This paper is from Session 6A: Conditional Probability, Part 1: Learning probability with visualisation
which comes under Topic 6: Innovations in teaching probability
Paper 6A2 (Monday 9th, 16:00-17:30)
The effect of visualizing statistical information in Bayesian reasoning problems
Presenter
- Markus Vogel (University of Education Heidelberg, Germany)
Co-author
- Katharina Böcherer-Linder (University of Education Freiburg, Germany)
Abstract
In this paper, we compare the effectiveness of the two visualizations tree diagram and unit square in increasing people’s performance in Bayesian reasoning tasks. We summarize the results of three experiments with randomized samples of about nearly 300 university students. In these experiments the students were asked to compute ratios of different set-subset relations included in Bayesian situations (experiment 1), to solve tasks by applying Bayes’ rule (experiment 2), and to assess the impact of changing base rates in Bayesian situations (experiment 3). In all experiments there was empirical evidence for the unit square being more supportive. On this base we conducted a replication study with about 140 students in six different classes of lower secondary level. The results will be presented at the conference.