TY - JOUR
T1 - On rewarded actions and punishment-avoidant inactions
T2 - The action–valence asymmetry in face perception
AU - van Lent, Tjits
AU - Bijlstra, Gijsbert
AU - Holland, Rob W.
AU - Bijleveld, Erik
AU - Veling, Harm
PY - 2025/7
Y1 - 2025/7
N2 - Although social interactions are ubiquitous, people often choose not to interact with others—for example, people may choose to not greet a stranger, to not talk to a colleague at work, or to ignore a text message from a friend. Here, we systematically investigate how people's actions, inactions, and their consequences (rewards and punishments) affect impressions. In four preregistered experiments (N = 240), we used a reinforcement learning go/no-go task, in which people learned to act or not act to images of fractals/faces to obtain rewards or avoid punishments. Findings replicated the action–valence asymmetry in learning (Experiments 1–4): People more easily learned to act when acting led to the attainment of rewards (vs. the avoidance of punishments), while people learned inactions more easily when these inactions led to the avoidance of punishments (vs. the attainment of rewards). Our experiments demonstrate that these action–valence asymmetries extend to social stimuli (Experiment 2 ingroup faces; Experiment 3 outgroup faces; Experiment 4 ingroup and outgroup faces) and that they affect subsequent impressions. That is, people evaluated faces most positively when acting had previously led to the attainment of rewards; people evaluated faces most negatively when not acting had previously led to the avoidance of punishment. We discuss our findings in light of the approach–avoidance literature. This work has implications for our understanding of the role of inactions in social contexts: It shows evidence that inactions lead to less positive impressions than actions, over and above the effect of punishment signals.
AB - Although social interactions are ubiquitous, people often choose not to interact with others—for example, people may choose to not greet a stranger, to not talk to a colleague at work, or to ignore a text message from a friend. Here, we systematically investigate how people's actions, inactions, and their consequences (rewards and punishments) affect impressions. In four preregistered experiments (N = 240), we used a reinforcement learning go/no-go task, in which people learned to act or not act to images of fractals/faces to obtain rewards or avoid punishments. Findings replicated the action–valence asymmetry in learning (Experiments 1–4): People more easily learned to act when acting led to the attainment of rewards (vs. the avoidance of punishments), while people learned inactions more easily when these inactions led to the avoidance of punishments (vs. the attainment of rewards). Our experiments demonstrate that these action–valence asymmetries extend to social stimuli (Experiment 2 ingroup faces; Experiment 3 outgroup faces; Experiment 4 ingroup and outgroup faces) and that they affect subsequent impressions. That is, people evaluated faces most positively when acting had previously led to the attainment of rewards; people evaluated faces most negatively when not acting had previously led to the avoidance of punishment. We discuss our findings in light of the approach–avoidance literature. This work has implications for our understanding of the role of inactions in social contexts: It shows evidence that inactions lead to less positive impressions than actions, over and above the effect of punishment signals.
KW - Impression formation
KW - Intergroup relations
KW - Pavlovian bias
KW - Reinforcement learning
KW - Value-based decision making
U2 - 10.1016/j.jesp.2025.104754
DO - 10.1016/j.jesp.2025.104754
M3 - Article
AN - SCOPUS:105002237064
SN - 0022-1031
VL - 119
JO - Journal of Experimental Social Psychology
JF - Journal of Experimental Social Psychology
M1 - 104754
ER -