Assessing Student Performance in a Computational-Thinking Based Science Learning Environment

摘要

Computational Thinking (CT) can effectively promote science learning, but K-12 curricula lack efforts to integrate CT with science. In this paper, we present a generic CT assessment scheme and propose metrics for evaluating correctness of computational and domain-specific constructs in computational models that students construct in CTSiM - a learning environment that combines CT with middle school science. We report a teacher-led, multi-domain classroom study using CTSiM and use our metrics to study how students' model evolution relates to their pre-post learning gains. Our results lay the framework for online evaluation and scaffolding of students in CTSiM.