If you are in a position to make any changes to your assessment then you could think of redesigning it. Redesign can help to either limit the use of AI through a different task design, a different distribution of weighting or, if the ability to utilize AI is an important skill you want your students to develop, to embed AI into the assessment. Some options for redesign at different levels are presented and discussed below but in any case it is important to also consider other factors that will influence your ability to make changes. When deciding on whether you are able to make changes consider the following questions:
- Will this redesign require minor/ major modifications?
- Does this change require a change to the module level learning outcomes?
- Does this change require a change to program level learning outcomes?
- Does this change require a change to assessment type?
- Does this change have a knock on effect on other assessments on the module? (for example in terms of redistribution of the weighting)?
- Will there be sufficient time left for those to be approved by relevant committees prior to the start of the module?
You can read more about major and minor modifications procedures here.
When introducing any changes to the assessment please ensure that:
- Students are appropriately briefed about what is allowed and what is not allowed in terms of AI use. An example of such briefing can be found in the downloadable resources section on the Reflective Essay case study.
- You consider a module/ programme level strategy to teach students about positive ways to engage with AI that are pedagogically helpful and do not raise concerns around academic integrity.
- There is a procedure in place for dealing with suspected plagiarism/ AI misconduct cases. There is no reliable way to detect AI in students’ work currently. There are certain pieces of software that might highlight some aspects of work that could come from AI but you cannot know for certain that is the case. In general, AI output can be formulaic in nature and AI is known to use false references. If you suspect AI was used to plagiarize you can introduce an additional assessment method to determine whether that was the case or not. This could include authenticity interviews, as mentioned in College guidance.
- Ensure that all stakeholders are consulted about potential changes
Redesign the assessment strategy at module level:
Redesign programme level assessment strategy
Modular assessment can lend itself towards overassessing as each module tends to be assessed independently often by more than one assessment making assessment strategy more compartmentalized. An alternative way of looking at assessment strategy is employing synoptic assessment. Synoptic assessment synthesises student learning from several modules within a programme. This means that students need to demonstrate their ability to make connections between learning on different modules. Such an approach also has the potential to reduce assessment burden as the need to assess on every module is removed. Such an approach has been introduced at Brunel and at Queen’s University Belfast.
Redesign an individual assessment
In circumstances when changes to assessment strategy on the module aren’t possible or desirable you can focus on one assessment and redesign it with AI in mind. This redesign will depend on your beliefs about the role of AI in students’ future so can focus on either attempting to eliminate any impact, or making it a part of assessment process or product.