The effects of an AWE-aided assessment approach on business English writing performance and writing anxiety: A contextual consideration

Abstract

In recent decades, AWE (Automated Writing Evaluation) systems have been increasingly used in English writing instruction. Numerous studies have examined whether AWE can help improve students’ writing performance, but most of the writing tasks involved are for general purposes. As a result, the effects of AWE remain unclear in some specific writing tasks, such as business writing. Besides, few studies have investigated whether the effects on writing performance are mediated by writing anxiety, an affective factor that can influence their writing performance. Seen from a broader perspective, the two questions are related to the context of using AWE systems. However, limited studies have examined the contextual factors within a unified framework. To address these issues, we conducted a one-semester quasi-experiment using an AWE-aided assessment approach in a business writing course. The experimental group (39 participants) received the AWE-aided assessment, while the control group (34 participants) obtained an instructor-only assessment. The participants’ writing performance and anxiety were measured by the B2 Assessment Scales and SLWAI (Second Language Writing Anxiety Inventory), respectively. Results showed that the experimental group’s writing performance improved significantly in language and organization. However, there was no significant mediation effect of writing anxiety on the relationship between the assessment approach and writing performance. Moreover, no significant effect of the approach was observed on the participants’ writing anxiety, except for a marginally significant reduction of their avoidance behavior. Finally, the research results were interpreted by placing the related factors within Martin’s register-genre-ideology context framework.