Proceedings Feb 2002

relations group, and were then, over the course of a week, given four sets of tasks: (1) pretest classical analogies; (2) matching relations; (3) post-test classical analogies; and (4) crossdomain analogies. Results: ANOVA showed an improvement in performance from preto postclassical analogies and transfer of analogical reasoning skills to the novel cross-domain analogies. However, planned comparisons showed children were better able to extract and transfer procedural knowledge from causal analogies to abstract analogies than they were from abstract to causal. Conclusions: Further analyses and interpretation of findings are in progress, and will be considered in terms of