This proposal addresses the problem that Stevens' power law for psychophysical matching judgments accurately describes data averaged over observers but often fails to do so for the data of individuals. Questions have been raised both about the form of the individual function--it departs systematically if idiosyncratically from a power relation--and about the individual exponent--it varies so widely across observers as to vitiate its interpretation as a reflection of sensory processes. This latter fact has serious implications for characterizing not only individuals with normal sensory functioning but also those with sensory deficits. It would be valuable, from both a diagnostic and treatment evaluation standpoint, to be able to describe accurately and reliably the growth of sensory magnitude in patients with visual, auditory, tactile, gustatory, and olfactory disorders. The general aim of the research proposed is to gain sufficient understanding and control of the variables operating on psychophysical marching judgments to allow general principles to emerge from the study of individual observers. The experimental paradigm is that of cross-modal matching, in which the observer produces values on a judgment continuum that match values on a target continuum. There are two variants--a few-stimulus, few-judgment procedure, used to study variability in exponent, in which 8-10 relatively widely-spaced target levels are repeated only once or twice--and a many-stimulus, many-judgment procedure, used to study the form of the function, in which 18-21 relatively closely-spaced target levels are repeated 50 to 100 times. Specific experiments concern the effect on the form and exponent of the individual function of 1) the observer's ability to recognize stimuli and responses previously presented or given, 2) the pool of numbers available to the individual, and the assignment of the first number to a target level, and 3) repeated judging, especially of many target values. Other experiments seek to identify the sources of idiosyncratic departures from a power function as sensory or response-based by 1) comparing matches of number to sound intensity to matches of sound intensity to number, and 2) comparing matches of number to luminance to those of number to sound intensity, to see whether, in each case, identical departures from the power relation characterize both scales, as would be expected on a response-bias hypothesis. A final set of experiments asks whether individual differences in response styles (different number ranges used) or cognitive strategies (absolute vs. ratio matching) can account for individual differences in power functions.
| Teghtsoonian, R; Teghtsoonian, M (1997) Range of acceptable stimulus intensities: an estimator of dynamic range for intensive perceptual continua. Percept Psychophys 59:721-8 |
