Sequential choice effects predict prevalence-induced concept change

Prevalence-induced concept change is the process by which the definition of a concept can shift as the prevalence of that concept changes. In a series of beautiful experiments, Levari and colleagues (2018) demonstrated that prevalence-induced concept change occurs across a range of situations, from color perception to threat assessment to ethical judgment. Here we show that in all of these cases the extent of prevalence-induced concept change is accounted for by sequential effects in which previous stimuli and responses bias the upcoming categorization. Crucially, these sequential effects are always active, even when the prevalence is does not change. Thus, by modeling these effects, we were able to predict the magnitude of prevalence-induced concept change based on participants’ behavior when the prevalence is fixed. This suggests that sequential effects are the mechanism by which prevalence-induced concept change occurs.