Date of Original Version
COGNITION AND INSTRUCTION 21(2) 2003 149-173
Abstract or Table of Contents
The widespread availability of computers in elementary schools makes them an appealing option for presenting instructional materials in laboratory science. However, there are neither theoretical nor empirical grounds for predicting whether virtual or physical presentation of instructional materials will be more effective. The definition of "active manipulation" is poorly specified and there are few studies that directly compare the two approaches unaccompanied by other potential confounds. In this study, 4th- and 5th-grade children were taught how to design simple unconfounded experiments using 1 of 2 instructional methods differing only in whether children manipulated physical or virtual materials. The 2 types of materials were equally effective in achieving several instructional objectives, including the design of unconfounded experiments, the derivation of correct predictions from them, and explicit reference to the need for experiments to be unconfounded.