Testing the race model inequality : a nonparametric approach

Abstract This paper introduces a nonparametric procedure for testing the race model explanation of the redundant signals effect. The null hypothesis is the race model inequality derived from the race model by Miller (Cognitive Psychol. 14 (1982) 247). The construction of a nonparametric test is made possible by a small change in the usual experimental procedure. This change involves that whenever only a single stimulus is presented, its modality is determined independently from the previous trials. It is shown that the test procedure is consistent against every violation of the null hypothesis. The test procedure is developed for data from a single participant, but it can easily be extended to the testing of the null hypothesis across participants, and this is also shown in the paper.