This paper is from Session 2B: Engaging young learners in using models and modelling in data investigations (Modelling Part 2)
which comes under Topic 2: Statistics education at the school level
Paper 2B3 (Thursday 12th, 14:00-15:30)
Young students’ ways of constructing and evaluating statistical models
Presenter
- Sibel Kazak (Pamukkale University, Turkey)
Co-authors
- David Pratt (University College, London, United Kingdom)
- Rukiye Gökce (Ministry of National Education, Turkey)
Abstract
We report on research in which we aimed to develop young students’ ideas about statistical models and modelling. In the study, students in a 6th grade classroom engaged in a data modelling task about jump lengths of two different paper frogs that involves formulating a statistical problem, identifying what and how to measure, deciding how to structure/represent data, comparing distributions and making predictions beyond experimental data. The data were analysed using progressive focussing. The study showed that most students tended to match “ups and downs” in the actual data when constructing their models. In evaluating models, students believed that the model data should look like the real data but had differing ideas about what made a good resemblance.