4 Data Analysis Method

4.1 Analysis of reliability

According to Nunnally (1978), a reliability value of 0.7 indicates high reliability. Cuieford (1965) believes that if Cronbach’s α value is above 0.7 then the reliability is high. If the value is acceptable between 0.7 and 0.35. The value lower than 0.35 then the data should be rejected.

Our study focuses on the reliability analysis of the dimensions and factors that impact on entrepreneurial cognition, entrepreneurial intention and entrepreneurial decision. As can be seen in Table 1, Cronbach’s α value exceeds 0.7, indicating the high reliability of variables and structure of research.

Table 1 Result of Reliability Analysis

4.2 Correlation analysis

We analyze the variable and variable dimension correlation coefficients. Table 2 showed that ability cognitive and prepared cognitive, entrepreneurial feasibility and entrepreneurial desirability, entrepreneurial feasibility and entrepreneurial decision are correlated (coefficient＜0.05). Also, other variables and variables dimensions are highly correlated (coefficient＜0.01). The evidence above points out the high correlation between the variables and dimensions. This finding suggests that entrepreneurial cognition and entrepreneurial intention have a significant impact on the entrepreneurial decision, and the mediating of entrepreneurial intention is obvious.

Table 2 The Correlation Coefficients

4.3 The regression analysis

Table 3 The Models of Regression Analysis

Table 3 The Models of Regression Analysis-Continued

We estimate four different models for each of separate regression analysis. Regression results are reported in Table 3.

4.3.1 The regression analysis of entrepreneurial cognition and entrepreneurial decision

Model 1 is the regression analysis of entrepreneurial cognition and entrepreneurial decision. The result shows that Adj.R² is 0.611, indicating that entrepreneurial cognition can explain 61.1% of the variation in the entrepreneurial decision. The effect of regression degree fit better. Meanwhile, the measured value β=0.783, sig.=0.000＜0.01. Therefore, entrepreneurial cognition has a significant positive effect on the entrepreneurial decision. So hypothesis H1 is supported.

Next, we analyze prepared cognitive and entrepreneurial decision; regression result shows Adj.R² is 0.430, indicating that prepared cognitive can explain 43.0% of the variation in the entrepreneurial decision, and the effect of regression degree fit better. Meanwhile, the measured value β=0.658, sig=0.000＜0.01, Therefore, prepared cognitive has a significant positive effect on the entrepreneurial decision. So hypothesis H1a is supported. The result of regression analysis between ability cognitive and entrepreneurial decision shows that Adj.R² is 0.565, indicating that ability cognitive can explain 56.5% of the variation in the entrepreneurial decision, the effect of the degree of regression fit better. The measured value β=0.753, sig.=0.000＜0.01, Therefore, cognitive ability has a significant positive effect on the entrepreneurial decision. So hypothesis H1b is supported.

4.3.2 The regression analysis of entrepreneurial cognition and entrepreneurial intention

Model 2 is the regression analysis of entrepreneurial cognition and entrepreneurial intention. The results show that Adj.R² is 0.382, indicating that entrepreneurial cognition can explain 38.2% of the variation in entrepreneurial intention, and the effect of regression degree fit better. Meanwhile, the measured valueβ=0.621, sig.=0.000＜0.01, Therefore, entrepreneurial cognition has a significant positive effect on entrepreneurial intention. So hypothesis H2 is supported.

Then we analyze the regression analysis between the cognitive dimension and entrepreneurial intention. Prepared cognitive and entrepreneurial intention regression results show Adj.R² is0.232, indicating that prepared cognitive can explain 23.2% of the variation in entrepreneurial intention, the effect of regression degree fit better. Meanwhile, the measured valueβ=0.486, sig.=0.000＜0.01, Therefore, prepared cognitive has a significant positive effect on entrepreneurial intention. So hypothesis H2a is supported. Secondly, the result of regression analysis between ability cognitive and entrepreneurial intention shows that Adj.R² is0.387, indicating that ability cognitive can explain 38.7% of the variation in entrepreneurial intention, the effect of regression degree fit better. The measured valueβ=0.624, sig.=0.000＜0.01, Therefore, cognitive ability has a significant positive effect on entrepreneurial intention. So hypothesis H2b is supported.

4.3.3 The regression analysis of entrepreneurial intention and entrepreneurial decision

Model 3 is the regression analysis of entrepreneurial intention and entrepreneurial decision; the result shows that Adj.R² is 0.440, indicating that entrepreneurial intention can explain 44.0% of the variation in the entrepreneurial decision, the effect of regression degree fit better. Meanwhile, the measured value β=0.665, sig.=0.000＜0.01, Therefore, entrepreneurial intention n has a significant positive effect on the entrepreneurial decision. So hypothesis H3 is supported.

Then we do the regression analysis of the relationship between entrepreneurial intention dimension and entrepreneurial decision. Entrepreneurial feasibility and entrepreneurial decision regression result show Adj.R² is 0.029, indicating that entrepreneurial feasibility can explain 2.9% of the variation in the entrepreneurial decision, and the effect of regression degree fit better. Meanwhile, the measured valueβ=0.184, sig.=0.011＜0.05, Therefore, entrepreneurial feasibility has a significant positive effect on entrepreneurial decisions. So hypothesis H3a is supported. Secondly, the result of regression analysis between entrepreneurial desirability and entrepreneurial decision shows that Adj.R² is 0.549, indicating that cognitive ability can explain 54.9% of the variation in the entrepreneurial decision, the effect of regression degree fit better. The measured value β=0.743, sig.=0.000＜0.01. Therefore, entrepreneurial desirability has a significant positive effect on entrepreneurial decisions. So hypothesis H3b is supported.

4.4 The mediating role of entrepreneurial intention

We use the validated methods in Model 4 (Kenny), this study will test the intermediary role of entrepreneurial intention. Firstly, we examine the relationship between entrepreneurial cognition and entrepreneurial decision. Model 1 data shows, β=0.783, sig.=0.000＜0.01. Therefore, entrepreneurial cognition has a significant positive relationship with the entrepreneurial decision. Secondly, examine the relationship between entrepreneurial cognition and entrepreneurial intention. Model 2 data shows,β=0.621, sig.=0.000＜0.01. Therefore, entrepreneurial cognition has a significant positive relationship with entrepreneurial intention. Thirdly, we examine the relationship between entrepreneurial intention and entrepreneurial decision. Model 3 data shows,β=0.665, sig.=0.000＜0.01. Therefore, the entrepreneurial intention has a significant positive relationship with entrepreneurial decisions. Fourthly, we verify the relationship among entrepreneurial cognition, entrepreneurial intention and entrepreneurial decisions. Model 4 data shows a regression coefficient of entrepreneurial cognition and the entrepreneurial decision is β=0.292, sig.=0.000＜0.01, this confirmed the positive relationship remained significant. However, the β value is reduced from 0.783 to 0.292, sig.=0.000＜0.01, this shows that entrepreneurial intention in entrepreneurship partially mediated the relationship between entrepreneurial cognition and entrepreneurial decision, so the Hypothesis H4 is confirmed.