ABSTRACT
Descriptions of practice are selective, therefore instructional designs based on alternative descriptions of practice likely support learning in different ways. This empirical exploration of the relationship between descriptions of experimental practice and the learning they support utilizes two concurrent sixth grade design experiments. Each design experiment was based on a practice theory of instruction stemming from one of two divergent descriptions of experimentation. One view of experimentation emphasizes the logic of controlling variables, whereas an alternative description locates instructional leverage in precisely those aspects of experimentation necessarily backgrounded by a focus on logic, that is, in the mundane actions of constructing experimental apparatus. Learning outcomes documented through multiple assessments identify ways in which divergent descriptions of practice support learning in different ways. The results suggest that logic may not play the role often assumed in learning to experiment and point to apparatus construction as an activity from which students glean some flexible capabilities that may be educationally fundamental to experimentation.
