Institute of Education

Publications

• Book

  Sorokin P., Mironenko I. A.

  Human Agency and the Problem of Replicability in the Dynamic Neo-Structured World

  Social reality is undergoing rapid transformations, giving rise to new challenges for both social theory and practice. One of these much-discussed challenges is the so-called replicability crisis, a notion that describes the failure of the social sciences and humanities to verify established research findings with new empirical investigations. In the related debates, it is argued that this “crisis” can be interpreted as a sign of the growing disconnect between, on one side, traditional perspectives on human nature and social structures that still dominate academic mainstream and, on the other side, the reality of an emerging new phase of societal evolution—neo-structuration, which emphasizes individual agency as a driver of development—in contrast to the well-known “structuration” argument by Antony Giddens, whereby human agency is seen as ontologically fused with the social structure, procuring stability of the latter rather than its transformation. Bridging the gap requires an interdisciplinary paradigm shift. This shift would involve moving away from considering human beings as changeless or predictably evolving subjects whose intentions and actions are determined both internally and externally by structures that supposedly follow a linear trajectory of “progress,” assuming the reproduction of the basic principles and forms of social interaction. Instead, in the context of intensifying neo-structuration, individual agency emerges as both an expression of the core of human nature and a catalyst for societal change, including the pressing social issues within the techno-material environment, which have radically changed in the first decades of the 21st century, especially after the COVID-19 outbreak in 2020. This presents a fundamentally new challenge for the social sciences and humanities: to develop methodological approaches and theoretical models capable of systematically analyzing the human capacity for creation and transformation—not just reproduction.

  Under conditions of neo-structuration, individual agency causes transformations of both social structures and the agentic subjects themselves, which makes the crisis of replicability inevitable. The crisis itself is evidence of changes in social reality and human nature as the objects of research, not only the mere sign of the imperfection of measurements. Thus, the replicability crisis relates not only to the epistemology of science but also to the ontology of its subject matter.

  Most important is to give an adequate theoretical and methodological account for the so-called strong agency, capable of a significant transformative impact on social environment. A special challenge refers to understanding the power of human agency in the context of technological advancement, including the opportunities related to artificial intelligence. Distinguishing between two types of such agency is necessary: on one hand, a conditionally “free” creative individual initiative, leveraging digital infrastructures as tools for forging new communities, transforming existing behavioral patterns, practices, and so on, and, on the other hand, a sphere of a “hybrid” strong agency, partially stimulated and augmented by targeted, customized systems of artificial intelligence’s recommendations, suggestions, navigation services, consultancy, and advice. Within these intersecting realms, the foremost theoretical and practical challenges emerge—particularly regarding the reconceptualization of human agency as a pivotal driver of institutional change under the condition of neo-structuration, which puts the problem of replicability in a principally new light.

  Oxford Research Encyclopedia of Science, Technology, and Society, 2026.

• Article

  Kuzminov Y., Koshel A., Kruchinskaia E.

  Non-discriminatory storefront of digital platforms and ecosystems: Consumer experience and regulatory frameworks

  The expansion of digital platforms and ecosystems is reshaping consumption and the structure of modern markets. In the Russian context, a representative survey conducted by the authors indicates frequent online purchasing driven primarily by time savings, broader assortment and price accessibility. Platforms thus generate consumer benefits, including non-monetary gains that are not fully captured by conventional macroeconomic indicators. Addressing this measurement and regulatory problem, the paper integrates transaction cost economics with the concepts of the non-discriminatory storefront and digital consumer surplus. Empirically, we use a three-stage sequential econometric strategy. Model 1 examines access to paid ecosystem subscriptions; Models 2–8 examine heterogeneity in subscription motives among paid-subscription users; Model 9 examines diversification beyond the primary ecosystem through third-party subscriptions. The results indicate that settlement type is not associated with paid-subscription use once age, education, income and other covariates are controlled for, while income and education remain significant correlates. Among subscribers, cost savings are the dominant motive; higher income is associated with bonuses/cashback and convenience, whereas trust is not significantly associated with income. In Model 9, trust is negatively associated with frequent third-party subscriptions, a pattern consistent with trust-based ecosystem loyalty, while bonuses/cashback are positively associated with subscription diversification. On this basis, the paper derives four legal dimensions of non-discrimination concerning information disclosure, ranking and pricing, manipulative interface design, and self-preferencing. The policy implication is not that ecosystem scale is benign per se, but that antitrust assessment should evaluate market power together with consumer access, transaction-cost reductions, and discriminatory platform practices.

  Russian Journal of Economics. 2026. No. 12(3). P. 311-334.

• Book chapter

  Mozhaysky A., Pichugina V.

  ‘Seven-gated Thebes’: Epic Tradition, Topography, and Evidence of Material Culture

  The paper by Andrej Mozhajsky (Moscow) and Victoria Pichugina (Moscow) studies the topography of the ancient city of Thebes. Numerous stories about the mythological history of ancient Thebes were assembled in poetry and prose, e.g. about the bloody struggle for power among king Oedipus’ sons and the siege of the city that accompanied the struggle gained much popularity in Classical literature (Aeschylus, Euripides, Pausanias, Pseudo-Apollodorus, Statius, Hyginus, and others). Thanks to Aeschylus, who had turned to the Mycenaean topography and stated the presence of the city’s seven gates, Thebes appeared as a city with a specific spatial organization. Even though ancient authors list the names of these gates and attempt to locate them, the issue of whether the city had exactly seven gates still causes controversy among researchers. The epithet “7- gated” for Thebes, known since the creation of Homer’s Odyssey (11.260–5), presumably reflects the Kadmeia’s real fortifications, which remains date back to Mycenaean times. Unlike some other scholars, Mozhajsky and Pichugina argue that the Mycenaean Thebes indeed possessed seven gates and set out to prove this through the evidence of archaeological research, as well as the arguments considering the topography and the functions of the fortifications. Mozhajsky and Pichugina also find confirmation in some later literary sources that directly connect the fortification of the Kadmeia with the monuments of the Mycenaean period. For instance, Aeschylus (Sept. 526–528) associates the Kadmeia’s Northern Gate with the Amphion Hill, which, in their opinion, reflects the development of urban space in the Archaic period and the further transfer of ancient cultural meanings linked to topography into the classical period.

  In bk.: Epic, Society, and Religion in the Ancient Near Eastern and Mediterranean Cultures. Cambridge Scholars Publishing, 2026. Ch. 10. P. 229-261.

• Working paper

  Lyu M., Omopekunola M., Wang F.

  LLM Difficulty Prediction Across Three L2 Task Types and Two Examination Forms

  Large language models (LLMs) have been proposed as a substitute for field trials in itemdifficulty estimation, but evidence about where they succeed is usually drawn from a single test form. We compare LLM difficulty predictions across three task types that make different language demands, grammar cloze, semantic cloze and reading comprehension, on two operational Grade 9 English examinations sat by the same cohort in China (1,199 and1,166 examinees). Three LLMs estimated difficulty for 60 MCQ items under the same standardized protocol. On Form 1, grammar cloze had the highest Pearson correlation for all three models; the Fisher-averaged correlations were .81 for grammar cloze and .34 for semantic cloze. On Form 2, reading had the highest correlation for all three models (Fisheraveraged .78), and the grammar–semantic difference was smaller and inconsistent in direction. With ten items per task type, a task-type advantage seen on one form needs testing on other forms before it is read as a property of the model.

  Proceedings of the Artificial Intelligence in Measurement and Education Conference (AIME-Con). 2026.aimecon-wip.55. The Association for Computational Linguistics (ACL), 2026