4.1 Reliability and correlation analysis
1.Sample processing. In this study, all 427 valid sample questionnaire data were subjected to validated factor analysis (CFA), and the results of the analysis of scale fit indices (RMSEA= 0.080, CFI= 0.915, TLI= 0.906, SRMR= 0.053) showed that the data and the model were well fitted. In addition, the variances of inflated factors (VIFs) were all below 2.5 (maximum), indicating that there is no problem of multicollinearity between the independent variables.
2.Reliability test. In this study, Cronbach's internal consistency coefficient was used to test the reliability of the scales, and the values of the four scales of entrepreneurship education, entrepreneurial attitudes, entrepreneurial policies, and entrepreneurial intentions were 0.920, 0.879, 0.892, 0.960, respectively, and the CR values of their combined reliabilities were all greater than 0.9, which indicated that the scales were highly reliable.
3. validity test. This study firstly conducted the aggregated validity test. According to the results of statistical analysis, the factor loading values of all scales are greater than 0.5; the average variance extracted (AVE) values are greater than 0.6, indicating that the scales have high convergent validity. Further discriminant validity tests were conducted. The AVE square root values of the scale factors are all greater than the correlation coefficients between the factor and other factors, which satisfies the criterion of discriminant validity.
From the correlation coefficients, entrepreneurship theory education (r = 0.500, P < 0.001), entrepreneurship practice education (r = 0.508, P < 0.001), entrepreneurship atmosphere education (r = 0.521, P < 0.001) and entrepreneurial intention have a more significant positive correlation; entrepreneurship education and entrepreneurial attitudes (r = 0.526, P < 0.001) are significantly positively correlated; and endogenous entrepreneurial attitude has a more significant positive correlation with entrepreneurial intention (r = 0.656, P < 0.001); and exogenous entrepreneurial attitude has a significant positive correlation with entrepreneurial intention (r = 0.564, P < 0.001), which is in line with the theoretical expectations of this study and provides initial support for the research hypotheses.
4.2 Direct effects of entrepreneurship education
In order to test the relationship between the independent variable entrepreneurship education (theory, practice, atmosphere) and the dependent variable entrepreneurial intention, and the relationship between entrepreneurship education and the mediator variable entrepreneurial attitudes (endogenous, exogenous), this study used path analysis to test the direct effect of the independent variables on the mediator and dependent variables. According to the results of the path analysis shown in Table 2, it can be seen that the path coefficient of entrepreneurship theory education to entrepreneurial intention is positive and significant (β = 0.466, p < 0.001), the path coefficient of entrepreneurship practice education to entrepreneurial intention is positive and significant (β = 0.467, p < 0.001), and the path coefficient of entrepreneurship atmosphere education to entrepreneurial intention is positive and significant (β = 0.487, p < 0.001). Therefore, hypothesis 1 is supported. The path coefficient of entrepreneurship education to endogenous entrepreneurial attitudes is positively significant (β = 0.482, p < 0.001) and the path coefficient of entrepreneurship education to exogenous entrepreneurial attitudes is positively significant (β = 0.467, p < 0.001), therefore, hypothesis 2 is supported.
Table 2. Path analysis of entrepreneurship theory, practice, ambient education and direct entrepreneurial intentions.
Variables
|
Entrepreneurial intention
|
Estimate
|
S.E.
|
P
|
Intercept
|
2.447***
|
0.243
|
0.000
|
Genders
|
0.059
|
0.042
|
0.163
|
Birthplace
|
-0.001
|
0.041
|
0.984
|
Grade
|
-0.058
|
0.043
|
0.174
|
Student Leadership Experience
|
-0.038
|
0.042
|
0.370
|
Entrepreneurial experience of family members
|
0.107
|
0.041
|
0.009
|
Ranking of professional grades
|
-0.106*
|
0.043
|
0.013
|
Entrepreneurial state
|
0.127**
|
0.042
|
0.003
|
Participation in entrepreneurship competitions
|
0.074**
|
0.042
|
0.001
|
Entrepreneurship Theoretical education
|
0.466***
|
0.046
|
0.000
|
Entrepreneurship Practical education
|
0.467***
|
0.043
|
0.000
|
Entrepreneurship Ambient education
|
0.487***
|
0.040
|
0.000
|
R2
|
0.338***
|
|
|
F-value
|
36.988***
|
|
|
Note: N = 427; * P < 0.05, ** P < 0.01, *** P < 0.001 (two-tailed test), below; Intercept indicates intercept.
Table 3. Path analysis of entrepreneurship education and direct endogenous and exogenous entrepreneurial attitudes.
Variables
|
Endogenous attitude
|
Exogenous attitude
|
Estimate
|
S.E.
|
P
|
Estimate
|
S.E.
|
P
|
Intercept
|
3.272***
|
0.246
|
0.000
|
2.552***
|
0.266
|
0.000
|
Genders
|
0.031
|
0.048
|
0.511
|
0.062
|
0.050
|
0.217
|
Birthplace
|
-0.038
|
0.047
|
0.414
|
0.019
|
0.049
|
0.691
|
Grade
|
-0.065
|
0.049
|
0.182
|
-0.061
|
0.051
|
0.230
|
Student Leadership Experience
|
0.021
|
0.048
|
0.668
|
-0.004
|
0.050
|
0.932
|
Entrepreneurial experience of family members
|
0.134**
|
0.047
|
0.004
|
0.085
|
0.049
|
0.082
|
Ranking of professional grades
|
-0.058
|
0.048
|
0.233
|
-0.011
|
0.050
|
0.831
|
Entrepreneurial state
|
0.061
|
0.048
|
0.206
|
0.051
|
0.050
|
0.313
|
Participation in entrepreneurship competitions
|
-0.010
|
0.048
|
0.837
|
0.046
|
0.050
|
0.362
|
Entrepreneurship education
|
0.482***
|
0.045
|
0.000
|
0.467***
|
0.046
|
0.000
|
R2
|
0.270***
|
|
|
0.239***
|
|
|
F-value
|
27.279***
|
|
|
11.446***
|
|
|
4.3 Mediating effects of entrepreneurial attitudes
Regarding the test of mediating effect, the stepwise test method is commonly used in existing studies, but its statistical efficacy is small. The test of mediating effect This study utilizes Stata 17.0 and Mplus 7.0 software, and adopts the integrated sequential test method and Bootstrap method (Wen Zhonglin and Ye Baojuan, 2014) in two steps. First, construct the structural equation model from the independent variable to the dependent variable; second, construct the structural equation model after incorporating the mediating variables. The results of the sequential tests and the confidence intervals of the Bootstrap method are finally obtained. In this study, this method is used to test the mediating effect of endogenous and exogenous entrepreneurial attitudes, and the bias-corrected nonparametric percentile Bootstrap method is used for the estimation of parameter robust standard errors and confidence intervals (the sample size of resampling is 2000).
In the first step, this study tested the total effect of entrepreneurship education. After controlling for gender, place of birth, grade level, student leader experience, family member entrepreneurial experience, major grade rank, entrepreneurial status, and entrepreneurial competition participation, entrepreneurship education (β = 0.453, p < 0.001) was a significant predictor of entrepreneurial intention with a 95% confidence interval of [ 0.380, 0.526 ] and 34.9% of the variance in entrepreneurial intention explained.
In the second step, this study added both mediator variables of endogenous entrepreneurial attitudes, exogenous entrepreneurial attitudes and all control variables to the original model, and the results showed that the model was also well-fitted (CFI = 0.895, TLI = 0.889, SRMR = 0.074, RMSEA = 0.069), with 90% confidence intervals for the RMSEA of [ 0.065, 0.074]. The explanations of endogenous entrepreneurial attitudes, exogenous entrepreneurial attitudes and entrepreneurial intention variance were 41.4% and 31.8%, respectively. As shown in Figure 2, entrepreneurship education was a significant predictor of endogenous entrepreneurial attitudes (β = 0.512, p < 0.001) and entrepreneurship education was a significant predictor of exogenous entrepreneurial attitudes (β = 0.479. p < 0.001); endogenous entrepreneurial attitudes were a significant predictor of entrepreneurial intentions (β = 0.505, p < 0.001) and exogenous entrepreneurial attitudes were a significant (β = 0.391, p < 0.001). All of the above indirect effects were significant. According to the testing process of mediation effect, the direct effect of entrepreneurship education on entrepreneurial intention was also positive and significant (β = 0.644, P < 0.001), thus, endogenous entrepreneurial attitudes and exogenous entrepreneurial attitudes have partial mediation effect between entrepreneurship education and entrepreneurial intention, and the ratio of mediation effect to the total effect (ab/c) was 49.3%, and hypothesis 3 was established.
4.4 Moderating effects of entrepreneurship policies
Due to the significant effect of entrepreneurship theory, practice, and atmosphere education on entrepreneurial intention, as shown in Table 2. adding the moderating variables entrepreneurship policy and interaction terms (entrepreneurship theory education entrepreneurship policy, entrepreneurship practice education entrepreneurship policy, and entrepreneurship atmosphere education entrepreneurship policy), constructing Models 1-model 3 through Stata 17.0 software to test the moderating effect of entrepreneurship policy, the results are shown in Table 4. Entrepreneurship policy has a significant positive moderating effect on entrepreneurship theory education and entrepreneurial intention (β = 0.235, P < 0.001), entrepreneurship policy has a significant positive moderating effect on entrepreneurship practice education and entrepreneurial intention (β = 0.236, P < 0.001), entrepreneurship policy has a significant positive moderating effect on entrepreneurship atmosphere education and entrepreneurial intention (β = 0.182, P < 0.001 ), and hypothesis 4 was verified.
Table 4. Results of regression analysis of moderating effects.
Variables
|
Entrepreneurial intention
|
Model 1
|
Model 2
|
Model 3
|
Genders
|
0.149
|
0.126
|
0.135
|
Birthplace
|
-0.000
|
-0.009
|
-0.011
|
Grade
|
-0.080
|
-0.062
|
-0.091
|
Student Leadership Experience
|
-0.297
|
-0.100
|
-0.102
|
Entrepreneurial experience of family members
|
0.216*
|
0.217*
|
0.203*
|
Ranking of professional grades
|
-0.081
|
-0.144*
|
-0.102
|
Entrepreneurial state
|
0.725*
|
0.485
|
0.696*
|
Participation in entrepreneurship competitions
|
0.099
|
0.065
|
0.158*
|
Entrepreneurship Theoretical education
|
0.466***
|
|
|
Entrepreneurship Practical education
|
|
0.467***
|
|
Entrepreneurship Ambient education
|
|
|
0.487***
|
Entrepreneurship Policy
|
-0.416*
|
-0.380*
|
-0.226
|
Entrepreneurship Theoretical education × Entrepreneurship Policy
|
0.235***
|
|
|
Entrepreneurship Practical education × Entrepreneurship Policy
|
|
0.236***
|
|
Entrepreneurship Ambient education × Entrepreneurship Policy
|
|
|
0.182***
|
R2
|
0.382
|
0.415
|
0.402
|
ΔR2
|
0.365
|
0.400
|
0.386
|
F-value
|
23.30
|
26.79
|
25.36
|