Reasoning from data: How students collect and interpret data in science investigations

Abstract This study explored the understandings of data and measurement that school students draw upon, and the ways that they reason from data, when carrying out a practical science inquiry task. The two practical tasks used in the study each involved investigations of the relationships between two independent variables (IVs) and a dependent variable (DV); in both tasks, one IV covaried with the DV, whereas the other did not. Each was undertaken by 10 students, aged 10, 12, and 14 years (total n = 60 students), working individually. Their actions were video‐recorded for analysis. In a subsequent interview, each student was asked to discuss and interpret data collected by two other students, undertaking a similar (but different) practical task, shown on a video‐recording. An analysis of the sample students' performance on the practical tasks and their interview responses showed few differences across task contexts, or with age, in students' reasoning, but significant differences in performance when investigating situations of covariation and non‐covariation. Few students in the sample displayed sufficient understanding of measurement error to deal effectively with the latter. Investigations of non‐covariation cases revealed, much more clearly than investigations of covariation cases, the students' ideas about data and measurement, and their ways of reasoning from data. Such investigations therefore provide particularly valuable contexts for teaching and research. © 2004 Wiley Periodicals, Inc. J Res Sci Teach 41: 748–769, 2004