Increased temporal sensitivity for threat: A Bayesian generalized linear mixed modeling approach

Abstract

People overestimate the duration of threat-related facial expressions, and this effect increases with self-reported fearfulness (Tipples in Emotion, 8, 127–131, 2008, Emotion, 11, 74–80, 2011). One explanation (Cheng, Tipples, Narayanan, & Meck in Timing and Time Perception, 4, 99–122, 2016) for this effect is that emotion increases the rate at which temporal information accumulates. Here I tested whether increased overestimation for threat-related facial expressions in high fearfulness generalizes to pictures of threatening animals. A further goal was to illustrate the use of Bayesian generalized linear mixed modeling (GLMM) to gain more accurate estimates of temporal performance, including estimates of temporal sensitivity. Participants (N = 53) completed a temporal bisection task in which they judged the presentation duration for pictures of threatening animals (poised to attack) and nonthreatening animals. People overestimated the duration of threatening animals, and the effect increased with self-reported fearfulness. In support of increased accumulation of pacemaker ticks due to threat, temporal sensitivity was higher for threat than for nonthreat images. Analyses indicated that temporal sensitivity effects may have been absent in previous research because of the method used to calculate the index of temporal sensitivity. The benefits of using Bayesian GLMM are highlighted, and researchers are encouraged to use this method as the first option for analyzing temporal bisection data.