On Measuring Discrimination and Response Bias: Unequal Numbers of Targets and Distractors and Two Classes of Distractors

Models underlying the measurement of accuracy (discrimination) and decision rule under uncertainty (response bias) are discussed in relationship to tasks in which there are unequal numbers of targets and distractors, 2 classes of distractors, or both. Problems with measures that do not adjust for unequal numbers of targets and distractors are described. Alternative models that meet the desiderata of theoretical independence of discrimination and bias and meaningful bias values at all levels of accuracy are given. These models, signal detection theory (with d' and C) and 2 versions of two-high threshold theory (with Pr and Br), are applied to a sample data set, the predictions of the models are presented, and recommendations for their appropriate use are made.