How Complex or Abstract Are Science Learning Outcomes? A Novel Coding Scheme Based on Semantic Density and Gravity

Abstract

There has been a longstanding interest in the kinds of scientific knowledge that primary science learners must know and be able to do, which comprise the intellectual demands in this subject. These prescriptions chiefly take guidance from national curriculum documents, especially in the form of their learning outcomes (LO) or learning standards. Using the concepts of semantic density (SD) and semantic gravity (SG), we formulate a novel coding scheme for primary science LO based on Semantics and Legitimation Code Theory. We demonstrate how SD and SG provide insights into the levels of complexity and abstraction respectively from a mix of qualitative and quantitative criteria that we devised. We empirically test the utility of this coding scheme by comparing present reformed primary science LO with their previous versions across three East-Asian regions. It was shown that their LO were not significantly different over versions in terms of SD/SG, had typically one to two learning points, favoured more context-dependent expressions, and were predominantly coded as SD-SG+. This research provides a complementary method of determining the intellectual demands of science curricula in terms of complexity and abstraction of LO that has implications for science teaching as well as the improvement of epistemological access for learners.