A Method to Evaluate Performance Reliability of Individual Subjects in Laboratory Research Applied to Work Settings.

摘要

This report presents a method that may be used to evaluate the reliability of performance of individual subjects, particularly in applied laboratory research. The method is based on analysis of variance of a tasks-by-subjects data matrix, with all scores standardized. If all tasks are parallel, then the average correlation among tasks is an inverse function of the within-subject variance, which may be computed for any individual subject or group of subjects. The formula for determining the relationship between within-subject variance and average correlation is developed and a method of testing the reliability of individual subjects against the general level of reliability is presented. Possible applications of the method are noted. (Author)