What determines the influence of attention on binocular rivalry?
56.522, Tuesday, 20-May, 2:45 pm - 6:45 pm, Pavilion
Kevin C Dieter1,2,3,4, Michael D Melnick3,4,5, Duje Tadin3,4,5; 1Vanderbilt Vision Research Center, Vanderbilt University, Nashville, TN, USA, 2Department of Psychology, Vanderbilt University, Nashville, TN, USA, 3Center for Visual Science, University of Rochester, Rochester, NY, USA, 4Department of Brain & Cognitive Sciences, University of Rochester, Rochester, NY, USA, 5Department of Opthalmology, University of Rochester, Rochester, NY, USA
INTRODUCTION: Binocular rivalry is relatively resistant to contextual and/or attentional modulations, distinguishing it from other forms of visual bistability that are readily influenced by attention or context. What explains this difference? Drawing from our recently developed framework (Dieter & Tadin, 2011), we predicted that the temporal dynamics of visual competition during rivalry determine the efficacy of attentional influences over rivalry. Specifically, we hypothesized that rivalry would be susceptible to attentional influence only during periods of unresolved conflict: at rivalry onset and near the end of individual rivalry percepts (Alais et al., 2010). Indeed, attention strongly affects initial rivalry (Mitchell et al., 2004), but there has been no work examining the temporal specificity of attentional effects during continuous rivalry. METHODS: Because percept durations during binocular rivalry are stochastic, we could not cue attention at a set time. Instead, while observers continuously viewed rival gratings, we occasionally presented transient, feature-based attentional cues with variable SOAs relative to the start of each percept. This ensured that some percepts were cued near their start, and some near their end. To establish whether these cues influenced observers’ perceptions, we compared dominance durations of cued percepts with their expected duration (based on un-cued percept durations). RESULTS: Attentional cues influenced percept durations in a content-specific manner. Cues whose features matched the currently dominant stimulus led to a 5% increase in percept duration (p <0.05), while cues matched to the currently suppressed stimulus shortened percepts by almost 20% (p <0.01). Crucially for our hypothesis, analyses of the temporal aspect of attentional effects in our experiment revealed that cues influenced perception only when presented during the final second of a percept’s dominance. Cues presented earlier had no influence. This indicates a temporally isolated effect of attention that corresponds to periods of unresolved competition during binocular rivalry.