During the past quarter century, over 1000 articles have documented changes in student behaviors related to participation in mastery learning programs. Although the results have been generally positive, a debate continues as to the cause for increased student performance: Are results due to changes in how students attend to the information, or simply due to increased study time as a result of remediation? In this study, a videodisc-based program in fractions was used with fifth-grade students. The videodisc-based instruction was chosen to help minimize differences in instructional materials, instructional time, and instructional delivery. A pretest-posttest, control-group design was used to compare academic achievement, locus of control, and goal setting scores of two groups (N=154). Both groups received instruction in fractions via the teacher-directed, videodisc-based Mastering Fractions program. Treatment 1 students (N=80) were knowledgeable that they were participating in a mastery-learning program and would be held accountable for their progress and remediation. Treatment 2 students (N=74) were not aware that their teacher was using mastery learning principles to determine progression and remediation. A control group (N=32) received their normal grade five mathematics program. Comparisons between Treatment 1 and Treatment 2 student scores, after adjustments for pretest results using analysis of covariance, revealed standardized mean difference effect sizes of +0.03 for achievement, +0.32 for locus-of-control, and +0.46 for goal setting mean scores favoring Treatment 1. A discrepancy in implementation of the videodisc program in two classes may have skewed results. When data from these two classes were omitted, the analyses showed adjusted standardized mean difference effect sizes of +0.63 for achievement, +0.75 for locus-of-control, and +0.55 for goal setting mean scores favoring Treatment 1. A two-way analysis of covariance with treatment groups and achievement levels was also conducted. Subsequent standardized mean difference effect sizes using adjusted mean scores were greatest for students from the lowest achievement level (+0.64 for internal locus-of-control and +0.55 for goal setting mean scores). When data from the two discrepant classes were omitted, the adjusted standardized mean difference effect sizes were found to be + 1.24 for internal locus-of-control and + 1.06 for goal setting mean scores favoring students from the lowest achievement level. Implications of these results for mastery learning programs in public schools are discussed.