Abstract
Reform in science and mathematics curricula has moved from rote learning of facts and computation skills toward learning a structure of a subject matter. While the reasons for this shift, such as greater understanding, transfer and intellectual excitement are admirable, there is little empirical evidence to support them. A first step in examining such claims is to develop methodology for representing structure in the to‐be‐learned instructional material and in students' memories. The purposes of this paper, then, are to present (a) a method for examining a subject‐matter structure in prose materials, (b) methods for examining a subject‐matter structure in students’ memories, and (c) data bearing on the validity of structural interpretations of these measures. If these structures can be represented objectively and independently — even if not in a complete form or error free — a beginning has been made in validating the reasons for teaching structure.