Research of the factorial knowledge assessment paradigm efficiency illustrated by an example of testing students in the automatic control theory

An algorithm for constructing test materials and the test results evaluation techniques that allow combining high estimate rate and test objectivity are developed. The test system in this case can be considered as an assessment tool for the knowledge level under investigating the appropriate modules. The decomposition of knowledge into lots of sections is conducted for the knowledge systematization. At that, each section is associated to a set of sub-sections which, in turn, consists of a subset of questions, the answers to which reveal the level of knowledge of each section. The test assignment is formulated in such a way that upon answer, a set of responses revealing the knowledge level of a certain subsection is to be given. Since the test development process is considered as a construction of multivariable dependence, it is proposed to approach the problem from the perspective of planning multifactor experiments. In this case, a set of answers is presented as a set of factors taking on certain values. With their varying on two levels, the problem is reduced to the construction of the experiment that implements all possible combinations of factors, i.e. to a complete factorial experiment. The technique of applying the technology of two-level multifactor experiment planning that allows formalizing the evaluation of partially incomplete knowledge, and implementing the formation of the multifactor tests of different kinds of complexity is offered. To illustrate this technique, the examples of the construction of tests for the discipline “Automatic control theory” assessing the degree of the knowledge level formation are considered. The statistical analysis of the test results has shown the knowledge level of the testees, as well as the credibility of this knowledge.