Abstract
This exploratory experimental study investigates the contribution of stimuluspresentation order to the regression effect in cross-modal matching of brightness andloudness. Previous research has largely relied on the magnitude production paradigm toobtain individual cross-modal matches, which induces distinct stimulus roles (i. e., afixed standard vs. an adjustable comparison). The cross-modal matches typicallyexhibit a role-dependent regression effect, that is, they gravitate toward the mean of theadjusted stimulus’ range. This regression effect can be explained by the theory of globalpsychophysics via role-dependent internal references. However, stimulus role andpresentation order were often confounded in previous experiments. To decouple thesevariables, this study employed a two-alternative forced-choice paradigm. Participantsjudged whether a noise sound or an achromatic disk appeared more intense. To hideexplicit stimulus roles and level selection across the luminance and sound pressurespectra, several adaptive staircases were run interleaved, balancing presentation orderacross a high number of trials. Results showed an order-dependent regression effect inthree of the seven subjects, suggesting these individuals used the first stimulus in a trialas an implicit standard. The remaining four subjects did not seem to exhibit thisstrategy. Irregularities in the adaptive staircases, deviations from the globalpsychophysical model, and future directions are discussed.