New data and prior belief: The two faces of scientific reasoning

Kuhn (1989) has argued that at the heart of the ability to reason scientifically is the process of differentiating existing mental models (i.e. theory) from new data. In this regard she has proposed a developmental sequence in which, in the early stages, theory and data are fully integrated and are used interchangeably. Later, when theory and data are compatible, they tend to be moulded together as ‘the way things are’, but when they are incompatible conflict is avoided by the use of strategies which bring the two into line: these strategies often include selective attention to the data. Only at the upper levels of this developmental spectrum are theory and data conscientiously differentiated, with each being used to reflect on the other. This paper analyses the responses made by Year 11 students to problems which required them to evaluate a prediction based on some provided data. The problems were set in two contexts, one scientific and one social, and the predictions to be evaluated combined plausibility/implausibility and validity/invalidity. The response patterns were very similar to those described by Kuhn, and the implications of this for teachers, especially those attempting to use conflict based teaching approaches, are developed. Specializations: science teacher education, scientific problem solving, changing students’ alternative conceptions. Specializations: psychological theories applied to science education.