Chang, Jina

Guiding students’ STEM problem solving entails dynamic processes driven by changes in real-world contexts. To understand these processes, we aimed to identify the formation and influence of ‘norms’ as shared behaviour patterns desirable in STEM problem-solving. To this end, 10 sessions of STEM lessons for secondary students were carried out, and we collected data comprising lesson observation videos, fieldnotes, interviews, and photographs of students’ artifacts. The data were analysed based on three features of norms: justifiability, sharing, and behaviours. The results report three epistemic norms in STEM problem-solving. First, the norm of ‘defining a real-world problem that is useful and accessible’ was found. While the students searched for useful problems in their everyday lives, they also needed to ensure that these problems were scoped such that they had the capacity to manage them. The second norm was ‘designing creative and concrete prototypes’. Students’ prototypes were expected to be creative in addressing the established problems and to be developed in a concrete manner. The last norm identified was ‘testing and revising in more iterative and feasible ways’. The students tested their models repeatedly in a manner aligning with their skills and the materials provided. Based on the findings, educational implications are discussed in terms of understanding and facilitating STEM problem-solving.

The purpose of this study is to identify the features of student-generated visual models in an online learning platform. To this end, online-based learning activities were designed and applied to 123 primary school students. Specifically, the students were guided to generate visual models for three phenomena related to light and shadow, following three steps: (1) observing a phenomenon in a video and constructing the first visual model of the phenomenon, (2) evaluating two different models in a concept cartoon and choosing the better model, and (3) observing another similar phenomenon and constructing the second model. Six visual models per student (738 visual models in total) were collected and analysed in terms of using proper symbols and conceptual understanding. In using symbols to visualise how light travels, students were found to employ increasingly higher levels of symbols over the course of constructing the six visual models. In terms of conceptual understanding, students demonstrated their conceptual development in the visual models they used to explain simple phenomena; however, for complex phenomena, the development of the conceptual levels of their models was challenged. Based on the above results, educational implications are discussed in terms of fostering students’ visual models in an online environment.