Multitheoretic research designs are increasingly seen as a useful approach to connect and contrast multiple theoretical perspectives on the same research settings or in relation to the same phenomena (Bikner-Ahsbahs & Prediger, 2014). This paper discusses the issues associated with the choice of unit of analysis in multitheoretic research designs with illustrative findings from the Social Unit of Learning project. The paper addresses the questions, “What was the decision-making process involved relating to the selection of the units of analysis in the project?" and "What are the possible consequences where the units of analysis employed in parallel analyses are the same or different?” Method The project was conducted in a laboratory classroom equipped with 10 built-in video cameras and up to 32 audio channels. Two classes of Year 7 students (12 to 13 years old; N = 50) each with their usual teacher participated in a 60-minute session involving three separate problem solving tasks. The laboratory classroom allows the generation of continuous video and audio records of every student in the class. All student written work was also collected. Several parallel analyses are currently being undertaken on the data collected, including analysis of the negotiative foci of student exchange (Clarke and Chan); sophistication in mathematical student-student exchange (Tran); student motivating desires (Tuohilampi); students' repertoires in participation (Moate); and dialogic talk between students (Díez-Palomar). The primary data source for all analyses was the transcript of all student interactions, supplemented by the video record of those same interactions and by student written material produced during those interactions. Findings Several analytical approaches were considered, including: coding using a fixed, common transcript unit as the unit of analysis (e.g., negotiative events or speaker turns); the interpretive annotation of interactions; or the narrative reconstruction of interactive sequences. Each of these approaches requires a different unit of analysis. However, some of the analyses (e.g., negotiative focus and mathematical sophistication) required the same or very similar units of analysis. The degree of correspondence between the different units of analysis has implications for the connections that might be made between the parallel analyses. Strategic consideration of the units of analysis can facilitate the identification of connection and distinction between multiple theories and therefore enhance our understanding of both the theories and the phenomenon of interest. The explication of the issues of the units of analysis in this paper should contribute to the further theorisation of multitheoretical research approaches.

Bikner-Ahsbahs, A., & Prediger, S. (Eds.). (2014). Networking of theories as a research practice in mathematics education. Cham, Switzerland: Springer.