This paper is from Session 2E: Linking research and practice in teaching and learning statistics at the school level
which comes under Topic 2: Statistics education at the school level
Paper 2E2 (Tuesday 10th, 16:00-17:30)
Students' reasoning about association of categorical variables
Presenter
- Stephanie Casey (Eastern Michigan University, United States)
Co-authors
- Lauren Ridley (Eastern Michigan University, United States of America)
- Rick Hudson (University of Southern Indiana, United States)
Abstract
Statistical association is an important concept in statistics. An exploratory study examined how students reason about statistical association of categorical variables using both numerical and graphical representations. Task-based interviews were conducted with thirteen students ages 11 to 13 prior to formal instruction. When prompted to make a graph to display the conclusions they reached numerically, successful students made side-by-side pie charts or segmented bar graphs, but most students were unable to produce a meaningful graph to compare the variables. When later asked to interpret previously constructed graphs, they were most successful with segmented bar graphs and least successful with eikosograms. When students failed to interpret the graph correctly, they often interpreted the percentages with incorrect referents. These results have curricular and software implications.