THE PROBLEM A pilot study had shown that below average learners have commonalities in their learning styles. Since the literature had demonstrated that matching students’ learning style with a compatible teaching approach could improve the academic performance of children, this study attempted to determine what the impact of teaching cooperatively, a teaching approach that closely matched the preferred learning style of the below average learners, would have on such learners. The impact of the teaching strategy was evaluated through changes in spelling scores, in self-concept and in locus of control. Further, the impact on boys and girls was monitored to determine if either gender would respond differently to this teaching approach. METHOD A non-equivalent control group design was used to organize the study. Six teachers were randomly assigned to either an experimental or control group from two public school sites. The experimental teachers received inservice in cooperative learning. At the same time, the students of both the experimental and control group teachers were pretested in terms of spelling achievement, self-concept and locus of control. The experimental teachers then taught spelling using cooperative learning during a six week treatment period. At the conclusion of the six weeks, the students were posttested. ANALYSIS OF DATA The data was analyzed using the analysis of covariance procedure in order to evaluate the three null hypotheses and their sub-hypotheses. RESULTS Two of the three null hypotheses were rejected. The results indicated that the experimental group significantly outperformed the control group in spelling achievement with both the experimental males and females showing significantly greater growth in spelling than the control males and females, respectively. On the locus of control measure, there was a significant difference between the experimental and the control groups, favoring the experimental group, due mainly to the performance of the experimental females. On the self-esteem measure, the experimental males scored significantly higher than control males. Analyses to determine the effect that the school site had on the performances of the students did not reflect a specific effect for school site. When Q1 students (those scoring below the twenty-fifth percentile on their overall score on the Comprehensive Test of Basic Skills) were compared with Q2 students (students scoring between the twenty-fifth percentile and below the fiftieth percentile on their overall score on the Comprehensive Test of Basic Skills) no systematic effect because of academic level (Q1 or Q2) was found.