metadiscourse in l2 writing: how do high-and low-rated essays differ?

The aim of this study was to compare the use of metadiscourse markers in high- and low-rated essays written by iranian efl learners. to this end, a 483115-word corpus of 1635 essays written by iranian efl learners was compiled. all the essays were rated by two raters based on the ielts public writing band descriptor. those essays with a score of 5 and lower were marked as low-rated, and those with a score of 6 and higher were categorized as high-rated. adopting hyland’s (2005) interpersonal model of metadiscourse and using antconc software, we explored the two subcorpora in search of metadiscourse instances. our findings revealed substantial differences between high and low-rated essays in the use of both interactive and interactional metadiscourse. in particular, interactive metadiscourse markers were used more frequently by less competent writers in comparison to their more competent counterparts. the difference between the two groups in the use of interactional metadiscourse was, however, not significant. further investigation of the data showed both similarities and differences in the use of each of the 10 subcategories of metadiscourse, indicating that appropriate use of metadiscourse is essential in achieving higher scores in writing. implications of our findings for eap, language assessment, and l2 writing pedagogy were discussed

Metadiscourse in L2 Writing: How Do High-and Low-Rated Essays Differ?

The aim of this study was to compare the use of metadiscourse markers in high- and low-rated essays written by Iranian EFL learners. To this end, a 483115-word corpus of 1635 essays written by Iranian EFL learners was compiled. All the essays were rated by two raters based on the IELTS public writing band descriptor. Those essays with a score of 5 and lower were marked as low-rated, and those with a score of 6 and higher were categorized as high-rated. Adopting Hyland’s (2005) interpersonal model of metadiscourse and using Antconc software, we explored the two subcorpora in search of metadiscourse instances. Our findings revealed substantial differences between high and low-rated essays in the use of both interactive and interactional metadiscourse. In particular, interactive metadiscourse markers were used more frequently by less competent writers in comparison to their more competent counterparts. The difference between the two groups in the use of interactional metadiscourse was, however, not significant. Further investigation of the data showed both similarities and differences in the use of each of the 10 subcategories of metadiscourse, indicating that appropriate use of metadiscourse is essential in achieving higher scores in writing. Implications of our findings for EAP, language assessment, and L2 writing pedagogy were discussed