A precise quantification of how prior experience informs current behavior

Human behavior does not exist in a bubble-it is influenced by countless forces, including each individual's current goals, preexisting cognitive biases, and prior experience. The current project leveraged a massive behavioral data set to provide a data-driven quantification of the relationship between prior experience and current behavior. Data from two different behavioral tasks (a categorization task and a visual search task) demonstrated that prior history had a precise, systematic, and meaningful influence on subsequent performance. Specifically, the greater the evidence for (or against) all aspects of the current trial, the more (or less) efficient behavior was on that trial. The robust influence of prior experience was present for even distracting and likely unattended information. The ubiquity and consistency of the effect for features both related and unrelated to stimulus presence suggests a domain-general mechanism that increases the efficiency of behavior in contexts that match prior experience. These findings are theoretically important for understanding behavioral adaptation, experimentally powerful for directly addressing effects of previous trials when designing and analyzing research projects, and potentially useful for optimizing behavior in various applied contexts. (PsycInfo Database Record (c) 2022 APA, all rights reserved).