GENERALIZABILITY OF NEW STANDARDS PROJECT 1993 PILOT-STUDY TASKS IN MATHEMATICS

The New Standards Project conducted a pilot test of a series of perfor mance-based assessment tasks in mathematics and English language arts at Grades 4 and 8 in the spring of 1993. This article reports the resu lts of a series of generalizability analyses conducted for a subset of the 1993 pilot study data in mathematics. Generalizability analyses f or completely crossed designs of Raters x Tasks x Pupils were conducte d for a total of nine collections of mathematics tasks. The results of those analyses were used to estimate standard errors of measurement f or absolute decision studies using various combinations of number of r aters and number of tasks. Consistent with results of previous analyse s of performance-based assessment tasks, sampling variability due to t asks was found to be substantially larger than that due to raters. Imp lications for assessment designs are discussed.