This article first characterises current conceptions that separate mathematics from STE by privileging either mathematics (ste|M) or STE (STE|m), whereas education should help students to develop joint reasoning (STE↔M). A theoretical framework is proposed, breaking with the way modelling is conceived in mathematics education research and considering the intermediation role of models between an empirical referent composed of objects and phenomena as perceived by the learner, and a theoretical referent. It is inspired by research into science teaching but it differs by promoting mathematical reasoning in the learner's theoretical referent. In this framework, covariations are models that give meaning to the processes perceived in the empirical referent. Modelling also allows covariations to be treated as dependencies between mathematical variables, thereby establishing a link with the theoretical referent. Epistemological considerations highlight two important dimensions: the plurality of models, and teamwork involving different approaches. School chemistry is chosen to test this framework, as it is an area where mathematical reasoning is overlooked. Two models of the process involved in a laboratory technique are proposed, one based on a mathematical function and the other on a white-box computer simulation. Innovative tasks for grade 12 students are designed and assessed in class.
Read full abstract