Date of Award:


Document Type:


Degree Name:

Doctor of Education (EdD)


Instructional Technology and Learning Sciences

Department name when degree awarded

Curriculum and Supervision (Instructional Technology)


Ron Thorkildsen


The study investigated the effect of a videodisc-based, teacher-controlled, direct instruction-based program in fractions content, on self-concept. Self-concept was operationally defined as scores on a slightly modified version of Marsh's Self-Description Questionnaire (Marsh, 1988). A quasi-experimental, nonequivalent control group design was used to compare the self-report self-concept of two groups of upper elementary students (N = 337). The treatment group (n = 171) received instruction in fractions via the teacher-directed, videodisc-based, Mastering Fractions program (Systems Impact, 1986a). The control group (n = 166) received their normal grade four or grade five mathematics program, but did not include common fractions.

Differences in achievement scores provided support for previous findings regarding the Mastering Fractions program. The treatment group covariance-adjusted mean on a criterion-referenced test was higher than that of the control (5.9 standard deviations). Differences in achievement test scores among the treatment classes varied directly with the levels of program implementation across classes.

The data were examined using both the student and the class as the unit of analysis. Using the student as the unit of analysis, the treatment group mathematics self-concept covariance-adjusted mean was 0.22 standard deviations above that of the control group. An analysis of raw gain scores yielded a standardized mean difference effect size between the treatment and control group scores of +.12. A statistically significant but small main effect was also noted across student pretest achievement levels. The posttest difference between low-achiever means treatment versus control students is slightly larger than the difference between high-achiever means. No statistically significant interaction was noted between student achievement level at pretest and treatment condition.

The class was also used as the unit of analysis. In this case the mean difference effect sizes between experimental groups were +0.86 (ANCOVA) and +0.34 (raw gain scores).

Differences were small to moderate, but consistent with the study hypotheses. Recommendations are presented regarding future research and the use of direct instruction in school settings. (197 pages)