Louca, Loucas T.

The purpose of this study was to investigate whole classroom discourse during modeling-based learning in science, seeking to describe the discourse's characteristics, its relation to the micro-context in which it took place and to the student-constructed models, and to ascertain when it becomes productive. Additionally, we aimed to describe how whole classroom modeling discourse in science may be better supported by the teacher as well as the role of the modeling tool. In doing so, we analyzed student conversations and student-constructed models from two groups of 11- to 12-year-old students. All students used a computer programming environment, namely Stagecast Creator, as a modeling tool. Findings revealed three distinct discourse types (modeling frames): (a) (initial) phenomenological description, (b) operationalization of the physical system's story, and (c) construction of algorithms. Finally, we discuss what we consider productive modeling discourse, as well as the contributions of the software and the teacher in this respect.

The purpose of this study was to develop a framework for analyzing and evaluating student-constructed models of physical phenomena and monitoring the progress of these models. Moreover, we aimed to examine whether this framework could capture differences between models created using different computer-based modeling tools; namely, computer-based programming environments which, in prior research, were found to differ in various aspects of the models constructed through them. We analyzed 220 computer-based models of physical phenomena developed by two groups of elementary-school students. Using open coding we developed a framework that includes five elements of scientific models that code for representations of: (i) the physical objects; (ii) the physical entities; (iii) the object behaviors; (iv) the interactions among physical objects, physical entities, and object behavior(s); and (v) the accuracy of the phenomenon depiction. The implementation of this framework confirmed that it can differentiate student-generated models according to their sophistication and structural components, independent of the computer-based programming environments used to create the models.