Utility of a dynamic termination criterion in the ZEST adaptive threshold method

Abstract

Bayesian adaptive threshold procedures may be run for a fixed number of trials, or may be stopped when the calculated confidence interval for the threshold reaches a selected limit (a dynamic termination criterion). This study used Monte-Carlo simulations to determine whether the confidence interval is a useful predictor of errors in the estimated threshold. No difference was found between the distribution of errors in a fixed trial procedure versus a dynamically terminated procedure of the same average number of trials. In addition, the width of the confidence interval failed to usefully predict observer variability arising from a shallow psychometric function slope or increased false positive response probabilities. This study suggests that dynamic termination criteria are of little use in Bayesian adaptive threshold procedures.