ABSTRACT
Science domains are typically understood through the use of multiple models, where by model we mean a representational and reasoning system that allows one to describe, predict, or explain domain phenomena. Different forms of models serve different purposes and are designed to capture different aspects of a physical system. Nonetheless, these models can be coherently linked with one another. Science teaching typically incorporates multiple models, but it seldom addresses differences in their nature and purposes or develops the conceptuallinkages among these alternative models. The result is a set of impediments to learning science due to the complexities associated with nonintegrated collections of models that have different forms, reasoning processes, and purposes.
