Abstract
With the continuous advancement of higher education reform, talent classification evaluation has become a new trend in the management and development of teachers in colleges and universities. As an important foundation for the development of teachers in colleges and universities, talent evaluation not only plays the role of the baton but also promotes the process of educational reform in colleges and universities. As the difficulty and key point of the reform of the personnel system in colleges and universities, the implementation of the classification and evaluation of talents in colleges and universities is conducive to mobilizing their enthusiasm and creativity and is an inevitable trend in the reform and development of higher education. However, at present, teaching, scientific research, and social services are mainly used as the important basis for teacher evaluation in college talent evaluation. The evaluation mechanism is not perfect, the evaluation standard is single, the evaluation methods are convergent, and evaluation mechanism of the academic circles on evaluation, etc. The research on the problem is still very weak, and the development of college teachers is increasingly affected. It is urgent to build a set of accurate classification and evaluation systems for college talents. In view of the many problems existing in the above-mentioned talent classification and evaluation mechanism in colleges and universities, this paper adopts AHP and fuzzy evaluation method, adopts this method to evaluate talents in the stage of talent introduction in colleges and universities, classifies talents according to the evaluation results, and finally uses the classification results as it is tested in practice to verify the effectiveness of the method constructed in this paper.
1. Introduction
With the rapid development of economy, the social demand for talents is also increasing, and the quality of higher education has become the focus of social attention. In order to better improve the teaching level of college teachers, we must strengthen the research of classification evaluation systems [1–3]. Because the traditional teaching quality research adopts qualitative analysis, the results are highly one-sided, which cannot accurately and comprehensively reflect the objective reality, and due to the limitations of technology and other reasons, the evaluation system is too cumbersome, the process is too complex, and the results lag is serious, which is not very helpful for the improvement of teaching quality [4–7]. The evaluation of teaching quality is mainly aimed at the comprehensive evaluation of teaching ability, teaching process, and teaching effect, which not only provides a certain data reference for teachers to improve the teaching content but also provides more basis for formulating a scientific and reasonable classification evaluation system [8, 9].
At present, China still has problems in talent evaluation methods and mechanism construction, such as weak classification evaluation management, convergence of evaluation standards for all kinds of talents, backward evaluation methods, and low integration of evaluation indicators with social needs. The existence of these problems is crucial to further deepen the reform of talent evaluation mechanisms [10–12]. As a large country of human resources, although China has unique advantages in the development and training of human resources, colleges and universities are gathered, scientific research institutes are numerous, and the industrial economy is booming, there are still various drawbacks in the evaluation of all kinds of talents; In terms of talent resource evaluation and use, the concepts and methods are not advanced enough, and there are many areas that need to be improved [13–16]. In the new era, China’s demand for independent innovation is more urgent than ever before. Therefore, more high-level, high-tech and interdisciplinary talents are needed to help the process of social development. However, under the influence of the contradiction between the continuous progress of demand and the stagnation of technology, China’s talent evaluation system is always in the development stage. It can be seen that a series of work such as talent training, evaluation, and management in China are facing great challenges, and it is also a key stage of comprehensive reform [17, 18]. Therefore, taking talent evaluation as the starting point and connecting point of reform, research, and explore the corresponding talent evaluation index system and evaluation methods and corresponding ways to improve, so as to improve the employment environment, enhance the attraction to talents, and promote the best use of talents, which is also one of the basic work in talent management [19].
In order to solve the problems of incomplete talent evaluation system and improper talent classification in our country, based on previous studies, this paper adopts AHP and fuzzy evaluation method to evaluate in the talent introduction stage of colleges and universities, and classifies talents according to the evaluation results. Finally, the classification results are used for practical testing, so as to promote the improvement of talent evaluation system in our country.
2. Problems in the Classification and Evaluation of Talents in Colleges and Universities at the Present Stage
2.1. Problems in the Evaluation Subject
(1)Alienation and dislocation of evaluation subject. Talent evaluation should be oriented by innovation, quality, contribution, and other indicators. However, in the actual implementation process, administrative power is often greater than academic power in China. The evaluation process is controlled by the leadership intention of the evaluation subject unit, and the correct and scientific evaluation method is alienated into a subjective evaluation method involving complex “relationships” [14, 20].(2)The professional quality of the evaluation subject is low. Compared with foreign professional evaluators who have received systematic training, although China also has some excellent professional evaluators of talents in different industries, the number of such professionals is insufficient, and the quality level is uneven, which seriously restricts the development and progress of talent evaluation in China.
2.2. Problems in the Evaluation Object
(1)The definition of scientific and technological talents is unclear. At present, academia, industry, and government departments have different understandings of the definition of talents, and lack an accurate definition of the deep-seated connotation and characteristics of talents. Ji x[12] pointed out that the definition of talent and related concepts in China is vague, and there is no unified and scientific definition. Hua et al. [14] pointed out that although many domestic scholars have made various interpretations of the connotation of innovative talents, they have not reached a consensus, which makes the definition of talents a form, which is not conducive to the accurate evaluation of all kinds of talents.(2)Insufficient talent classification and evaluation: at present, the domestic talent evaluation system is general, lacking the classification mechanism and corresponding evaluation indicators for various types of talents. Wang et al. [21] pointed out that the lack of classified evaluation is one of the difficulties faced by the evaluation of scientific research talents in China. Wang [22] pointed out that the classification and evaluation of talents is the requirement of talent science and technology concept, but at present, the classification standards of talents are different, which leads to the evaluation results cannot truly reflect the value and contribution of talents, and the guiding role is not obvious.
2.3. Problems in Evaluation Standards
(1)The pertinence of talent evaluation standards is poor. At present, the evaluation criteria of talents in China are often determined by the evaluation subject, and will change with the change of evaluation objectives. Therefore, the existing evaluation criteria are not reasonable, not only lack scientificity but also have strong fuzziness. Cui [23] pointed out that the existing talent evaluation standards do not emphasize systematic innovation enough, which is not conducive to promoting talent output and transformation. Powell et al. [18] evaluated high-level scientific and technological talents from multiple comprehensive standards such as talent quality level, academic performance, and influence, avoiding the untrue evaluation results caused by a single evaluation standard.(2)The evaluation index is lack of operability. At present, the selection of talent evaluation indicators in China has not overcome the shortcomings of traditional evaluation indicators, and still focuses on “morality, ability, diligence, performance” and other aspects. On the one hand, static indicators lack adjustability, and cannot well meet the needs of actual evaluation; On the other hand, the selection of evaluation indicators is divorced from the principle of seeking truth from facts, resulting in the measurement of some indicators cannot be operated, and the subjectivity is large.
2.4. Problems in Evaluation Methods
(1)There are few original theories, which copy the theoretical framework of foreign countries. The weakness of theoretical construction makes the scientific foundation of talent evaluation in our country insufficient, especially the talent evaluation theory with Chinese characteristics that is suitable for the actual situation of our country has not been fully formed. If we copy the theoretical evaluation method that is not suitable for the development of colleges and universities in our country, it will cause the risk of “going astray” to the domestic talent evaluation work.(2)The evaluation method is single and subjective. At present, China mainly adopts the weighted talent evaluation method, evaluates talents through the evaluation index system formulated in advance, and obtains the total score of talents according to the weight of each evaluation index. Although this method pays attention to the comprehensive evaluation of scientific and technological talents, it also ignores the correlation between various indicators, and does not consider the systematic effect of the talent subject formed by various indicators in the talent evaluation process, so that the evaluation results cannot accurately reflect the quality, ability, and skills of talents.(3)The evaluation system is unreasonable, such as a single evaluation standard, no differential treatment of scientific and technological personnel at different levels and scientific research posts of different nature, the complexity and variability of teachers’ work are not effectively reflected, the phenomenon of seniority and thesis orientation is serious in the evaluation process, and the phenomenon of ignoring or avoiding qualitative indicators such as the morality and scientific research quality of the evaluated personnel occurs from time to time.(4)It is easy for some teachers to ignore the research of teaching art and the improvement of teaching quality.(5)Emphasize quantity and neglecting quality, that is, focusing on quantity in the evaluation of scientific research achievements such as papers, works, fund projects, and patents. This evaluation without quality as the core is easy to lead to the emergence of a large number of low-level achievements.(6)The classification of scientific researchers is general, and even the differences among the most basic scientific research teachers, teaching teachers, and teaching and scientific research teachers are not reflected, and the differences between disciplines and majors are impossible to talk about.
2.5. Problems in Evaluation Procedures
(1)The evaluation process lacks impartiality. First of all, government departments and organizations will have an impact on talent evaluation, which destroys the objectivity and impartiality of the talent evaluation process. Secondly, the evaluation process is easily affected by economic factors, such as bribery, which not only affects the authenticity of talent evaluation but also leads to a bad atmosphere in the process of talent evaluation. Finally, talent evaluation is vulnerable to interpersonal relationships in the process, and “colleague” and “leadership” evaluation procedures are inevitable, making the evaluation results unable to truly reflect the actual situation of talents.(2)The evaluation process lacks effective supervision. The unreasonable talent evaluation process directly leads to the distortion of the evaluation results, which is far from the actual situation. However, at present, there is no effective supervision mechanism or relevant regulatory agencies in China, and the function of external supervision is often a mere formality, which leads to the fact that the supervision procedure does not improve the objectivity and impartiality of the talent evaluation process.
2.6. Problems in Evaluation Results
(1)The application of evaluation results is insufficient. The application field of domestic talent evaluation results is relatively narrow, and the value of evaluation results cannot be effectively transformed into market value. This drawback is mainly manifested in two cases: one is formalism, which regards the evaluation process as a process or procedure, and the subsequent evaluation results are shelved; The other is to overemphasize the role of talent evaluation, resulting in the resistance of the evaluated to talent evaluation.(2)The evaluation results are lacking follow-up evaluation and supervision feedback. From the actual operation of the domestic innovative talent selection project, it can be seen that the government departments strictly control the whole process of the selection of various talent projects according to the system, which means that excellent individuals or entrepreneurial teams will be able to obtain high social funds, strong support, and good social reputation, but there are obvious deficiencies in the follow-up implementation of performance appraisal and market supervision of the construction of various talent selection projects.
3. Construction of Talent Classification and Evaluation System in Colleges and Universities from the Perspective of AHP
3.1. AHP Theory
AHP is a qualitative and quantitative evaluation and analysis method proposed by Professor satti, an American operational research scientist. Its principle is relatively simple, that is, the evaluator decomposes the problem into several levels and elements by analyzing the governing relationship, internal essence or influence factors of a complex problem, and compares, judges, and calculates among the elements at the same level, so as to obtain different ranking schemes and provide decision-making basis for selecting the optimal scheme. The advantage of this analysis method is that it makes use of less quantitative information and mathematicizes the decision-making thinking process, which has high reliability and small error. The talent classification using this method mainly includes four steps as shown in Figure 1.
3.1.1. Establish Hierarchical Structure Model
When using analytic hierarchy process to make decisions, we should first clarify the purpose, scope, and requirements of the evaluation object of the decision-making goal, as well as the original information we have, and analyze the relationship between various elements; Secondly, the system problem is divided into levels and layers, and a hierarchical structure model is constructed. In this model, the problem is decomposed into several elements, and these elements are grouped according to attributes to form a mutually disjoint hierarchy. In the hierarchy, the upper layer dominates the elements of the lower layer, forming a layer by layer domination relationship from top to bottom according to the hierarchy:(1)The highest level: the target level, which usually has only one element, is the desired result of the decision or the goal to be achieved;(2)Intermediate layer: in order to achieve the intermediate link covered by the goal, it can be composed of several levels such as criteria and sub criteria, so the intermediate layer is also called the criteria layer;(3)The bottom layer: also known as the scheme layer, mainly includes the solutions and measures to be taken to achieve each goal.
Generally speaking, the number of levels of a hierarchical structure model is related to the complexity of system problems and the level of detail that needs to be analyzed. In addition, in order to avoid the difficulty of comparing and judging because of too many dominant elements, the next level of elements dominated by each element in each level is required to be no more than nine.
3.1.2. Build the Judgment Matrix of Pairwise Comparison
With the establishment of hierarchical structure, the membership relationship between the upper and lower elements is also basically determined. Next, we need to build a pairwise comparison judgment matrix. That is, the elements at the upper level are the evaluation criteria, and the elements at this level are compared and weighted according to the relative importance of the evaluation criteria. The form of judgment matrix is shown in Table 1.
In order to quantify and compare the results, the 9-level scaling method is introduced, that is, the natural number 1–9 and its reciprocal are used to compare the relative importance of the two elements, so the evaluator’s thinking decision process can be quantified more accurately. The 9-level scaling method and its significance are shown in Table 2.
3.1.3. Hierarchical Single Sorting and Consistency Verification
The maximum eigenvalue and its corresponding eigenvector W are calculated for the judgment matrix A, so the weight vector corresponding to the evaluation element is obtained.
The steps of using the square root method to calculate and W are as follows.
First, multiply the weight value obtained when each element in the judgment matrix A is compared with other elements, and open the final result to the power of n to obtain a new set of vectors:
Then, normalize the new vector to obtain the feature vector:
Calculate the maximum eigenvalue of the judgment matrix:
Because the judgment matrix is obtained by comparing the elements in pairs, the weights of the elements are estimated. If the deviation of valuation is too large, and there is a serious inconsistency of thinking and judgment, the matrix must be corrected. The purpose of consistency verification is to verify whether such a deviation exists in the matrix. The matrix is valid only if it passes the consistency verification. The steps are as follows:(1)Calculate the consistency index CI: where , according to Table 3, since the dimension of matrix A is n = 4 and RI = 0.9. Table 3 shows the RI values of 1–10 dimensional matrices.(2)Calculate the random consistency ratio CR: The average random consistency index RI is obtained based on the arithmetic mean of the eigenvalues of the random judgment matrix after repeated calculations. When CR < 0.1, this judgment matrix is determined to meet the consistency requirements. According to the above, , the CR of each judgment matrix is less than 0.1, so the consistency test is passed.
Similarly, according to the above steps, calculate the weight of layer C indicators relative to layer B, and the weight of layer D indicators relative to layer C in turn, and conduct consistency test. After calculation, all indicators also pass the consistency test.
3.1.4. Combination Weight Calculation and Total Hierarchy Sorting
(1)In order to choose a better place, after calculating the weight vector of a group of elements relative to a certain criterion element of the upper level, we also need to synthesize the weights of each single level element to obtain the combined weight of the lowest level relative to the highest level target. If the lowest layer is set as K layer, the weight of its index I relative to the index of the previous layer is , and the weight of the index of this layer relative to the index of the next higher layer is , and so on, then the weight of the index of the lower layer relative to the highest layer can be obtained:(2)Consistency verification shall be carried out after the total ranking of the levels. When CR < 0.1 inch, the overall ranking consistency verification of the judgment level passes:
3.2. Fuzzy Comprehensive Evaluation Theory and Steps
(1)Determine evaluation factors: set as m factors (i.e., evaluation index) to describe the evaluated object; M is the number of evaluation factors, which is determined by the specific index system.(2) Determine the evaluation level: let be n kinds of decisions (i.e., evaluation grade) that characterize the evaluated object; N is the number of comments, and each comment can correspond to a model subset. The specific level can be described in appropriate language according to the evaluation content. For example, the evaluation of talents in this paper v = {Excellent, Good, General, Poor, Bad}.(3) Construct evaluation matrix: first of all, for the single factor in the key factor set, from the , determine the affiliation of the thing to the decision level , so as to obtain the single factor evaluation set of the ith factor : . In this way, the evaluation set of m such factors constitutes a general evaluation matrix R, that is, each evaluated object determines the fuzzy relationship R from U to V, which is a matrix, as shown in equation (9). where means that starting from the factor , the evaluation object can be rated as the membership of , . Specifically, represents the frequency distribution of the ith factor Ui on the j-th comment , which is generally normalized to meet .(4)Determine the fuzzy weight a of evaluation factors: after obtaining the fuzzy relation matrix, it is still unable to evaluate things. The status and function of each factor in the evaluation factor set in the “evaluation objective” are different, that is, each evaluation factor occupies a different proportion in the comprehensive evaluation. A fuzzy subset a on u is to be introduced, which is called the weight or weight assignment set, . Where , and , it reflects a tradeoff of various factors.(5)A is synthesized in the R of each evaluated object by using a suitable synthesis operator to obtain the fuzzy comprehensive evaluation result vector B of each evaluated object.(6)Analyze the result vector of fuzzy comprehensive evaluation.
In order to make the evaluation results more realistic, we can make full use of the information brought by B and consider the rating parameters of various grades and the evaluation result B as a whole. In this way, we can set the parameter column vector relative to each level as
Then, the evaluation result of grade parameters is
P is a real number, which reflects the inductive information brought by hierarchical fuzzy subset B and hierarchical parameter vector C. In many daily applications, it is a very effective comprehensive parameter.
3.3. Construction of AHP Fuzzy Comprehensive Evaluation Model
The AHP fuzzy comprehensive evaluation system structure comprehensively constructed by the above steps is shown in Figure 2. The comprehensive classification evaluation model constructed in this paper has the following advantages:(1)Because colleges and universities have the characteristics of complex personnel, great uncertainty and many complex factors when introducing talents, one research method is no longer suitable for the classification and evaluation of talents, so it needs the combination of multiple analysis methods.(2)In the specific work of talent classification and evaluation, the process of talent introduction itself is very uncertain. Although analytic hierarchy process can solve the problem of qualitative and quantitative multi-objective decision making, it has limitations and needs to be combined with the comprehensive evaluation results of fuzzy comprehensive evaluation methods.(3)Increase the convenience of operation and calculation of the evaluation method, and make the results more simple and clear.(4)The evaluation index is more comprehensive, and the index evaluation that cannot be quantified is transformed into quantitative evaluation.
4. Case Analysis
4.1. Questionnaire Survey and Data Statistics
4.1.1. Investigation Purpose
In order to highlight the applicability of the AHP fuzzy comprehensive evaluation model constructed in this paper to the talent evaluation of ordinary colleges and universities, the author conducted a sample survey of teachers and managers from four higher vocational and undergraduate colleges in China. A total of 400 relevant talents were selected for this sampling.
4.1.2. Investigation Form
This paper adopts the method of interview and questionnaire. Interviews and surveys were conducted with some teachers and administrators of national and autonomous regional backbone schools and universities with obvious professional characteristics. The main issues involved include: the current situation of talent training in Colleges and universities, the construction of internship and training projects inside and outside schools, the construction of “double qualified” teachers, the enthusiasm and investment of enterprises in the process of talent training, the role of enterprises in the training process, the role of enterprises in the training of talents, the employment status of students, and the satisfaction of students with teachers and colleges.
The main evaluation indicators included in the questionnaire are shown in Table 4.
4.1.3. Investigation Implementation
It took half a year to complete the distribution, recovery and data statistics of the three questionnaires. A total of 400 questionnaires were distributed, 367 of which were recovered, with a recovery rate of 91.75%, 3 invalid questionnaires, 364 valid questionnaires, and an effective recovery rate of 91%. See Figure 3 for details.
4.2. Evaluation Effect of AHP Fuzzy Comprehensive Evaluation Model
It is known that the weight set of talent evaluation indicators of four colleges and universities can get W = {0.154, 0.234, 0.186, 0.157, 0.269} through calculation in Chapter 3, and the P calculated by Excel formula MMULT is a 364 × 1, and the scores and ranking of all survey objects can be observed from the results. On this basis, according to Table5, the talent level of the above four universities is shown in Table 6. V = {Excellent, Good, Average, Poor, Bad}
This result shows that the talent level of the above four universities is in a normal distribution. The “excellent” and “poor” students account for 5.7% and 3.3% of the total sample, respectively, and the students at the intermediate level account for 91% of the total sample. The score difference is small, and the distribution is relatively average. It can be seen from this that the talent level of the above four universities is high, the overall quality of talents is good, and they generally have the basic conditions to become high-quality university talents; A few students have outstanding talent characteristics; There are also fewer teachers whose overall level is poor, and individual teachers do not reflect the development characteristics of talents.
4.3. Research Based on Principal Component Analysis
In addition, this paper uses 364 effective samples for principal component analysis. By dimensionality reduction, a few principal components are extracted, and the linear function of the original variables is established. Calculate the comprehensive score of each survey factor according to the variance contribution rate of each variable. On the basis of scoring, the overall distribution was observed hierarchically to verify the evaluation results. Overall, 364 groups of survey data were imported into SPSS software to build the database, and the results were obtained by principal component analysis. See Figures 4 and 5 to show the correlation between the components.
As can be seen from Figure 4, the cumulative variance rate of some principal components is more than 80%. The eigenvalue of the first principal component is 4.872, which can explain 29.322% of the data samples, and the latter eigenvalue and variance rate gradually decrease. Therefore, we select 8 principal components for analysis, and calculate the scores of 8 principal components, respectively, by obtaining the principal component regression equation, as shown in Table 5.
As can be seen from Table 6, according to the principal component score, the sample sizes of “excellent” and “poor” are 19 and 18, respectively, accounting for 5.22% and 4.95% of the total sample, which are at a relatively low level; The “medium,” “poor,” and “poor” of the middle level are 78, 78, and 66, respectively, and the overall middle level accounts for more than 90% of the total. It can be seen from this that the results of principal component analysis are consistent with those of AHP fuzzy analysis and evaluation method, that is, the talent level of the above four universities is at the medium level, and the distribution is mainly in the “good” and “middle” range. The number of teachers with basic quality is large, and the foundation for talent introduction is good.
In addition, the factor load of the principal component can also be observed from the coefficient of the principal component equation, that is, the interpretation degree of the principal component to each explanatory variable. The first principal component can better explain these four variables, and has a better ability to explain the indicators of various dimensions such as vocational skills, innovation ability, and diligence. Among them, the third, fifth, and sixth principal components have a better ability to explain work attitude, responsibility, etc., combined with the other principal components, the extracted eight principal components can better explain all variables. Therefore, the first principal component better reflects the scientific research level of the introduced talents; the second principal component can identify the education level of the introduced talents; the third principal component can identify the application of the learned theory transformation in discipline competitions and academic papers; and the fourth principal component can measure their work, respectively.
In the AHP expert scoring and weighting, the scientific research level ability and vocational skill ability are given a large weight, which are, respectively, in the first and second place of the total weight ranking, representing the full attention of colleges and universities to teachers’ scientific research level and teaching and educating ability. This is the basis for higher education to achieve quality improvement and the most fundamental task for colleges and universities to cultivate professional talents for various industries. Therefore, the first and second principal components can be used to measure the overall sample.
In order to observe the results of the two methods more clearly, we use the histogram sub table to represent the international talent level distribution obtained by the AHP fuzzy comprehensive evaluation method and the principal component analysis method, as shown in Figures 6 and 7.
From the pie chart, the distribution similarity of the comprehensive scores of the two methods is very high, the number of samples of the maximum and minimum values is very small, and the samples are mainly distributed in the representative middle level interval; The sample size of each adjacent interval distribution has little difference, and the distribution is relatively uniform; The maximum sample size calculated by the two methods is within the range of “good” and “general,” and the estimation conclusions of the samples are consistent, indicating that the evaluation results are relatively robust.
It is worth noting that in the results of principal component analysis, the sample size in the “excellent” and “poor” range is more differentiated than the same kind of AHP fuzzy evaluation method, and compared with the principal component analysis method, the screening of excellent and poor talents is more obvious, and it is more in line with the talent introduction strategy of colleges and universities. Therefore, in practical application, the AHP fuzzy evaluation method constructed in this paper has more practical application value. This verifies the accuracy of the model constructed in this paper again.
5. Conclusion
In order to establish a correct employment orientation, encourage and guide the professional development of talents, mobilize the enthusiasm of talents for innovation and entrepreneurship, and speed up the construction of a strong country with talents, this study investigates and analyzes the situation and current situation of the existing talent evaluation system in China in terms of evaluation objectives, evaluation indicators, evaluation methods, evaluation cycle, evaluation procedures, and the use of evaluation results. It also summarizes the problems in the current talent evaluation system of colleges and universities and the various constraints behind the problems, and then clarifies the key elements for different talent evaluation. Based on the different scientific research stages of scientific and technological talents, the domestic talent evaluation indicators are divided into four first level indicators and 12 second level indicators; This paper uses AHP fuzzy comprehensive evaluation method to establish a comprehensive classification evaluation model of college talents, and analyzes and evaluates the results of the two models based on the talent classification index. The example shows that the talent evaluation system based on AHP fuzzy comprehensive evaluation method is more suitable for the talent introduction strategy of colleges and universities at the present stage. This study can not only promote the progress of related model building theory but also provide theoretical guidance and practical reference for the construction of similar systems.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that they have no conflicts of interest.