Abstract:

Given the high demands in knowledge and practice of written language conventions of academia and of specific disciplines, research traditions, and accepted approaches to thesis writing, doctoral students face a daunting array of challenges in writing a thesis. Here we discuss some ideas for automated analysis of low-level features of a thesis and preliminary work using Correspondence Analysis showing differences across chapters in theses from four fields (Biology, Linguistics, Tourism, and Film Studies) according to the presence of the three types of reporting verbs studied by Hyland (2002), i.e. those expressing research acts, cognitive acts, and discourse acts. The analysis illustrates the method and is suggestive of its potential for pointing up differences in thesis structure that might be of value for thesis students and their supervisors.

Author: Oliver Mason, Martha C. Pennington