Feuerstein's Representational Stencil Design Test: A Comparison of Two Scoring Methods

Abstract

The Representational Stencil Design Test, one of the tasks from Feuerstein's Learning Potential Assessment Device, was administered to 370 seventh- and eighth-grade students from regular public school classrooms. Two scoring methods were used to investigate factorial and psychometric differences that would relate to problem-solving processes used by the students. The dichotomous method scored each design as correct or incorrect (possible range = 0 to 20 points). The non-dichotomous method gave 2 points for each stencil chosen and put in correct juxtaposition to other stencils or 1 point if placed incorrectly (possible range = 0 to 160 points). Coefficient alpha and split-half reliability values for both methods ranged from .88 to .96. Both scoring methods were subjected to principal components factor analyses with Varimax rotation. For the dichotomous method, four factors were identified; the first two accounted for about 36% of the variance. For the nondichotomous method, however, three factors were derived, the first of which accounted for about 52% of the variance. These results suggest that different problem-solving skills are revealed when the two methods are used. Implications for theory and future research are discussed.