Adaptive Testing Of Knowledge by Methods of Logical Networks

Abstract

Adaptive test control is a computerized system of scientifically based verification and evaluation of learning outcomes, which is highly effective by optimizing the procedures for generating, presenting and evaluating the results of adaptive tests, based on methods of building and optimizing logical networks. Algorithms for selection and presentation of tasks are based on the principle of feedback, when the correct answer of the subject of training is the next difficult task, and the wrong answer causes the presentation of the next easier task than that to which the subject of training the wrong answer was given. It is also possible to ask additional questions on topics that the subject does not know very well to clarify the level of knowledge in these areas. The choice of testing algorithms is currently actually limited by the forms of presentation of test tasks and algorithms for evaluating test results. Achieving higher results and increasing the motivation to learn is ultimately the main goal of testing knowledge. To determine the basic algorithm, it is necessary to provide a scenario of the system. It is based on the model of taking the exam by a teacher as a model of adaptive testing. This choice of the scenario of the system is due to the fact that, firstly, this procedure is historically well formalized, and secondly, when designing tests, their developer must rely on common, known and used methods with minimal modification.