Measuring laterality effects in dichotic listening

Abstract

The standard dichotic 2-response paradigm, as well as a number of indices of the degree of ear advantage [human] proposed in the literature, are described. The numerical range of most of these indices is constrained by performance level; only one particular index avoids these constraints. This index is not necessarily the optimal one. A correction for guessing is proposed and analogies to signal-detection theory are discussed, as well as theoretical and empirical criteria for choosing the correct index of laterality. The index called here eg is proposed as the best solution at the present state of knowledge. The phenomenon of dichotic fusion and the dichotic single-response paradigm, which offers many methodological advantages over the 2-response paradigm, is discussed. The factors of ear dominance and stimulus dominance in the perception of fused stimuli are also discussed. An index of ear dominance is derived by again taking advantage of analogies to signal-detection theory. Stimulus intelligibility, guessing and selective attention, blend responses, test reliability, validity and homogeneity are also examined.