Analogy events: How examples are used during problem solving

Research output: Contribution to journalArticlepeer-review

80 Scopus citations


In attempting to fit a model of analogical problem solving to protocol data of students solving physics problems, several unexpected observations were made. Analogies between examples and exercises (a form of case-based reasoning) consisted of two distinct types of events. During an initialization event, the solver retrieved an example, set up a mapping between it and the problem, and decided whether the example was useful. During a transfer event, the solver inferred something about the problem's solution. Many different types of initialization and transfer events were observed. Poor solvers tended to follow the example verbatim, copying each solution line over to the problem. Good solvers tried to solve the problem themselves, but referred to the example when they got stuck, or wanted to check a step, or wanted to avoid a detailed calculation. Rather than learn from analogies, both Good and Poor solvers tended to repeat analogies at subsequent similar situations. A revised version of the model is proposed (but not yet implemented) that appears to be consistent with all the findings observed in this and other studies of the same subjects.

Original languageEnglish (US)
Pages (from-to)347-388
Number of pages42
JournalCognitive Science
Issue number3
StatePublished - 1998
Externally publishedYes

ASJC Scopus subject areas

  • Experimental and Cognitive Psychology
  • Cognitive Neuroscience
  • Artificial Intelligence


Dive into the research topics of 'Analogy events: How examples are used during problem solving'. Together they form a unique fingerprint.

Cite this