Using spatial analogy to facilitate graph learning

Linsey A. Smith*, Dedre Gentner

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Scopus citations


Graphical depictions of complex interactions pose a challenge to spatial reasoning. In this research, we tested whether analogical processes can be harnessed to help students learn to solve complex graphical reasoning problems. Specifically, we asked whether a brief training experience using spatial analogies could help students learn about stock-and-flow graphs. The basic idea of our intervention was to juxtapose contrastive graphs and encourage students to compare them. In two studies, we test the following predictions derived from structural alignment theories of analogy: (1) comparing contrastive graphs during training will lead to better performance in a graph-understanding task than will studying the same exemplars sequentially; and (2) comparing high-similarity pairs will lead to better performance than will comparing low similarity pairs. The results support both of these predictions, indicating that even a brief analogical comparison task can confer relational insight. Further, these results corroborate prior evidence that a structural alignment process underlies analogical comparison.

Original languageEnglish (US)
Title of host publicationSpatial Cognition VIII - International Conference, Spatial Cognition 2012, Proceedings
Number of pages14
StatePublished - Nov 6 2012
EventInternational Conference on Spatial Cognition, SC 2012 - Kloster Seeon, Germany
Duration: Aug 31 2012Sep 3 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7463 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


OtherInternational Conference on Spatial Cognition, SC 2012
CityKloster Seeon


  • Analogical Comparison
  • Analogy
  • Graphical Reasoning
  • Spatial Learning
  • Structural Alignment

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)


Dive into the research topics of 'Using spatial analogy to facilitate graph learning'. Together they form a unique fingerprint.

Cite this