Abstract
The literature on measurement reliability shows the general consensus that examinee group heterogeneity with regard to the trait being measured affects the score reliability. Potentially, the performance level of the examinee group may also affect score reliability because of its potential effect on the relative magnitude of error variance. This article empirically examines the effects of these two examinee sample characteristics on score reliability of optimal-performance measurement. Two large extant data sets (criterionand norm-referenced, respectively) were used in the investigation. The results suggest that both performance variability and group performance level affect score reliability, and measurement error tends to be smaller for high-performance groups than for low-performance groups.
Get full access to this article
View all access options for this article.