Second-graders' predictive reasoning strategies

This paper reports predictive reasoning strategies used by ten second-graders in a classroom design study. A modelling activity based upon real data required students to predict maximum monthly temperatures for the current year using the natural variation provided by readings from the previous six years. The development of reasoning strategies was documented throughout the lesson sequence by analysis of responses to written prompts, videos of interviews and student drawn graphs. Student predictions reflected an emerging understanding variability, clusters and mean. Reasoning strategies became increasingly sophisticated using TinkerPlots, and with repeated opportunities for students to observe, represent, reflect upon trends in data.