Effects of Activity-based Elementary Science on Student Outcomes: A Quantitative Synthesis

TLDR

This meta‑analysis synthesized research on the effectiveness of three federally funded activity‑based elementary science programs (ESS, SAPA, SCIS) across 57 controlled studies in over 900 classrooms. Across 57 controlled studies, the meta‑analysis found a moderate overall effect size of .35, with larger gains in science process (.52) than content (.16) or affective (.28) outcomes, and additional benefits in creativity, intelligence, language, and mathematics, with disadvantaged students benefiting most, unbiased tests yielding smaller effects, published reports showing higher effects, and program‑specific curricular emphases influencing outcomes, while follow‑up studies found no lasting advantage of early activity‑based programs.

Abstract

In this study, meta-analysis techniques were used to synthesize research on the effectiveness of three major activity-based elementary science programs (ESS, SAPA, and SCIS), which were developed with federal support. In 57 controlled studies, outcomes were measured in over 900 classrooms; the overall mean effect size for all outcome areas was .35. The mean effect size was .52 for science process tests, .16 for science content, and .28 for affective outcomes. On the average, gains also were realized in creativity, intelligence, language, and mathematics. Only 3 of 14 coded study features were related to reported effects: Disadvantaged students derived greater benefits than other students; tests not biased in favor of the activity-based programs resulted in positive but lower effects than those favoring the activity-based approach; and published reports had higher effects than unpublished reports. The effects of particular programs reflect their relative curricular emphases. In three followup studies, student groups that had had activity-based programs in elementary school and had later experienced traditional science programs during middle school years could not be consistently distinguished from control groups.