Few studies have been published that have examined the psychometric properties of Gickling's model of curriculum-based assessment (CBA), even though this model is among the few educational assessment models that include both performance assessment and cognitive learning theory. The current study examined the delayed alternate-form reliability of measuring sight-word acquisition and retention rates using CBA with 91 students from grades 1, 3, and 5. Each student was twice taught a different group of words from the Esperanto International Language, with a 14-day retest interval. The total number of words acquired, number of words retained, and percentage of words retained were correlated between the two sessions. The reliability of the first two rates was above established minimum levels of adequacy for grades 3 and 5, but the retention rates expressed as a percentage were not.
Copyright 2017 Elsevier B.V., All rights reserved.