Psychophysical Methodology: Comparisons within the Method of Limits

Abstract

Assume that, in a “Yes”—“No” psychophysical experiment, the probability of a “Yes” response increases as the stimulus intensity increases. Then, on the basis of probability considerations (a) the mean threshold of the descending method of limits (DML) is greater than the mean threshold of the ascending method of limits (AML) and (b) σ of the DML threshold distribution may be greater than, equal to, or less than σ of the AML threshold distribution. The implications of these and other deductions are considered with respect to analyses of psychophysical data. For example, one deduction indicates that the present method for evaluating errors of habituation and expectation is wrong. Data that support the deductions are included.