Figure 3. Effects of instructions and learning on expected pain and pain ratings prior to reversal.

(A) Expectancy ratings prior to reversal. Participants in the Instructed Group (Top Left) expected higher pain in response to the Original High Cue relative to the Original Low Cue at baseline (left) and differences in expectations grew larger following conditioning and the first test phase (right). Participants in the Uninstructed Group did not report differences prior to the task (left), consistent with the fact that they were not instructed about specific cue-outcome contingencies. Following conditioning and the first test phase, Uninstructed Group participants expected higher pain in response to the Original High Cue, relative to the Original Low Cue. Cue-based differences in expectancy ratings were larger in the Instructed Group. (B) Predictive cue effects on pain prior to reversal. We measured the effects of predictive cues on perceived pain prior to the first reversal (see Table 3). Both groups reported higher pain when medium heat was preceded by the high pain cue (black) relative to the low pain cue (gray) and this effect was present in nearly all participants. (C) Cue effects increase over time. Both groups show larger cue-based differences in perceived pain on medium heat trials as a function of experience prior to the first reversal, but effects of time were larger in the Instructed Group. Data were visualized using the R toolboxes ggplot2 (Wickham, 2016) and Raincloud plots (Allen et al., 2021). Error bars and shaded areas denote standard error of the mean (n = 20 per group).