Evaluation of teacher performances with fuzzy logic method

In this study, it is aimed to examine the use of the fuzzy logic approach in performance processes, to design a model that evaluates teacher performance with the fuzzy logic approach. By comparing the evaluation results which are obtained from the traditional (classical) method with the evaluation results obtained from the fuzzy logic method it is intended to reveal positive and negative aspects of both methods in terms of evaluation. The data of the survey includes the principals, teachers, and students who take active roles in high schools in the central districts of Van Province (İpekyolu, Edremit, and Tuşba) during the spring term of the 2016-2017 academic year. The study is conducted on 375 teachers, 42 principals and 1500 students on simple random sampling. The data are collected by applying the evaluation form created by the Ministry of National Education to determine teacher performance. In order to determine the teacher performance, a three-input and one-output system which is based on fuzzy logic are designed by using the Fuzzy Logic toolbox tool in MATLAB program. The data collected from teachers, principals, and students are determined as input variables and the final performance score of the teacher is determined as the output variable. 27 fuzzy logic rules are created by taking advantage of expert opinions. By using the Mamdani fuzzy inference model after the clarification process teacher performance value is obtained. Teacher performances are first evaluated by the classical method and then by fuzzy-based model. In the analysis of study data, the results are compared using descriptive statistics (standard deviation, mean) T-Test and Pearson Correlation. All analysis studies are carried out with the SPSS package program. According to the results, the average teacher performance scores obtained by the classical method is higher than the average of the scores obtained by the fuzzy method. It is found that there is a significant difference between the scores of the teachers' performance according to both methods and this difference is in favor of the classical method. In this study, by determining the correlation value that expresses the relation between the data obtained by fuzzy and classical methods, it is concluded that the relationship between the two methods is positive and high.