Architecture of Assessment Systems in Higher Education: Multilevel Model and Diagnostic Criteria

Assessment systems in higher education are typically described through scales, points, and formal procedures. However, can one understand what function assessment actually performs – supporting learning or administrative control – by relying solely on the type of scale or tools? Why do formally similar systems in different universities lead to fundamentally different educational practices and student behavioral strategies? Assessment cannot be adequately analyzed through individual elements, since identical solutions can perform different functions depending on the mechanism of their formation, combination, and use. The purpose of the article is to develop an architectural analytical framework for designing and diagnosing assessment systems in higher education as a multilevel configuration of interrelated decisions. To achieve this, heterogeneous characteristics of assessment systems have been systematized as decision variants at individual architectural levels, typical system configurations for different institutional purposes have been modeled, and diagnostic criteria for analyzing the constructive coherence of a specific assessment system have been developed. The scientific novelty lies in the proposed architectural approach: the conceptual level defines the space of variants for the methodological level, which is modified by regulatory parameters and implemented through organizational forms. Typical assessment trajectories and diagnostic tools for evaluating their coherence are presented. The model operationalizes the principle of balance between pedagogically desirable, technically feasible, and administratively permissible, to design systems for mindful learning even under institutional constraints. The main conclusion: assessment systems are valid only under the condition of internal architectural coherence between levels.