In the concurrent rating task, stimuli are constructed by factorially combining several levels of separate physical components. Subjects give simultaneous judgments of their confidence that each component is presented at the highest possible level. The concurrent rating task has great potential for testing hypotheses and estimating parameters of multidimensional signal detection theory. In this article, the relationship between the parameters of the distribution of rating responses and the underlying perceptual distributions is detailed. The product moment correlation between ratings is shown to seriously underestimate the degree of perceptual dependence. More accurate estimators of the correlation between perceptual effects are suggested. The relation between the interrating correlation and measures of perceptual dependence developed for complete identification tasks is discussed. A powerful test of perceptual separability is then developed. Finally, these techniques are applied to some representative data. © 1988 Psychonomic Society, Inc.
CITATION STYLE
Ashby, F. G. (1988). Estimating the parameters of multidimensional signal detection theory from simultaneous ratings on separate stimulus components. Perception & Psychophysics, 44(3), 195–204. https://doi.org/10.3758/BF03206288
Mendeley helps you to discover research relevant for your work.