Group-vote rules for adaptive psychophysical testing

Publisher Website
Google Scholar
Abstract
A procedure is described that allows groups of subjects to be tested using adaptive psychophysical procedures when equipment limitations or experimental design requires each member of the group to be presented with the same stimuli. Group-vote rules are used to classify the group response on any trial as correct or incorrect, and this classification is used as input to standard, single-subject, adaptive procedures. The relationship between the probability of a correct group response and the probability of a correct individual response is derived by assuming members of the group respond independently and have identical psychometric functions. This relationship is used to choose a group-vote rule and an adaptive procedure that leads to a probability of correct individual response at convergence that is close to that which is desired. The usefulness of this procedure is limited to experiments where individual subjects have similar psychometric functions. It has, however, been used extensively for small groups of subjects to collect simple auditory discrimination data.

This publication has 0 references indexed in Scilit: