Teaching quality improvement method based on correlation analysis of teaching evaluation and subject scores

Abstract

The invention discloses a teaching quality improvement method based on correlation analysis of teaching evaluation and subject scores. The method comprises the following steps: firstly, collecting teacher teaching evaluation data and corresponding student subject score data, and completing data preprocessing through Z transformation; thirdly, analyzing correlation coefficients between different teachers, different teaching evaluation indexes and student subject scores by applying a Pearson correlation coefficient analysis method; and finally, through a k-means clustering analysis method, the teachers are automatically grouped according to the teaching evaluation index characteristics, and a teaching quality improvement scheme of teacher individuals and groups is formulated. According to the invention, the case library is constructed by integrating teaching evaluation and subject scores, so that on one hand, teachers can accurately locate own teaching advantages and weak links and carry out targeted teaching ability improvement training; and on the other hand, a scientific basis is provided for teaching management personnel, and a personalized teaching quality improvement scheme can be formulated according to different clustering results, so that the configuration of teaching resources is optimized, and the teaching quality is improved.

Description

Technical Field

[0001] The present invention belongs to the field of education and teaching, and specifically relates to a method for improving teaching quality based on correlation analysis between teaching evaluation and subject performance. Background Art

[0002] In today's digital age, the rise of data mining technology has opened up new avenues for monitoring teaching quality in universities. With the deepening of educational informatization, universities have accumulated a vast amount of data related to teaching activities in their daily operations. This data covers the structure and teaching capabilities of faculty, the design and implementation of professional courses, the multi-dimensional evaluation of teaching effectiveness, and student learning outcomes. If this data can be fully mined and utilized, a dynamic and comprehensive teaching quality monitoring system can be established, accurately identifying weak links in the teaching process and improving teaching quality in a targeted manner to meet the needs of higher education development in the new era. The era of big data requires more refined capture of dynamic data from all aspects, as well as the complex relationships between data. This will help teaching decision makers gain a clearer understanding of the current situation, effectively regulate factors affecting teaching quality, and ultimately achieve the goal of improving teaching quality.

[0003] When improving teaching quality across university disciplines, some institutions have established internal teaching quality monitoring systems and leveraged feedback to make quality improvement decisions. However, significant deficiencies remain in the processing of teaching monitoring data. Currently, most universities still rely on traditional manual methods to collect, aggregate, and analyze teaching evaluation data. This approach is not only labor-intensive, cumbersome, and lacks specificity, but is also prone to errors. The long data processing time also prevents timely feedback from teaching quality evaluations, making it difficult for teaching quality monitoring to effectively guide the actual teaching process, and thus limiting the effectiveness of teaching quality improvement. To address these issues, there is an urgent need to explore more scientific and efficient methods for improving teaching quality. Summary of the Invention

[0004] The purpose of the present invention is to provide a teaching quality improvement method based on the correlation analysis between teaching evaluation and subject performance. By analyzing the correlation between teaching evaluation and subject performance, the k-means clustering algorithm is used to automatically group teachers with similar teaching characteristics to improve the efficiency and accuracy of teaching quality analysis, thereby improving teaching quality.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] The teaching quality improvement method based on the correlation analysis between teaching evaluation and subject performance includes the following steps:

[0007] Step S1: collecting teacher teaching evaluation data and corresponding student subject performance data, and preprocessing the data, wherein the preprocessing includes standardizing the data to generate a standard Z score;

[0008] Step S2: Based on the teaching evaluation and subject performance data processed in step S1, the Pearson correlation coefficient analysis method is used to calculate the correlation coefficients between the teaching evaluation index scores of different dimensions and the subject performance, and screen out the indicators with significant correlation;

[0009] Step S3: construct a feature space based on the indicators screened in step S2, use the k-means clustering algorithm to group teachers according to the characteristics of the teaching evaluation indicators, and classify teachers with similar teaching evaluation characteristics into the same cluster;

[0010] Step S4: Based on the correlation analysis and clustering results, analyze the differences in the teaching evaluation index scores of different teachers horizontally, analyze the changing trends of the teachers' teaching evaluation index scores over multiple semesters vertically, and formulate teaching quality improvement plans for individual teachers and groups.

[0011] Furthermore, in step S1, the teaching evaluation data is collected to collect relevant data from multiple semesters of the college, and the teaching evaluation data includes indicators of the following dimensions: course arrangement and teaching progress, teaching content and teaching methods, classroom management and organization of teaching, teaching communication and homework assignments, teaching effectiveness and overall evaluation.

[0012] Furthermore, the preprocessing includes deleting data entries in the data table whose missing values exceed a preset threshold, and normalizing the remaining original data to generate a standard Z score.

[0013] Furthermore, the conversion formula for generating the standard Z score through standardization is: Where Z is the standard Z score, X is the original data value, is the total mean score of all analyzed data, σ X is the population standard deviation.

[0014] Furthermore, step S2 specifically includes the following process:

[0015] S21. Express the teaching evaluation data vector group after Z transformation as Where k represents the number of teaching evaluation indicators to be analyzed;

[0016] S22. Select the fitting vector that needs to be analyzed by Pearson correlation coefficient from the above vector group. in It represents the scores of m teachers on the teaching evaluation indicator, and determines the corresponding subject score sequence y={y1,y2,...yn}, where n represents the number of subjects;

[0017] S23. Use the Pearson correlation coefficient analysis method to calculate the correlation coefficient r between the teaching evaluation index score and the subject performance: In the formula, β is the teaching evaluation index score sequence, y is the subject score sequence, represents the score of the i-th teacher among m teachers on the teaching evaluation index, is the average value of the teaching evaluation index score sequence β, y j represents the student's score in the jth subject among n subjects, is the average value of the subject score sequence y; when r>0, β and y sequence are positively correlated, and when r<0, β and y sequence are negatively correlated;

[0018] S24. According to steps S22 and S23, the correlation coefficients between the scores of different teaching evaluation indicators and the subject grades are obtained in sequence.

[0019] Furthermore, in step S2, indicators with significant correlation are screened out, and the screening condition is |r|≥0.4.

[0020] Furthermore, the implementation process of the k-means clustering algorithm in step S3 includes:

[0021] S31. According to the analysis results of step S2, select teaching evaluation indicators and construct a feature space for k-means clustering;

[0022] S32. In the constructed feature space, randomly select T initial cluster centers, denoted as u1, u2, ..., u T ;

[0023] S32. Calculate the Euclidean distance between each teacher's feature vector and each centroid According to the calculated distance, each teacher is assigned to the cluster with the nearest centroid. Where x i represents the characteristic vector of the i-th teacher, u t represents the t-th cluster center, t=1,2,...T, S represents the number of characteristic dimensions of teaching evaluation indicators owned by each teacher, s represents the index of the characteristic dimension, x is represents the value of the i-th teacher on the s-th feature dimension, u ts represents the value of the t-th cluster center on the s-th feature dimension, c i Represents the data point x i The cluster in which it is located;

[0024] S33. For each cluster, recalculate its centroid. The calculation formula is: Where C t represents the set of data points in cluster t, |C t | is the number of data points in cluster t;

[0025] S34. Repeat steps S32 to S33 until the centroid position is stable or the maximum number of iterations is reached.

[0026] Furthermore, in step S4, the horizontal analysis compares and analyzes the scores of teachers on various teaching evaluation indicators based on the semester dimension, presenting the differences in teachers' scores on different teaching evaluation indicators in different semesters; the vertical analysis tracks the changing trends of teachers' teaching evaluation indicator scores over multiple semesters, presenting the fluctuations in teachers' teaching levels in each semester; based on the results of the horizontal and vertical analyses, teaching management personnel formulate targeted teaching suggestions and improvement measures for teachers in the same cluster to improve teaching quality.

[0027] The beneficial effects of the above scheme are:

[0028] (1) The present invention improves the scientificity and reliability of the analysis through data standardization, providing a solid foundation for subsequent accurate analysis. The original teaching evaluation data is standardized into Z scores, making the teaching evaluation results of different semesters and different courses comparable, and their relative positions in the overall distribution can be clearly understood, eliminating errors caused by different evaluation scales, and making the comparison more objective.

[0029] (2) The present invention uses the Pearson coefficient analysis method to explore the correlation between teaching evaluation and subject performance. By calculating the correlation coefficients between teaching evaluation indicators of different dimensions and subject performance, it can accurately find out the teaching factors that have a greater impact on performance. If a certain dimension is found to have a high positive correlation with subject performance, it indicates that strengthening teaching in this aspect plays an important role in improving performance. It helps to guide teaching management personnel and teachers to clarify the focus of teaching improvement, reasonably allocate teaching resources, improve the teaching input-output ratio, and improve teaching quality in a targeted manner.

[0030] (3) The present invention uses the k-means clustering method to automatically group teachers and achieve personalized development of teachers. After classifying teachers according to the similarity of teaching evaluation, personalized teaching improvement plans can be formulated according to the characteristics of different categories of teachers. For example, for a group of teachers with low evaluation in a certain teaching dimension, training on optimizing course arrangements is provided. This avoids the "one-size-fits-all" management model, and enables teaching management to shift from "extensive" to "precise", improves the utilization rate of training resources, meets the differentiated development needs of teachers, promotes the professional growth of teachers, and thus improves the overall teaching quality.

[0031] (4) The present invention organically combines data preprocessing, correlation analysis, and cluster analysis to form a complete set of teaching quality improvement methodologies. From data collection and processing to relationship mining and then to teacher classification management, each link is closely linked. This multi-method integration and innovation comprehensively and systematically solves the problem of improving teaching quality, provides scientific and efficient management methods for education and teaching, and promotes the transformation of the education and teaching model from experience-based to data-driven. It upgrades traditional experience-based fuzzy management to data-driven precise decision-making, reduces management costs, and improves decision-making efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is the correlation coefficient diagram between different teaching evaluation indicators and subject scores;

[0033] Figure 2 This is a cluster comparison analysis chart of teacher evaluation results in each semester. DETAILED DESCRIPTION

[0034] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] It should be noted that unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0036] The teaching quality improvement method based on the correlation analysis between teaching evaluation and subject performance includes the following steps:

[0037] Step S1: collecting teacher teaching evaluation data and corresponding student subject performance data, and preprocessing the data, wherein the preprocessing includes standardizing the data to generate a standard Z score;

[0038] Step S2: Based on the teaching evaluation and subject performance data processed in step S1, the Pearson correlation coefficient analysis method is used to calculate the correlation coefficients between the teaching evaluation index scores of different dimensions and the subject performance, and screen out the indicators with significant correlation;

[0039] Step S3: construct a feature space based on the indicators screened in step S2, use the k-means clustering algorithm to group teachers according to the characteristics of the teaching evaluation indicators, and classify teachers with similar teaching evaluation characteristics into the same cluster;

[0040] Step S4: Based on the correlation analysis and clustering results, analyze the differences in the teaching evaluation index scores of different teachers horizontally, analyze the changing trends of the teachers' teaching evaluation index scores over multiple semesters vertically, and formulate teaching quality improvement plans for individual teachers and groups.

[0041] The following is a detailed description of the implementation process of each step:

[0042] Step S1: Collect teaching evaluation data and student subject scores to obtain original sample data and pre-process them.

[0043] The teachers' teaching evaluation scores and the subject scores of the students participating in the evaluation were collected, and the data entries with too many missing values in the data table were deleted. In order to make the teaching evaluation results comparable and facilitate the observation of the relative position of the teaching evaluation scores in the overall distribution, the original data were standardized to generate standard Z scores.

[0044] In this example, we first collected the college's teaching evaluation data and corresponding student subject performance data for five semesters, and performed a preliminary data cleansing process to delete duplicate records and remove invalid data. The teaching evaluation data includes indicators in the following dimensions: course arrangement and teaching progress, teaching content and teaching methods, classroom management and teaching organization, teaching communication and homework assignments, and teaching effectiveness and overall evaluation. As shown in Table 1: Table 1 Description of teaching evaluation data fields

[0045] Then, delete the data entries in the data table with too high missing value ratio. In this embodiment, the missing value ratio threshold is set to 20%-30%, and the data entries in the data table with missing values exceeding this threshold are deleted. The missing value ratio threshold is preferably 25%. Of course, the missing value ratio threshold can also be dynamically adjusted according to the sample size. If the sample size is large, such as the sample size is greater than 800, it can be set to 20% to retain higher quality data; if the sample size is small, such as the sample size is less than 400, it can be set to 30% to avoid a sharp drop in the sample size. For example, if a teacher's teaching evaluation data table contains 5 teaching evaluation indicators, if 2 of the indicators are missing, the missing ratio is 40%, which exceeds the threshold, and the record is deleted.

[0046] Next, the remaining raw data is used to calculate the mean and standard deviation, and then standardized to generate the standard Z score. The conversion formula is:

[0047] Where Z is the standard Z score, X is the original data value, is the total mean score of all analyzed data, σ X is the population standard deviation. The standardized Z score has a mean of 0 and a standard deviation of 1. The positive and negative values indicate the direction of the original data's deviation from the mean.

[0048] The output results are then verified and exception handling is performed to ensure there are no significant skew or kurtosis anomalies. Finally, data storage and output are completed, with the standardized Z scores stored as a structured data table with fields such as teacher ID, Z scores for each indicator, and Z scores for subject performance. This allows for direct access to subsequent steps (correlation analysis and clustering) to ensure data consistency and traceability.

[0049] Step S2: Correlation analysis.

[0050] S21. Express the teaching evaluation data vector group after Z transformation as Where k represents the number of teaching evaluation indicators to be analyzed; each fitting vector to be analyzed Contains the scores of m teachers on this teaching evaluation indicator.

[0051] S22. Select the fitting vector that needs to be analyzed by Pearson correlation coefficient from the above vector group. in It represents the scores of m teachers on the teaching evaluation indicator, and determines the corresponding subject score sequence y={y1,y2,...y n}, where n represents the number of subjects;

[0052] S23. Use the Pearson correlation coefficient analysis method to calculate the correlation coefficient r between the teaching evaluation index score and the subject performance:

[0053] In the formula, β is the teaching evaluation index score sequence, y is the subject score sequence, represents the score of the i-th teacher among m teachers on the teaching evaluation index, is the average value of the teaching evaluation index score sequence β, y j represents the student's score in the jth subject among n subjects, is the average value of the subject score sequence y; when r>0, β and y sequence are positively correlated, and when r<0, β and y sequence are negatively correlated;

[0054] S24. According to steps S22 and S23, the correlation coefficients between the scores of different teaching evaluation indicators and subject grades are obtained in turn, such as Figure 1 shown.

[0055] Based on statistical standards and educational scenario requirements, the screening condition for significant indicators is |r| ≥ 0.4. The screening process is as follows: all calculated teaching evaluation indicators are ranked by the absolute value of the correlation coefficient, indicators with |r| ≥ 0.4 are retained, and indicators with weak correlation, such as |r| < 0.4, are eliminated.

[0056] Because the teaching process is influenced by a variety of factors, moderately correlated indicators are sometimes worth exploring in depth when analyzing the correlation between teaching evaluation and subject performance. This is because multiple factors work together in teaching, and while a single factor may not have a strong correlation, the combined effect may have a significant impact on teaching quality. Therefore, the correlation coefficient screening criteria should not be too strict.

[0057] During specific implementation, the correlation coefficient screening conditions should be adjusted based on factors such as the research purpose, characteristics of the teaching field, input resources, and sample size. For example, if the focus is on strongly correlated indicators, the screening condition can be set to ∣r∣≥0.6, and the teaching factors that play a key role in subject performance can be identified and analyzed in depth to optimize teaching. If one wants to fully understand the relationship between all aspects of teaching and performance and explore potential influencing factors, the standards can be appropriately relaxed, such as setting the screening condition to ∣r∣≥0.3.

[0058] This method analyzes the selected indicators with significant correlations to understand which teaching evaluation indicators are positively correlated with subject performance and which are negatively correlated. For example, if the "teaching method diversity" indicator is positively and significantly correlated with subject performance, then further promotion of diverse teaching methods can be considered. Based on the analysis results, targeted suggestions and decision-making basis are provided for subsequent teaching improvements and teaching quality enhancement.

[0059] Step S3: Grouping is performed using the k-means clustering algorithm to construct a case library.

[0060] S31. Construct a feature space: Based on the analysis results from step S2, select teaching evaluation indicators that are significantly correlated with subject performance, such as |r| ≥ 0.4. This filter selects two indicators: course schedule and teaching progress, and teaching content and teaching methods. Alternatively, the criteria can be relaxed to select three indicators. Then, construct a k-means clustering feature space, representing each teacher as a multidimensional feature vector. The dimensionality of the feature space is the number of selected indicators.

[0061] S32. Determine the number of clusters and initialize the centroid: In the constructed feature space, randomly select T initial cluster centers, denoted as u1, u2, ..., u T ; If the school wants to divide teachers into three groups: "excellent, good, and average", then T=3.

[0062] S32. Assign teachers to the nearest centroid: Calculate the Euclidean distance between each teacher's feature vector and each centroid. The calculation process is as follows:

[0063] Where x i represents the characteristic vector of the i-th teacher, u trepresents the t-th cluster center, t=1,2,...T, S represents the number of characteristic dimensions of teaching evaluation indicators owned by each teacher, s represents the index of the characteristic dimension, x is represents the value of the i-th teacher on the s-th feature dimension, u ts Represents the value of the t-th cluster center on the s-th feature dimension;

[0064] According to the calculated distance, each teacher is assigned to the cluster with the nearest centroid, as follows:

[0065] Where c i Represents the data point x i If the distances between teacher 1 and the three centroids are 1.2, 2.5, and 0.9 respectively, then teacher 1 is assigned to the cluster where centroid 3 is located.

[0066] S33. Update the centroid position: For each cluster, recalculate its centroid and calculate the mean of all teacher feature vectors in the cluster as the new centroid. The calculation formula is:

[0067] Where C t represents the set of data points in cluster t, |C t | is the number of data points in cluster t;

[0068] For example, if cluster 1 contains teacher 1 (x1 = [1.2, -0.5]) and teacher 2 (x2 = [0.8, -0.3]), the new centroid is: u1 = [1.0, -0.4].

[0069] S34, iterative optimization: Repeat steps S32 to S33 until the centroid position stabilizes or the maximum number of iterations is reached. If the change in the distance between the new centroid and the old centroid is less than a threshold (e.g., <0.001), the centroid position is stable. To prevent infinite loops, the maximum number of iterations is set to 100.

[0070] After the iterative optimization is completed, each teacher is assigned a label of the cluster to which he belongs, such as "excellent, good, or average", and the automatic grouping of all teachers is completed.

[0071] The specific process for building a case library is as follows: Based on the teacher information within each cluster (including basic information, teaching evaluation index scores, subject performance, etc.), summarize the common characteristics and teaching features of the teachers within that cluster. For example, teachers within a cluster generally score high on "teaching content and teaching methods" and low on "classroom management and teaching organization." Then, based on the cluster characteristics and teaching features, representative teaching cases are generated. These cases can include the successful experiences of the teachers in the cluster, existing problems, and possible improvement measures. All cases are stored in a database or file system to form a case library, which is convenient for teaching management personnel and teachers to query and reference.

[0072] Step S4: comparative analysis.

[0073] The horizontal analysis compares and analyzes teachers' scores on various teaching evaluation indicators based on the semester dimension, presenting the differences in teachers' scores on different teaching evaluation indicators in different semesters; the vertical analysis tracks the changing trends of teachers' teaching evaluation indicator scores over multiple semesters, presenting the fluctuations in teachers' teaching levels in each semester; based on the results of the horizontal and vertical analyses, teaching management personnel formulate targeted teaching suggestions and improvement measures for teachers in the same cluster to improve teaching quality.

[0074] The horizontal analysis data consists of the Z scores and cluster grouping labels for all teachers' teaching evaluation indicators within the same semester. The vertical analysis data consists of historical teaching evaluation data for teachers over multiple semesters. The data must be normalized to Z scores to ensure comparability. The horizontal analysis data table includes: teacher ID, cluster label, Z scores for each indicator, and Z scores for subject performance; the vertical analysis data table includes: teacher ID, semester number, Z scores for each indicator, and Z scores for subject performance.

[0075] The purpose of horizontal analysis is to identify teachers’ relative strengths and weaknesses, e.g. Figure 2 The output visualization is shown in the figure. Longitudinal analysis is used to assess the long-term trends in a teacher's teaching ability. Specifically, the Z scores of the same teacher's teaching evaluation indicators over the past three to five semesters can be extracted and the changes in these indicators over each semester can be displayed using a line chart.

[0076] Then, teaching quality improvement plans are developed for individual teachers or groups. For individual teachers, improvement focuses on low-scoring indicators based on horizontal analysis. Specific measures include recommending training courses, providing relevant case studies of high-scoring teachers in the same group, and setting monthly improvement targets. For teachers within the same cluster, teaching management personnel design improvement strategies based on cluster characteristics. For example, if Cluster A has weak classroom management, regular "classroom interaction workshops" or "classroom management training" will be organized. If Cluster B has insufficient homework assignments, a tiered homework template library will be developed for weekly use. Dedicated mentors or teaching resource packages will also be assigned to different groups.

[0077] At the end of each semester, steps S1 to S4 are re-executed to update the clustering grouping and improvement plan. During this process, the changes in Z scores before and after the improvement are compared to calculate the improvement in teaching quality. The results of data analysis are converted into feasible teaching improvement actions, and personalized plans are tailored to the specific needs of different individuals and groups, abandoning the previous "one-size-fits-all" extensive management model. Quantitative indicators are used to replace traditional subjective evaluation methods to greatly improve the scientific nature and accuracy of teaching improvement plans. Through a periodic re-evaluation mechanism, changes in teaching quality are continuously tracked and evaluated to ensure the long-term effectiveness of teaching quality improvement work.

[0078] Finally, it should be noted that the parts of the present invention that are not described in detail are all prior art. Those skilled in the art will understand that the above description is only a preferred embodiment of the invention and is not intended to limit the invention. Although the invention has been described in detail with reference to the above examples, those skilled in the art can still modify the technical solutions described in the above examples or replace some of the technical features therein with equivalents. Any modifications, equivalent replacements, etc. made within the spirit and principles of the invention should be included in the scope of protection of the invention.

Claims

1. A teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance, characterized by: The following steps are involved: Step S1: collecting teacher teaching evaluation data and corresponding student subject performance data, and preprocessing the data, wherein the preprocessing includes standardizing the data to generate a standard Z score; Step S2: Based on the teaching evaluation and subject performance data processed in step S1, the Pearson correlation coefficient analysis method is used to calculate the correlation coefficients between the teaching evaluation index scores of different dimensions and the subject performance, and screen out the indicators with significant correlation; Step S3: construct a feature space based on the indicators screened in step S2, use the k-means clustering algorithm to group teachers according to the characteristics of the teaching evaluation indicators, and classify teachers with similar teaching evaluation characteristics into the same cluster; Step S4: Based on the correlation analysis and clustering results, analyze the differences in the teaching evaluation index scores of different teachers horizontally, analyze the changing trends of the teachers' teaching evaluation index scores over multiple semesters vertically, and formulate teaching quality improvement plans for individual teachers and groups.

2. The teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance according to claim 1 is characterized in that: In step S1, the teaching evaluation data is collected to collect relevant data from multiple semesters of the college. The teaching evaluation data includes indicators in the following dimensions: course arrangement and teaching progress, teaching content and teaching methods, classroom management and teaching organization, teaching communication and homework assignments, teaching effectiveness and overall evaluation.

3. The teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance according to claim 1 is characterized in that: In step S1, the preprocessing includes deleting data entries in the data table whose missing values exceed a preset threshold, and normalizing the remaining original data to generate a standard Z score.

4. The teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance according to claim 3 is characterized in that: In step S1, the conversion formula for generating the standard Z score by standardization is: Where Z is the standard Z score, X is the original data value, is the total mean score of all analyzed data, σ X is the population standard deviation.

5. The teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance according to claim 4 is characterized in that: Step S2 specifically includes the following process: S21. Express the teaching evaluation data vector group after Z transformation as Where k represents the number of teaching evaluation indicators to be analyzed; S22. Select the fitting vector that needs to be analyzed by Pearson correlation coefficient from the above vector group. l=1,2,...k where It represents the scores of m teachers on the teaching evaluation indicator, and determines the corresponding subject score sequence y={y1,y2,...y n }, where n represents the number of subjects; S23. Use the Pearson correlation coefficient analysis method to calculate the correlation coefficient r between the teaching evaluation index score and the subject performance: In the formula, β is the teaching evaluation index score sequence, y is the subject score sequence, represents the score of the i-th teacher among m teachers on the teaching evaluation index, is the average value of the teaching evaluation index score sequence β, y j represents the student's score in the jth subject among n subjects, is the average value of the subject score series y; When r>0, β and y series are positively correlated, and when r<0, β and y series are negatively correlated; S24. According to steps S22 and S23, the correlation coefficients between the scores of different teaching evaluation indicators and the subject grades are obtained in sequence.

6. The teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance according to claim 5 is characterized in that: In step S2, indicators with significant correlation are screened out, and the screening condition is |r| ≥ 0.

4.

7. The teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance according to claim 1 is characterized in that: The implementation process of the k-means clustering algorithm in step S3 includes: S31. According to the analysis results of step S2, select teaching evaluation indicators and construct a feature space for k-means clustering; S32. In the constructed feature space, randomly select T initial cluster centers, denoted as u1, u2, ..., u T ; S32. Calculate the Euclidean distance between each teacher's feature vector and each centroid According to the calculated distance, each teacher is assigned to the cluster with the nearest centroid. Where x i represents the characteristic vector of the i-th teacher, u t represents the t-th cluster center, t=1,2,...T, S represents the number of characteristic dimensions of teaching evaluation indicators owned by each teacher, s represents the index of the characteristic dimension, x is represents the value of the i-th teacher on the s-th feature dimension, u ts represents the value of the t-th cluster center on the s-th feature dimension, c i Represents the data point x i The cluster in which it is located; S33. For each cluster, recalculate its centroid. The calculation formula is: Where C t represents the set of data points in cluster t, |C t | is the number of data points in cluster t; S34. Repeat steps S32 to S33 until the centroid position is stable or the maximum number of iterations is reached.

8. The teaching quality improvement method based on correlation analysis between teaching evaluation and subject performance according to claim 1 is characterized in that: In step S4, the horizontal analysis compares the scores of teachers on various teaching evaluation indicators by semester, presenting the differences in scores of teachers on different teaching evaluation indicators in different semesters; the vertical analysis tracks the trend of changes in teachers' teaching evaluation indicator scores over multiple semesters, presenting the fluctuations in teachers' teaching levels in each semester; Based on the results of horizontal and vertical analysis, teaching management personnel formulate targeted teaching suggestions and improvement measures for teachers in the same cluster to improve teaching quality.

Images