Figure 3

Best-fit trajectories of mean preference strength generated from the Markov (left), quantum (middle), and hybrid open system (right) models for each of the experiments. Data from the experiments are presented as red + (choice condition) or blue \(\times\) (no-choice condition). For Experiment 2, the preference strength data are presented as a simple running average of responses within a \(\pm 3\) s window—this pattern does not change substantially if the window is made wider or narrower, until the window is so wide that it washes out cross-time variability in mean preference. Model predictions (blue/red lines) are computed for every second after the initial response through the last time point.