Effects of supporting lexical bundles on the quality of argumentative writing

[ABSTRACT ONLY; NO FULL TEXT] In this study, I explored the effects of supporting lexical bundles (e.g., on the other hand and in order to) on the quality of second language writing. Specifically, I examined how student writing performance can be supported by the provision of lexical bundles that are selected from argumentative essay corpora. I used two argumentative writing tasks, which I manipulated the task condition with lexical bundles (e.g., task with the support of lexical bundles and task without the support of lexical bundles). I collected essays from 20 international university students, and each student wrote two essays (i.e., a total of 40 essays). These essays were evaluated by two master's students majoring in TESL. With the result of evaluation, I could compare how students' writing performance differed between the two writing tasks. The result of essay quality scores showed that the writers in the task without lexical bundles support get higher scores than the task with lexical bundles support in terms of development and task completion. This finding explained that the support of lexical bundles interferes with learners' task performance since the writers in the task with lexical bundles support more focus on the use of given lexical bundles than on the task completion. I also found that the writers in task with lexical bundles support get slightly higher scores than the writers in the task without lexical bundles support in terms of the range of vocabulary. It explained that the actual use of lexical bundles might have led to an actual enhancement of writing quality. Nevertheless, as the score of the vocabulary area increases, the decrease of the score in other areas was found. Thus, it still supported the interpretation that lexical bundles support interferes with task performance. The final conclusion is that the supporting of lexical bundles had no effects on the quality of second language writing since the differences in the essay quality score between the two tasks were not notably significant. This study explored a new area of writing task manipulations with a focus on the useful lexical bundles.