Neural and behavioral correlates of change detection following training with complex auditory scenes

Abstract

The effects of training on an auditory flicker task were examined in this study, where auditory scenes alternated until a participant responded “change” or “same.” Change scenes differed in the location (on the horizontal plane) of two or more sounds. Half of participants were trained with auditory scenes on Day 1, and half were trained on a visual task. On Day 2, all participants were tested on auditory scenes containing either trained sounds or novel sounds within scenes; EEG was collected during testing. Participants trained with auditory scenes performed better overall than control participants. Both groups had lower reaction times to correct response change trials than same trials. Trained participants performed no better on trained sounds than novel sounds, but did perform better on trained sounds than control participants. Electrophysiologically, there were differences in ERP components based on the type of trial (change versus same auditory scene). For same trials, N1 and P2 amplitudes were significantly higher, and P3b amplitudes were significantly lower than change trials. Additionally, P2 showed a general decrease in amplitude as scene presentations neared the response. These data show that training reduces change deafness, and improvements are not limited to the sounds experienced during training.