The odd couple? Gen AI and post-entry language assessment (PELA)

Post-entry language assessments (PELAs) have been commonly used in the Australian and New Zealand university context to identify students who may require academic language and literacies support (for review, see Read, 2015). Many PELAs, such as Bond English Language Assessment (BELA; Lydster & Brown, 2017), often require students to submit a piece of academic writing and are self-administered by students online. This raises potential academic integrity concerns. In response, BELA-AI, a formative assessment item incorporating Generative AI, was developed for the Bachelor of Medicine program at Bond University. BELA-AI, a 60-minute assessment, contains three tasks: 1. Students write an academic essay, 2. Students generate an essay using Generative AI, and 3. Students write a brief critical reflection analysing the two essays. Initial discussions with raters (i.e. Bachelor of Medicine academic staff) indicate preliminary evidence that BELA-AI has reduced the academic integrity concerns associated with the previous version of the PELA. However, further research is required to confirm students are maintaining academic integrity and using Generative AI ethically. This case study presents BELA-AI and how it has been used as a “conversation starter” with Learning Advisors at Academic Skills Centre, to discuss aspects of academic writing including maintaining academic integrity. It is argued that Generative AI and language assessment are not such an odd pairing, and embracing the technology is encouraged.