Entropy and the Observer: How Beliefs, Biases, and Social Pressure Influence the Choice of the Level of Description (Methodological article for philosophers, sociologists, and methodologists of science)

This paper continues the methodological analysis of entropy initiated in the first work of the series, where entropy was presented as a function of two factors: system properties (Ω) and the choice of the level of description (G), also called coarse‑graining. In contrast to the physics paper, where G was determined only by rational factors (the observer's task and available information), the present work investigates subjective and social factors influencing the choice of G. An extended definition G = G(Z, I, B, P, Cog, M, …) is introduced, where additional variables include biases, social pressure (including mainstream pressure and paradigmatic filters), cognitive limitations, and the researcher's motivation. Special attention is paid to the phenomenon of data fitting, where an observer discards "inconvenient" results, considering them measurement errors, in order to maintain a convenient point of view. This phenomenon is linked to confirmation bias. The paper draws on classic works in the philosophy and sociology of science (Kuhn, Hanson, Merton, Longino) and contemporary research on cognitive biases (Nickerson). The author emphasizes that recognizing the influence of subjective factors on the choice of G does not lead to relativism, but rather requires reflexivity in science — an awareness of these factors and an effort to minimize their distorting effect. The paper is addressed to methodologists of science, philosophers of science, sociologists of science, and all those interested in the influence of the human factor on the process of cognition. It invites interdisciplinary dialogue and offers a tool for analyzing how beliefs and social pressure shape "observed" reality.