Video research in a laboratory classroom: Drawing on complex multimodal data sets to characterise collaborative problem solving in mathematics

Year: 2017

Author: Esther, Chan Man Ching, Clarke, David, Petterson, Andreas, Tran, Dung, Moate, Josephine

Type of paper: Abstract refereed

Team-based problem solving and learning are regarded by governments, industry, and educational theorists as essential for effective functioning in society in the 21st century (DET, 2012; Griffin et al. 2012). However, these processes are highly complex and challenging to investigate. This paper reports research undertaken in a laboratory classroom facility at The University of Melbourne in which 10 video cameras and up to 15 audio inputs recorded the interactions of intact classes of Year 7 students and their mathematics teacher as they engaged in collaborative problem solving activities.

This paper addresses the question: How can the multimodal data generated by the facility be used to characterise and construct hypotheses regarding collaborative problem solving? A case study is reported, based on the analysis of multimodal data (student written products, videos, and transcripts) generated from the laboratory classroom facility.

Two classes of Year 7 students (12 to 13 years old; 50 students) participated in a 60-minute session in the laboratory classroom involving three separate problem solving tasks that required them to produce written solutions individually, in pairs and in groups of four students, respectively. As an entry point for data analysis, an analytical frame was developed to characterise the student written solutions in terms of response sophistication. This was supplemented by an analysis of the transcript data generated from the audiovisual record of each group with respect to sophistication in mathematical argumentation, and connections were constructed between the written product and the transcribed student speech. Analysis of the video data allowed further characterisation of the collaborative problem solving process in terms of multimodal communication and cognition.

Beginning with the structured characterisation of the collaborative problem solving process and product/s of a single group of students, hypotheses can be developed, which are consistent with analyses of the multimodal data in relation to this group. A process of hypothesis refutation and revision can then be pursued iteratively, focusing in turn on each collaborative group of four students, including the data concerning its constituent pairs and individuals, progressively refining or discarding hypotheses as each group is analysed. In this presentation, we report the first stage of this iterative process. The findings generated in the multi-layered analytical process, outlined above, are combined to form a case study of a single group. Our overall aim is to model collaborative problem solving in a way that recognises the fundamentally social nature of collaborative activity and associated learning.