Choosing to avoid: The evaluative impact of autonomous selection of approach and avoidance behaviors

Much of everyday learning affords us autonomy over the information we learn about, but evaluative learning research often strips participants of this autonomy. The present research assesses the effects of autonomy in the form of sampling decisions in the traditionally passive approach-avoidance (AA) training. In three experiments, participants completed an AA task, in which half of the participants were given genuine autonomy over how frequently they interacted with each stimulus, and the other half were yoked to engage in the AA behaviors of a high-autonomy participant. In all conditions, we found that people positively shifted their evaluations toward approached stimuli and negatively shifted their evaluations toward avoided stimuli. We also found that people used their sampling decisions to inform their attitudes about stimuli. In Experiment 1, sampling decisions were explicitly linked to the AA behaviors, and an increased number of interactions strengthened the AA effect among high-autonomy participants. In Experiment 2, participants first sampled stimuli and then approached or avoided them; here, an increased number of interactions predicted a positive shift in evaluation among high-autonomy participants, regardless of AA behaviors. Experiment 3 replicated both patterns of effects by randomly assigning high-autonomy participants to a version of the task that framed sampling decisions as AA at the time of sampling (as in Experiment 1) or separated sampling and AA into two steps (as in Experiment 2). Thus, we demonstrate that sampling decisions can positively impact evaluations, but the effect is sensitive to how those decisions are framed at the time of sampling. (PsycInfo Database Record (c) 2026 APA, all rights reserved).