Entrepreneurial education and intention: fear of failure, self-efficacy and gender

New venture creation has been broadly defined as a cognitive process of discovering, evaluating and exploiting the business opportunities through which potential entrepreneurs seek to create additional value by engaging in entrepreneurial and innovative activities (Nguyen, 2023; Vasilev, 2022; Nguyen and Ahmed, 2023). They do so in order to make novel contributions to their personal prosperity, the country's economic development and job creation (Pelegrini and Moraes, 2021). The psychological understanding of why individuals engage in entrepreneurial activities although the start-up failure stories are countless (Asante and Affum-Osei, 2019), resulting in an increasing interest from scholars (Cacciotti et al., 2020). From the psychological aspect, researchers have consistently sought to explore how psychological attributes (i.e. entrepreneurial self-efficacy (ESE)) and psychological barriers (i.e. fear of failure FOF) affect individuals' entrepreneurial career choice (Van Trang et al., 2019; Duong, 2022; Asante and Affum-Osei, 2019).

Recently, scholars, policymakers and practitioners have become increasingly interested in entrepreneurial education (EE) due to its role in nurturing and fostering learners' entrepreneurial activities (Maheshwari, 2022). It has thus been identified as the foremost strategy to enhance a country's innovative and creative capacity (Duong, 2021). Most young people who are making the important decision of a career choice can obtain the knowledge and skills required to create and manage their own business venture through EE programs (Uddin et al., 2022). However, a question that has received scant attention is whether and why some learners might benefit more from EE than others. The person-environment fit theory (Kristof, 1996) showed that individuals' attitudinal outcomes are formed by their perceptions of both basic psychological factors and environmental factors. Moreover, many scholars argue that the effectiveness of EE in inspiring individuals' entrepreneurial activities depends significantly on psychological factors and personal traits (Nowiński et al., 2017). Unfortunately, in the entrepreneurship literature, little is known about how psychological barriers (i.e. FOF) and personal traits (i.e. gender) can weaken or enforce the impact of EE on entrepreneurial intention (EI) via underlying mediation mechanisms of cognitive state.

FOF is a common theme in the entrepreneurial literature. It is considered a primary reason why some individuals hesitate to pursue entrepreneurship altogether (Dutta and Sobel, 2021). Not only does it discourage people from creating and running a business venture but also it acts as a significant obstacle to various entrepreneurial activities (Darden et al., 2022). In particular, FOF can impede the transition from entrepreneurial ideas and knowledge, which are acquired through education and training programs, to an actual intention to run a business venture (Duong, 2022; Guelich, 2022). Although numerous prior studies have examined the direct and/or mediation effect of FOF (Stroe et al., 2020 and Hunter et al., 2020), little attention has been paid to the possibility that FOF can serve as a detrimental moderator of the process of entrepreneurship (Duong, 2022). This suggests that FOF not only has a direct impact on entrepreneurial activities but also that it negatively moderates the effect of EE on individuals' intentions to become entrepreneurs (Guelich, 2022). Although scholars have recently shown a growing interest in examining gender-based disparities in entrepreneurship research (Santos-Jaén et al., 2022), little is known about how gender can moderate the effects of EE on individuals' intentions to engage in entrepreneurial activities (Ramadani et al., 2022; Nowiński et al., 2017), especially via the mediation impact of ESE.

The social cognitive career theory (SCCT) (Lent and Brown, 2013), is an extension of the social cognitive theory (SCT) (Bandura, 2011) that focuses on how contextual support can enhance a person's career self-efficacy and intentions to pursue certain careers. The SCCT is broadly known as one of the most influential and empirically supported models in this field (Lent and Brown, 2013, 2019). It has thus been used as a framework in the current research to explain why and how some individuals become entrepreneurs while others prefer to be employed with the assurance of a salary (Meoli et al., 2020). The SCCT conceptualizes how external environments (i.e. entrepreneurial ecosystems and EE) interact with psychological attributions or barriers (i.e. FOF) to influence a particular action (i.e. new venture creation) (Lent and Brown, 2019). Our study therefore adopts the lens from the SCCT to draw a moderated moderated mediation model to address the above-mentioned research gaps and contribute to the entrepreneurship by answering the following research questions (RQs):

Entrepreneurial cognition research has generated numerous theories and methodologies that allow scholars to gain a better comprehension of the impact of fundamental beliefs and values on the inclination to engage in entrepreneurial activities (Calza et al., 2020). Within this stream, the theory of planned behavior (TPB) (Ajzen, 1991), the entrepreneurial event model (EEM) (Shapero and Sokol, 1982) and the SCCT (Lent and Brown, 2013), are the most frequently employed for explaining the entrepreneurial phenomenon (Duong, 2022; Meoli et al., 2020). Our study built on the SCCT to draw a moderated mediation model to examine whether FOF and gender can significantly moderate the direct and indirect impacts of EE and ESE on EI. The rationale for this choice is as follows. First, although many scholars argue that the primary theories utilized to predict EI are the TPB or the EEM (Duong, 2022), the adoption of the SCCT can offer a viable alternative to these two theories since it provides a career-oriented perspective on entrepreneurship and enables us to connect it with the broader field of career literature, which has been a key recommendation in this area of research (Meoli et al., 2020). Second, while the TPB and EEM help to explain the intentions and actions derived from beliefs and motivations, they neglect the role of external factors (Uskuri and Sesen, 2022) such as EE. Moreover, Pérez-López et al. (2019) argued that the SCCT in itself incorporates the key components of the EI models (i.e. TPB and EEM), such as self-efficacy (similar to perceived behavioral control in the TPB and perceived feasibility in the EEM), intentions or goals. Although the SCCT, TPB and EEM contain several similar points (Uskuri and Sesen, 2022), the SCCT is better suited to our study because it conceptualizes how contextual support accompanies internal and psychological motives and/or barriers and can thus affect a specific decision (Lent and Brown, 2019). As such, it offers a comprehensive theoretical framework that helps to determine under which conditions an individual would become an entrepreneur (Meoli et al., 2020), and how psychological barriers and personal traits could hinder or strengthen the cognitive process of entrepreneurship (Cui and Bell, 2022). In other words, according to the career approach, the SCCT can provide a sufficient framework to examine how EE (the contextual environment) directly increases ESE and EI, as well as how the indirect effects of EE on EI via ESE could be moderated by FOF and gender.

The SCCT framework highlighted the ways in which individuals experience personal agencies in the process of career development, as well as how motivational and psychological factors can reinforce or weaken these agencies (Osorio Tinoco et al., 2020). Self-efficacy is a crucial construct in the SCCT in terms of explaining career choice goals/intentions and actions. The SCCT comprises a framework with which one can investigate the causal nexus of self-efficacy and career goals/intentions with the influence of others (Lent and Brown, 2019). In the entrepreneurship literature, prior studies state that the constructs of the SCCT strictly refer to the components that contribute to the development of EIs, such as ESE (Uysal et al., 2022). Indeed, ESE is considered to be one of the most important psychological constructs that can explain why individuals consider entrepreneurship as a career choice (Neneh, 2020; Tsai et al., 2016a).

Self-efficacy is a psychological concept that relates to an individual's perception of their ability to control events in their life. This includes their beliefs about their capacity to generate expected outcomes, as well as their evaluation of their capability to mobilize motivation and cognitive resources and take appropriate actions (Bandura, 2011). People develop attitudes that influence their anticipated outcomes and select tasks that align with their skills and abilities to address challenges. Therefore, individuals tend to engage in behaviors that they believe will enable them to overcome the obstacles and difficulties they encounter (Neneh, 2020). In line with the SCCT, ESE refers to an individual's confidence in their capability to perform entrepreneurial tasks (Tsai et al., 2016a). Thus, it is hypothesized that the more a person perceives himself as being capable of creating and running his own business venture successfully, the higher their intent to become an entrepreneur (Hsu et al., 2019).

Fayolle et al. (2006) defined EE as “any pedagogical program or process of education for entrepreneurial attitudes and skills” (p. 702). Since the major purpose of EE and training is to train and educate students to become entrepreneurs, and to impart the essential knowledge and skills required to establish and run their own business ventures, it focuses on how to increase students' entrepreneurial knowledge and ability, help them to be aware of entrepreneurship as an career choice, and encourage them to take risks related to entrepreneurial activities (Cui and Bell, 2022). In other words, entrepreneurial competency, which fosters new value creation and ESE, can be nurtured via pedagogical start-up programs. EE can be instrumental to the formation of EIs since it equips students with specific skills and knowledge for establishing and developing their own business ventures (Cui and Bell, 2022). In support of this, current studies affirm that students who have received EE have high ESE (Maheshwari, 2022). However, the relationship between EE and EI is inconsistent. Indeed, while some studies report that EE can significantly increase EIs (Cui and Bell, 2022), others claim that EE is not significantly correlated with EIs (Maheshwari and Kha, 2022; Duong, 2021) or even that EE reduces EIs (Martínez-Gregorio et al., 2021). In the context of Vietnam, EE could have positive impacts on master's students' ESE and EI. Therefore, we hypothesize that:

EE acts as an effective catalyst for sculpting individuals' entrepreneurial skills and behaviors (Otache et al., 2022). Some academicians argue that EI can be developed when students have strong ESE, which can be nourished through EE and training. As mentioned above, the direct effect of EE on EI is rather weak (Duong, 2021; Maheshwari and Kha, 2022). This weak relationship can result from the absence of one or several mediators between two of these constructs (Otache et al., 2022). Hair et al. (2020) argue that the relationship between independent and dependent variables can be better explained through a mediator. In line with the SCCT framework, self-efficacy can convert the effects of different antecedents, such as support and barriers (Lent and Brown, 2019), ability/past performance (Lee et al., 2015) and personality and contextual influences (Lent et al., 2021), into choice goals/intentions. In other words, self-efficacy and career interests/attitudes can serve as mediators of the relationships between various predictors and career goals/intentions (Rogers and Creed, 2011). In the entrepreneurship research, prior studies have illustrated that ESE can serve as a potential mediator in the EE–EI link (Maheshwari and Kha, 2022; Saptono et al., 2021). The capacity to recognize and explore entrepreneurial opportunities requires a high level of self-confidence and great perceived behavior control (Loan et al., 2021), which are defined as ESE (Tsai et al., 2016a and Maheshwari, 2022). Thus, it is hypothesized that EE can increase ESE, and in turn, reinforce EI among Vietnamese master's students.

Nascent entrepreneurs are continuously assessing the opportunities they are pursuing within the context of the current business environment while they are engaging in the early stages of the entrepreneurial process. They are attentive to signals that suggest that their planned business venture may not succeed (Kollmann et al., 2017). It is true that challenges and barriers that jeopardize the success of new ventures are quite common during the initial stages of entrepreneurship (Cacciotti et al., 2016). Encountering obstacles and challenges during the early stages of entrepreneurship can trigger an FOF, which can significantly impact entrepreneurial activity. This FOF arises because the nascent entrepreneurship phase is a crucial period, where success is highly valued (Morgan and Sisak, 2016). When faced with obstacles that threaten the competitive advantage of the planned business, individuals may experience a FOF as an entrepreneur, namely FOF (Cacciotti et al., 2020). FOF is considered to be a psychological barrier that can cause potential entrepreneurs to avoid engaging in entrepreneurial acts (Darden et al., 2022). Cacciotti et al. (2020) defined FOF as “a negative effective reaction based on cognitive appraisals of potential for failure in the uncertain and ambiguous performative context of entrepreneurship”.

Some studies illustrate that an FOF can be detrimental to the entrepreneurial process (Cacciotti et al., 2020), such as the transition from initial EI to organizational emergence (Duong, 2022). Indeed, FOF can adversely affect EI (Duong, 2022), opportunity recognition (Hunter et al., 2020) and reengagement in entrepreneurial pursuits (Stroe et al., 2020). However, almost all prior studies fail to investigate how FOF could reduce the effectiveness of EE. In other words, FOF could act as a negative moderator of the transformation of students' knowledge and skills (gained from EE) into positive beliefs about their capacities and their EI (Guelich, 2022). FOF runs in parallel with risk aversion (Darden et al., 2022). Therefore, FOF can decrease the likelihood that individuals will turn the entrepreneurial knowledge and skills obtained from EE into the intention to create a new venture when they are highly risk averse (Cacciotti et al., 2020). In addition, when individuals are afraid of failing, they tend to be more inclined to avoid situations that may lead to failure to prevent experiencing the negative outcomes of failure (Cacciotti et al., 2016). FOF can cause individuals who are otherwise confident in their skills and capabilities to second-guess their willingness to engage in entrepreneurial activities because they are afraid of the possibility of failure, which in turn, can lead to a decrease in their motivation to pursue entrepreneurial opportunities (Ng and Jenkins, 2018). Consequently, although individuals may have a strong sense of ESE, their FOF can still prevent them from taking action toward achieving their entrepreneurial goals. This fear can hinder their ability to act on their intentions, even if they feel confident in their abilities. This means that the impact of EE on ESE and EI is presumed to vary in strength, depending on the level of FOF.

Males and females often act in dissimilar ways, a difference that leads to the different pathways of socialization accomplished by each gender (Blocker and Eckberg, 1993). Gender socialization theory highlighted that males and females experience the socialization process differently, even in the earliest phase of their lives, and hence their approaches to entrepreneurship can be different (Ramadani et al., 2022). The role of gender in entrepreneurship has been emphasized by a body of previous studies (St-Jean et al., 2022; Nguyen and Pham, 2021; Pelegrini and Moraes, 2021; Ettis, 2021). Ramadani et al. (2022) argue that the effect of EE on EI can differ significantly due to sociocultural differences. Some recent studies report that the proportion of female entrepreneurs is gradually increasing (Ojong et al., 2021; Chaudhary, 2017). Although several studies indicate that males are more likely to become entrepreneurs than females (St-Jean and Jacquemin, 2022), many others argue that females are encouraged more than males to become entrepreneurs in the emerging ecosystem (Pelegrini and Moraes, 2021). This is because they stive for a better income and empowerment (Aparicio et al., 2022) or just simply for financial independence (Ojong et al., 2021). In the context of the emerging economy of Vietnam, it is hypothesized that the impacts of EE and ESE on EI are stronger among females than among males.

If FOF and gender significantly moderate the direct impact of EE on the intention to create one's own business venture, it is also feasible to assume that individuals' FOF and gender can hypothetically affect the level of influence of the indirect effect of EE on EI via ESE. In other words, the mediating effect of ESE originating from the effect of EE on EIs can be moderated by FOF and gender, thereby demonstrating a moderated mediation impact. Some of the prior studies that adopted a moderated mediation model showed that when a moderator significantly moderated the effect of an independent variable on the outcome variable, it could also conditionally affect this link through a mediator (Duong, 2022). Our research thus assumes that the impact of EE on ESE and EI will be weakened when the level of FOF increases and that the effects of EE and ESE on EI are higher among females than among males. Therefore, it is hypothesized that FOF and gender significantly moderate the mediation effect. That is, the mediation effect of EE on EI through ESE will increase when FOF decreases and vice versa, and these mediation effects are always higher among females than among males.

The hypothesized model is demonstrated in Figure 1.

Vietnam's economic achievements can be attributed to its implementation of the “Đổi Mới” (Reform) policy in 1986, which marked the beginning of progressive economic reforms, as well as to the favorable global trends that have supported its growth (Nguyen et al., 2021; Le Thi Kim et al., 2021). Within a single generation, the country has undergone a remarkable transformation, progressing from being one of the world's poorest nations to becoming a middle-income economy (Le and Finch, 2021). Particularly, Vietnam achieved impressive economic growth between 2002 and 2021, with GDP per capita increasing nearly 3.6 times to reach US$3,743. Moreover, poverty rates, measured at US$3.65 income or less per day based on 2017 purchasing power parity, have significantly decreased from 14% in 2010 to just 3.8% in 2020 (GSO, 2020). The extent of Vietnam's transformation is underscored by the growth of operating firms in the country (Nguyen, 2023). While around 40,000 new enterprises were created between 1991 and 1999, the number increased dramatically by almost 18 times from 2000 to 2018, with over 700,000 new enterprises established during this period (Vuong et al., 2020). Considering the high number of small and medium-sized enterprises (SMEs), it is justifiable to view entrepreneurship as the cornerstone of the economy (Duong, 2022). Over the years, the number of SMEs has grown considerably, from roughly 38,000 in the 2000s to more than 785,000 in 2022, so that they comprised 98% of all enterprises (Nguyen, 2023). Additionally, between 2016 and 2019, there was a significant increase in the number of investments made in tech startups in Vietnam, growing by a factor of eight and reaching a record high of US$861 million in 2019. Moreover, the number of Fintech startups in the country grew from 44 in 2017 to 123 in 2020, suggesting that Vietnam has the potential to become a leading start-up hub in Southeast Asia (Hoang et al., 2022).

There has been a notable upswing in entrepreneurial activities worldwide, and Vietnam has gained recognition for providing favorable conditions for business startups (Maheshwari and Kha, 2022). In 2018, the country was recognized as one of the countries with the highest level of early-stage entrepreneurial activity (23.3%) and a significant entrepreneurial spirit index (0.26) (Vuong et al., 2020). Given the essential role that entrepreneurship has played in Vietnam's economic progress and its relevance to ongoing policy debates, it is important to assess the research trends of the past decade (Duong, 2022). In addition, the Vietnamese government has placed a greater emphasis on the significance of entrepreneurship in the country's economy in recent years, resulting in increased efforts to enhance entrepreneurship, particularly by integrating entrepreneurship education into the programs offered by national universities (Maheshwari and Kha, 2022). Nguyen et al. (2021) found that Vietnam had one of the lowest proportions of women entrepreneurs in 2021. Nevertheless, in 2018, women owned nearly 33% of businesses in the country, with 42% of them being SMEs and only 1% being large companies, placing Vietnam in the sixth position out of 53 surveyed countries. This ranking surpassed numerous developed nations/region such as Europe, the USA and China. Thus, Vietnam constitutes excellent research setting for our study.

In lieu of using real-world entrepreneurs, most previous researchers adopted undergraduate students as the targeted sample to investigate EI and/or entrepreneurial behavior (Gieure et al., 2020; Cui and Bell, 2022), since they are often inclined to become involved in entrepreneurial activities (Uddin et al., 2022). Nevertheless, a body of scholars argues that undergraduate students' EIs are easily altered, and that they often choose other career paths when they graduate from university (Meoli et al., 2020; Calza et al., 2020; St-Jean and Labelle, 2018). One of the most noticeable explanations for this is that undergraduate students often have no real working and/or business experience; hence, although they demonstrate a high intention to become entrepreneurs, these initial intentions are often changed when they face obstacles and challenges in the real-world business environment (Fayolle and Liñán, 2014). Moreover, according to the findings from a survey of 2,118 respondents and insights from 36 experts, GEM (2017/2018) offers a comprehensive perspective on entrepreneurship traits in Vietnam across various stages-from initial intentions to start a business through the early stages of starting up, running a new or established enterprise and even to the point of discontinuing a business. Perceived entrepreneurial capabilities appear to have declined overtime. In 2017, the rate of perceived entrepreneurial capabilities was 53%, ranking 19th out of 54 economies, which is lower than the 56.8% recorded in 2015, where it ranked 19th out of 60 economies. Notably, the average rate of perceived entrepreneurial capabilities in factor-driven economies stands at 53.8%. However, the rate for fear of entrepreneurial failure in Vietnam remained at the high level. For example, the percentage of adults experiencing fear of entrepreneurial failure decreased from 56.7% in 2013 to 45.6% in 2015. There was a slight increase to 46.6% in 2017. This places Vietnam at the 10th position out of 54 economies, indicating a higher fear of entrepreneurial failure compared to the average rate of 36.6% in factor-driven economies (Van Trang et al., 2019). In Vietnam, the rate of adults expressing start-up intentions for the next three years exhibited an upward trend, rising from 18.2% in 2014 to 22.3% in 2015 and further reaching 25% in 2017. Despite this progress, Vietnam's ranking stand at the 19th position out of 54 economies, indicating that it is still lower than the average rate of 30.3% observed in factor-driven economies. Consequently, a sample of master's students who had already at least two years' working/business experience was employed in this research to test how FOF and gender significantly moderate the mediation impact of ESE on the linkage between EE and EI.

To collect the data, a three-phase stratified random sampling approach was employed. In the first phase, based on their historical and geographical characteristics, the three main regions, Northern, Central and Southern Vietnam, were chosen for the survey. In the next phase, based on the number of master's students, we chose the three largest universities/institutions in each region for the distribution of the survey questionnaire. In the final phase, lecturers and coordinators who were responsible for master's programs and courses were contacted to ask for their support during the process of data collection. Particularly, we contacted lecturers and coordinators at nine universities and higher education institutions verbally over the phone before meeting them directly and asking them for help with delivering the questionnaires to their master's students. With their support, the hard-copy questionnaires were then distributed to students directly in the classroom. To increase the representativeness of the research sample, only master's students with at least two years' working and/or business experience were invited to take part in the survey. For this, we relied on the information provided by the master's program coordinators as well as the students' self-reporting. All of the participants were openly notified that their participation was completely voluntary, that they had the right to withdraw from the survey at any time if they wished and that their responses would be kept totally in confidence and only used for academic purposes.

The data collection was performed between 20 February and 20 April 2022. Of about 300 master's students were contacted at each university/institution and thus 2,700 respondents were targeted in total. Nevertheless, only 1,982 master's students participated in our study. After eliminating the questionnaires with missing responses, the valid sample consisted of 1,890 master's students. The demographic profiles of the master's students are presented in Table 1.

More than one-half of the participants are male (52.6%) and the majority were aged between 27 and 30 years old (49.8%), followed by those aged over 30 years old (24.4%) and those aged 23–26 years old (22.8%). The largest proportion of students was studying economics and business management, accounting for 56.5%. This high percentage was due to the fact that some universities in the chosen regions (i.e. the National Economics University and the University of Economics Ho Chi Minh city) only research and teach economics and business management. Moreover, 46.4% of the masters' students had actual business experience while 49.9% of them had a family business background.

To serve the research purposes, the scales utilized in our research were borrowed from prior research. All the scales have been widely utilized and validated in a body of recent research (Duong, 2022; Maheshwari, 2022; Cui and Bell, 2022). To specify, a six-item EI scale was borrowed from the study of Liñán and Chen (2009), a five-item scale of ESE was adopted from the study of Tsai et al. (2016a), a five-item scale of EE was adopted from the studies of Walter and Block (2016) and Adekiya and Ibrahim (2016), and a 21-item scale of FOF was developed by Cacciotti et al. (2020). Based on a seven-point Likert-type format, all of the items were scored from 1 to 7, representing “strongly disagree” to “strongly agree”, respectively. The questions related to social demographic information (i.e. gender, age, educational fields, prior business experience and family business background) were placed in the last section of the questionnaire. All of the items are listed in Table 2.

In this study, a strict three-phase process of developing the questionnaire was employed to avoid common method bias. First, two native English and two native Vietnamese experts were invited to facilitate translation between the two languages, and they engaged in discussion and comparison before releasing a final version of the questionnaire together. Second, the items of the borrowed scales were randomly ordered in order to minimize common method bias. A concise explanation of the concepts was also given to ensure that the participants all had the same understanding of them. Finally, a pilot sample was conducted with 98 respondents in order to test the validity and reliability of the questionnaire before developing the final version.

The normality of the constructs and items was examined through the skewness and kurtosis values (as shown in Table 2). The results showed that the skewness and kurtosis of the items related to four constructs–EE, FOF, ESE and EI–were within the acceptable range, with all skewness values less than 3 and all kurtosis values less than 8, illustrating the satisfactory normality of the constructs (Hu and Bentler, 1999; Hair et al., 2020).

Cronbach's alpha and CFA were then employed to examine the consistent reliability and discriminant validity of the constructs. The results of the measurement model reported excellent level of fit indices (see Figure 2): χ²(582) = 2,210.022; χ²/df = 3.797; p < 0.001; GFI = 0.938 > 0.9; AGFI = 0.925 > 0.8; CFI = 0.978 > 0.9; TLI = 0.975 > 0.9; NFI = 0.970 > 0.9 and RMSEA = 0.038 < 0.05 (Hair et al., 2020). The t-tests of all the items were significant at the 0.001 level, while the standardized regression weights (factor loadings) of all the items were much higher than 0.5 (Hair et al., 2020). Table 3 displays the correlation matrix and assesses the constructs' reliability and validity. The values of Cronbach's alpha, average variance extracted (AVE), composite reliability (CR) and maximal reliability (MaxR(H)) for the constructs surpassed the respective threshold values of 0.63, 0.5, 0.7 and 0.8 (Henseler et al., 2014). Furthermore, the maximum shared variance (MSV) of all the constructs remained below their corresponding AVE (Cheung et al., 2023), providing confirmation of the reliability and validity of the constructs within our conceptual model.

In our study, we took a comprehensive approach to address the common method bias issue. Firstly, a three-stage approach was implemented to control this bias. Additionally, we utilized specific statistical methods to verify the absence of common method bias. Specifically, Harman's single-factor method with an unrotated solution was applied. The results of the analysis showed that when all the items were forced to load on a single common factor, only 41.39% of the variance could be explained. This value falls below the established cut-off threshold of 50%, suggesting that a single common factor may not adequately capture the variability present in the data (Fuller et al., 2016). This result indicates that common method bias was not a significant concern in our research. Furthermore, we conducted a single-factor CFA in which all the items of the four constructs were constrained to load on a single common factor. The results of this analysis indicated a very poor fit of indices, with the following values: χ²(629) = 37,876.172; GFI = 0.315; AGFI = 0.234; CFI = 0.494; TLI = 0.464; NFI = 0.490 and RMSEA = 0.177. For further investigation, a common latten factor was performed and then the standardized regression weights of the items were compared. The difference in the standardized regression weights was not above 0.2; common method variance was thus not identified as a major problem in this study (Fuller et al., 2016).

To test the formulated hypotheses, including the moderated mediation impacts, the PROCESS macro (model 4 and model 29)–bootstrapping with 95% confidence intervals–was utilized (Hayes, 2018). As illustrated in Table 4, the findings supported the first premise (H1) that ESE had a positive impact on EI (β_ESE-EI = 0.2958, p < 0.001, 95% CI [0.1786, 0.4129]). The results also showed that EE had positive impacts on ESE (β_EE-ESE = 0.6041, p < 0.001, 95% CI [0.4877,7205]) and EI (β_EE-EI = 0.6206, p < 0.001, 95% CI [0.4728, 0.7683]), which supported H2 and H3. To verify the mediation role of ESE, we used model 4 in the PROCESS macro. The findings revealed that ESE significantly mediated the EE–EI linkage (β_EE-ESE-EI = 0.1868, 95% CI [0.1597, 0.2135]), supporting H4.

The results indicated that FOF negatively moderated the effects of EE on ESE (β_FOF*EE-ESE = −0.0364, p < 0.01, 95% CI [−0.0619, −0.0108]) and EI (β_FOF*EE-EI = −0.0373, p < 0.001, 95% CI [−0.0587, −0.0158]). Figures 3 and 4 demonstrate the negative moderation effects of FOF. The line increases more sharply with a low level of FOF in both figures. Hayes (2018) stated that low and high FOF were identified based on 1.S.D. ± Mean. As presented in Table 4, the conditional effects of EE on ESE and EI were higher at the lower levels of FOF (β = 0.4902 and β = 0.4623, respectively, with all p-values <0.001) and significantly lower at the higher levels of FOF (β = 0.3914 and β = 0.3612, respectively, with all p-values <0.001). H5a and H5b were therefore supported.

Considering the moderation role of gender, our results revealed that gender insignificantly moderated the effect of EE on EI (p > 0.05). H6a was thus not supported. However, gender significantly moderated the impact of ESE on EI (β_{Gender*ESE-EI} = 0.0863, p < 0.05, 95% CI [0.0108, 0.1617]). The conditional effect of ESE on EI was higher among females (β_females = 0.4683, p < 0.001) and significantly lower among males (β_males = 0.3820, p < 0.001). Figure 5 illustrates that the line increased more sharply among females than among males. Thus, H6b was supported by the data.

There would be a significantly moderated mediation of FOF and gender if EE had an indirect impact on EI via EE, regarding the various levels of FOF and gender. The results showed that the index of moderated mediation was statistically significant (index = −0.0031, Boot SE = 0.0021, 95% CI [−0.0080, −0.0002]). The conditional indirect effects of EE on EI through ESE were statistically significant at various values of FOF (1.S.D. ± Mean) and with gender (1 = males, 2 = females). Interestingly, for any level of FOF (low, mean or high), the effect of EE on EI via ESE was always higher among females than among males. To specify, for both males and females, the conditional indirect impact of EE on EI through ESE was higher with a lower degree of FOF (β_females = 0.2295, 95% CI [0.1888, 0.2709] and β_males = 0.1873, 95% CI [0.1536, 0.2237], respectively) and significantly lower with a higher degree of FOF (β_females = 0.1833, 95% CI [0.1435, 0.2258] and β_males = 0.1495, 95% CI [0.1156, 0.1855], respectively). However, these indirect impacts were always higher among females than among males. More importantly, pairwise contrasts between conditional indirect effects (effect 1 minus effect 2) showed that for both females and males, the indirect effects with a low FOF were higher than the same effects with the mean FOF (Contrast_females = −0.0462; 95% CI [−0.0817, −0.0110], Contrast_males = −0.0189; 95% CI [−0.0333, −0.0043], respectively), and they were the same with a high FOF (Contrast_females = −0.0231; 95% CI [−0.0409, −0.0055], Contrast_males = −0.0377; 95% CI [-0.0665, −0.0086]). The effects with a mean FOF were also higher than the same effects with a high FOF (Contrast_females = −0.0231; 95% CI [−0.0409, −0.0055], Contrast_males = −0.0189; 95% CI [−0.0333, −0.0043]). This finding demonstrates that the conditional impacts of EE on EI via ESE at different levels of FOF were significantly different. This supported the hypothesis that the mediation effect of ESE on the EE–EI link was moderated by FOF and gender. H7 was therefore supported.

The SCCT has attracted interest from academicians wishing to explain why individuals select business venture creation as a career choice (Meoli et al., 2020; Cui and Bell, 2022; Uysal et al., 2022). However, almost all the previous research has only investigated the antecedents of EI (Pérez-López et al., 2019; Cui and Bell, 2022) and researchers have neglected the negative effects of psychological factors on the relationship between antecedents and EI. Moreover, our extant understanding of how gender can significantly moderate these relations is scant (Ramadani et al., 2022). In this research, we thus adopted a moderated mediation model to investigate the moderation impacts of FOF and gender on the direct and indirect relationships between EE, ESE and EI. The results of the hypothesis testing are summarized in Figure 6.

Our results revealed that EE significantly and positively contributed to the increase in ESE and EI. These findings were in line with those of some prior studies (Duong, 2021; Maheshwari and Kha, 2022) that illustrated the importance of EE in fostering individuals' entrepreneurial activities, especially in youths (Nguyen et al., 2019). Indeed, Yeh et al. (2021) reported that EE not only significantly increases ESE but also fosters entrepreneurs' Internet entrepreneurial performance. Furthermore, Saptono et al. (2021) argued that under the effect of an outdoor learning environment, ESE is inspired by EE. Moreover, ESE was found to be significantly and positively associated with EI. This finding is consistent with previous studies (Elnadi and Gheith, 2021), including research conducted in the context of Vietnam (Maheshwari and Kha, 2022). The SCCT suggested the mediation role of self-efficacy in transforming the impacts of environmental and personal determinants into career goals/intentions (Lent and Brown, 2019). For example, adopting the SCCT, Dheer and Castrogiovanni (2023) stressed the mediating role of ESE in transferring the effects of cognitive adaptability to EI, whereas Nwosu et al. (2022) argued that ESE can mediate the relationship between placement learning in business firms and EI. In this study, our findings demonstrated the mediation impact of ESE on the EE–EI link, suggesting that when Vietnamese master's students' received EE, they believed strongly in their ability to create and run their own business ventures, which led to an increase in their intention to become entrepreneurs. This also means that EE first affects ESE, which in turn positively influences EI.

The findings illustrated that FOF negatively moderated the impacts of EE on both ESE and EI. In line with the SCCT, prior research has suggested that it is crucial to examine how supportive and hindering factors moderate the cognitive processes involved in making career choices (Meoli et al., 2020; Lent and Brown, 2019). However, there has been limited investigation into the adverse moderating influence of psychological obstacles on the entrepreneurial process (Hunter et al., 2020). FOF, which is considered to be a main psychological barrier in the entrepreneurial process, can help to explain why many individuals have a low level of belief in their entrepreneurial capacity and ability and hesitate to become entrepreneurs after they have participated in EE programs (Maheshwari and Kha, 2022; Kollmann et al., 2017). The findings of our study therefore provide a significant contribution to the entrepreneurship literature, confirming that the positive relationships between EE, ESE and EI become weaker if FOF is high. In other words, many students, despite receiving EE, had a low level of confidence in their entrepreneurial abilities and did not intend to engage in entrepreneurial activities because their FOF was high. More importantly, our results also revealed that the influence of ESE on EI was significantly higher among females than among males, suggesting that when females believe in their entrepreneurial capacity and ability, they have a higher intention to become entrepreneurs than males.

A moderated mediation analysis illustrated that the mediation impact of FOF on the relationship between EE and EI was weakened when the level of FOF was high, yet this mediation relationship was stronger among females than among males. This means that the indirect effect of EE on EI was significantly moderated by FOF and gender. This finding demonstrates that EE plays a more important role among women who have a low FOF (Guelich, 2022) than among men (Ojong et al., 2021). According to Ramadani et al. (2022), male and female students who take entrepreneurship courses and engage in related events such as seminars, expos and business practices, are united in their aspiration to turn their business ideas into new enterprises. This means that students are motivated to establish new businesses regardless of their gender and without facing any discriminatory practices. Vamvaka et al. (2020) suggest that male and female entrepreneurs are driven by a desire for success and unique experiences in their entrepreneurial endeavors. This motivation helps them to develop the skills to overcome external obstacles such as gender-based discrimination and cultural gender norms from the beginning. Consequently, it is important to offer equal access to entrepreneurship education opportunities to both genders. In the context of Vietnam, Duong and Vu (2023) reported that women face several more challenges than men when participating in entrepreneurial activities. The gendered nature of entrepreneurship is a significant factor, as scholars have pointed out that social beliefs influence gender-specific behaviors in various cultures. Indeed, in many cultures, women are expected to prioritize their family responsibilities over their professional goals, making it difficult for them to dedicate the necessary time and effort to entrepreneurial activities (Chatterjee and Ramu, 2018). Additionally, institutional barriers such as lack of access to finance, networks and training programs can further impede women's ability to engage in entrepreneurial activities (St-Jean and Jacquemin, 2022). These factors can contribute to the gender gap in entrepreneurship and highlight the importance of addressing gender-specific challenges in entrepreneurship education programs. Moreover, we found that individuals who receive EE do not necessarily have high self-efficacy or a strong intention to become entrepreneurs, especially if they are males and their FOF is high. This is in line with earlier studies that reported that entrepreneurship is characterized by high risks and uncertainty; therefore, some individuals hesitate to consider new venture creation as a career choice or to implement their obtained knowledge and skills in business when they fear entrepreneurial failure (Duong, 2022).

By addressing the above-determined research gaps, our theoretical contributions constitute significant progress in the extant entrepreneurship literature as follows. First, built on the SCCT, our research makes a critical contribution to entrepreneurship research by examining the role of EE in the development of ESE and EI. It also explains the underlying mediation mechanism of ESE in the EE–EI linkage. This also provides an important prerequisite for adopting the SCCT in the entrepreneurship context in order to determine whether the process of entrepreneurial career choice might be inspired or hindered by psychological factors (Tsai et al., 2016b). Second, while little attention has been paid to the negative catalytic effect of the FOF or to the moderation role of gender in the entrepreneurial process, our study enhances our understanding. It does this by firstly employing a moderated mediation model to confirm that the indirect correlation between EE and EI via ESE becomes stronger when the level of FOF is lower. More importantly, the findings of our study also narrow the gender gap in the entrepreneurship research by revealing that the mediation impact of ESE on the EE–EI link at low, medium or high levels of FOF is always stronger among females than among males.

The results of this research offer some practical and managerial insights for policymakers, educators and practitioners. Firstly, our findings could provide recommendations for policymakers. Instead of a range of strategies, such as inspiring cultural values (Shiri et al., 2017), enhancing contextual influences (Meoli et al., 2020) and employing human capital and role models (Van Trang et al., 2019), we recommend that educators should pay attention to the promotion of EE programs. This strategy will not only foster learners' intentions, but by equipping learners with essential knowledge and skills about entrepreneurship, it will also help them to have a greater self-confidence in their ability to engage in entrepreneurial activities. Nevertheless, the results also indicate that EE should be incorporated into experiential learning, real-life business cases and extracurricular activities (Nguyen et al., 2019). Students could reap a host of benefits from the prioritization of EE as a strategy. Apart from honing their entrepreneurial aptitude and self-efficacy, it could also cultivate their career interests and intentions, thanks to the valuable knowledge, skills and exposure they acquire in entrepreneurship (Uddin et al., 2022). Furthermore, EE can also heighten students' ESE, or their belief in their capacity to accomplish entrepreneurial tasks (Qudsia Yousaf et al., 2022). By means of hands-on learning, practical exercises and exposure to authentic entrepreneurial problems, students can cultivate self-assurance in their entrepreneurial skills, which can have a favorable effect on their entrepreneurial attitudes and conduct (Bae et al., 2014). Secondly, policymakers should employ appropriate strategies to facilitate the start-up process and to help individuals overcome FOF, for example, by enhancing institutional environments and entrepreneurial ecosystems (Soluk et al., 2021), providing financial funding for ideal business plans (Duong, 2022) or supporting new enterprises (Maheshwari and Kha, 2022). Finally, policymakers and educators should have a priority strategy for female entrepreneurs and/or female students. For example, they could provide females with free EE and training programs in order to increase their entrepreneurial capacity and self-efficacy and to encourage them to consider entrepreneurship as a career choice. Through entrepreneurship, women can enhance their social status and construct a new social identity (Ojong et al., 2021).

There are some limitations to this study that could offer opportunities for future research. First, the results cannot be universally applied to demonstrate the ideal function of EE in promoting learners' entrepreneurial actions since the data were exclusively gathered in Vietnam, although efforts were made to collect data from diverse universities and institutions in different areas. To offer a more widely applicable understanding, further research should broaden the research setting. Second, this research only investigated the antecedents of EI, but not actual start-up behavior; further studies should therefore expand the research model and adopt a longitudinal approach to bridge the intention-behavior gap in entrepreneurship. Third, the scale of FOF developed by Cacciotti et al. (2020) has different constructs (i.e. potential to idea, opportunity cost and ventures' capacity to execute). Nevertheless, in the context of the emerging economy of Vietnam, all the items of the subscales were loaded on one common factor in the exploratory factor analysis. Additionally, the CFA analysis confirmed the reliability and validity of this construct when seven subscales are loaded on one common scale. Further studies could divide this construct into subscales and test the reliability and validity of this construct in other research contexts. Finally, another limitation is that our research's cross-sectional design solely permits examination of the relationships between EE, ESE and EI under the moderation effects of FOF and gender at a specific point in time. However, this approach presents challenges concerning the causal relationship between these variables, and it does not provide a means of tracking the temporal progression of transferring the outcome of EE on ESE and EI. Thus, future studies could utilize longitudinal survey data to obtain more in-depth insights into the role of EE on ESE and EI.

This research is funded by National Economics University, Hanoi, Vietnam.