МОДЕЛЬНЕ ПОДАННЯ ТЕХНОЛОГІЇ АСОЦІАТИВНОГО ТЕСТУВАННЯ ДЛЯ ОЦІНКИ КОГНІТИВНИХ КОМПЕТЕНТНОСТЕЙ В ОСВІТІ
Radioelectronic and Computer Systems · 12.08.2026 · 2026 (2) · 173-185
DOI: 10.32620/reks.2026.2.12 ↗Kurz gefasst
Fachartikel in der Zeitschrift Radioelectronic and Computer Systems, online erschienen am 12.08.2026. Verfasst von Ihor Shevchenko und 2 weiteren Autor:innen.
Abstract (Originalsprache)
The subject of the article is the development of a methodology for designing and utilizing Word Association Tests (WAT) to assess competencies within academic disciplines. The goal of this study is to establish the theoretical and technological foundations of an information system for associative testing, aimed at enabling a more structurally informed and diagnostically sensitive evaluation of the theoretical competence of learners. These tests are constructed on the basis of associative fields related to specific aspects of a discipline, allowing for the assessment of the structure and quality of cognitive representations in a learner’s mind. This study focuses on developing a comprehensive set of models and methods that underpin the technology of associative test construction and defines the methodological mechanisms through which such tests enhance the analytical capabilities of theoretical knowledge assessment. The proposed tests enable the evaluation of learners' depth of understanding of specific theoretical components of a discipline. The methods incorporate an approach for identifying underlying patterns and hypotheses relevant to the construction of WATs tailored to particular thematic areas. A systematic analysis is performed to explore the local cognitive features of learners and to generate micro-thesauri corresponding to individual disciplinary aspects. The frame theory is employed as a structural foundation for modeling the testing procedure. The following key results were obtained: (1) a conceptual model of the associative testing domain was developed to systematize all phases of the testing process; (2) an ontology model was created to integrate semantic elements of test questions into a coherent system; (3) a micro-thesaurus model was designed to represent specific disciplinary aspects and support the generation of associative test matrices across various academic fields; (4) a novel method for assessing associative test performance was proposed, leveraging an additional feature space defined by auto-association map coefficients, providing a formal basis for more fine-grained analytical evaluation of associative relations; (5) a frame-based model of the associative testing procedure was developed, offering a structured framework for implementation and facilitating software development, allowing software developers to quickly master the subject area; and (6) a general methodology for test construction was described, requiring minimal web resources. The proposed approach demonstrates a measurable improvement in assessment adequacy, as evidenced by a reduced discrepancy between automated test results and expert evaluation. In particular, WAT-based assessment approximates expert judgment more closely than conventional testing methods, indicating reduced assessment bias, higher diagnostic precision, and greater ability to accurately capture the learner’s cognitive structure. Conclusions: The scientific novelty of this research lies in the formalization and integration of a comprehensive methodological framework for associative testing that is applicable across educational and research contexts. This framework includes a system-level conceptual model, an ontology model that organizes semantic elements into a unified structure, and a micro-thesaurus model for a specific discipline-specific aspect that reflects the learner’s mental map and enables the generation of associative test matrices. The proposed method for evaluating associative responses differs from existing approaches in that it introduces a new metric in the form of auto-association map coefficients, providing a formalized representation of associative alignment rather than an empirical performance indicator. This enables a more precise structural analysis of the learner’s associative connections within a thematic micro-thesaurus compared to the instructor’s thesaurus. The frame-based model further supports structured implementation, and the overall methodology for determining cognitive structures (mental maps) has potential applications in psychology and sociology. The results confirm that the proposed approach improves the reliability and validity of competence assessment by more closely aligning automated evaluation with expert judgment.
Zitieren
APA
Shevchenko, I., Poddubei, O., & Amosov, V. (2026). МОДЕЛЬНЕ ПОДАННЯ ТЕХНОЛОГІЇ АСОЦІАТИВНОГО ТЕСТУВАННЯ ДЛЯ ОЦІНКИ КОГНІТИВНИХ КОМПЕТЕНТНОСТЕЙ В ОСВІТІ. Radioelectronic and Computer Systems, 2026(2), 173-185. https://doi.org/10.32620/reks.2026.2.12
Vancouver
Shevchenko I, Poddubei O, Amosov V. МОДЕЛЬНЕ ПОДАННЯ ТЕХНОЛОГІЇ АСОЦІАТИВНОГО ТЕСТУВАННЯ ДЛЯ ОЦІНКИ КОГНІТИВНИХ КОМПЕТЕНТНОСТЕЙ В ОСВІТІ. Radioelectronic and Computer Systems. 2026;2026(2):173-185. doi:10.32620/reks.2026.2.12
BibTeX
@article{Shevchenko2026,
author = {Shevchenko, Ihor and Poddubei, Olena and Amosov, Vladyslav},
title = {МОДЕЛЬНЕ ПОДАННЯ ТЕХНОЛОГІЇ АСОЦІАТИВНОГО ТЕСТУВАННЯ ДЛЯ ОЦІНКИ КОГНІТИВНИХ КОМПЕТЕНТНОСТЕЙ В ОСВІТІ},
journal = {Radioelectronic and Computer Systems},
year = {2026},
volume = {2026},
number = {2},
pages = {173--185},
doi = {10.32620/reks.2026.2.12}
}
EUVIMED