The role of model comparison in young learners’ reasoning with statistical models and modeling

Michal Dvir, Dani Ben-Zvi

Research output: Contribution to journalArticlepeer-review

Abstract

The goal of this study is to explore the role of model comparison, which is a key activity of young learners’ informal reasoning, with statistical models and modeling in the context of informal statistical inference. We suggest a framework to describe this reasoning (the RISM framework), and offer an illustrative case study of two-sixth graders showcasing its utility. In particular, we illustrate the benefit of untangling the informal modeling process into three separate, though not independent, modeling processes: modeling a conjecture, modeling data, and comparing them by means of a comparison model. This case study shows the possible progression of a comparison model, and its potential role as a catalyst for the development of the other two modeling processes. Finally, an expansion of our initial framework is discussed, highlighting the centrality of model comparisons.

Original languageEnglish
Pages (from-to)1183-1196
Number of pages14
JournalZDM - International Journal on Mathematics Education
Volume50
Issue number7
DOIs
StatePublished - 1 Dec 2018

Bibliographical note

Funding Information:
We thank the University of Haifa, the I-CORE Program of the Planning and Budgeting Committee and the Israel Science Foundation Grant 1716/12 for supporting this research, as well as the Connections research team.

Funding Information:
Program of the Planning and Budgeting Committee and the Israel Science Foundation Grant 1716/12 for supporting this research, as well as the Connections research team.

Publisher Copyright:
© 2018, FIZ Karlsruhe.

Keywords

  • Informal statistical inference
  • Model comparison
  • Statistical model
  • Statistical modeling

ASJC Scopus subject areas

  • Education
  • Mathematics (all)

Fingerprint

Dive into the research topics of 'The role of model comparison in young learners’ reasoning with statistical models and modeling'. Together they form a unique fingerprint.

Cite this