Buus S, Florentine M, Zwicker T
Department of Electrical and Computer Engineering (409 DA), Northwestern University, Boston, Massachusetts 02115-5096, USA.
J Acoust Soc Am. 1995 Aug;98(2 Pt 1):853-61. doi: 10.1121/1.414450.
The purpose of this study was to determine whether the form of the psychometric functions for level discrimination (also known as intensity discrimination) measured in a two-interval, two-alternative forced-choice paradigm changes with hearing impairment. Measurements of sensitivity, d', as a function of delta L (= 20 log[(p+delta p)lp], where p is pressure) were obtained in six listeners with cochlear impairments, five normal listeners tested in the quiet, and two masked-normal listeners. Stimuli were chosen to encompass a wide range of conditions and difference limens. Results show that the d' is nearly proportional to delta L over the entire range of stimuli. This simple relationship is not changed by cochlear impairments or masking noise. These findings indicate that if the transformation from stimulus intensity to decision variable is affected by hearing impairment or masking, the change affects both the mean and the standard deviation in the same manner--i.e., the standard deviation may be proportional to the derivative of the transformation as was suggested by Zwislocki and Jordan [J. Acoust. Soc. Am. 79, 772-780 (1986)]. The results also lend support to the notion that delta L--plotted on a logarithmic scale--is an appropriate representation of level-discrimination performance.
