Research on Talent Cultivation Paths for High Quality Employment in the Perspective of Industry-Education Integration

Under the strategic guidance of accelerating the development of education, how to highlight the main line of quality in the accelerated pace, how to better understand the quality of education and talent training, and constantly improve the ability to train talents, enhance the ability to guarantee and promote the fast, good and healthy development of education has become the focus of the community’s common concern. To write a good education, we should adopt various quality assurance measures and management measures according to the focus of stakeholders’ needs, and improve the satisfaction of talent training quality through the satisfaction of key points of stakeholders’ needs [28-29]. Based on the above analysis, the quality management and quality improvement of talent training is carried out along the logic of “stakeholder needs - quality assurance”.

Talent training quality assurance QFD model is a kind of quality assurance method starting from the micro-level demand and mapping the demand into quality characteristics through matrix transformation, which is the application of quality function theory in the field of education quality management. From the perspective of understanding stakeholders’ needs for talent training, the QFD model of quality assurance for talent training is constructed to explore their deep-seated needs for quality. The quality house and its correlation matrix play a crucial role in converting stakeholder needs to quality assurance characteristics in the talent training quality assurance model. The overall conceptual structure of the QFD model for talent training quality assurance is shown in Figure 1.

Figure 1.

The general concept structure of QFD model of talent training

1)

Analysis of demand element matrix and its weight matrix

On the basis of the in-depth analysis of the needs of talent development stakeholders, a preliminary demand element development table was obtained, and through the questionnaire, the test was conducted to check whether the needs of the stakeholders were comprehensively and accurately summarized, so as to obtain the development table of the demand elements of the talent development stakeholders, i.e., the matrix of demand elements, which contains i elements.

Secondly, we analyze the weight matrix of demand elements in two parts, on the one hand, we use the questionnaire to test how important the demand of each item is to each stakeholder, and invite them to score each demand of each item, and the large-sample survey will help to better guarantee the validity and credibility of the weight data obtained. In the other part, five relevant experts are invited to give a ratio score between the two demand evaluation indicators based on their empirical knowledge according to the 1-9 scale method, and the scores are processed to obtain the demand importance data. The average value of the scores for the two parts of the demand can be synthesised to obtain the weight matrix of the demand elements.

The following descriptions of the arithmetic methods mainly focus on the relevant arithmetic involved in the conceptual model, including the description of the arithmetic method for the stakeholder demand importance (wTi)$$\left( {w_T^i} \right)$$, the description of the arithmetic method for the correlation matrix between demand elements and quality assurance characteristics (Aij)$$\left( {{A_{ij}}} \right)$$, and the arithmetic method for the weights of quality assurance characteristics (hj)$$\left( {{h_j}} \right)$$.

In order to improve the validity and reliability of the constructed model from the source, it is necessary to firstly determine the stakeholder demand element weights in a reasonable way, which is one of the possible innovations considered in this paper [30]. Firstly, the results of the scores obtained through the traditional questionnaire survey are analyzed, and the mean value of the scores of the ith demand element is represented by αi(i=1,2,⋯⋯n)$${\alpha _i}\left( {i = 1,2, \cdots \cdots n} \right)$$, and the weight of the ith demand element is calculated by the following formula w1Ti$$w_{1T}^i$$: (1)w1Ti=αi/∑i=1nαi$$w_{1T}^{i}={{{\alpha }_{i}}}/{\underset{i=1}{\overset{n}{\mathop \sum \nolimits}}\,{{\alpha }_{i}}}$$

Secondly, combining the previous analysis of the integration of Quality Function Deployment (QFD) with other methods and the application of QFD in the field of education, this study sets and rough hierarchical analysis (RAHP) and Decision Making Experiment and Evaluation Experiment (DEMATEL) to measure the degree of importance of requirements. The rough hierarchy analysis method is suitable for dealing with complex requirements analysis due to its superiority of not being limited by any prior knowledge. The needs of college talent training stakeholders have unstructured qualities such as diversity and complexity, and the use of RAHP to analyze the needs is conducive to better meeting the requirements of the characteristics of the needs. The specific steps for measuring the importance of college talent cultivation stakeholders’ needs based on RAHP are as follows:

STEP1: Assuming that m relevant expert is involved in scoring, pairwise comparison of indicators is performed to establish AHP pairwise comparison matrix Y_i. and the consistency of the results is tested by the consistency judgment formulas: CI=(λmax−n)/(n−1)$$CI=\left( {{\lambda }_{\max }}-n \right)/(n-1)$$ and CR = CI/RI, and the judgment results are considered to be consistent if the CR is less than 0.1.

STEP2: The obtained pairwise comparison matrix Y_i is further transformed into rough group decision matrix Y^*, and the roughness numbers and roughness boundary intervals in this matrix are calculated, where: (2)Y*=[ 1 A12* ⋯ A1n* A21* 1 ⋯ A2n* ⋮ ⋮ ⋯ ⋮ An1* An2* ⋯ 1]$${Y^*} = \left[ {\begin{array}{*{20}{c}} 1&{A_{12}^*}& \cdots &{A_{1n}^*} \\ {A_{21}^*}&1& \cdots &{A_{2n}^*} \\ \vdots & \vdots & \cdots & \vdots \\ {A_{n1}^*}&{A_{n2}^*}& \cdots &1 \end{array}} \right]$$

Where: A12*={a121,a122,⋯,a12m}$$A_{12}^* = \left\{ {a_{12}^1,a_{12}^2, \cdots ,a_{12}^m} \right\}$$, and so on. Taking A12*$$A_{12}^*$$ as an example, it corresponds to the roughness number: (3)RN(a12i)=[ a12i−, a12i+]$$RN\left( {a_{12}^i} \right) = \left[ {\begin{array}{*{20}{l}} {a_{12}^{i - },}&{a_{12}^{i + }} \end{array}} \right]$$ (4)RN(A12*)={[a12i−,a121+],[a122−,a122+],⋯,[a12m−,a12m+]}$$RN\left( {A_{12}^*} \right) = \left\{ {\left[ {a_{12}^{i - },a_{12}^{1 + }} \right],\left[ {a_{12}^{2 - },a_{12}^{2 + }} \right], \cdots ,\left[ {a_{12}^{m - },a_{12}^{m + }} \right]} \right\}$$

Eq. i ∈ [1, m].

The average roughness interval is obtained from Eq: (5)RN(A12)=[a12−,a12+]$$RN\left( {{A_{12}}} \right) = \left[ {a_{12}^ - ,a_{12}^ + } \right]$$

STEP3: Create a rough pairwise comparison matrix Y, where: (6)Y=[ [1,1] [a12−,a12+] ⋯ [a1n−,a1n+] [a21−,a21+] [1,1] ⋯ [a2n−,a2n+] ⋮ ⋮ ⋮ [an1−,an1+] [an2−,an2+] ⋯ [1,1]]$$Y = \left[ {\begin{array}{*{20}{c}} {[1,1]}&{\left[ {a_{12}^ - ,a_{12}^ + } \right]}& \cdots &{\left[ {a_{1n}^ - ,a_{1n}^ + } \right]} \\ {\left[ {a_{21}^ - ,a_{21}^ + } \right]}&{[1,1]}& \cdots &{\left[ {a_{2n}^ - ,a_{2n}^ + } \right]} \\ \vdots & \vdots &{}& \vdots \\ {\left[ {a_{n1}^ - ,a_{n1}^ + } \right]}&{\left[ {a_{n2}^ - ,a_{n2}^ + } \right]}& \cdots &{[1,1]} \end{array}} \right]$$

STEP4: Split the rough pairwise comparison matrix Y into rough lower boundary A⁻ and rough upper boundary A⁺, and solve the eigenvalues and eigenvectors of A⁻ and A⁺ to obtain the respective demand importance: (7)A−=[ 1 a12− ⋯ a1n− a21− 1 ⋯ a2n− ⋮ ⋮ ⋮ an1− an2− ⋯ 1]$${A^ - } = \left[ {\begin{array}{*{20}{c}} 1&{a_{12}^ - }& \cdots &{a_{1n}^ - } \\ {a_{21}^ - }&1& \cdots &{a_{2n}^ - } \\ \vdots & \vdots &{}& \vdots \\ {a_{n1}^ - }&{a_{n2}^ - }& \cdots &1 \end{array}} \right]$$ (8)A+=[ 1 a12+ ⋯ a1n+ a21+ 1 ⋯ a2n+ ⋮ ⋮ ⋮ an1+ an2+ ⋯ 1]$${A^ + } = \left[ {\begin{array}{*{20}{c}} 1&{a_{12}^ + }& \cdots &{a_{1n}^ + } \\ {a_{21}^ + }&1& \cdots &{a_{2n}^ + } \\ \vdots & \vdots &{}& \vdots \\ {a_{n1}^ + }&{a_{n2}^ + }& \cdots &1 \end{array}} \right]$$ (9)E−=[e1−,e2−,⋯,en−]T$${E^ - } = {\left[ {e_1^ - ,e_2^ - , \cdots ,e_n^ - } \right]^T}$$ (10)E+=[e1+,e2+,⋯,en+]T$${E^ + } = {\left[ {e_1^ + ,e_2^ + , \cdots ,e_n^ + } \right]^T}$$

was obtained by normalization: (11)wi−=ei−/∑i=1nei−$$w_{i}^{-}=e_{i}^{-}/\underset{i=1}{\overset{n}{\mathop \sum \nolimits}}\,e_{i}^{-}$$ (12)wi+=ei+/∑i=1nei+$$w_{i}^{+}=e_{i}^{+}/\underset{i=1}{\overset{n}{\mathop \sum \nolimits}}\,e_{i}^{+}$$

From the above results it can be concluded that the importance of stakeholder need T_i for this layer is: (13)w(Ti)=12(|wi−|+|wi+|)$$w\left( {{T_i}} \right) = \frac{1}{2}\left( {\left| {w_i^ - } \right| + \left| {w_i^ + } \right|} \right)$$

1)

Talent quality function development

Talent quality indicators, different from the quality indicators of specific products, so at the beginning of determining talent quality indicators, according to the current situation of the entrepreneurial market, experience conditions, relevant research, expert insights, etc., to establish broad indicators, and then through the entrepreneurs, entrepreneurial personnel and entrepreneurial market research to further analyse and screen, select the highest recognition of a few indicator items, and through the questionnaire pre-testing to finally determine the quality indicators. In the process of decomposition of quality, using the hierarchical analysis method, the quality function is subdivided and unfolded at three levels. Finally, it is determined that the decomposition of talent quality indicators is mainly carried out in five aspects: entrepreneurial knowledge literacy, innovation and entrepreneurial awareness, entrepreneurial practice skills, entrepreneurial comprehensive quality, and entrepreneurial character. The requirements for high-quality employment talent competencies are shown in Table 1.

The survey is divided into three parts: first, the survey on the quality of entrepreneurial talents, and the talent quality survey includes the evaluation of the entrepreneurial market on the importance of talent quality indicators and the evaluation of the entrepreneurial market on the competitiveness of the entrepreneurial talent market. The entrepreneurial market mainly investigates past entrepreneurs from our college, school-enterprise cooperation units, and small and medium-sized enterprises founded by alumni. Secondly, the survey on the competitiveness of talent cultivation quality is based on the competitiveness of talent cultivation technology through the interviews with former entrepreneurial graduates, school-enterprise cooperation units, alumni-founded small and medium-sized enterprises, entrepreneurship instructors, and experts on entrepreneurship education in similar institutions. Thirdly, the survey on the correlation between talent quality indicators and talent cultivation quality characteristics (cultivation technology), specifically to find out the degree of correlation between each talent quality indicator and each cultivation quality characteristic. 1)

Questionnaire design and distribution

In order to ensure the rationality and practicability of the questionnaire survey, and also according to the needs of quality control principle, the steps of involving, testing in advance, distributing and recovering the questionnaire are also carried out before the questionnaire collection formally, to further ensure the scientificity and truthfulness of the survey results. (1)

Design of the questionnaire

In accordance with the principle of customer demand development of QFD theory, firstly starting from the entrepreneurial market, the original voice of the entrepreneurial market’s requirements for the quality of entrepreneurial talents is converted into the quality indicators of entrepreneurial talents’ quality through scientific methods, and then the results are processed through the methods of expert interviews, university research and visits to entrepreneurial personnel, utilizing the Customer Response Tool (VOC), and the specific measures of quality assurance for the cultivation of talents are established. On the basis of reference to previous research, combined with the actual situation of this paper’s research, a questionnaire was designed, and in order to objectively reflect the respondents’ degree of agreement with the survey indicators, and at the same time to facilitate the operation and statistical measurement of data, the questionnaire was based on the Richter’s Five Level Scale method, which is often used in marketing research activities, in which all the survey items in the questionnaire were in the form of a list, and the respondents were asked to use the five-level numbers to directly rate the various survey scale for evaluation.

The weight results of the two groups of evaluation indicators on talent training quality are calculated using the hierarchical analysis method (AHP) and the entropy weight method, respectively. In view of the possible differences in weights caused by subjective and objective assignment methods, the linear combination assignment method is adopted to determine the final comprehensive weights as the weight values of talent training quality evaluation indicators. The weight results are summarized and combined as shown in Table 6. As can be seen from the table, entrepreneurial practice skills have the highest proportion of 0.296 in the AHP weights.

In the application of the model to the training activities of talents, we obtained the complete demand-quality characteristic correlation matrix and calculated the demand element weights. Ten experts have been invited to score the correlation degrees in the demand-quality characteristic correlation matrix respectively. A consistent rating summary table was obtained by combining the procedures of the Delphi method, followed by the calculation of the importance and ranking of talent quality characteristics. Due to the existence of certain autocorrelations among different quality characteristics, this paper uses the DEMATEL method to analyze the mutual influence of each quality characteristic on talent development through expert scores. The relationship between these characteristics was assessed by experts using a scale of 0 to 3, which represents different degrees of influence. The average of these ratings formed the direct influence matrix. After drawing conclusions these data were processed using R language software in order to arrive at the final calculations and the results of the DEMATEL calculations for the talent quality characteristics are shown in Table 7. The correlation between the demand and quality characteristics elements is shown in Table 8.

The main goal of this dissertation is to utilize quality function unfolding to conduct research on talent training quality pathways rather than direct quality improvement. Since there is no mature and perfect talent training service system for reference, this thesis does not conduct competitive assessments during the construction of a quality house. Based on the analysis of the weight of talent cultivation demand and the centrality of quality characteristics, this thesis constructs the HoQ of talent cultivation services using the QFD framework.

Step 1: According to the calculated demand weights, the demand for talent development and its weights are located in the left part of the HoQ.

Step 2: Calculate the relationship between talent development needs and quality characteristics through the matrix. The strength of the relationship was finalized by averaging the ratings of ten experts.

Step 3: Combining the matrix and demand weights between talent development calculated in Step 2, the initial weights of quality characteristics can be calculated.

Step 4: The autocorrelation and centrality between quality characteristics are calculated using the DEMATEL method, and the final weights of the quality characteristics are calculated and ranked by centrality. The final weight ranking of the elements of talent training characteristics is shown in Table 9. From the table, it can be seen that the weight of program consulting and tracking services in entrepreneurship education management services has the highest proportion of 2.238.