Improve of the adequacy and reliability of no standardized processing model datasets in studies of educational systems method of experiment planning

Abstract. The article is devoted the questions of improvement of the accuracy of statistical mathematical models, used to investigate the effectiveness of the educational process. The authors suggest changing standard approaches to information processing large datasets, using the method of experiment planning. This takes into account features of data sets characteristic of pedagogical processes. Using these proposals one can improve the accuracy of the model and the value processing, and hence the accuracy of machining.