4.2.2. Calculating Weight Coefficients

Because the hierarchical structure model we constructed had more elements at the program level and the generated judgment matrix order was greater than nine, we used a combination of the AHP and entropy methods for data processing. We formed the evaluation indexes by decomposing the problem and comparing the judgment. We calculated the weightings to obtain the comprehensive weight values of the elements at the program level. We first calculated the AHP-based weight.

Check the consistency of matrix *B*. Calculate:

$$\text{CR} = \text{CI}/\text{RI} \tag{7}$$

where CI is the consistency index; CR is the consistency ratio; and RI is the common random consistency index.

$$\mathbf{CI} = (\lambda\_{\text{max}} - n) / (n - 1) \tag{8}$$

From Equation (8), CR can be calculated. CR < 0.1 indicates that the calculation of matrix B is qualified and valid. If CR > 0.1, the matrix needs to be corrected [52].

Based on the above-mentioned ideas, we constructed the judgment matrix and calculated the weights of the impact factor (Tables 6–10).

From the outcomes in Tables 6–10, we found that the CR values of the judgment matrices were all <0.1, so we skipped the consistency test. From this, we calculated the weighting for the program-level elements to obtain the comprehensive weight values of the program-level elements (Table 11).

**Table 6.** Target layer judgment matrix and weight value of influencing factor.



**Table 7.** PU judgment matrix and weight values.

**Table 8.** PEOU judgment matrix and weight values.


**Table 9.** Perceived cost judgment matrix and weight values.


**Table 10.** Personal motivation judgment matrix and weight values.


**Table 11.** Comprehensive weight values of criterion-layer elements.

