Please use this identifier to cite or link to this item:
Metadiscourse in research articles : a comparative study across disciplines and research paradigms
Issue Date: 
Research into academic writing has shown that discourse conventions are shaped by a variety of complex socio-cultural factors. As part of discourse conventions, metadiscourse in academic writing is found to be affected by the knowledge-making practices of different disciplinary fields. While writers from the natural sciences and the humanities differ in their use of metadiscourse, it is largely unclear whether those writers from individual disciplines within the social sciences may show any variability in this interpersonal dimension. In addition, due to the importance of research paradigms in knowledge creation and representation, it is expected that a particular research paradigm and its associated epistemological assumptions may exert influences on the academic discourse in reporting research in that paradigm. This study has set out to investigate whether and how the use of metadiscourse in quantitative, qualitative, and mixed methods research articles (RAs) may vary across
the three social science disciplines of applied linguistics, education and psychology. The study adopts a mixed methods design and combines a primary corpus-based analysis with a complementary semi-structured interview study. The corpus was comprised of the post-method sections of 180 published RAs sampled from a number of internationally prestigious journals in applied linguistics, education, and psychology. Using an analytical framework adapted from Hyland’s (2005b) metadiscourse model, I manually coded all metadiscoursal features in the post-method sections of the RAs and performed both quantitative and qualitative analyses. On the basis of the corpus-based research results, a semi-structured interview study was conducted with six specialist informants from the three selected disciplines. The interview data were used to supplement the discussion of the corpus findings. The results showed that there are important disciplinary and paradigmatic differences in the use of both interactive and interactional metadiscourse in the post-method sections of the RAs. Across the disciplines, the applied linguistics RAs
differed significantly from the psychology RAs in the use of a number of interactive and interactional metadiscourse while the education RAs took a middle position, showing convergence and divergence from the applied linguistics and the psychology RAs in metadiscourse use. Such disciplinary discrepancies can be accounted for by specific knowledge-knower structures prevailing in each of the three disciplines. Across the research paradigms, quantitative RAs used a range of interactive and interactional metadiscoursal features significantly more frequently than qualitative RAs. Furthermore, the mixed methods RAs occupied a middle ground, sharing both similarities and differences with both the quantitative and the qualitative RAs in the use of metadiscourse. These paradigmatic differences could be explained by the
contrasting epistemological stances between quantitative and qualitative research paradigms as well as paradigm-specific knowledge-making practices. These general patterns of metadiscourse use were corroborated by the insider accounts from my specialist informants. A number of implications can be drawn from this study for both novice writers of RAs and for teaching and learning academic writing in courses such as English for academic/specific purposes. The thesis concludes with possible avenues for further research.
Issued Date: 
Call Number: 
P128.P37 Cao
Appears in Collections:Doctor of Philosophy (Ph.D.)

Files in This Item:
File Description SizeFormat 
CaoFeng-PHD.htmAbstract6.96 kBHTMLView/Open
CaoFeng-PHD.pdfFull Text3.62 MBAdobe PDFView/Open
Show full item record

Page view(s) 5

Last Week
Last month
checked on Dec 9, 2018

Download(s) 50

checked on Dec 9, 2018