ABSTRACT
Although many computational models have successfully simulated the results of controlled psychological experiments, few researchers have attempted to apply their models to complex, realistic phenomena. In this study, MAC/FAC (“many are called, but few are chosen”), which models two stages of analogical reasoning (Forbus, Gentner, & Law, 1995), was applied to our experimental data. In our experiment, subjects were presented a cue story and asked to retrieve cases learned from everyday life. Next they rated the inferential soundness (goodness as an analogy) of each retrieved case. For each retrieved case, we used the algorithms of the MAC/FAC to compute two kinds of similarity scores: content vectors and structural evaluation scores. As a result, the computed content vectors explained the overall retrieval of cases well, whereas the structural evaluation scores had a strong relation to the rated scores. These results support the MAC/FAC’s theoretical assumption – different similarities are involved in the two stages of analogical reasoning.
