Complex dynamics of preference-based decision-making patterns in rats

Abstract

According to the ‘revealed preference’ and the ‘expected utility theory’, preference can be observed by choice and is invariant across context; thus individuals have been considered to choose the optimal option and make rational choices. However, recent evidence, which opposes this classical concept of preference, shows that choices do not always reflect preference, such as in the status quo bias, and that choices change depending on the description of the problem, leading to suboptimal choices and thereby resulting in irrationality. Herein, we investigated preference-based decision-making in rats. We observed choice behavior based on the rats’ own subjective preference between different flavored foods or different amounts of food by applying the framing effect. Thus, we monitored the macroscopic behavior of decision dynamics and preference changes when the alternatives were framed. In the first study, we measured choices among 4 different flavored pellets during a long-term period to mimic choice dynamics in real life. We analyzed temporal and sequential components in choice patterns with a distribution analysis to find the governing role within those patterns. We found the temporal and sequential choice components that respectively reflect the “wanting” and “liking” attributes of preference followed heavy-tailed distribution in which extreme observations were discovered far more frequently than in randomly arising systems. Temporal dynamics were modeled by cascading non-human Poisson processes and sequential dynamics were explained by dual evaluation processes consisting of goal-directed and habit systems. This study provides evidence for a universal bursting property of animal behavior in both timing and sequential choices. In the next study, we examined the role of the nucleus accumbens (NAc), known to participate in hedonic responses and reward process in choice dynamics. First, we damaged the NAc shell by electrolytic lesions, and next we downregulated...