Visual rhythm perception improves through auditory but not visual training

Memory research has shown that test performance is optimal when testing and practice occur in identical contexts [1]. However, recent research in object recognition and perceptual learning has shown that multisensory practice leads to improved test performance, even when the test is unisensory [2,3]. It is also known that different sensory modalities can have differing proficiencies in a given domain. For instance, research shows that, compared to the auditory modality, the visual modality is significantly less proficient at discriminating the rhythms of temporal sequences [4,5]. Although rhythm perception is typically thought of as residing in the auditory domain, instances of visual rhythm perception abound in daily life, for example, when one watches a dancer or a drummer, or when a doctor examines a patient’s breathing or heart rate on a monitor (such as when diagnosing arrhythmia). However, no previous study has examined whether visual rhythm discrimination is a trainable perceptual skill. In light of this, we examined the extent to which visual rhythm perception can be improved through two sessions of visual, auditory, or audiovisual training. We found that visual rhythm discrimination was significantly improved in the auditory and audiovisual training groups, but not in the visual training group. Our results show that, in certain tasks, within-modality training may not be the best approach and that, instead, training in a different sensory modality can be a necessary approach to achieve learning. In performing a visual task, one would expect that training in the visual modality should be superior to training in any other modality. Barakat et al. show the opposite in the context of visual rhythm discrimination, such that auditory training produces significant learning while visual training fails.