Article Preview
TopIntroduction
For most Chinese EFL learners, their speaking ability remains the weakest link (Yu & Dai, 2019). Wei et al. (2011) state that students' poor oral ability could be triggered by their incorrect pronunciation and poor grammar. According to Gimson (1980), the key to master any language is that one needs to master only 50%-90% of its grammar, 1% of its vocabulary and 100% of its knowledge on pronunciation (qtd in Liu et al., 2018, p.333). It is reasonable to note that grammar and pronunciation are the foundation of mastering English, and that improving these two aspects could be one focus of language teaching and learning. However, large-class teaching not only reduces students' opportunities to speak, but also makes it impossible for teachers to evaluate and correct each student's oral practice, thereby hindering students’ oral performance, grammar accuracy, and pronunciation.
Since classroom oral training is relatively limited, one good remedial measure towards it is to engage students into oral training after school. However, when students intend to practice speaking after class, they can easily find themselve in the dilemma of having no partner around to practice with, thus making it hard to carry out oral training, let alone obtaining feedback from partners or teachers. The development of a chatbot offers a solution. According to Brennan (2006), a chatbot is an artificial construct designed to chat with human beings by means of using natural language for input and output. Many studies have been conducted that show chatbots can be used for language practice (Fryer & Carpenter, 2006; Jia, 2004; Jia & Chen, 2008; Shawar & Atwell, 2007). Microsoft Xiaoying (hereafter called Xiaoying), a chatbot launched by Microsoft on WeChat in 2016, is targeted at helping Chinese speakers learn English. This study applies Xiaoying as a language learning tool in after-class oral training, with the aim to locate students’ accuracy errors in dialogues and check the effect of this tool on improving learners’ oral accuracy, mainly grammar and pronunciation.