Abstract
Evolutionary theories suggest that humans can evaluate and prioritise emotionally salient stimuli, irrespective of whether they reach awareness. Many behavioral observations support this idea. Studies using the masked visual probe (MVP) paradigm suggest that threatening stimuli, rendered invisible by backward masking, can nonetheless capture spatial attention. Other studies suggest that threatening stimuli are preferentially processed, even when suppressed from awareness via binocular rivalry (BR) or continuous flash suppression (CFS). We provide a meta-analysis of the evidence for a threat-related advantage provided by these three paradigms. The CFS paradigm produced inconsistent effects; some types of threat stimuli were prioritised, whereas others were disadvantaged, relative to neutral stimuli, yielding an overall null effect. In the BR paradigm, the pooled effect size was medium, but was associated with conditions likely to induce response biases (e.g. piecemeal rivalry). The overall effect size from the MVP paradigm was small, with substantial heterogeneity explained by stimulus presentation time. Moreover, this dependency (possibly reflecting unintended stimulus visibility) was larger when the study failed to objectively measure stimulus awareness. This suggests that inadequate awareness measures and partial stimulus visibility may have contributed to the reported threat-related advantages. To investigate these issues, we conducted a well-powered visual probe study (N=41) with stimuli presented under normal conditions (conscious viewing: 500 ms), under backward masking (17 ms) and under CFS (500ms). Awareness of stimuli was assessed via stringent signal detection criteria. Although emotionally salient stimuli elicited attentional cuing effects under normal conditions, we detected none under masked or CFS conditions. Moreover, increased awareness of the stimuli in CFS and masked trials (assessed via a 2-alternative forced choice location task) predicted increased attentional cuing. Considered alongside our meta-analyses, our data suggest that attentional orienting to emotionally salient stimuli is dependent on, or at least strongly modulated by, their conscious registration.
Meeting abstract presented at VSS 2016