Visual and behavioral responses to social and non-social threats: A multi-site replication

Abstract

Researchers have proposed that humans have evolved psychological mechanisms that facilitate the detection, rapid response, and subsequent avoidance of potential threats. However, some inconsistent results in the literature have called into question the robustness of these responses. Here, we sought to replicate previous findings on the rapid detection of both social (i.e., angry faces) and nonsocial (i.e., snakes) threats within a large and diverse sample of adults, and to examine the robustness of our effects across three data-collection sites using two response metrics-visual latency to detect threatening versus non-threatening stimuli and motor (i.e., button press) responses to indicate that threatening versus non-threatening targets had been detected. We found an advantage for both social (angry facial configurations) and non-social (snakes) threats across all three data collection sites, demonstrating that the phenomenon is both replicable and robust. Further, we found that the threat advantage was only significant for visual latency to first detect threatening stimuli and not for subsequent motor responses-the overall replication effect was driven by first fixations-suggesting that biases for threat might be perceptually based.

Keywords: Attention; Attentional bias; Rapid attention; Threat detection; Visual search.

Fig. 1.

Sample matrix schematic with an angry target present.

Fig. 2.

Means for button-press responses by condition. Participants more quickly detected both social (first set of bars) and nonsocial threats (second set of bars) faster than non-threats, and they detected nonsocial threats (snakes) more quickly than social threats (angry faces) overall (third set of bars). Error bars represent standard errors.

Fig. 3.

Means for button-press responses by condition by site.