?
СПОСОБЫ СВЯЗЫВАНИЯ ШКАЛ ДЛЯ ИЗМЕРЕНИЯ ОБРАЗОВАТЕЛЬНОГО ПРОГРЕССА В РАЗНЫХ ПАРАДИГМАХ АНАЛИЗА ДАННЫХ ОБРАЗОВАТЕЛЬНОГО ТЕСТИРОВАНИЯ
Educational progress monitoring programs can provide valuable data to all users of test results, from test-takers to educational authorities. This is due to the enormous amount of psychometric work needed to establish comparability of scales from different measurement occasions. Comparability means that a single test score is assigned precisely to the same ability level. Then, it can be interpreted consistently across different measurement occasions. Contemporary psychometrics offers a vast list of different procedures allowing to establish comparability of scales. However, they only can be applied if the tests are developed using the same operationalization of the (changing) construct. This paper proposes a classification of methodological approaches for measuring educational progress. We introduce two types of programs for educational progress monitoring: (i) educational trends monitoring (that provide group-level information like NAEP) and (ii) individual educational progress monitoring (that provide individual-level information like Pupil Monitoring System). Then, we describe different approaches to conceptualizing educational progress — as a quantitative gain and as a qualitative transition to a new cognitive state. Finally, we review the most popular procedures for linking scales that provide a common interpretation against which educational progress can be measured within each of these approaches. Among special cases of quantitative progress understanding, we describe vertical equating and longitudinal item response theory modeling. Among methods based on understanding educational progress as quantitative transition, we describe the vertically moderated standard-setting, latent transition analysis, and cognitive diagnostic modeling.