Reviewing Existing Assessments

Greenwich AI Risk Measure Scale (ARMS) – University of Greenwich have develop this rating tool which allows you to review your existing and rate from very low to very high the potential risk that a student could use Generative AI to produce their assessment.  This short video demonstrates how to use the scale. 

Assessment Types that are at most risk

Francis & Smith have identified assessment modalities that are most at risk including essays, lab reports, literature reviews and presentations as susceptible to the use of GenAI tools.  They propose the following strategies to make assessment more robust.

• Process orientated implementing the move from final product to process output

• Oral formats although there is still the potential that GenAI could be used to prepare by having a conversation or questioning the student will have to demonstrate their knowledge and understanding

• Base around novel areas of study by using unique scenarios that require higher level skills than is possible in GenAI tools 

• Real word situations enables the student to use and reflect on their personal situations and experiences. If asking students to work through an issue don’t use contexts which they may be able to obtain through GenAI. For example, don’t use big companies such as Apple or Amazon.

• Use collaboration and peer review

• Competency based where practical skills can be demonstrated such as observations and debating skills

• Portfolios where the assignment is built up and developed over time.  Again this is very process focussed and can be used to encourage reflection