PURPOSE: The reaction time (RT) of the visual priming task provides a window into the rich dynamics of brain activities. The task is to report the identity of a visual target that is preceded by a cue. The cue and the target may be congruent or incongruent. The RT is a function of the SOA (the length of the cue-to-target interval) with three regularities: (1) The RT decreases exponentially with the SOA. (2) The congruent RTs are shorter than the incongruent RTs (i.e. positive priming) at short SOAs, but are longer than the incongruent RTs (i.e. negative priming) at long SOAs. (3) Both the congruent and incongruent RTs change periodically with SOA in the theta-band (~4 Hz) but in the reverse phases. Here we developed the oscillated temporal-expectation model (OTEM) to account for the observed RT-SOA function and tested the model in a new experiment.
METHODS: OTEM assumes that the human brain has a constantly updated expectation for the onset time and the identity of the target during its waiting for the target. The temporal expectation, which determines the RT via its influence on the motor preparing rate, is based on previous experience and modulated by an attentional gain that oscillates with time. We performed a visual priming experiment to test OTEM, in which SOA and temporal uncertainty—two factors that usually co-vary—were disentangled.
RESULTS: We found that the RT of visual priming was influenced by temporal uncertainty as well as by SOA. Temporal uncertainty exerted pronounced effects on the slow trend of the RT-SOA function but more subtle effects on the theta-band oscillation. OTEM could quantitatively predict the effects of both SOA and temporal uncertainty.
CONCLUSION: OTEM provides a unified explanation for positive and negative priming and connects the oscillation in RT to attentional oscillation.