Analysis and optimization of psychophysical procedures in audiology

The minimal standard deviations achievable for concurrent estimates of thresholds and psychometric function slopes as well as the optimal target values for adaptive procedures were calculated as functions of stimulus levels and track length on the basis of the binomial theory. Two approaches to adaptive procedures that converge at the optimal target values were introduced and tested with Monte–Carlo simulations. Both approaches utilize the fact that in some psychophysical tasks more than one statistically independent Bernoulli trial is performed per trial, as it is done, e.g., in sentence intelligibity tests that score each word separately. The first approach is a generalization of the adaptive procedure for sentence tests proposed by Hagerman and Kinnefors (1995). The second approach is a modification of the transformed up/down procedures proposed by Levitt (1971). The first approach yields higher efficiency. When the target discrimination value at which the adaptive procedure should converge is set to 50 %, a biasfree SRT estimate with a standard deviation only about 1.3 times higher than the theoretically optimally achievable value is reached using 20 trials. When the adaptive procedure converges at the target discrimination values 20 and 80 % in a randomized interleaved order, it achieves standard deviations in both SRT and slope estimates which are about 1.5 times higher than the optimal achievable values, with a small bias in slope estimates, using 30 trials. The adaptive procedures according to Hagerman and Kinnefors were evaluated