On quagmires, philosophical and otherwise: A reply to Phillips

Abstract

This reply addresses three related problems in research on cognitive structure identified by Phillips: (a) the difficulty (impossibility for Phillips?) of inferring cognitive structure from observed behavior, (b) the confounding of cognitive structure with subject‐matter structure, and (c) the validity of construct interpretations of cognitive‐structure measurements. Inference from behavior to cognitive structure can be made if done carefully and cautiously under adequate constraints placed on the data. Cognitive structure is not confounded with subject‐matter structure; to the contrary, cognitive structure has been defined as the student's public understanding of the subject‐matter structure. Extensive prior research has supported cognitive‐structure interpretations of word association, similarity judgment, card sorting, and graph building measurements. Additional criticisms also are discussed.