LAUSR

The affordances offered by automated writing evaluation systems (AWES) have been widely reported (Wilson and Roscoe, 2020). In the university English-as-a-foreign-language (EFL) setting, Pigai has been widely used as one of AWES (Bai andHu, 2016).With backup frommultiple corpora (e.g., scored essays), Pigai could help users (e.g., student writers) quickly identify and correct grammatical errors and lexical inappropriateness at the sentence level, improving the accuracy of the evaluated essay (Bai and Hu, 2016). Pigai includes multiple dimensions of checking, such as typos, collocation, punctuation, article use, and word choice, whose assistance primarily lies at the sentence level (Yao, 2021; see also Figure 1). However, effective and accurate writing is not merely static or mechanical (Derewianka and Jones, 2012). Instead, it is a meaning-making process, where the choice of language resources (including grammar and vocabulary) interacts with the context in which it is embedded—for example, the target audience—and with various cultural expectations (Schleppegrell, 2004). This means that Pigai needs to be additionally able to help diagnose such issues. For example, when a verb such as knows occurs with a third-person singular she, the verb may not be underlined by Pigai. Although the answer provided by AWES is correct for the sentence in terms of grammatical agreement between subject and verb in English, the automated feedback may be not correct when the writing is documenting an uncertain event. In other words, the appropriate sentence should be she may know. Existing negligence of such issues and inability to provide feedback regarding them are understandable. One primary reason may be a lack of technicians who understand the mechanism of effective writing in relation to writing process; if this is the case, technicians may not know how to incorporate context-based feedback into AWES. As Reinhardt and Oskoz (2021) also argued, in terms of technology-based teaching practices, we need to focus on situating “their uses and designs in commensurate theories of learning and pedagogy” (p. 2). To make better use of Pigai as one of AWES, the present paper discusses solutions and references for Pigai designers. It is hoped that Pigai designers may then be able to engage in empirical tests, collaborating with writing experts and taking actions to make further improvements of Pigai.