Efficient use of peripheral information for temporal prediction

Abstract

We adapt to the environment by predicting subsequent events. Generally, intervals between predictions and events make it difficult to predict the events accurately. Previous studies reported that using peripheral information is useful for maintaining predictions of subsequent events; however, it remains unclear how this information maintains the accuracy of the prediction. I presented peripheral visual stimuli in a discrimination task and manipulated the number of times these stimuli were presented while participants were waiting for a task-relevant visual stimulus, and compared participants’ response times and event-related brain potentials in Experiment 1. In addition, the influence of the difficulty of predicting the task-relevant visual stimulus was examined in Experiment 2. In both experiments, contingent negative variation (CNV) amplitude immediately before the task-relevant visual stimulus appeared was larger under the condition where many peripheral visual stimuli were presented, and the response time was shorter under this condition. In addition, the largest CNV amplitude under this condition was elicited by the third peripheral visual stimulus, followed in order by the first and second peripheral visual stimuli. These results show that we can predict the timing of events that occur with a delay after the prediction by using peripheral information. Moreover, this peripheral information is processed according to the importance of predicting a task-relevant stimulus, and attentional resources are allocated efficiently. These results provide evidence of the predictive function for temporal prediction of using peripheral information and the allocation of cognitive resources.