Learning by analogical bootstrapping

Analogies are typically drawn from a well-understood situation to a situati on that is poorly understood. In this research, we investigate a different route to analogical insight. We suggest that mutual alignment-that is, comp arison between 2 partially understood situations-can act to promote compreh ension and abstraction. We presented participants with 2 analogous scenario s depicting heat flow. They were given processing tasks that varied in the degree to which comparison was required. We then measured insight into the common structure in 3 ways. Participants were asked to (a) specify differen ces between the 2 pictured scenarios, (b) write scenario descriptions, and (c) rate the similarity of the 2 scenarios. The results show that carrying out comparison promotes greater insight into the common causal structure, b ut only when the comparison is intensive. The best results were obtained wh en participants were asked to jointly interpret the scenarios and to list s pecific correspondences. In a second experiment designed to further pinpoin t the source of the comparison advantage, participants were asked to make c orrespondences between the elements of the 2 scenarios. These results sugge st that mutual alignment is an effective means of promoting insight.