Trends for third-and fourth-grade students were identified from data collected in a longitudinal study of how children learn science concepts. Approximately 325 children from three school districts participated in this study from their first week in kindergarten through the end of sixth grade. A heuristic model of science learning was developed to represent entering ability, home background, home support, classroom processes, instructional material characteristics, and end-of-year performance. Data were then collected in each of these areas. LISREL analyses were performed to produce a structural model for science learning at each grade level. Results showed that at the third-grade level, only the amount of school activities brought home played an active role in the model. At the fourth-grade level, no measure of home support affected student performance. Classroom instructional variables were endogenous in the third- and fourth-grade models. Independent work made a significant positive difference in end-of-year performance at the third-grade level but a significant negative difference at the fourth-grade level. In addition, at both levels, general instructional variables affected end-of-year verbal performance but neither time nor other instructional variables (save individual criticism and terminal feedback) made a difference in science learning. Results are discussed generally in terms of the shift away from the significant difference in science performance made by home support after second grade and the impact instruction had on verbal ability but not on science performance.
|Name||CSR Technical Report|