A comparative analysis of research abstracts written by novice and professional writers :a synergy of genre-based and corpus-based approaches

IPhD ThesisResearch abstracts can serve as both informational and promotional tools to attract readers’ attention. Although numerous studies have confirmed that there are disciplinary differences in the rhetorical patterns commonly found in research abstracts, little attention has been paid to the comparison of abstracts written by professional and novice writers. In this study research abstracts by professional and novice writers in the Education and Applied Linguistics fields were analyzed. The English for Specific Purposes (ESP) approach to textual analysis was used to investigate the rhetorical move patterns of the abstracts, while a corpus-based approach was adopted to examine lexical choices. Two corpora were compiled: a Novice Abstract Corpus (NAC) consisting of 150 abstracts written by Master’s degree students, and a Professional Abstract Corpus (PAC) consisting of 378 abstracts taken from four international peer-reviewed journals. By using the five-move analysis framework developed by Hyland (2000), similarities and differences in the abstracts written by the two groups were examined in relation to three different issues: (1) the various combinations of moves that can be involved in the composition of an abstract; (2) the different structural patterns or orders of the moves; and, (3) the status of the various kinds of moves, in terms of whether they are optional, conventional or obligatory. The findings revealed that a greater variety of move compositional types were found to occur in the PAC. The result of a chi-square test showed that there is a statistically significant difference between the two corpora in the frequency of the seven compositional types found in both corpora. Similarly, the results showed that there is a greater variety of structural patterns found in the PAC than in the NAC (163 patterns: 33 patterns). In addition, variation in the frequency of occurrence of the five types of moves across the two corpora demonstrated that the NAC and the PAC exhibited differences in the status of only one move; that is, the Product Move was found to be obligatory in the NAC, but only conventional in the PAC. After the move analysis, all the abstracts in both corpora were tagged according to their moves. The lexis of the two corpora and their constituent moves was analyzed, using corpus software to generate wordlists, keyword lists, and lists of recurring n-gram clusters (or lexical bundles). The word frequency lists revealed some similarities and differences between the top twenty most frequent lexical words in each corpus. The degree of similarity and difference ii was evaluated using an effect size measure (log-ratio). In comparing the frequency of words in the two corpora in order to identify keywords in each, NAC keywords with zero occurrence in the PAC were identified that represent characteristic aspects of the research focus of the novice writers. The results of the lexical bundles analysis revealed a number of distinctive features in each specific move, such as how novice and professional researchers differently motivate their research in the Introduction Move. Pedagogical implications are discussed, in terms of how the results of the study can be used to raise teachers’ and students’ awareness in relation to conventional structural patterns and the use of lexis in abstract writing, to help prepare novice researchers for scholarly publications.Thammasat Universit