Bounded rationality: Fallible decisions in unbounded decision space

TLDR: In this paper, the authors present two related models of decision making by individual persons: the first model describes how nonrational mechanisms serve drastically to shrink problem space to a size enabling boundedly rational deciders to apply their limited, and highly fallible, rational/empirical strategies.

Abstract: This article concerns decision making by individual persons. Current theoretical and research challenges to classical decision theory provide the foundations for mapping the general structure of a revised theory of decision making. The main task facing the emerging theory is to make explicit provision for both nonrational and rational/ empirical sources of variance. This paper presents two related models. The first describes how nonrational mechanisms serve drastically to shrink problem space to a size enabling boundedly rational deciders to apply their limited, and highly fallible, rational/empirical strategies. The second model describes how confidence can be readily developed and maintained in fallible and invalid decisions made in typical problem domains At a metalevel of analysis the theory assumes a quasi-stable nested hierarchy, with the balance between equilibrium and stress at any given level related to the plasticity of its upper and lower boundary conditions. In the case of human systems boundary stability, within and between levels, is seen to be mainly a function of nonrational vectors (e.g., bias, confidence, fatigue, etc.), rather than cognitive vectors (e.g., information, rationality, probability, etc.). The article focuses on the individual decision maker functioning within an organizational system against the backdrop of historical and cultural constraints.