothesized effects of entrepreneurial performance on the components of passion in terms of positive feelings and identity centrality through entrepreneurial self-efficacy over shorter and longer cycle times, respectively. The results replicated the findings from Study I and Study II. Replicating findings is important to provide evidence for the validity and generalizability of results (Open Science Collaboration, 2015).

General Discussion

We tested a theoretical model of the development of passion in entrepreneurship in three longitudinal studies and found support for our hypotheses that entrepreneurial self-efficacy mediates the effects of entrepreneurial performance on positive feelings over a shorter time frame and the effects of entrepreneurial performance on identity centrality over a longer time frame. Furthermore, we found support for the reciprocal nature of the relationships between passion, entrepreneurial self-efficacy, and performance.

Theoretical Implications

Based on self-regulation theory, we present a theoretical model of the development of passion that considers the time horizons over which the components of passion develop. Previous research on the development of passion has predominantly focused on explaining changes in positive feelings (Collewaert et al., 2016; Dalborg & Wincent, 2015; Gielnik, Spitzmuller et al., 2015), or has examined the composite without disentangling the two components of passion (Cardon & Kirk, 2015). Going beyond these studies, we considered the nuances of the different time horizons of the development of each component. Our model explains that the components of passion develop as a result of the cognitive evaluation of one’s entrepreneurial performance, emphasizing the importance of entrepreneurial self-efficacy for the development of passion. Furthermore, our model specifies the time horizons over which the components of passion develop, consistent with the notion of longer cycle times on a high level (i.e., identity centrality) and shorter cycle times on an intermediate level (i.e., positive feelings). Accordingly, our theoretical model provides a self-regulatory explanation for the fundamentally different properties with respect to the stability and time horizons over which the components of passion develop.

Our theoretical model, which proposes shorter and longer cycle times over which positive feelings and identity centrality develop, offers a starting point for future research to identify exact time horizons of these developments. Scholars pointed to the importance of specifying the time horizons over which phenomena unfold, such as the development of passion, in order to improve the predictive validity of theories (Mitchell & James, 2001). In case there is little theory about the time horizon of certain phenomena, scholars suggest using empirical findings to begin formulating theories on the timing of effects (Zaheer et al., 1999). Our study draws on self-regulation theory and demonstrates effects over shorter and longer cycle times, in terms of weeks and months. We thus provide a theoretical and empirical basis for future research to develop hypotheses regarding the time interval of effects to unfold in the development of passion (Dormann & Griffin, 2015).

Our study adds to research on the formation of identities in entrepreneurship by providing new insights into the development of identity centrality as the component of passion that operates on a high regulatory level. Our theoretical model holds that growth in entrepreneurial self-efficacy, and thus becoming more efficacious, is critical to the internalization process of developing an entrepreneurial identity. Accordingly, an entrepreneurial identity develops after proving viable in feedback loops in which individuals appraise their performance and develop the belief that they have mastered the domain of entrepreneurship. Our study thus supports notions that people’s entrepreneurial identity becomes more central in the hierarchy of self-concepts as the result of a development process that involves continuously internalizing the meanings of the entrepreneurial role (Murnieks et al., 2014). Furthermore, our study adds to previous research on identity formation early in the entrepreneurial process, suggesting that entrepreneurs establish and confirm their identities while working on their ventures and ideas (Grimes, 2018; Powell & Baker, 2014). Our study demonstrates that an increase in the belief to be competent in entrepreneurship reflects a mechanism through which entrepreneurs reassure and verify their identity.

Finally, our model of the development of passion proposes a reciprocal instead of a unidirectional view of the relationships between passion, entrepreneurial self-efficacy, and performance. The view of reciprocal relationships follows from a self-regulation perspective on the components of passion. This perspective suggests that emotions and identity are both outcome and antecedent of behavior, albeit these cycles operate on different self-regulatory levels (Lord et al., 2010). A self-regulation perspective on passion thus consolidates previous findings on the relationship between passion, entrepreneurial self-efficacy, and performance, which provides evidence for the respective unidirectional relationships (e.g., Cardon & Kirk, 2015; Dalborg & Wincent, 2015; Drnovsek et al., 2016; Murnieks et al., 2014). Our model integrates the fragmented perspectives on the antecedents and outcomes of passion, showing that previous findings are complementary and result in a recursive model of passion when considered simultaneously. Accordingly, asking the question “What comes first, passion or performance?” might be as futile as pondering the chicken-egg dilemma. Our recursive model suggests that passion and performance develop jointly and iteratively over time in a circular manner, not necessarily with a starting point inherent to unidirectional relationships.

It is important to note, however, that our model suggests that the two sides of the cycle are not simply mirror images of each other. Rather, our model demonstrates that a more fine-grained perspective is necessary in order to consider the different regulatory levels on which the components of passion operate. Positive feelings and identity centrality operate on different levels and thus develop over different time frames. Nevertheless, they can jointly operate to influence entrepreneurial self-efficacy and performance over a short cycle time of several days. Identity centrality, which operates on a high level, gives meaning and direction and can thus enhance the effect of positive feelings on entrepreneurial self-efficacy and performance at an intermediate regulatory level. This reasoning is in line with the conceptualization that the interaction of positive feelings and identity centrality exert positive effects on entrepreneurial performance (Cardon et al., 2009, 2013; Drnovsek et al., 2016). Positive feelings only are not sufficient for passion to unfold its motivational potential; rather, both components must be addressed. Accordingly, interventions that primarily focus on boosting individuals’ positive feelings without strengthening their entrepreneurial identity might not result in long-term entrepreneurial success (cf., Souitaris et al., 2007). Furthermore, our model depicts a cycle that could be self-reinforcing, resulting in continuous growth in entrepreneurial self-efficacy over time, and, ultimately achieving 100% confidence. However, it is important to note that temporary reductions in entrepreneurial self-efficacy that are reflected in variability and a level of self-efficacy below 100% confidence (e.g., around 90%) are positively related to goal accomplishment in entrepreneurship (Gielnik et al., 2020). Accordingly, only experiencing growth in entrepreneurial self-efficacy is not necessarily positive; experiencing setbacks and temporary dips in entrepreneurial self-efficacy are eventually beneficial for long-term success because they prevent overconfidence and overshooting in entrepreneurial self-efficacy (Gielnik et al., 2020).

Strengths and Limitations

A limitation of our research design is that we assessed all study variables using self-reports. In particular, our self-report measurements of performance might assess subjective rather than objective performance. However, there are two reasons why measuring entrepreneurs’ perceived performance was adequate for our studies. First, most of our participants were in early phases of the entrepreneurial process, in which financial outcomes are oftentimes not available and in which an entrepreneur’s success is better reflected by their individual performance and progress in starting a new venture (Davidsson, 2016; Uy et al., 2015; Wall et al., 2004). Second, affective and cognitive experiences—such as positive feelings and self-efficacy—depend on an individual’s perceived performance rather than objective success (Bandura, 1989; Holland & Garrett, 2015). Accordingly, subjective measures of performance better capture the mechanisms underlying our theoretical reasoning. Notwithstanding, we provide evidence that the subjective performance of our participant entrepreneurs was significantly correlated with the profit made during the study period, indicating that the subjective and objective measures of performance are related.

Moreover, assessing all variables with self-reports could inflate the correlations among study variables, due to common method variance (Podsakoff et al., 2003). Most problems associated with common method variance were resolved in our studies because we controlled for earlier levels of the dependent variable in all analyses, thus ruling out constant sources of common method variance such as negative affectivity and response biases (Spector, 2006). In addition, our results regarding passion should not be affected by common method variance because interaction terms are not biased when controlling for the main effects of the respective variables (Siemsen et al., 2010).

We note that we did not follow Cardon et al.’s (2013) differentiation of passion for the specific roles of inventing, founding, and developing. Rather, we captured passion for inventing in Studies I and II, and passion for the business idea in Study III. This approach is in line with previous research that conceptualized passion for entrepreneurship broadly (Murnieks et al., 2016). Furthermore, research suggests that entrepreneurs can be passionate about different objects, such as entrepreneurship, their venture, or a business opportunity (Cardon et al., 2017; Warnick et al., 2018). Future research should examine the reciprocal relationships between passion, entrepreneurial self-efficacy, and performance by relying on different facets of passion, including passion for founding and developing. We further note that we measured target-specific affective experiences that focus on positive feelings toward a specific entrepreneurial role or business idea. Due to the target-specific nature, these positive feelings will vary to a lesser extent than other nontarget specific affective experiences that can fluctuate more heavily over time. Finally, we did not control for economic status and general competence in our analyses. However, these constructs are relatively stable and, therefore, less likely to influence the dynamic covariation of the main variables of our model.

Another factor to take into consideration is that we conducted our studies in Tanzania, which may limit the generalizability of our findings. However, there are two reasons why limited generalizability may be less of a concern for our findings. First, previous research conducted in more developed countries led to similar results with respect to the effects of performance and self-efficacy on passion (Gielnik, Spitzmuller et al., 2015; Murnieks et al., 2014). These findings indicate that our results are applicable to more developed countries as well. Second, it is important to consider that people living in less developed countries, such as Tanzania, constitute the majority of the world and thus represent an important yet chronically understudied population (Reynolds, 2012). In fact, research needs to include such countries to develop representative theories (Bruton, 2010). Against this background, scholars have explicitly called for adopting Africa as a research context in order to extend existing theoretical perspectives in the fields of management and entrepreneurship (George et al., 2016).

Our research relies on data collected in entrepreneurship training programs. This raises the question whether our findings are generalizable to ventures operating outside of a training context. In particular, it is possible that identity centrality may take a longer time to develop outside of a training context in which participants may be especially receptive to cues associated with the development of their identity as an entrepreneur (Miscenko et al., 2017). If true, this would suggest that the time horizon of the long cycle time of passion development through identity centrality may even be longer in other venture settings. This possibility further accentuates the value of a more fine-grained conceptualization of temporal dynamics in the development of passion in entrepreneurship. We further note that our samples in Study II and Study III included more male than female participants, which might limit the generalizability of our findings. However, this issue might be less of a concern because of the equivalent results in Study I, which were based on a more balanced sample.

During our studies, we investigated the role of entrepreneurial performance for the development of passion. We note that the factors that shape entrepreneurial performance can be different from those that lead to failure. In fact, organizational success and failure are predicted by different factors. For example, inertia is a key factor in the decline of once successful companies, whereas the absence of inertia is not sufficient to drive corporate success (Habersang et al., 2019). We thus encourage future research on cyclical processes of passion to consider entrepreneurial failure in addition to entrepreneurial performance.

We also encourage future research to revisit the differentiation between harmonious and obsessive passion (Vallerand et al., 2003) in entrepreneurship. Entrepreneurs are known to invest all their effort into their ventures, often at great personal costs (Spivack & McKelvie, 2018). There is a potential danger for entrepreneurs to become obsessively focused on the development of the venture, especially when early successes provide reinforcement for an exclusive focus on the venture. There is even the possibility of entrepreneurship addiction as an extreme form of commitment (Spivack & McKelvie, 2018). We therefore point to the possibility of a dark side to the cycle that we describe in our research—the same factors that drive the cycle of passion and entrepreneurial performance in our research may become self-destructive over time when coupled with an obsessive passion for the venture.

We also encourage future research to investigate boundary conditions for the effects in our recursive model. As with any model that contains recursive effects, it is important to understand the conditions under which the cycle can weaken or break down. Specifically, we encourage research that investigates how the perceived difficulty of the entrepreneurial task could moderate the effect of entrepreneurial performance on self-efficacy, and vice versa (Vancouver et al., 2008). Similarly, we view research on the role of interpersonal relationships among members of the venture team for self-regulation in venture teams as a promising venue for future research (Klotz et al., 2014).

Conclusion

Passion and its components of positive feelings and identity centrality develop over time as a function of entrepreneurial self-efficacy and performance. Positive feelings can change as a consequence of entrepreneurial self-efficacy and performance over shorter cycle times, while changes in identity centrality require growth in entrepreneurial self-efficacy over a longer cycle time. Furthermore, passion has reverse effects on entrepreneurial self-efficacy and performance. Overall, the results of our research suggest that the development of passion unfolds over both shorter and longer time horizons that can be understood from a self-regulation perspective.

Footnotes

Acknowledgements

This research was supported by German Academic Exchange Service (Deutscher Akademischer Austausch Dienst [DAAD]; ID 50020279 and ID 54391079). Furthermore, we would like to express our gratitude toward the German Commission for UNESCO and the BASF Stiftung for supporting the STEP project. We are grateful to Miriam Stark and Kim Marie Bischoff for their support in the set-up of this study. We also thank Tabea Sarah Müller, Martine Jelden, Daniel Müller, Tiziano Gabriele Tonin, Johannes Max Hilck, and Jan Paap for their support in collecting the data. Special thanks go to the STEP team at University of Dar es Salaam, in particular Lemayon Melyoki and Cosmas Moanja, for their support in implementing the STEP training.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This research was supported by German Academic Exchange Service (Deutscher Akademischer Austausch Dienst [DAAD]; ID 50020279 and ID 54391079).

ORCID iD

Michael Frese

Author Biographies

Maike Lex is a PhD graduate from the Leuphana University of Lueneburg, Germany, and currently works as Executive Assistant in Manufacturing Engineering Daimler Trucks at Daimler Truck AG. Her research interests encompass entrepreneurship, in particular entrepreneurial passion and entrepreneurship trainings, as well as creativity.

Michael M. Gielnik is a Professor of HR Development at the Leuphana University of Lueneburg. He studied psychology at the University of Giessen and received his PhD from the Leuphana University of Lueneburg. He was a Visiting Senior Fellow at the National University of Singapore Business School. His research focuses on entrepreneurship from a psychological perspective, in particular entrepreneurial learning and training. He has taken a special interest in entrepreneurship in developing countries.

Matthias Spitzmuller is an Associate Professor of Organizational Behavior at the Smith School of Business at Queen’s University. His research investigates self-regulatory processes at the within-person, between-person and group level and in different contexts including leadership, helping behaviours, and entrepreneurship.

Gabriel Henry Jacob is a research staff at the Nanyang Technological University and a faculty member at the Asian Pastoral Institute. He received his PhD from the National University of Singapore. His research focuses on entrepreneurship, in topics such as psychology of entrepreneurship, new venture teams, entrepreneurship training, and entrepreneurship in developing countries.

Michael Frese has a joint appointment as Professor for Organizational Behavior at the Asia School of Business (in collaboration with MIT Sloan Management) (ASB) in Kuala Lumpur, Malaysia, and Professor for Psychology and Entrepreneurship at the Leuphana University Lueneburg, Germany. He studies the psychology of entrepreneurship, innovation, training entrepreneurs, personal initiative and proactivity, and learning from errors and experience; using action theory as a meta-theory.

References

Ashforth

B. E.

Schinoff

B. S

. (2016). Identity under construction: How individuals come to define themselves in organizations. Annual Review of Organizational Psychology and Organizational Behavior, 3(1), 111–137.doi:10.1146/annurev-orgpsych-041015-062322

Bandura

. (1989). Human agency in social cognitive theory. American Psychologist, 44(9), 1175–1184.doi:10.1037/0003-066X.44.9.1175

http://www.ncbi.nlm.nih.gov/pubmed/2782727

Baum

J. R.

Locke

E. A

. (2004). The relationship of entrepreneurial traits, skill, and motivation to subsequent venture growth. Journal of Applied Psychology, 89(4), 587–598.doi:10.1037/0021-9010.89.4.587

http://www.ncbi.nlm.nih.gov/pubmed/15327346

Baumeister

R. F.

Vohs

K. D

. (2004). Handbook of self-regulation: Research, theory, and applications. Guilford Press.

Baumeister

R. F.

Vohs

K. D.

DeWall

C. N.

Zhang

. (2007). How emotion shapes behavior: Feedback, anticipation, and reflection, rather than direct causation. Personality and Social Psychology Review, 11(2), 167–203.doi:10.1177/1088868307301033

http://www.ncbi.nlm.nih.gov/pubmed/18453461

Beck

J. W.

Schmidt

A. M

. (2012). Taken out of context? Cross-level effects of between-person self-efficacy and difficulty on the within-person relationship of self-efficacy with resource allocation and performance. Organizational Behavior and Human Decision Processes, 119(2), 195–208.doi:10.1016/j.obhdp.2012.06.009

Bliese

P. D.

Ployhart

R. E

. (2002). Growth modeling using random coefficient models: Model building, testing, and illustrations. Organizational Research Methods, 5(4), 362–387.doi:10.1177/109442802237116

Bosco

F. A.

Aguinis

Singh

Field

J. G.

Pierce

C. A

. (2015). Correlational effect size benchmarks. Journal of Applied Psychology, 100(2), 431–449.doi:10.1037/a0038047

http://www.ncbi.nlm.nih.gov/pubmed/25314367

Breugst

Domurath

Patzelt

Klaukien

. (2012). Perceptions of entrepreneurial passion and employees’ commitment to entrepreneurial ventures. Entrepreneurship Theory and Practice, 36(1), 171–192.doi:10.1111/j.1540-6520.2011.00491.x

10.

Bruton

G. D

. (2010). Business and the world’s poorest billion - The need for an expanded examination by management scholars. The Academy of Management Perspectives, 24(3), 6–10.

11.

Burke

P. J

. (2006). Identity change. Social Psychology Quarterly, 69(1), 81–96.doi:10.1177/019027250606900106

12.

Burke

P. J.

Stets

J. E

. (2009). Identity theory. Oxford University Press.

13.

Callero

P. L

. (1985). Role-identity salience. Social Psychology Quarterly, 48(3), 203–215.doi:10.2307/3033681

14.

Cardon

M. S.

Glauser

Murnieks

C. Y

. (2017). Passion for what? Expanding the domains of entrepreneurial passion. Journal of Business Venturing Insights, 8, 24–32.doi:10.1016/j.jbvi.2017.05.004

15.

Cardon

M. S.

Gregoire

D. A.

Stevens

C. E.

Patel

P. C

. (2013). Measuring entrepreneurial passion: Conceptual foundations and scale validation. Journal of Business Venturing, 28(3), 373–396.doi:10.1016/j.jbusvent.2012.03.003

16.

Cardon

M. S.

Kirk

C. P

. (2015). Entrepreneurial passion as mediator of the self-efficacy to persistence relationship. Entrepreneurship Theory and Practice, 39(5), 1027–1050.doi:10.1111/etap.12089

17.

Cardon

M. S.

Wincent

Singh

Drnovsek

. (2009). The nature and experience of entrepreneurial passion. Academy of Management Review, 34(3), 511–532.doi:10.5465/amr.2009.40633190

18.

Chen

C. C.

Greene

P. G.

Crick

. (1998). Does entrepreneurial self-efficacy distinguish entrepreneurs from managers? Journal of Business Venturing, 13(4), 295–316.doi:10.1016/S0883-9026(97)00029-3

19.

Collewaert

Anseel

Crommelinck

De Beuckelaer

Vermeire

. (2016). When passion fades: Disentangling the temporal dynamics of entrepreneurial passion for founding. Journal of Management Studies, 53(6), 966–995.doi:10.1111/joms.12193

20.

Dalborg

Wincent

. (2015). The idea is not enough: The role of self-efficacy in mediating the relationship between pull entrepreneurship and founder passion—A research note. International Small Business Journal: Researching Entrepreneurship, 33(8), 974–984.doi:10.1177/0266242614543336

21.

Davidsson

. (2016). The dependent variable. In Davidsson

(Ed.), Researching entrepreneurship: Conceptualization and design (2nd ed.,pp. 195–215). Springer.

22.

Davidsson

Honig

. (2003). The role of social and human capital among nascent entrepreneurs. Journal of Business Venturing, 18(3), 301–331.doi:10.1016/S0883-9026(02)00097-6

23.

Donnellon

Ollila

Williams Middleton

. (2014). Constructing entrepreneurial identity in entrepreneurship education. The International Journal of Management Education, 12(3), 490–499.doi:10.1016/j.ijme.2014.05.004

24.

Dormann

Griffin

M. A

. (2015). Optimal time lags in panel studies. Psychological Methods, 20(4), 489–505.doi:10.1037/met0000041

http://www.ncbi.nlm.nih.gov/pubmed/26322999

25.

Drnovsek

Cardon

M. S.

Patel

P. C

. (2016). Direct and indirect effects of passion on growing technology ventures. Strategic Entrepreneurship Journal, 10(2), 194–213.doi:10.1002/sej.1213

26.

Enders

C. K.

Tofighi

. (2007). Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychological Methods, 12(2), 121–138.doi:10.1037/1082-989X.12.2.121

http://www.ncbi.nlm.nih.gov/pubmed/17563168

27.

Ewing

D. R.

Ewing

R. L

. (2017). Leveraging experiential learning to encourage role transition from “student” to “professional”: Insights from identity theory. Journal of Marketing Education, 39(3), 132–144.doi:10.1177/0273475317724844

28.

Farmer

S. M.

Yao

Kung-Mcintyre

. (2011). The behavioral impact of entrepreneur identity aspiration and prior entrepreneurial experience. Entrepreneurship Theory and Practice, 35(2), 245–273.doi:10.1111/j.1540-6520.2009.00358.x

29.

Foo

M.-D.

M. A.

Baron

R. A

. (2009). How do feelings influence effort? An empirical study of entrepreneurs’ affect and venture effort. Journal of Applied Psychology, 94(4), 1086–1094.doi:10.1037/a0015599

19594247

30.

Frese

Gielnik

M. M.

Mensmann

. (2016). Psychological training for entrepreneurs to take action: Contributing to poverty reduction in developing countries. Current Directions in Psychological Science, 25(3), 196–202.

31.

Frese

Krauss

Keith

Escher

Grabarkiewicz

Luneng

S. T.

Heers

Unger

Friedrich

. (2007). Business owners’ action planning and its relationship to business success in three African countries. Journal of Applied Psychology, 92(6), 1481–1498.doi:10.1037/0021-9010.92.6.1481

http://www.ncbi.nlm.nih.gov/pubmed/18020791

32.

George

Corbishley

Khayesi

J. N. O.

Haas

M. R.

Tihanyi

. (2016). Bringing Africa in: Promising directions for management research. Academy of Management Journal, 59(2), 377–393.doi:10.5465/amj.2016.4002

33.

Gielnik

M. M.

Bledow

Stark

M. S

. (2020). A dynamic account of self-efficacy in entrepreneurship. Journal of Applied Psychology, 105(5), 487–505.doi:10.1037/apl0000451

http://www.ncbi.nlm.nih.gov/pubmed/31535872

34.

Gielnik

M. M.

Frese

Kahara-Kawuki

Wasswa Katono

Kyejjusa

Ngoma

Munene

Namatovu-Dawa

Nansubuga

Orobia

Oyugi

Sejjaaka

Sserwanga

Walter

Bischoff

K. M.

Dlugosch

T. J

. (2015). Action and action-regulation in entrepreneurship: Evaluating a student training for promoting entrepreneurship. Academy of Management Learning & Education, 14(1), 69–94.doi:10.5465/amle.2012.0107

35.

Gielnik

M. M.

Spitzmuller

Schmitt

Klemann

D. K.

Frese

. (2015). “I put in effort, therefore I am passionate”: Investigating the path from effort to passion in entrepreneurship. Academy of Management Journal, 58(4), 1012–1031.doi:10.5465/amj.2011.0727

36.

Gielnik

M. M.

M. A.

Funken

Bischoff

K. M

. (2017). Boosting and sustaining passion: A long-term perspective on the effects of entrepreneurship training. Journal of Business Venturing, 32(3), 334–353.doi:10.1016/j.jbusvent.2017.02.003

37.

Grimes

M. G

. (2018). The pivot: How founders respond to feedback through idea and identity work. Academy of Management Journal, 61(5), 1692–1717.doi:10.5465/amj.2015.0823

38.

Habersang

Küberling‐Jost

Reihlen

Seckler

. (2019). A process perspective on organizational failure: A qualitative meta-analysis. Journal of Management Studies, 56(1), 19–56.doi:10.1111/joms.12341

39.

Holland

D. V.

Garrett

R. P

. (2015). Entrepreneur start-up versus persistence decisions: A critical evaluation of expectancy and value. International Small Business Journal, 33(2), 194–215.doi:10.1177/0266242613480375

40.

Huber

Baum

Frese

. (in press). Contagion of entrepreneurial passion: Effects on employee outcomes. Entrepreneurship Theory and Practice, .doi:10.1177/1042258719883995

41.

Humberg

Dufner

Schönbrodt

F. D.

Geukes

Hutteman

van Zalk

M. H. W.

Denissen

J. J. A.

Nestler

Back

M. D

. (2018). Enhanced versus simply positive: A new condition-based regression analysis to disentangle effects of self-enhancement from effects of positivity of self-view. Journal of Personality and Social Psychology, 114(2), 303–322.doi:10.1037/pspp0000134

http://www.ncbi.nlm.nih.gov/pubmed/28333473

42.

Ibarra

. (1999). Provisional selves: Experimenting with image and identity in professional adaptation. Administrative Science Quarterly, 44(4), 764–791.doi:10.2307/2667055

43.

Ilies

Judge

T. A

. (2005). Goal regulation across time: The effects of feedback and affect. Journal of Applied Psychology, 90(3), 453–467.doi:10.1037/0021-9010.90.3.453

http://www.ncbi.nlm.nih.gov/pubmed/15910142

44.

Klapper

L. F.

Parker

S. C

. (2011). Gender and the business environment for new firm creation. The World Bank Research Observer, 26(2), 237–257.doi:10.1093/wbro/lkp032

45.

Klotz

A. C.

Hmieleski

K. M.

Bradley

B. H.

Busenitz

L. W

. (2014). New venture teams: A review of the literature and roadmap for future research. Journal of Management, 40(1), 226–255.

46.

J. J.

Chen

X.-P.

Kotha

Fisher

. (2017). Catching fire and spreading it: A glimpse into displayed entrepreneurial passion in crowdfunding campaigns. The Journal of Applied Psychology, 102(7), 1075–1090.doi:10.1037/apl0000217

28333500

47.

Liden

R. C.

Wayne

S. J.

Stilwell

. (1993). A longitudinal study on the early development of leader-member exchanges. Journal of Applied Psychology, 78(4), 662–674.doi:10.1037/0021-9010.78.4.662

48.

Lord

R. G.

Diefendorff

J. M.

Schmidt

A. M.

Hall

R. J

. (2010). Self-regulation at work. Annual Review of Psychology, 61(1), 543–568.doi:10.1146/annurev.psych.093008.100314

http://www.ncbi.nlm.nih.gov/pubmed/19534590

49.

McCall

G. J.

Simmons

J. L

. (1978). Identities and interactions: An examination of human associations in everyday life. Free Press.

50.

Miscenko

Guenter

Day

. (2017). Am I a leader? Examining leader identity development over time. The Leadership Quarterly, 28(5), 605–620.doi:10.1016/j.leaqua.2017.01.004

51.

Mitchell

T. R.

James

L. R

. (2001). Building better theory: Time and the specification of when things happen. Academy of Management Review, 26(4), 530–547.doi:10.5465/amr.2001.5393889

52.

Mueller

B. A.

Wolfe

M. T.

Syed

. (2017). Passion and grit: An exploration of the pathways leading to venture success. Journal of Business Venturing, 32(3), 260–279.doi:10.1016/j.jbusvent.2017.02.001

53.

Murnieks

C. Y.

Cardon

M. S.

Sudek

White

T. D.

Brooks

W. T

. (2016). Drawn to the fire: The role of passion, tenacity and inspirational leadership in angel investing. Journal of Business Venturing, 31(4), 468–484.doi:10.1016/j.jbusvent.2016.05.002

54.

Murnieks

C. Y.

Mosakowski

Cardon

M. S

. (2014). Pathways of passion: Identity centrality, passion, and behavior among entrepreneurs. Journal of Management, 40(6), 1583–1606.

55.

Nakagawa

Schielzeth

. (2013). A general and simple method for obtaining R² from generalized linear mixed-effects models. Methods in Ecology and Evolution, 4(2), 133–142.doi:10.1111/j.2041-210x.2012.00261.x

56.

Open Science Collaboration . (2015). Estimating the reproducibility of psychological science. Science, 349(6251), 1–7.

57.

O’Brien

R. M

. (2007). A caution regarding rules of thumb for variance inflation factors. Quality & Quantity, 41(5), 673–690.doi:10.1007/s11135-006-9018-6

58.

Pearce

C. L.

Sims Jr

H. P

. (2002). Vertical versus shared leadership as predictors of the effectiveness of change management teams: An examination of aversive, directive, transactional, transformational, and empowering leader behaviors. Group Dynamics: Theory, Research, and Practice, 6(2), 172–197.doi:10.1037/1089-2699.6.2.172

59.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J.-Y.

Podsakoff

N. P

. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879–903.doi:10.1037/0021-9010.88.5.879

14516251

60.

Powell

E. E.

Baker

. (2014). It’s what you make of it: Founder identity and enacting strategic responses to adversity. Academy of Management Journal, 57(5), 1406–1433.doi:10.5465/amj.2012.0454

61.

Rauch

Frese

. (2007). Let’s put the person back into entrepreneurship research: A meta-analysis on the relationship between business owners’ personality traits, business creation, and success. European Journal of Work and Organizational Psychology, 16(4), 353–385.doi:10.1080/13594320701595438

62.

Reynolds

P. D

. (2012). Entrepreneurship in developing economies: The bottom billions and business creation. Foundations and Trends® in Entrepreneurship, 8(3), 141–277.doi:10.1561/0300000045

63.

Schutte

N. S.

Malouff

J. M

. (2016). General and realm-specific self-efficacy: Connections to life functioning. Current Psychology, 35(3), 361–369.doi:10.1007/s12144-014-9301-y

64.

Selig

J. P.

Preacher

K. J

. (2008). Monte Carlo method for assessing mediation: An interactive tool for creating confidence intervals for indirect effects. http://quantpsy.org/

65.

Siemsen

Roth

Oliveira

. (2010). Common method bias in regression models with linear, quadratic, and interaction effects. Organizational Research Methods, 13(3), 456–476.doi:10.1177/1094428109351241

66.

Sitzmann

Yeo

. (2013). A meta-analytic investigation of the within-person self-efficacy domain: Is self-efficacy a product of past performance or a driver of future performance? Personnel Psychology, 66(3), 531–568.doi:10.1111/peps.12035

67.

Smilor

R. W

. (1997). Entrepreneurship: Reflections on a subversive activity. Journal of Business Venturing, 12(5), 341–346.doi:10.1016/S0883-9026(97)00008-6

68.

Souitaris

Zerbinati

Al-Laham

. (2007). Do entrepreneurship programmes raise entrepreneurial intention of science and engineering students? The effect of learning, inspiration and resources. Journal of Business Venturing, 22(4), 566–591.doi:10.1016/j.jbusvent.2006.05.002

69.

Spector

P. E

. (2006). Method variance in organizational research: Truth or urban legend? Organizational Research Methods, 9(2), 221–232.

70.

Spivack

A. J.

McKelvie

. (2018). Entrepreneurship addiction: Shedding light on the manifestation of the “dark side” in work-behavior patterns. Academy of Management Perspectives, 32(3), 358–378.doi:10.5465/amp.2016.0185

71.

Stets

J. E.

Burke

P. J

. (2000). Identity theory and social identity theory. Social Psychology Quarterly, 63(3), 224–237.doi:10.2307/2695870

72.

M. A.

Foo

M.-D.

Aguinis

. (2010). Using experience sampling methodology to advance entrepreneurship theory and research. Organizational Research Methods, 13(1), 31–54.doi:10.1177/1094428109334977

73.

M. A.

Foo

M.-D.

Ilies

. (2015). Perceived progress variability and entrepreneurial effort intensity: The moderating role of venture goal commitment. Journal of Business Venturing, 30(3), 375–389.doi:10.1016/j.jbusvent.2014.02.001

74.

Vallerand

R. J.

Blanchard

Mageau

G. A.

Koestner

Ratelle

Léonard

Gagné

Marsolais

. (2003). Les Passions de l’ame: On obsessive and harmonious passion. Journal of Personality and Social Psychology, 85(4), 756–767.doi:10.1037/0022-3514.85.4.756

http://www.ncbi.nlm.nih.gov/pubmed/14561128

75.

Vancouver

J. B.

K. M.

Yoder

R. J

. (2008). Self-efficacy and resource allocation: Support for a nonmonotonic, discontinuous model. Journal of Applied Psychology, 93(1), 35–47.doi:10.1037/0021-9010.93.1.35

http://www.ncbi.nlm.nih.gov/pubmed/18211133

76.

Wall

T. D.

Michie

Patterson

Wood

S. J.

Sheehan

Clegg

C. W.

West

. (2004). On the validity of subjective measures of company performance. Personnel Psychology, 57(1), 95–118.doi:10.1111/j.1744-6570.2004.tb02485.x

77.

Warnick

B. J.

Murnieks

C. Y.

McMullen

J. S.

Brooks

W. T

. (2018). Passion for entrepreneurship or passion for the product? A conjoint analysis of angel and VC decision-making. Journal of Business Venturing, 33(3), 315–332.doi:10.1016/j.jbusvent.2018.01.002

78.

Zaheer

Albert

Zaheer

. (1999). Time scales and organizational theory. Academy of Management Review, 24(4), 725–741.doi:10.5465/amr.1999.2553250

79.

Zhao

Seibert

S. E.

Hills

G. E

. (2005). The mediating role of self-efficacy in the development of entrepreneurial intentions. Journal of Applied Psychology, 90(6), 1265–1272.doi:10.1037/0021-9010.90.6.1265

http://www.ncbi.nlm.nih.gov/pubmed/16316279