ELEMENTARY SCHOOL PEER TUTORING IN MATHEMATICAL VERBAL PROBLEM SOLVING

Abstract

Analyses of the subtest scores on standardized tests in mathematics revealed a general weakness in the area of mathematical verbal problem solving (MVPS). It was, therefore, reasoned that only through the total individualization of instruction could the needs of students be met. The purpose of this study was to determine the extent to which it was feasible to use an individualized programmed instructional format in MVPS.;The population in the study was drawn from a school in the Ocean-Hill Brownsville section of Brooklyn, New York. The sample consisted of 160 fourth-grade students, 80 of whom comprised the experiment, and 80 of whom were in the control group. Eighty sixth-grade volunteer students were trained as tutors. These students were divided evenly into two groups, half experimental and half control.;The data were analyzed through a series of single-factor analyses of covariance utilizing pretest criterion variate raw scores as the covariates so as to adjust for any differences in initial MVPS ability. Separate analyses of covariances were performed on post and post-posttest scores for each of the dependent variables (Metropolitan Achievement Test--MAT; and Childhood Attitude Toward Problem Solving Scale--CAPS). The analyses of covariance of performance of the fourth-grade tutees (as compared to the non-tutees) and sixth-grade tutors (as compared to the non-tutors) revealed significant differences between treatment and control groups in favor of the treatment group on the MVPS portion of the Metropolitan Achievement Test (MAT). The performance of the tutors and tutees on the Childhood Attitude Toward Problem Solving (CAPS) scale was also significantly higher than that for the non-tutors and non-tutees.;The findings of this study indicated that a program utilizing trained tutors to teach mathematical verbal problem solving (MVPS) would be beneficial to both tutors and tutees.