| Abstract | Proposes the phi correlation coefficient as an index of ear differences in dichotic listening tests, specifically for the 2-response paradigm, where, as an index of ear difference over all trials, it would be statistically appropriate for correlation with overall performance. Using the same computational formula, the phi index may also be applied to the results of a 1-response, directed-recall listening test. The interest of the index lies in the fact that for a constant size of response set and number of dichotic trials, its values may be directly compared for statistical significance |