Metadiscourse in undergraduate students' essays

Abstract

Metadiscourse refers to linguistic resources that are used to organise a text and also connect with readers. The use of metadiscourse markers in academic writing helps students produce coherent texts. This study explored and analysed the types, frequency, and appropriate use of metadiscourse markers in the essays of undergraduate students who were second-language learners of English. To achieve these objectives, mixed-methods and corpus-analysis approaches were employed. Quantitative data from the results of a cloze test and qualitative data from two corpora (a learner corpus and a native speakers’ corpus) were used. The learner corpus contained 197 essays written by undergraduate students at the University of Ghana, and the native speakers’ corpus contained 207 essays written by undergraduate students at selected American and British universities. The main purpose for using the cloze test data was to ascertain the learners’ receptive knowledge of metadiscourse use, and the learner corpus was analysed to determine their productive use of the markers in relation to the types, frequency, and appropriate use. The students were aged 17 years and older, came from various home-language backgrounds (Akan, Ewe, Ga-Adangbe, Northern Ghana languages, and English), and were from four academic disciplines (BA Arts, BA Education, BSc Applied Sciences, and BSc Health Sciences). The learner corpus comprised written, end-of-semester academic essays, which were classified into three performance levels: high-rated, average-rated, and low-rated. The learner corpus was compared to the native speakers’ corpus to analyse similarities and differences in metadiscourse use by the learners and the native speakers. Hyland’s (2005a) classification of metadiscourse markers served as the theoretical framework to identify and discuss metadiscourse markers employed in the students’ essays. The software tools used for the analyses were #LancsBox and SPSS. The findings revealed that although the learners were able to select the correct and appropriate metadiscourse markers to complete the cloze test, they were not able to use some of the markers independently in their essays. It was also observed that the learners who wrote the high-rated essays used metadiscourse markers in appropriate proportions. Regarding how the learners’ use of metadiscourse compared to that of the native speakers, the findings revealed that although both groups of students employed more interactive resources than interactional resources, metadiscourse markers were denser in the learner corpus than in the native speakers’ corpus. However, the analysis revealed that the two groups of undergraduate students showed a similar pattern in their use of interactive resources. The findings of this study may be helpful in the teaching and learning of academic writing. It is also hoped that the findings may be useful to other researchers who are interested in the field of metadiscourse, especially among Ghanaian undergraduate students.