Howell-Richardson, Christina; (2004) Towards a framework for the analysis of CSCL (computer supported co-operative learning) discourse. Doctoral thesis , Institute of Education, University of London.

	Text 411660.pdf - Accepted Version Access restricted to UCL open access staff Available under License Creative Commons Attribution Non-commercial Share Alike. Download (25MB)
Preview	Text HOWELL-RICHARDSON, C._Redacted.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial Share Alike. Download (18MB) | Preview

Abstract

The thesis aims to develop a possible description of electronic discourse in CSCL through a data-driven description of the linguistic behaviour and discourse strategies of 4 groups of postgraduate students engaged in an asynchronous CSCL task during-February 2000 and February 2001. The study develops an analytic framework for the coding of the messages. The framework consists of three levels, with a default inheritance relationship between these levels. The top level concerns the aim of the messages, identified within the broad context of Levinson's Activity Type. The mid-level consists of the traditional conversational analysis categories, with some minor adaptations to the CMC medium. The third level is based on a neo-Gricean approach to utterance interpretation, with special attention to Levinson's (2000) theory of generalised conversational implicature. The analysis was conducted through intensive reading of the coded data to identify categories of speaker behaviour. The categories were then collated to address the research question. 19 categories were identified, covering 4 aspects of discourse behaviour. As an additional test of the discourse analysis framework, the coded output was used as data for a separate theory-driven question. The question was to seek evidence of behaviour typical of the iterative dialogue that characterises Laurillard's (2002) model of learning through conversational dialogue. The research study found that the majority of the discourse categories identified by the framework are valid, although some need to be refined. In particular, 4 basic message structure types, and distinctive patterns in the use of indirect and direct forms of expression are clearly identified in this data. There are also clear indicators of strategies used to maintain cohesion and coherence. In the test case, the coded data was used to identify six types of critical learning behaviour that are broadly consistent with Laurillard's model of learning.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as XMLCSVJSON

Downloads by country - last 12 months

Export as XMLCSVJSON

Archive Staff Only

View Item