The effect of guided Bilateral Eye Movements and Changing Visual Input on the detection of gradual changes in emotional faces

Bilateral Eye Movements (BEM) performed concurrently with subjective reports about ongoing imagery, memory, and recollection of emotional visual material affect their accuracy and vividness. It has been suggested that visual representations are degraded because bilateral eye movements during perception and subjective reports tax working memory capacities and attention. It needs to be clarified, however, if these effects are attributable to the eye movement (BEM), to a change in the viewpoint of the visual input (CVI), or a change in the contents of the display during a fixation condition (CVC).
This study investigated the effects of three different visual manipulations on a subsequent change detection task. We used a pre-post-treatment design and assigned participants to one of three computer-guided manipulations. One group was asked to follow a moving dot from left to right (BEM); a second group was asked to keep eyes stationary at fixation while the background switched viewpoint from left to right (CVI); and a third group was asked to keep eyes stationary at fixation while the fixation dot changed colour (CVC). All changes happened at a frequency of 1.2Hz, and all treatments lasted for 16 minutes.
Before and after the manipulations, participants saw videos with naturalistic scenes. Either an actor’s face morphed from neutral to angry or happy, or an object changed colour (and back). Changes were subtle and restricted to a small part of the display. Participants had to press a button as soon as they detected a change, followed by a confidence rating.
Because all manipulations entail changes in either display or content, we hypothesised that change detection and subjective reports would be affected in all three groups. However, because of the added visual and motoric effort of the bilateral Eye Movements, we expected this group to be affected the most, followed by the CVI.
Results demonstrated that the BEM and the CVI conditions affected detection and confidence reaction times similarly, depending on whether changes involved expressive changes in Anger and Happiness. Participants in the Fixation (CVC) condition became slower in their detection and confidence responses involving Anger, but became faster in their detection and confidence responses involving Happiness. Participants in the BEM and CVI conditions, however, only presented faster responses at detection for Happiness, and significantly slower RTs in their confidence responses for Anger, with no time changes in anger detection RTs, and happiness confidence RTs; thus demonstrating an asymetry in the effects that depended on valence. Crucially, only the BEM condition affected the relationship between relative accuracy and confidence. Participants in this group presented a reduced level of low confidence for missed trials, suggesting that they became less sensitive to their misses.