COMBINING EVIDENCE PRESENTED SIMULTANEOUSLY TO EYE AND EAR - COMPARISON OF SOME PREDICTIVE MODELS

Data from a sustained monitoring experiment involving auditory, visual and combined audio-visual signal recognition were used to assess the predictive validity of 5 models of human bisensory information processing. Satisfactory predictions of the dual-mode performance levels were made only by 2 models, neither of which assumes that the auditory and visual systems operate independently, and correlations which attest to this nonindependence are presented. One of these models explicitly assumes that the 2 systems are associated so that their judgments tend to coincide; the othera ssumes that the visual system alerts the auditory system to the presence of a signal. Both models accurately predict the levels of d'' [index of detectability] and .beta. [bias] in the dual-mode condition, and the alerting one also accounts for the observed reduction in response latencies.