Conservative Criterion Explains The Non-Conscious Perception of Facial Expression Under Continuous Flash Suppression
Ali Pournaghdali, Bennett L Schwartz
Journal of Vision September 1, 2018 DOI: 10.1167/18.10.945 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Conference abstract (meeting abstract presented at VSS 2018) Peer reviewed |
|---|---|
| Measures | d' (signal detection sensitivity index), response criterion |
| Key points | Sensitivity (d') did not differ significantly between the conscious detection and non-conscious 2AFC tasks, but the response criterion was higher for detection than for 2AFC. The authors suggest that non-conscious perception of facial expressions under CFS may stem from a more conservative criterion in detection rather than from dissociable processes, supporting the use of bias-free measures in consciousness research. |
Abstract
A main question in the scientific study of conscious perception is the nature of the dissociation between conscious and non-conscious perception. That is, it is crucial to use bias-free measures to evaluate perceptual sensitivity and response criterion of participants in the conscious and non-conscious tasks to evaluate this dissociation. The aim of this study is to evaluate sensitivity and criterion of participants, using signal detection methods, in a conscious detection and a non-conscious 2-alternative forced-choice (2AFC) tasks for facial expression perception while employing continuous flash suppression (CFS). We hypothesized that non-conscious perception of facial expression under CFS demonstrates real differences between conscious and non-conscious perception. We predicted that participants' sensitivity in the detection task will be significantly lower than their sensitivity in the 2AFC task. After rendering invisible images of faces with different facial expression (fearful vs. neutral) for five hundred milliseconds using CFS, participants judged the presence/absence of the faces with a yes/no detection task and the emotion of faces with the 2AFC task. After acquiring data, we evaluated participants' ability to discriminate signal from noise using d' (an index of sensitivity in signal detection theory) and their criterion for detection and 2AFC task. Our results indicate that there is no significant difference between sensitivity of participants in detection and 2AFC task, but we found higher criterion value for detection compared to 2AFC task. Our results indicate that participants' ability to discriminate signal from noise is diminished for both detection and 2AFC tasks while using CFS. Therefore, non-conscious perception of facial expression with CFS may be the result of a more conservative criterion in the detection task as compared to 2AFC task rather than from dissociable processes. These results provided additional evidence for the importance of using bias-free measures over the accuracy-based methods in visual consciousness research. Meeting abstract presented at VSS 2018