Monday, February 27, 2023

ChatGPT and assessments

 One of the main challenges facing educators when digital tools can be used for a host of writing tasks, is that of assessments. How can the integrity of assessments be safeguarded when students are able to use nefarious means to complete their assessments.

Over the last two months, there have been a large number of articles discussing the issue. 

- FENews proposes the use of a variety of question types, effective proctoring systems and different testing types e.g. using oral presentations or practice-based assessments instead of written exams.

- The conversation advices the need to rethink assessments. In particular, to ensure they are authentic i.e. connected to personal context or the course specific material. Project-based, group and inquiry/problem based learning are recommended.

- workhe.com recommends that educators trust their students and to take on the possibility of discussing with students, the pros and cons of using AI to augment, or scaffold their work.

- Pulse in the context of teacher education reminds us that learning is a process, not a product or artefact. Hence authentic assessments are the way to go.

- uts provides good ideas on how students and teachers can draw on AI to support their work. Good ideas include asking students to use ChatGPT to generate a response to the assessment and then set up the criteria to critique the response and provide feedback to improve on it. 

In all, it is important to include context into assignments and to ensure students take ownership and responsibility for their learning. AI in the form of Chatbots able to provide responses which mimic those of humans, are here to stay. They will also improve rapidly as they learn from the feedback to responses provide by users. Therefore, it is important to ensure educators understand the implications and work through the consequences. The opportunity is now provisioned to improve learning and assessment processes, so that assessments are for learning and not of learning. 

No comments: