Abstract
The present study was designed to test the effectiveness of a constructivist–historical teaching strategy in changing students' misconceptions about the earth's shape and gravity at the upper elementary and middle school levels. The experimental treatment was an astronomy unit in the series Great Explorations in Math and Science (GEMS), entitled Earth, Moon, and Stars, which provides opportunities for students to clarify and articulate their understanding of the earth's shape and gravity concepts, and to modify their ideas through class discussions, creative activities, observations of the sky, and manipulating concrete models to help them explain the phenomena that they observed. The study included 539 students from 18 classrooms in 10 different states. Most of the teachers learned to present the treatment and administer and score the assessment instruments at a summer institute sponsored by the National Science Foundation. The primary experiment was a treatment-group-only design, in which the teachers administered the same test to all students before and after the treatment. The purpose of this experiment was to determine the impact of the treatment on students' understanding of the earth's shape and gravity concepts. Data were analyzed in three age groups: fourth and fifth graders; sixth graders; and seventh and eighth graders. As expected from previous studies, on the pretest all classes displayed a wide variety of conceptions about the earth's shape and gravity. After the unit, however, the number of subjects who held misconceptions was far fewer. Chi-square analyses showed that a significant number of the students at all grade levels shed their misconceptions concerning both the earth's shape and gravity. A surprising finding was that younger subjects responded more positively to the experimental treatment than older students, so that, after instruction, fourth and fifth graders were as knowledgeable as seventh and eighth graders concerning the earth's shape and gravity. An additional study was conducted at two of the sites, in which pre- and posttests were administered to all subjects, but some subjects received the treatment and others did not. The treatment group showed significant gains, whereas the experimental group showed no significant difference, thus ruling out the effect of maturation and the effect of the test itself as alternative explanations for improvement. The results of this study support a constructivist–historical approach as fruitful in developing instructional methods and materials to help students change core concepts that are normally highly resistant to instruction. The study concludes with suggestions for further research, and for placing instruction on this topic at a strategic point in the school curriculum.
