Abstract
Changes in prevalence can induce changes in conceptual boundaries (Levari et al., 2018). For instance, if participants see fewer and fewer blue colored dots, their conceptual boundaries for what counts as blue expands. Effectively this means that more non-blue dots are categorized as being blue. While this change in conceptual boundaries is remarkably robust and generalizes to a number of different stimuli, little is known about how the decision unfolds within a trial and across trials. One possibility is that prevalence shifts the criteria for what counts as blue but the decision process remains the same. Another possibility is that prevalence changes the decision process itself. In order to differentiate between these two possibilities, we presented participants with a spectrum of blue and purple dots and used computer mouse-tracking to examine the dynamics of their responses and how they may (or may not) change over time. In the experimental condition, the prevalence of blue dots steadily decreased over time. In the control condition, the prevalence of blue dots remained stable over time. Replicating the original finding, we found that the criteria for what counts as blue shifted over time for participants in the decreased prevalence condition. Importantly, we also found a trend in the mouse-tracking trajectories, b = 152.2, SE = 92.2, t = 1.654, p = 0.098. Visual inspection of the trajectories suggests that ambiguous colors near the middle of the blue-purple spectrum were more confidently categorized as blue in the decreasing condition than in the stable condition. This suggests that prevalence does not just change our conceptual criteria, it also changes the process by which we evaluate stimuli.