Uncovering the Cognitive Mechanisms Underlying the Gaze Cueing Effect

Abstract

The gaze cueing effect is the tendency for people to respond faster to targets appearing at locations gazed-at by others compared to locations gazed away from by others. However, although the gaze cueing effect has been studied extensively and has implications for a variety of theoretical and clinical contexts, much is still unknown about the cognitive mechanisms underlying its emergence. Formal evidence accumulation models provide the dominant account of the cognitive processes underlying speeded decisions, however, they have never been applied to understand the gaze cueing effect. In this study we applied the Diffusion Decision Model (DDM) to gaze and arrow cueing data for the first time in order to 1) better understand the cognitive processes underlying the gaze cueing effect, and 2) see whether these processes could be considered the same as those that underlie arrow cueing effects. To achieve these aims, we applied variants of the DDM to three preexisting gaze cueing data sets and one preexisting arrow cueing data set (n = 171, 139,001 trials total), using a combination of individual-level and hierarchical computational modelling techniques. At the group level, people were best described by an attentional orienting mechanism rather than higher-order decision bias or information processing mechanisms. However, we found evidence for individual differences such that not everyone was best described by an attentional orienting mechanism. Further, the same people who were best described by an attentional orienting mechanism for gaze cues tended not to be best described by that same mechanism for arrow cues, suggesting these cueing effects may induce different responses within the same people – although we interpreted this finding cautiously.

The computational substrates of holistic face processing in cognitive control

“…Importantly, the analysis of SSP parameters show significant evidence for a faster preconscious perceptual encoding speed (ter) and a more caution decision criterion (a) as well as a stronger automatic activation (sd0) and/or a weaker suppression (sdr) in gaze judgment compared to arrow judgment. While non-decision time (ter) is generally thought to represent perceptual encoding of cues and motor preparation for decision (White et al, 2018), some studies have suggested that the effects for arrow and gaze stimuli seen here might reflect an advantage in preconscious perceptual encoding (Alister, McKay, Sewell, & Evans, 2022;Besner et al, 2021;Vainio et al 2014). Consequently, the finding that arrow judgment is more automatic than gaze judgment is not surprising since the perceptual encoding speed of arrow is faster than that of gaze.…”

What can evidence accumulation modelling tell us about human social cognition?

“…In a recent application of this approach to individual differences in the gaze and arrow cueing effects, a model comparison approach revealed that while the gaze cueing effect was often characterised by a difference in nondecision time for most individuals, this was not the case for all individuals. There was a small subset of participants whose gaze-cueing effect was best characterised as a difference in start point (Alister et al, 2022). Importantly, these results highlight how the same pattern of accuracy and/or reaction times in a task may not necessarily represent the operation of the same cognitive process for all individuals.…”

What can evidence accumulation modelling tell us about human social cognition?

“…In a recent application of this approach to individual differences in the gaze- and arrow-cueing effects, a model comparison approach revealed that while the gaze-cueing effect was often characterised by a difference in non-decision time for most individuals, this was not the case for all individuals. There was a small subset of participants whose gaze-cueing effect was best characterised as a difference in start point ( Alister et al, 2022 ). Importantly, these results highlight how the same pattern of accuracy and/or RTs in a task may not necessarily represent the operation of the same cognitive process for all individuals.…”