How can I assure academic integrity?

The University is committed to continually enhancing our teaching and assessment practices based on pedagogy and evidence of effectiveness of learning. This means supporting the development and introduction of assessment that is meaningful, allows students to build their understanding across their programme of study, and enables them to demonstrate their understanding and skills as well as their knowledge (see Methods of Assessment). In-person exams can assure academic integrity and are useful for assessing knowledge recall and similar but are not suited to evidencing critical thinking. By designing assessment in line with the Learning Through Assessment Framework (March 2023), the risk of AI use in a way that breaches academic integrity will be considerably reduced.

For example:

  • Meaningful and Inclusive assessment: Developing or simply rewording marking criteria in collaboration with students to ensure that students fully understand the assessment criteria and their relationship to the ILOs. Alternatively, allow students to propose and submit assignment formats of their own choosing, with clear explanations for their selected format and how it evidences the learning outcomes.
  • Iterative assessments: Using course-long assessment activities, such as portfolios, and regularly advise students on building their portfolio content as they progress through the course.
  • Programmatic assessment: Help students create learning plans by identifying more advanced courses in the programme which are of interest to them, and designing assessments that give them the opportunity to use that interest and demonstrate the connection between, for example, the two courses.

AI tools like ChatGPT are language learning models, and nothing more. They generate the most likely text by determining word probability through analysing and interpreting data. They use an algorithm that establishes rules for natural language. What they cannot do is create personalised answers to assignments using reflections that are unique for individual students, for example: AI cannot, therefore, support students where we ask them to:

  • Build portfolios.
  • Demonstrate course concepts by synthesising up-to-date cited literature with their own personal experiences.
  • Incorporate aspects of students’ individual goals, backgrounds, and challenges.