Further tests of dual-process theory: A reply to Yonelinas (1999)

In a preceding article (M. Glanzer, K. Kim, A. Hilford, & J. K. Adams, 1999), the authors reported the results of testing two predictions of dual-process theory concerning the shape of the receiver-operating characteristic (ROC). The theory predicts that item recognition ROCs in standard score form, z-ROCs, will, in general, be concave. It also predicts the degree of concavity that will be present. The two predictions were tested. The theory failed on both. They were tested again in this article: the first prediction with an extended set of data, the second with a new set of predicted concavity measures. The first test found again that contrary to the theory, z-ROCs were not, in general, concave. The second test found again that dual-process theory failed to predict the degree of concavity of specific z-ROCs.