Seemingly discrepant data from hippocampectomized macaques are reconciled by detectability analysis.

Abstract

Wide differences in the effects of hippocampectomy on the visual memory performances of monkeys have been reported in the literature. These differences have been attributed to the extent of preoperative training. Previous calculations, however, have always utilized data expressed as a percentage correct. A reanalysis based on detectability, the d' of signal detection theory, suggests that the differences may be illusory. The losses in visual memory performance due to hippocampal lesions as measured in different laboratories are about equal in terms of d'. This points out a potential trap in making direct comparison of absolute differences in percentage values.