Abstract
Psychophysical studies have shown that temporal processing is capacity-limited and that the limit may depend on the stage of processing used for encoding. For example, perception of flicker has a temporal limit of 30-80 Hz, whereas identification of the individual states of flicker has a lower limit of 7-10 Hz. Here, we investigated the temporal limit for individuation of a face using steady-state visual evoked potentials (ssVEPs). The ssVEP method is ideal for measuring temporal processing because responses evoked by the transition between visual events occur at frequencies that are exact multiples of the alternation rate. Stimuli comprised a phase-scrambled and an intact face image alternating at a fixed temporal frequency. Face-specific responses occurred at the odd harmonics of the alternation rate. Stimuli were presented at four locations centered on fixation. The images at the "target" location flickered 180° temporally out-of-phase with those at the other three locations. High-density EEG was recorded while 10 adult observers performed a phase-judgment task identifying the "out-of-sync" location. Thresholds were defined as the temporal frequency yielding 62.5% correct performance. Perceptual limits for face individuation ranged from 2.95-6.31 Hz (M = 4.33 Hz, SD = 1.34). A vector average of the first harmonic response was calculated for electrodes over left, right, and medial occipital regions, for each temporal frequency. All observers showed decreasing responses as a function of temporal frequency, independent of electrode region, and responses were positively correlated with performance. Linear extrapolation of individual amplitude curves produced ssVEP limits ranging from 4.33-18.58 Hz (M = 10.47 Hz, SD = 4.02), consistent with the low perceptual limits. Together these results support the theory that perceptual access to visual information processed at higher stages is limited to slow temporal rates, and demonstrate that ssVEP is a valid tool for measuring temporal limits.
Meeting abstract presented at VSS 2012