Assessing distinctiveness test space of a computerized testing system

Abstract

The concept of information education involves the use of new learning technologies and leads to the development of new forms of knowledge using modern computer training aids. However, nowadays the simple computerization and informatization are insufficient to ensure the quality and effectiveness of the educational process. Learning outcomes, defined, in particular, as the level of students training, are considered as some measurable value for assessment which should involve scientifically based methods, and the values are accompanied by an assessment of their accuracy. The quality of the learning measurement is directly determined by the quality of the tools that were involved in the study and evaluation of the training level. Nowadays, the computerized testing system (CTS) is considered to be the most effective and objective means of the learning assessment, as evidenced by their prevalence in various fields of learning validation. At the same time, only the CTS that was scientifically justified under the development and testing quality indicators can give objective characteristics of the students training levels. A lot of publications are devoted to the issues such as the selection list of indicators that should characterize the CTS quality, assessment methods, as well as their practical value. However, despite the diversity of opinions and approaches the general concept of these issues has not been created. According to the author computerized testing system quality is laid on the stages of project development and CTS strategic calculation and verified on the stage of a CTS pilot experiment and its monitoring. This control verifies the CTS quality score by criteria reliability and validity of the test objectives and test difficulty and distinctiveness test tasks - based on statistics - test results. The evaluation of these indicators should be made using the developed information-measuring system that implements the task of collecting statistical data, their processing, combination, check with the established norms, issuance of reasonable recommendations to improve CTS. The article deals with the existing and proposed by the author evaluating methods of one of the computerized test quality indicator - indicator of the distinctive character test space problems. The indicator of the distinctive character test and CTS space test tasks is particularly important by applying the legal-oriented tests used in the classroom. Using the criterion-oriented tests (in the final control) the figure is not always decisive. However, by creating the criterion-oriented tests distinctiveness test objectives should be taken into account at the design stage of the test space, because it determines the sensitivity of the testing procedure. Information about the importance of the distinctive character can increase the sensitivity of the testing procedures on a range of scales investigated the students properties (the level of students preparedness) by choosing the appropriate test tasks. Therefore, along with the test difficulty index the distinctiveness tasks can be used for the CTS testing in adaptive mode. The article examined, systematized and proposed methods for evaluating the distinctive character of CTS test space tasks, submitted recommendations on the choice of one or another metrics that improve the efficiency of information-measuring system for CTS quality control indicators.