ABSTRACT
A model is proposed to describe the decision process engaged in by subjects making judgments on a trial-by-trial basis under the procedural constraints imposed by the experimenter in magnitude estimation tasks. The chief assumption is that subjects attempt to maintain rank order among successive responses that match rank order among stimulus intensities. The implications of the model were explored by comparing the output of a computer program written to simulate magnitude estimation with empirical results produced by a single subject who judged the loudness of a 1000-Hz tone (Green, Luce, & Duncan, 1977). The model mimicked the empirical data very well in terms of a variety of dependent measures. However, the assumption that subjects keep rank order among their responses on successive trials was not supported. As it stands, the theoretical approach is promising but incomplete.
