Effect of graphic representations on Bayesian inference problem solving among junior high school students: the role of problem context and spatial ability
摘要
The natural frequency facilitate effect in Bayesian reasoning is well-established, but the role of graphic representation remains debated. Probability is a core component of mathematics education, and enhancing probability literacy in junior high students is crucial. In specific contexts, these students can grasp Bayesian rules, with their graphic comprehension closely linked to spatial ability. This study involved junior high students using Bayesian tasks on breast cancer and lie problem, coupled with spatial ability questionnaires. Two experiments examined the impact of graphic representation on students’ cognitive mechanisms and reasoning strategies. The findings reveal that graphic representation significantly influences probabilistic reasoning, with better performance in familiar contexts. In unfamiliar contexts, graphics facilitate reasoning, particularly with unit square, where students outperform those using tree diagrams. High spatial ability students excel in interpreting unit square. Predominantly, their reasoning aligns with Bayesian and conservatism strategies, though it varies by context and spatial ability.