Test of the Accuracy of Crossman's Confusion-Function

Abstract

This study was done to test the accuracy of Crossman's Confusion-Function in predicting response patterns of Ss in discriminating between stimuli of known physical dimensions. The Confusion-Function, C = 1/[log₂ (x₁/x₂)], where x₁ and x₂ are some physical dimensions of stimuli S₁ and S₂, provides a method of quantifying the “discriminability” of stimuli. Crossman's work demonstrated a linear relationship between disjunctive reaction time (choice time) and his Confusion-Function. In this study 24 college sophomores were used as Ss and data were collected using two different tasks and two different kinds of stimuli. Ss discriminated between circles varying in diameter and patches varying in reflectance (amount of reflected light). The difficulty of discrimination was varied using the Confusion-Function to assign confusion values to the sets of stimuli. Response time was measured for two tasks, a card-sorting task and responding to tachistoscopic presentations. The data analysis provided striking confirmation of Crossman's hypothesis of linearity and added evidence to the possibility that the Confusion-Function might apply to discrimination in general.