This study assessed the usefulness of the dynamic testing approach to optimize testing procedures by reducing or eliminating bias, conceived of as error in measurement attributable to factors entering into performance which were not the target of the assessment. The study examined: (1) whether the dynamic assessment approach yields information which is a more appropriate indicator of cognitive competence than standard approaches based on static test theory; (2) why and for whom the dynamic assessment approach is effective; and (3) for what reason the procedures should be used. The sample of 147 fourth- and fifth-grade students from 15 elementary schools in the San Bernardino, California, area included 67 Anglo Americans, 37 Black Americans, and 43 Hispanic Americans randomly assigned to standard, verbalization, or elaborated feedback test conditions. Measures of cognitive and perceptual functioning, cognitive style and personality (e.g., impulsivity-reflectivity, motivation and planning, and school achievement) included Raven's Coloured Progressive Matrices, the Matching Familiar Figures Test, the Cattell Culture Fair Test of "g," Harter's Perceived Competence Scale for Children, and Trail Making and Visual Search Tasks. It was found that the testing condition which seemed to be most effective in increasing performance of all groups was verbalization; a type of compensatory effect for initially observed differences was brought about by subject verbalization. (CM)