Effects of an Experiment-Based Physical Science Program on Cognitive Outcomes

This study tested the effects of an experiment-based physical science program on cognitive outcomes with 556 urban students in Grades 4 through 8.1 A program theory model of evaluation with a quasi-experimental design was used in which science content knowledge and science process skills were viewed as functions of race, sex, grade, pretest scores, and program implementation, indexed by the number of experiments completed. Results indicated that the number of experiments completed had a significant influence on science process skills but not on science content knowledge. The program had limited implementation, which may explain why the program did not influence content skills. Factors affecting program implementation varied somewhat by race, but generally, teachers who experienced early implementation problems were less likely to fully implement the program. Despite some limitations, the study indicates that the program can be effective for students of relatively motivated teachers, especially if it is fully implemented. The study also shows how a program theory model of evaluation can be advantageous for testing hypothesized causes and effects.