摘要：Today, considerable attention is paid to the higher education quality issues. The problem is solved by using tests that should provide a reliable student evaluation. The article presents the technology for improving test tasks. It includes functional procedures that specify the test and test task improvement sequence. It is found that it is better to use specialized computer applications for their implementation, that is why this technology involves the use of the author program “Statistical Analysis of Test Results”. This program calculates the indicators – the item difficulty, discrimination, reliability and validity – according to empirical student testing data. The indicators help identify unsatisfactory quality test tasks and improve the student assessment means, as the program derives the recommendations. The steps set out by the testing result processing technology with the help of a statistical package increase the improvement process efficiency. The correlation and factor analyses help identify the tasks that put the highest load into the test score. These procedures influence on making a decision on the test task review need. The technology involves repeated checking procedures. The presented technology has been tested at Zaporizhzhya National University and Zaporizhzhya Regional Institute of Postgraduate Teacher Education. ANOVA has helped prove its effectiveness.
