The work sampling system: Reliability and validity of a performance assessment for young children

Abstract

Performance assessment, an alternative approach to assessing students' achievements in school, refers to assessment methods that allow students to demonstrate their skills, knowledge, behavior, and accomplishments across a wide variety of classroom domains on multiple occasions. This article presents data concerning the reliability and validity of the Work Sampling System with 100 kindergarten-age children. A psychometric design was implemented in which children were enrolled in classrooms where the Work Sampling System was used and were also given individually-administered norm-referenced assessments in the fall and spring; in addition, their teachers completed a behavior rating scale in the spring. Results show that the Work Sampling checklist and summary report have very high internal and moderately high interrater reliability. The Work Sampling System accurately predicts performance on the norm-referenced achievement battery, even when the potential effects of gender, maturation (age), and initial ability are controlled. These data provide empirical support for the reliability and criterion validity of this performance assessment system as a measure of children's overall school achievement in kindergarten. The discussion covers issues raised by the study's design and by the use of performance assessment in general.