Home > Research > Publications & Outputs > Using corpus methods to investigate classroom i...

Electronic data

  • 2020SmithPhD

    Final published version, 2.65 MB, PDF document

    Available under license: CC BY-NC-ND: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License

Text available via DOI:

View graph of relations

Using corpus methods to investigate classroom interaction and teacher discourse in special educational needs (SEN) classrooms: an investigation of methodological possibilities

Research output: ThesisDoctoral Thesis

Published
  • Gillian Smith
Close
Publication date2020
Number of pages342
QualificationPhD
Awarding Institution
Supervisors/Advisors
Thesis sponsors
  • ESRC
Award date17/07/2019
Publisher
  • Lancaster University
<mark>Original language</mark>English

Abstract

This thesis uses corpus methods to investigate classroom interaction in SEN classrooms. Typically research in the field of teacher talk takes a pedagogic or psychological perspective and has therefore utilised experimental or observational data on a much smaller scale than this corpus-based analysis. The advantages of such a corpus analysis is considered, including the benefits of a larger and empirical data set and automated analyses. The SEN Classrooms Corpus created for purpose of this study amounts to 52,813 words of spoken teacher-pupil interaction. Data comes from 16 lessons from two classes with two different teachers in a single SEN school over a two-year period. All interactions involve at least one teacher and groups of between three and nine pupils engaging in literacy classes with a focus upon shared reading.
As features of teacher discourse were often only vaguely defined by function in the relevant literature, a methodological process was adapted to translate these into automatic corpus queries. First, definitions were combined with definitions from contemporary English grammars in order to provide a linguistic form for each teacher discourse feature. These forms were then translated into CQP advanced syntax queries, allowing us to retrieval examples of each feature from the SEN Classrooms Corpus. Analyses in this thesis focuses upon the four most common features of teacher discourse as identified in the literature and based upon the pilot study (Smith, 2015): questions, directives, augmentative and alternative communication and feedback. Following the creation of queries, corpus methods including frequency, distribution and concordancing were used in order to assess both how often and in what contexts individual features were used within the SEN Classrooms Corpus. This, in turn, allows us to investigate exactly how teacher discourse occurs within these classrooms.
This thesis provides three major conclusions regarding the use of corpus methods to assess teacher scaffolding in SEN classrooms. First, it demonstrates how a corpus of such interactive data might be created, including important methodological considerations. Second, it provides a framework by which we might move from ill-defined features in literature to complete corpus queries that aid automated corpus analyses. Finally, the use of this unique corpus and this set of methods and queries allows us to investigate how different features of teacher discourse are used by teachers within the SEN Classrooms Corpus, including whether these uses confirm or challenge the findings of previous empirical research.