APA Dictionary of Psychology

in analyses estimating the relationship between two variables, an adjustment for error introduced during the process of obtaining the measures, where such error serves to underestimate the measured effect.