The Effect of Entrepreneurship Education on Entrepreneurial Intention: Mediation of Entrepreneurial Self-Efficacy and Moderating Model of Psychological Capital

Abstract

Based on planned behavior theory (TPB), this study aims to explore the direct or indirect impacts of entrepreneurship education on entrepreneurial intentions through entrepreneurial self-efficacy and explore the moderating role of psychological capital. Sample data were collected by sending online electronic questionnaires to university students in some universities in Guangxi. A structural equation model was used to test the 757 valid sample data. The results showed that: (1) college students of different genders and those with or without family business experience have significant differences in terms of their entrepreneurial intentions; (2) entrepreneurship education has a significant positive impact on entrepreneurial intentions; (3) entrepreneurial self-efficacy plays a complete mediating role; and (4) higher psychological capital can positively regulate the impact of entrepreneurial self-efficacy on entrepreneurial intention. The findings help explain the need for entrepreneurship education. In order to increase students’ participation in entrepreneurship education courses, different innovative technology-based curricula and educational methods can be used at higher educational levels. In addition, this study constructs a mediation and moderation model influencing entrepreneurial intention based on TPB, which further tests and enriches the research perspective of this theory from the perspective of positive psychology.

1. Introduction

There is a growing belief that entrepreneurial activity is an act that should be encouraged and supported, as it not only boosts the overall economy of the country [1,2], but also makes an important contribution to the economic development of a region [3]. In addition to this, entrepreneurial activities can also create more jobs for society [4]. Therefore, entrepreneurship is being studied all over the world, especially in Western countries with developed market economies and entrepreneurial ecosystems [5].

Entrepreneurs are an indispensable factor in undertaking entrepreneurial activities, and younger generations are undoubtedly an important source of entrepreneurs [6]. College students, as an important part of the youth group, have the characteristics of possessing a strong independent learning ability and easily accepting new things, and thus they show higher entrepreneurial potential and are more likely to participate in entrepreneurial activities [7]. China is a developing country, and there is still a lot of room for economic improvement. In addition, approximately 10 million graduates in China need to enter employment every year [8]. Improving the entrepreneurial behavior of college students can not only effectively enhance China’s economic development, but it can also alleviate the employment pressure of universities [9]. Unfortunately, the current entrepreneurship rate of Chinese college students is low, with 1.6% of undergraduate graduates and 3.4% of higher vocational graduates being self-employed in 2019, far lower than the 20–30% doing so in Europe and the United States [10]. Therefore, it is necessary to explore the issue of promoting entrepreneurship among university students.

Entrepreneurial behavior is formed from entrepreneurial intention, which is a long-term and challenging process [11]. The formation of entrepreneurial intent (EI) is an important step in becoming an entrepreneur and starting and developing new ventures or businesses [12]. It is the best predictor translating into entrepreneurial behavior, without which any further entrepreneurial steps would not exist [13,14]. Some scholars define EI as the willingness or inclination of an individual to become an entrepreneur in the near future [15]. In the field of entrepreneurship research, a large body of literature shows that EI is regarded as a decisive factor in engaging in entrepreneurial behavior [16,17,18,19]. Therefore, exploring the factors that trigger individual entrepreneurial intentions is a key topic in academia, and many factors have already been verified [20,21].

Entrepreneurship education (EE) is one of the important influencing factors of EI [22]. In fact, EE has become a policy tool to accelerate entrepreneurship and increase individuals’ willingness to start a business [23]. However, in previous studies, no further consistent conclusions have been reached [24]. Although most studies have found a positive correlation between EE and EI [25,26], there are also some studies that have come to different conclusions. For example, Oosterbeek et al. (2010) found that the effect of EE on EI was negative [27], while Iwu et al. (2021) found that there was no significant correlation between the two [28]. The reasons for the diversity of outcomes between EE and EI may be as follows: first, it may be due to different sample sizes or a lack of science and standardization when analyzing data [29]; second, there may be limitations to the measurement tool, i.e., the evaluation tool is only applicable to samples in a certain place and is not generalizable [30]; third, entrepreneurship education has not been standardized at present, and there are certain differences in teaching methods and training in different regions, which in turn leads to differences in entrepreneurial intentions [31,32]; fourth, entrepreneurship education has certain differences in different areas of knowledge and therefore shows different impacts [33,34].

EE encompasses a range of educational approaches designed to motivate, nurture, and promote students’ entrepreneurial skills and attitudes [35]. From an educational perspective, providing learners with in-depth knowledge, motivation, and skills through learning plans, learning experiences, and methodologies to enhance entrepreneurial intent success in various contexts, thereby enhancing the entrepreneurial behavior of college students [36,37,38]. Therefore, a key factor to consider when discussing EI, especially among students, is EE [38]. Although EE plays an important role in the entrepreneurial process of college students, and there is extensive empirical research evidence of it, through the above discussion, it can be found that the impact of EE on EI is still different, which also shows that the promotion of EE lacks a certain degree of replicability in different regions and different groups. Therefore, it is necessary to study the impact of EE on EI according to local conditions.

Entrepreneurial self-efficacy (ESE) is the most important personal factor that has a significant impact on EI [21,39,40]. ESE is “the strength of a person to believe that he or she can successfully complete various roles and tasks of entrepreneurship” [41], and is a measure of the degree of belief in a person’s ability to implement entrepreneurship on their own [42].

In entrepreneurial activities, planned behavior theory (TPB) is an important theoretical basis to explain entrepreneurial behavior [43]. In particular, empirical research on the EI of college students has been widely and effectively applied in TPB [44,45]. For example, Jena (2020) discusses the impact of EE on EI in India based on TPB [46]; Elnadi and Gheith (2021) explore the impact of Saudi Arabia’s entrepreneurial ecosystem on college students’ EI based on TPB and introduce gender as a moderating variable [11]; Anjum et al. (2021), based on the theory of planned behavior, takes Pakistani college students as the research group and the paper constructs a model of the relationship between perceived creativity position and entrepreneurial intent, and proposes the perception of university support moderation mechanism [47]. Although TPB-based entrepreneurship research is very rich, no conceptual framework containing positive psychology has been found. Starting a business is considered a difficult and challenging act, and Tang (2020) argues that mental capital has a better mechanism of action in both behavioral intentions and the positive evaluation of negative events [48]. In other words, psychological capital (psycap) enables one to show a more positive and optimistic attitude in the process of generating relevant behavioral intentions, which in turn promotes the possibility of intention-formation [49]. Therefore, the introduction of the concept of psycap into TPB has a practical significance, and this is still a gap in the field of entrepreneurship.

Based on the above arguments, this study aims to explore methods of enhancing college students’ entrepreneurial intentions. The research will take the influence of EE on EI as the main path and will take this as the main clue to construct a model that can effectively enhance college students’ entrepreneurial intentions. Since there is no consensus on the influence of EE on EI, the results of this study can only play a supplementary role in the field of entrepreneurship research. In addition, the previous research on EI based on the TPB model was mainly carried out on the basis of the original TPB model concept. In view of the fact that entrepreneurship is a challenging behavior, this study introduces the variable of psycap, with positive psychological characteristics, to form a model that contains both mediation and regulation, and applies psycap as a moderating variable in the TPB model, which is still a blank in the field of entrepreneurship research. In terms of theoretical significance, this study helps to clarify the key factors affecting college students’ EI. In terms of practical significance, the research results can provide a new reference direction for higher education to enhance college students’ entrepreneurial intentions. Based on the above discussion, this study will ask the following research questions:

Question 1: Does EE affect EI?

Question 2: Does EE affect ESE?

Question 3: Can ESE play a mediating role between EE and EI?

Question 4: Can psycap regulate the pathway effect of ESE and EI?

The structure of this paper is as follows: the next section will introduce the theoretical background, research model, and related assumptions of the study variables; research methods and data analysis are then described in Section 3 and Section 4; Section 5 and Section 6 are the discussion and conclusions of the study; and finally, its limitations and recommendations for future research are highlighted.

2. Theory and Research Hypothesis

Entrepreneurship is considered to be a planned act [41]. The planned behavior theory (TPB) states that intent is the most immediate premise of a relevant behavior, and the stronger the intent, the greater the likelihood of translating intent into action [50]. According to the TPB, an entrepreneur’s EI is influenced by entrepreneurial attitudes, perceived behavioral controls, and subjective norms [43]. However, EI is seen as a complex, multidimensional and interdisciplinary phenomenon [44]. Therefore, it is necessary to explore other dimensions that can extend the TPB base model, such as gender [51,52] and psychological characteristics [53], among others.

2.1. The Relationship between EE and EI

The TPB points out that people’s attitudes of positive or negative responses to behavior are highly predictive of behavioral intention [54]. EE in higher education focuses on the development of an entrepreneurial mindset, and when EE is effectively implemented, college students’ perception of entrepreneurship can be improved, which in turn can change students’ attitudes towards entrepreneurship [55]. Therefore, based on the TPB, EE has a positive impact on EI.

EE is an important method of improving the comprehensive quality of individual entrepreneurship. It can improve the ability of entrepreneurs, and it can also impart knowledge regarding how to create a new business [28], and stimulate students’ desire to choose self-employment after graduation [56]. Many universities are igniting the flame of entrepreneurial intent among students by improving their curricula and teaching methods; for example, by implementing extracurricular entrepreneurial activities [57]. It turns out that students who receive EE have a higher willingness to start a business [58]. Therefore, EE is crucial in the first stage of the entrepreneurial process [59]. In addition, studies have found a positive correlation between EE and EI, and empirical studies from scholars in many different countries have demonstrated that EE has a positive impact on EI [28,60,61,62]. Therefore, this study proposes the following assumption:

H1. 

EE can positively and significantly influence EI.

2.2. The Relationship between EE, ESE, and EI

The TPB states that perceived behavioral control (PBC) is the direct determinant of intention, and conceptually there is no difference between PBC and self-efficacy, as both refer to one’s belief that one is capable of completing a particular behavior [50]. However, the two factors are somewhat different in terms of their metrics. A study on the concept of self-efficacy by Bandura (1977) typically defines a range of potential barriers to behavioral performance, and participants were asked to rate their likelihood of overcoming each obstacle [63]. To assess PBC, participants are typically asked to rate the extent to which they are capable of performing the behavior, to what extent the behavior is under their control, etc. [50]. This study focuses on the ability of individuals to solve problems in the entrepreneurial process, so the concept of self-efficacy is adopted. In addition, attitudes, subjective norms, and self-efficacy are conceptually independent predictors of intention. Empirically, however, they can be freely correlated with each other [50], and, in general, self-efficacy can act as a mediator in terms of attitude and intention.

Bandura (1977) identifies vicarious experiences, verbal persuasion, physiological states, and role models as the four main sources of information for self-efficacy [63]. Entrepreneurship education can be an effective input to these four sources of information; for example, you can set an example through the entrepreneurial stories of alumni and successful entrepreneurs; encouragement between teachers and classmates also increases self-efficacy [62]. Studies have also shown that ESE can be cultivated through EE [64]. The main reason is that EE can improve the knowledge and skills required by individuals in the field of entrepreneurship, which will give individuals a positive evaluation of entrepreneurial activities, leading to an increase in their self-confidence in entrepreneurship and improving their self-efficacy [65]. Moreover, in the field of entrepreneurial research, ESE is considered to be one of the strongest predictors of EI [66]. ESE has a positive effect on EI, and a high level of ESE can develop a high level of EI [67]. In addition, a large number of empirical studies have shown that a mediating effect of ESE exists [68]. For example, the empirical results of Souitaris et al. (2007) showed that the level of entrepreneurial attitude and entrepreneurial willingness of science and engineering students who received entrepreneurship education was higher than before acceptance and that entrepreneurial self-efficacy played a mediating role [69]. Other studies have also pointed out that the mediating role of ESE enhances the direct impact of antecedent variables on EI [66,70]. Based on the above analysis, this study speculates that EE can indirectly affect EI by improving individuals’ ESE.

H2. 

EE can positively and significantly affect ESE.

H3. 

ESE plays a mediating effect between EE and EI.

2.3. The Moderating Role of Psycap

Based on trait theory, psycap is defined as the personality characteristics of individuals with relative stability and persistence [71], which involve a combination of the positive psychological states of individuals, mainly including the characteristics of self-confidence, optimism, hope, and resilience [39]. Optimistic people are hopeful and confident about the successful outcomes of a future event and can also be considered as an individual’s confidence in their ability to improve a situation [72]. If people have both the will to achieve a certain goal and a plan regarding how to achieve it, then they will be full of hope, and hope can also be understood as a positive psychological expectation that has a certain motivational effect [73]. Resilience refers to one’s ability to bounce back and recover quickly in the face of difficulties, and resilient individuals are able to develop their coping skills and thrive in the face of obstacles to achieve their desired goals [74].

Individuals often assess whether they have the skills needed to start a business before deciding to start a business. Wardana et al. (2020) argues that while individuals who receive EE may enhance their ESE, it still needs to be strengthened to translate ESE into EI [75]. The reason for this is that entrepreneurship is a challenging and risky behavior that comes with many risks and uncertainties [76], and introducing a positive factor that strengthens the transformation of entrepreneurial self-efficacy into entrepreneurial intention will show better results [77]. Psycap is described as a positive interpretation of events that is based on effort and persistence to stimulate prosperity and success [78]. Psycap, as a combination of positive mental states, can keep entrepreneurs optimistic in the face of adversity and setbacks and makes them believe that they are capable and resilient [48], making them likely to continue to develop EI even in difficult circumstances [77]. Therefore, this study speculates that the variable of psycap has the ability to regulate the path of ESE and EI. Therefore, the following hypothesis is proposed:

H4. 

Psychological capital can play a moderating role in the influence path of ESE on EI.

In summary, based on the TPB and hypothetical inference, a model containing mediating and regulating variables is designed and is shown in Figure 1.

3. Methodology

3.1. Participants and Procedure

Collecting sample information through electronic questionnaires has become a regular method of scientific research in China. However, this method is prone to the phenomenon that the test is not filled in carefully. Therefore, this study adopted the convenient sampling method. Considering that class teachers, as direct managers of students, have a certain affinity and prestige among students, class teachers from some universities in Guangxi were selected to assist in collecting questionnaires. First, we trained the class teacher to inform the students of the purpose of the survey and the specifications for filling out the questionnaire. The class teacher then relayed the relevant information about the survey to the class and sent out the electronic questionnaire generated by the Questionnaire Star application to encourage students to fill in the answers but did not force the students to fill in the answers. The questionnaire was completed from July 5 to July 10, 2022.

In order to prevent the phenomenon of common method deviation in the process of filling out students’ answers, this study used the “interview consultation concealment method” and the “psychological isolation method” to test the sample; that is, the anonymous survey was used, and the standard subscales in the questionnaire were independent so that the respondents had less personal bias tendency [79]. Participation was voluntary, and the data were anonymized.

In the survey, participants provided background information, such as their gender, campus entrepreneurship competition experience, family members’ business experience, etc., and scored the measurement items of EE, ESE, EI, and psycap. Of the 812 students who completed the questionnaire, 55 were deleted due to a short answering time, and 757 questionnaires were retained, with an effective rate of 93%. Among them, 324 (43%) were boys and 433 (57%) were girls; 266 (35%) had experience in entrepreneurship competitions and 491 (65%) did not; 288 (38%) had family business experience and 469 (62%) did not.

3.2. Instruments

The questionnaire contained four scales, EE, ESE, psycap, and EI, all of which used the five-point Likert scale scoring method.

EE: This study was based on a scale developed by Shi (2018) [80]. The content of the scale was reviewed by three academics in the field of entrepreneurship, focusing on the content of the questions, as well as the acceptability and completeness of the content of each question. In the end, four questions were selected, for example, “University education provided me with a systematic entrepreneurship theory course”. The total reliability of the original scale was 0.794.

ESE: Using the entrepreneurial self-efficacy scale developed by Tang (2009), there are 22 questions, such as “I often actively communicate with others”. A higher score represents a higher entrepreneurial self-efficacy. The reliability of each dimension was 0.86, 0.71, 0.76, 0.76, and 0.73, respectively [81].

Psycap: Zhang et al. (2010) created the positive psychological capital questionnaire (PPQ) with an overall reliability of 0.9 on the original scale [82]. It contains 26 questions, such as “When I encounter setbacks, I can recover quickly”.

EI: Using the entrepreneurial intention survey designed and developed by Liñán and Chen (2009), the scale contains a total of six questions, for example, “I’m ready to be an entrepreneur” [83]. The scale has a Cronbach’s α of 0.94.

3.3. Data Analysis

SPSS 22.0 was used to evaluate the reliability of the scale, and the independent sample t-test and common method variance test were performed on the samples. AMOSS 22.0 was used to test the convergence validity, discrimination validity, and model fit of the samples. Structural equation modeling (SEM) is commonly used to measure the effect between multivariate and model structural relationships, i.e., to assess the validity of a theory or hypothesis through the use of data [84]. Over the past few decades, SEM as multivariate relationship modeling has become an important research tool in psychology and pedagogy [84,85]. A major advantage of SEM over other multivariate methods is that it reduces the impact of measurement errors on these regression estimates, as these errors have already been “partially eliminated” from the factor model. Another advantage is that the model’s fit to the data can be tested and compared to the fit of competing models, allowing researchers to see which theoretical concepts are supported empirically [86]. Therefore, this study will create a relationship model of EE, ESE, and EI to test the direct effect and mediating effect of variables. Then, on the basis of this model, the psycap high and low group models on the ESE to EI path were set, and the moderating effect of psycap was tested by comparing the data of the two sets of models.

3.4. Data Preparation

3.4.1. Item Analysis

In order to ensure the measurement quality of subsequent results, this study first tested all the questions of the scale through project analysis, and the items that did not meet the standards would be deleted. The study will use two indicators: critical ratio (CR) and question correlation with the overall score. Yang and Yang (2010) believe that if a CR greater than 3 and reaches a significant level, the question item should be retained; otherwise, one should consider deleting or modifying the question item [87]. In addition, Wu (2018) pointed out that it is recommended to delete the question item when the correlation coefficient with the total score does not reach more than 0.4 [88]. After the project analysis of each variable, the 4 questions of the EE scale, 22 questions of the ESE scale and 6 questions of the EI scale all met the criteria and were retained. In the psycap scale, 5 questions did not meet the standard value and were deleted, because the question item did not reach the threshold of 0.4 related to the total score. The deleted psycap scale contains 21 items.

3.4.2. Exploratory Factor Analysis (EFA)

EFA is a multivariate statistical method that has become an essential tool for developing and validating psychological theories and measurements [89]. Lloret et al. (2017) argues that factor analysis is not available until the KMO value of the scale ≥0.70 [90]. Wu (2018) believes that the factor loading of each EFA item should be greater than 0.5, and the total explanatory variation >50% can indicate that the scale has good validity [88]. The EFA results of each variable in this study are as follows: the KMO value of the EE scale is 0.795 (p = 0.000), the factor loading of each item is between 0.807~0.835, and the explanatory variation is 66.519%. The KMO value of the ESE scale was 0.957 (p = 0.000), the factor loading of each item was between 0.632~0.799, and the explanatory variation was 64.592%. The KMO value of the psycap scale was 0.964 (p = 0.000), the factor loading of each item was between 0.573~0.806, and the explanatory variation was 59.146%. The KMO value of the EI scale was 0.909 (p = 0.000), the factor loading of each item was between 0.862~0.907, and the explanatory variation was 78.596%.

3.5. Common Method Variance (CMV) Test

As the sample data were self-reported, the samples were evaluated by HARMAN univariate testing to determine the quality of the data collection. The test data showed that the Kaiser–Meyer–Olkin (KMO) value was (0.945 > 0.8), and Bartlett’s test of sphericity was significant (p < 0.001). A total of nine non-rotational factors were extracted, the first of which had an explanatory force of 39.290%, which is less than the 50% critical value, i.e., the CMV problem was not serious [91].

3.6. Difference Analysis

Whether there was a significant difference in college students’ EI under different background variables was examined via the independent sample t-test. The results (

Table 1. Summary of difference analysis results.

Background Variable		Mean	SD	t	p
Gender	Male	3.289	0.978	8.478	0.000
	Female	2.702	0.912
Family business experience	Yes	3.147	0.968	4.293	0.000
	No	2.834	0.976

) showed that there were significant differences in the EI of college students of different genders (t = 8.478, p = 0.000), and the EI of male students were significantly higher than those of female students. In addition, the entrepreneurial experience of family members also made a significant difference to the EI of college students (t = 4.293, p = 0.000), and the EI of college students with family members with business experience was significantly higher than that of college students with no business experience.

4. Results

4.1. Normality Test

The estimation method for selecting the structural equation model was based on data allocation, with the maximum likelihood estimate (MLE) as the main criterion if the sample data were normal allocations, and the asymptotic distribution freedom method (ADF) if the data allocation were non-allocations normal [92]. Studies have shown that when both the skewness and kurtosis absolute values of the observed variable meet the condition of less than 2, the observed variable can be considered to be normal [93]. As shown in

Table 2. Summary of reliability and convergence validity.

	Mean	SD	Skewness	Kurtosis	Cronbach’s α	CR	AVE
Threshold	-	-	−2–2	−2–2	>0.70	>0.60	>0.50
EE	3.066	0.938	−0.481–0.055	−0.958–−0.376	0.832	0.833	0.555
ESE	3.453	0.658	−0.770–−0.105	−0.412–0.814	0.954	0.970	0.601
EI	2.953	0.985	−0.749–−0.036	−0.596–−0.486	0.945	0.945	0.743
Psycap	3.558	0.630	−0.481–−0.05	−0.519–0.750	0.951	0.965	0.572

, all indicators of the observed variables for each variable are between plus or minus 2, so the model estimation method used MLE to measure the data.

4.2. External Model Evaluation

Studies have pointed out that the validation of structural models is meaningless if the reliability and validity of the external models is not ideal [94]. External model evaluation is usually measured in terms of three aspects: internal consistency and reliability, convergence validity, and discriminating validity [95]. The results of the evaluation are shown in Table 2.

Hair et al. (2010) argues that when the Cronbach’s α value of a variable reaches 0.7 or above, it is considered to have good reliability [96]. The Cronbach’s α coefficients for each variable were EE: 0.832; ESE: 0.954; EI: 0.945; and psycap: 0.951. All scales met the threshold criteria, indicating that the scales had good reliability.

Average variance extracted (AVE) and composite reliability (CR) are the two main indicators used to measure convergence validity. The reference index for AVE must be greater than 0.5, and the CR value must be greater than 0.6 [97,98]. The AVE of each variable was between 0.555 and 0.743, and the CR value was between 0.833 and 0.970, so the EE, ESE, EI, and psycap scales all had good convergence validity.

Fornell and Larcker (1981) argued that when the AVE square root of each variable is greater than the number of correlation coefficients of each variable when accounting for more than 75% of the overall number of comparisons, the variables are considered to have good differential validity [99]. According to the data in

Table 3. Discriminant validity data.

	EE	ESE	Psycap	EI
EE	0.744
ESE	0.427 ***	0.775
Psycap	0.362 ***	0.775 ***	0.756
EI	0.310 ***	0.607 ***	0.452 ***	0.861

Note 1: *** p < 0.001; Note 2: The diagonal value is the square root of AVE.

, the validity of the distinction between variables is within acceptable limits.

4.3. Hypothesis Testing

4.3.1. Model Fit Analysis

The fit of the model was judged by 11 reference indexes, including GFI, AGFI, RMR, SRMR, NFI, NNFI, CFI, IFI, RFI, PNFI, PGFI, etc., among which the RMR and SRMR values should be less than 0.80, the PNFI and PGFI values should be greater than 0.5, and the remaining seven indexes should be greater than 0.80 [96,100]. As is shown in

Table 4. Summary of fits.

Threshold		Overall Sample	Psycap Low Grouping	Psycap High Grouping
GFI	>0.800	0.911	0.907	0.899
AGFI	>0.800	0.877	0.871	0.86
RMR	<0.080	0.046	0.039	0.054
SRMR	<0.080	0.040	0.050	0.052
NFI	>0.800	0.937	0.91	0.915
NNFI	>0.800	0.935	0.929	0.919
CFI	>0.800	0.946	0.941	0.933
RFI	>0.800	0.924	0.891	0.898
IFI	>0.800	0.946	0.941	0.933
PNFI	>0.500	0.776	0.754	0.758
PGFI	>0.500	0.660	0.657	0.652

, all indicators met the reference standard; that is, the overall sample model and the high and low grouping models of psychological capital show a good fit.

4.3.2. Analysis of Direct and Indirect Effects

We assessed the direct impact of EE on EI, as well as the mediating effect of ESE. The bootstrap method was repeatedly sampled 2000 times to calculate the 95% confidence interval to further verify the mediating effect of ESE. Cheung and Lau (2007) point out that if the 95% confidence interval of the indirect effect value does not contain “0”, it is significant; that is, there is an intermediary effect. At the same time, if the 95% trust interval of the direct effect contains “0”, it means that the direct effect is not significant [101].

The direct effect of EE on EI is shown in Model A in Figure 2 and

Table 5. Bootstrap verification.

Path		Estimate	95% Confidence Interval Bias-Corrected Percentile Method
			Lower Bounds	Upper Bounds
Indirect effect	EE→ESE→EI	0.315 ***	0.251	0.388
Direct effect	EE→EI	0.036 ns	−0.060	0.132
Total effect	EE→EI	0.351 ***	0.262	0.434

Notes: *** p < 0.001; ^ns = not significant.

: its direct effect (total effect) is β = 0.351 ***, p < 0.001. That is, EE can positively and significantly influence EI, and Hypothesis 1 is supported.

The mediating effect of ESE is shown in Model B in Figure 2 and Table 5: the indirect effect is β = 0.315 (p < 0.001), and the 95% confidence interval is 0.251–0.388. The interval does not contain “0”, and thus the mediation effect is significant. In addition, the direct effect of EE on EI in this model was β = 0.036 ^ns (p > 0.05), and the 95% confidence interval was (−0.060–0.132). It includes “0”, and thus the direct effect was not significant. Through model A, we know that the effect of EE on EI is significant; however, when ESE is introduced as a mediating variable, this direct effect decreases and becomes less significant. In addition, according to Cheung and Lau (2007), regarding the determination method of a mediation effect, it can be seen that ESE plays a complete mediating role in the path of EE and EI [101]. Hypothesis 2 is thus validated.

4.3.3. Modulating Effect

Firstly, the psychological capital is divided into a high-score group and a low-score group via the K-mean classification method of SPSS22.0, among which the low group consisted of 313 samples and the high group consisted of 444 samples. Then, the independent sample t-test was used to examine the differences between the two groups, and the results of the difference analysis showed p < 0.001; that is, the groups were valid [102].

The model is based on the original EE, ESE, EI model. On the basis of this model, the low-score group and high-score group paths of ESE on EI are set up, in which the low grouping path reads the low grouping sample of psychological capital, and the high-grouping path reads the high-grouping sample of psychological capital. It is assumed that the influence coefficient of ESE on EI is equal in the two groups to create the interference model. When comparing the degrees of freedom and chi-square values of the two models, if the difference between the chi-square values of the two is not significant, this means that there is no adjustment effect, and vice versa [102].

The test results are shown in

Table 6. Single-path identity coefficients (ESE→EI).

Model		χ2	DF	Δχ2	ΔDF	p
Model 1	baseline model	613.908	174	4.101	1	0.043 *
Model 2	interference model	618.010	175

Note. * p < 0.05.

. The chi-square value of model 1 (basic model) is 613.908 (DF = 174), and the chi-square value of model 2 (interference model) is 618.010 (DF = 175). The difference between the two models is 1 degree of freedom, and the difference between the chi-square values is 4.101. The study shows that when the difference in degrees of freedom between the two models is 1, if the difference between the chi-square values is greater than 3.84, this means that the difference in chi-square values is significant; that is, the chi-square difference in this study is significant.

Since only the restriction “low grouping = high grouping” was added to model 2 in this study, the restriction (the assumption that the path coefficients of the two groups are equal) is not valid when the chi-square values of the two models differ significantly. That is, a high psycap and low psycap are not equal in the estimation of the path of EI in ESE, and so psychological capital has a moderating effect.

When the path coefficient sizes of the two groups of high psycap and low psycap are further compared, the results are shown in

Table 7. Interference path coefficient analysis table (ESE→EI).

Path	High Grouping of Psychological Capital			Low Grouping of Psychological Capital
	Estimates	t	p	Estimates	t	p
ESE→EI	0.635 ***	10.237	0.000	0.494 ***	6.603	0.000

Note. *** p < 0.001.

and Figure 3. On the ESE-to-EI path, the path coefficient is 0.635 (p < 0.001) in the high psycap group model and 0.494 (p < 0.001) in the low psycap group model. That is, the influence of ESE on EI on the low psycap group is lower than its influence on the high psycap group. Therefore, psychological capital plays a regulating role in the path of ESE’s influence on EI. That is, Hypothesis 3 is verified.

5. Discussion

5.1. Gender Differences in EI

The results showed that there were significant differences in the EI of college students of different genders, and male students showed stronger EI than female students. Similar to the findings of Hassan et al. (2020) and Pan and Lu (2022) [103,104], boys display more confidence than girls in the cognition of entrepreneurship. However, studies have also shown that there is no significant difference in entrepreneurial intentions between men and women in the West [105]. The reason for this phenomenon may be the cultural differences. Under the influence of Confucianism in China, women are more inclined to choose a stable job so that they can have time to take care of their families, while men spend more energy on their careers and engage in challenging work [106].

Therefore, in the guidance on entrepreneurial intentions, the feelings of female groups should be taken into account. Colleges and universities should establish a hierarchical and classified innovation and entrepreneurship education system to meet the demands of different subjects and stimulate entrepreneurial willingness in different subjects so that students of different genders can choose corresponding entrepreneurship courses to study according to their own needs. In this way, they can choose different entrepreneurial practice projects to participate in so that they can play to their strengths in specific situations.

5.2. Differences in College Students’ EI in Terms of Family Business Experience

Another result of the differential analysis showed that college students with family business experience showed higher EI. The conclusion of this study validates the view of Bloemen-Bekx et al. (2019) that family background is an important area for developing career intentions and that encouragement from parents with entrepreneurial experience of college students in the process of transitioning from school to society significantly improves young people’s entrepreneurial intentions [107]. Parents’ business experience will expose children to business knowledge earlier, and they can learn a wealth of practical business experience from their parents. In addition, parents who engage in corporate behavior will act as role models for college students. This impact is usually positive, and this continuous family atmosphere will gradually affect the child’s value orientation [108]. Therefore, for the process of entrepreneurship education, a cooperative mechanism between school and family should be considered.

5.3. EE and EI

The results show that EE can positively and significantly affect EI, which is consistent with the findings of Hoang et al. (2020), Shah et al. (2020) [23,109], and others. However, this result is not uniform, as we mentioned earlier, and there are some differences in the impact on entrepreneurial intentions due to the differences in entrepreneurship education delivery in different regions [31,32]. As far as entrepreneurship education in China is concerned, the main reasons for this phenomenon may be as follows: first: entrepreneurship education can integrate traditional classroom teaching and entrepreneurial practice, and stimulate students’ enthusiasm for participating in entrepreneurship education; second: teachers of entrepreneurship education have high professional qualities and can provide high-quality entrepreneurship education teaching for individuals; and third: colleges and universities provide support for students’ entrepreneurial theory teaching and practice, such as appropriately postponing graduation for students who participate in entrepreneurial practice, or introducing relevant policy resources to them, providing entrepreneurial experimental funds and providing preferential or even free venues and even tax incentives after establishing enterprises, etc., which may promote individual entrepreneurial behavior. EE helps to develop students’ entrepreneurial attitudes, abilities, and skills, as well as their ability to seek new entrepreneurial opportunities, thereby enhancing their willingness to start a business [110]. We suggest that colleges and universities adopt a combination of theoretical teaching and practice to provide entrepreneurship education for college students, among which theoretical teaching should mainly take the form of basic entrepreneurship courses taught in the classroom and inviting training institutions or entrepreneurs to give entrepreneurship lectures on campus, which has proven to be an effective means for students to master entrepreneurial skills [111]. In terms of practice, the characteristics of the university should provide students with an entrepreneurial practice platform. In addition, we should take the national college students’ innovation and entrepreneurship competition as an opportunity to encourage students to participate in discipline competitions and improve the comprehensive quality of college students’ problem-solving abilities, thinking abilities, etc.

5.4. Discussion on the Mediating Role of ESE

Studies have shown that ESE plays a complete mediating role in the influence paths of EE and EI. ESE is mostly presented as a partial mediating effect in the field of entrepreneurship research [23,112], and the results of this study enrich the empirical research on the mediating effects of ESE. The TPB states that favorable attitudes provide the motivation to engage in behavior but that intentions towards related behaviors are formed only when the control over one’s behavior is strong enough [50]. In other words, ultimately the decisive role is actually the student’s self-efficacy in the field of entrepreneurship. EE is an effective way to enhance ESE; the stronger the ESE, the more effectively college students can exert their innate entrepreneurial ability, enhance their entrepreneurial potential, stimulate entrepreneurial confidence and passion, and generate EI [113]. Therefore, by participating in activities such as related discipline competitions, college students can discover their advantages and characteristics, affirm their own progress, and fully feel the joy and sense of gain of harvest, improving their degree of subjective effort. Secondly, some young entrepreneurial stars can also be invited to hold lectures in colleges and universities, share their entrepreneurial journey, play the role of role models, and stimulate college students’ interest in learning and entrepreneurial motivation.

5.5. Discussion on the Moderating Role of Psychological Capital

The results show that psycap plays a regulating role in the influence path of ESE on EI; that is, a higher psycap can strengthen the influence of ESE on EI. This fully confirms that psycap is a psychological factor that can lead to positive behavior in individuals [114]. This finding is very rare in the field of entrepreneurial research. Entrepreneurship is a challenging process, the process of forming EI may encounter many obstacles and difficulties, and psycap has positive, optimistic, resilience, and other high-quality characteristics. Psycap can also always be full of hope when encountering difficulties and setbacks, is characteristic of not giving in, facing difficulties with a positive attitude, and believing that they can overcome difficulties, so they can maintain a long-term spirit of endeavor [78,115]. Positive psychology believes that there are two forces in the inner world of people: positive forces and negative forces. If positive forces are stimulated, nurtured, and strengthened, negative forces are suppressed or eliminated [116]. Therefore, college students with a high psycap can reduce the negative psychological burden of these adverse factors. In addition, individuals with a high psycap have the ability to mitigate the negative effects of adverse circumstances and accept challenges rather than react negatively, and people with high levels of psycap are more likely to embark on new adventures [117], which stems from their confidence and perseverance to explore different options and tackle challenges as they work to achieve challenging goals [118]. More importantly, psycap is measurable, developed, and managed, not a personal trait that is difficult to change [116]. Based on this, actively cultivating and developing psycap will help enhance the EI of college students.

6. Conclusions

EE provides students with a different way of looking at the world, even if they do not end up choosing to start their own business. The main purpose of this study is to explore the impact of EE on college students’ EI. The study examines the mediating role of ESE and the moderating role of psycap, and explores the differences in EI between gender and family members’ business experiences. All the assumptions in this study are significantly supported. The results show that EE has a significant positive impact on EI. ESE plays a complete mediating role between EE and EI; psycap can positively regulate the influence of ESE on EI. Different genders and whether family members have business experience will show significant differences in college students’ EI.

Entrepreneurial activity has long been considered a case of planned behavior. Therefore, it can be stimulated through educational programs that increase EI. Based on the conceptual framework of TPB, this study explores the impact of EE on EI, and the results show that EE has a high explanatory power on EI, and plays a very important role in improving EI. The findings are consistent not only with the perceptions of traditional psychology [43], but also in the field of entrepreneurship [23,109]. In addition, the results also prove that the TPB model proposed by Ajzen (1991) is suitable for explaining the entrepreneurial activities of individuals [43]. Therefore, this study provides evidence for the relationship between EE and EI in China.

Theoretical significance: Studies have shown that EE plays a crucial role in stimulating college students’ willingness to start a business. Although the previous model of entrepreneurial psychology has similar results, we must be clear that the development of entrepreneurship education in colleges and universities should be combined with the actual situation of each region, adapting measures to local conditions and teaching according to aptitude to achieve the expected results. This study takes college students from Chinese universities as a research sample, which has certain a reference significance for EE in Chinese higher education. In addition, this study constructs a mediating and moderating model influencing EI based on TPB, which further tests and supplements the TPB from the perspective of positive psychology. The results of this study provide an empirical evidence to support the existing theory and provide a valuable reference for subsequent research.

Solving employment problems through entrepreneurship and improving the country’s economic development are considered to be an important means. EI is an important precursor variable for generating entrepreneurial behavior, and understanding the mechanism of influence on EI can better improve the entrepreneurial behavior of college students. Therefore, from a practical point of view, the results of this study will help answer the need for EE in universities. The results also show that universities with different gender and family business backgrounds have significant differences in EI. In the process of carrying out education, attention should be paid to distinguishing the demands of college students of different genders, and stimulating the EI of college students of different genders. In addition, as a positive intervention variable, psycap can effectively regulate the entrepreneurial mindset of college students in the challenging and risky entrepreneurial field, so that they have a stronger sense of self-efficacy, and then increase the generation of EI.

In conclusion, the results confirm the positive impact of EE on EI, so it is necessary for universities to establish a curriculum for entrepreneurship education. Therefore, in order to increase students’ participation in entrepreneurship education courses, different courses and educational methods based on innovative technologies can be used at higher educational levels. However, EE should not only stop at the theoretical stage as providing students with a certain entrepreneurial platform during school can effectively enhance their practical abilities [119]. The reform of the curriculum and the establishment of the platform require a lot of funds and policies to support it, which also means that the government should provide practical support for EE in colleges and universities [120,121]. In addition, colleges and universities should not only pay attention to the improvement of college students’ entrepreneurial skills but should also cultivate a positive attitude as this is also an important part of the entrepreneurial field.

7. Research Limitations and Future Study

This study achieved its purpose according to the expected design and obtained satisfactory results, but there are still certain limitations in terms of the overall view, and several suggestions are put forward here for future research:

First, because the sample was collected in the form of a self-reported questionnaire, participants’ evaluation of the questions may be exaggerated or conservative due to excessive subjectivity. In the future, the objectivity of data sources can be increased through diversified information collection channels.

Secondly, this study only uses the cross-sectional method, which can only confirm the correlation between variables and cannot establish a causal relationship. We recommend the use of longitudinal research to further verify the transformation process from EI to entrepreneurial behavior in the future.

Finally, although this study confirms that there is a significant difference in college students’ entrepreneurial intentions in the context of gender and family business experience, the results of studies by Western scholars have shown mixed conclusions [82,83]. Therefore, the difference in entrepreneurial intention under different background variables may be different for Chinese and Western college students, which may be caused by different cultural backgrounds. Future research should explore the differences in entrepreneurial intention under different background variables in China and the West.