Developing Data Visualisation Techniques for Enhancing Professional Development: Viewing Patterns of Discourse and Disciplinarity in Classroom Corpus Data

Project Number
OER 38/08 PD

Project Duration
April 2009 - February 2012

Status
Completed

Abstract
Data visualisation consists of a range of technologies designed to reveal the latent patterns and structure in large data sets, such as textual corpora, in a way that is extremely difficult or impossible to do by other means. As such, we envisage a role for data visualisation techniques in mining the corpus of classroom interaction data collected by CRPP in its first round of funding (Hong, 2006). Two groups of people will benefit from this: (1) academic colleagues in NIE interested in researching classroom interactions, and (2) teachers in schools wishing to explore and critically reflect on their own performance. This exploratory project will comprehensively review a wide spectrum of visualizations of the rich data existing in the Singapore Corpus of Research in Education (SCoRE), ranging from relatively simple distribution plots for elements such as open and closed questions, to more sophisticated mappings of the distribution of lexical items indicative of academic and 'everyday' registers, collocational patterns around key words and higher level textual patterns (Scott & Tribble, 2006). We intend to supplement this analysis with one that implements these textual patterns as computer generated visualizations, drawing on work in topic mapping and knowledge web construction (Orbst & Liu, 2003). A central focus will be on developing ways of visualising the patterning of linguistic elements to show their variance across subjects, levels and grades such that they contribute to the investigation of pedagogic strategies and disciplinarity. In its early stages, the project will explore ways of matching the technologies already developed for oweb-based corpus query tool 1 with others available for visualising large data sets. The deliverable will be a web-based interface: a software tool that enables users to select parts of the corpus (from a single transcript to a set of transcripts representing a significant subset of the corpus, e.g. Primary 5 Maths lessons) and view the patterns therein. In subsequent phases of the project, a version of this tool will be trialled with a small group of teachers who, following a recording of their lesson on digital video, would be able to view their lesson in terms of its discourse, pedagogic and knowledge patterns within 48 hours. They would also be able to compare their lesson with the data of the many lessons held in the corpus, thus enabling norm-based comparisons. We envisage this tool, therefore, being an aid to critical reflection. The second deliverable will be a set of structured interviews leading to an evaluation of the tool and a set of practitioner articulations of pedagogic and disciplinarity knowledge which are likely to feed back into a refinement of the tool and possibly the theoretical base of other research projects dealing with disciplinarity. The project can therefore be envisaged in three phases. Hong, H. 2006. Patterns of Students' Talk and Teachers' Modes of Participation across Curriculum: a corpus-based study. Proceedings of the International Conference of the American Association of Applied Corpus Linguistics (AAACL), Flagstaff, Arizona, USA, October 20 -22, 2006. Orbst, L. & Liu, H. 2003. Knowledge representation, ontological engineering, and topic maps. In J. Park and S. Hunting (Eds.), XML Topic Maps, Creating and Using Topic Maps for the Web. Boston: Addison Wesley. Pearson Education, 103-14. Scott, M. & Tribble, C. 2006. Textual patterns. Key Words and corpus analysis in language education. Amsterdam, John Benjamins 1. A Corpus Based Interactive Vortal for Reflective Practice and Professional Development, Funded Project CRP6/07DH, 2006-2007.

Research Themes
ICT Integration

Funding Source
NIE

Related Links
ReEd Vol 6: Visualizing Teacher Talk

Related Projects