Проблема оценки эквивалентности компьютерных и традиционных версий психодиагностических методик

Abstract

Introduction. This paper (a) discusses the theoretical and methodological evidence for the equivalence of computerized and conventional versions of psychological tests, (b) analyzes the studies investigating psychometric parameters of computerized versions of conventional tests, and (c) examines contradictions in approaches to assessing the equivalence of the two test forms. This paper represents a first effort in structuring the main problems in establishing the equivalence of paper-and-pencil and computer-based testing, as well as in finding ways and means to overcome them. Much attention is devoted to minimal and sufficient mathematical and statistical tools for assessing the equivalence of the two test forms. Theoretical Basis. The main problems associated with assessing the equivalence of computerized and conventional tests include the following: (a) the level of cultural and informational competence of respondents, (b) anxiety, (c) social environment, (d) motivation for testing, (e) difficulties in creating the same conditions for paper-and-pencil and computer-based testing (L. N. Babanin, Y. P. Chua, M. Russell, P. Květon et al.). Researchers most often choose the following procedures for assessing the equivalence of the two versions of psychological tests: (a) comparison of the main statistical parameters (mean values, variances, etc.) and (b) assessment of construct validity and reliability of the computer-based versions of psychological tests. Results and Discussion. The analysis of research methodology for assessing the equivalence of computerized and conventional versions of psychological tests focuses on a variety of approaches to (a) the use of mathematical and statistical methods for assessing psychometric parameters of computer-based versions of conventional tests, (b) the choice of research design, (c) considering specific characteristics of the situation of computer-based testing. It is necessary to formulate specific and structured requirements for the procedure for assessing the equivalence of computerized and conventional versions of psychological tests. The author suggests recommendations for the main sections of such requirements related to (a) the procedure for conducting empirical research, (b) mathematical and statistical methods, and (c) control of the factors specific to computer-based testing that may have impact on the results of equivalence assessment.