Distributed representation of misconceptions

Zachary A. Pardos, Scott Farrar, John Kolb, Gao Xian Peh, Jong Ha Lee

Research output: Contribution to journalConference articlepeer-review

4 Scopus citations

Abstract

Tutoring systems deployed at scale present an opportunity to reinvigorate the study of how misconceptions or partial understanding develops in a wide range of STEM domains by connecting to critical pedagogical theories from the learning sciences by way of a distributed representation of the learner. Using answer sequence data from three Khan Academy exercises, we generate high-dimensional vector representations of incorrect student answers using a model of distributed representation more commonly applied to natural language. After clustering wrong answers in the learned vector space, we use these clusters as the basis for analysis of student misconceptions with a quantitative comparison to manual coding and a deeper qualitative discussion based on a constructivist framework. The result is a demonstration of how big data from conventional tutoring systems can act as a bridge to more critical pedagogies from the learning sciences via a distributed, connectionist model of student concept formation.

Original languageEnglish (US)
Pages (from-to)1791-1798
Number of pages8
JournalProceedings of International Conference of the Learning Sciences, ICLS
Volume3
Issue number2018-June
StatePublished - 2018
Externally publishedYes
Event13th International Conference of the Learning Sciences, ICLS 2018: Rethinking Learning in the Digital Age: Making the Learning Sciences Count - London, United Kingdom
Duration: Jun 23 2018Jun 27 2018

Bibliographical note

Funding Information:
We thank Khan Academy for sharing their anonymized exercise data and Alan Schoenfeld for his assistance in identifying exercise topics in which misconceptions have been well studied. This work was supported, in part, by a grant from the National Science Foundation (#1547055).

Publisher Copyright:
© ISLS.

Keywords

  • Big data
  • Distributed representation
  • Misconception analysis
  • Tutoring systems

Fingerprint

Dive into the research topics of 'Distributed representation of misconceptions'. Together they form a unique fingerprint.

Cite this