Comparative Analysis of Different Correction Methods for Measuring Rod-and-Frame Test Performance

Abstract

A comparative analysis was conducted of five different methods for scoring the Rod-and-Frame Test to examine the convergent validities of each of the five methods with performance on the Embedded Figures Test for a sample of 408 13- to 16-yr.-olds. The scoring methods included the absolute error, the algebraic error, the absolute value of the algebraic error, the frame effect, and an assessment of relative field dependence. Correlations obtained among the five methods and the Embedded Figures Test indicated different patterns for boys and girls. In general, the method which employed the absolute value of the algebraic error yielded the strongest convergent validity coefficients with scores on the Embedded Figures Test and Group Embedded Figures Test for boys (.62 and –.54, respectively), although, for girls, the absolute algebraic error and absolute error both yielded coefficients of .50 with Embedded Figures Test scores and –.46 with Group Embedded Figures Test scores.