The use of AI in exams requires a clear framework

1. How do you use artificial intelligence in your teaching?

I use AI in my own preparation and as a topic in my teaching. In my own preparation, I use it to refine teaching materials, draft slides, structure workflows and things like that. This frees up time for in-depth reading and for working on the learning activities that form the foundation of the teaching.

When it comes to the first-year students I teach, the most important thing has been to clarify how they are allowed to use AI in their study process and in the written assignments they submit for their exams.

I expect that many students are using AI, whether we talk about it or not. That is why I tell them that they are welcome to use AI but that they must follow SDU’s guidelines and keep up to date with developments because things are moving quickly.

At the same time, they need to consider what will happen to their learning if they outsource parts of their written work. How will they process their own academic reflections? How will they engage in their psychological and academic craft? These are some of the questions I’d like them to work on.

2. What impact does AI have on exams and assessment?

Because of AI, we have to be clearer about what the students are expected to demonstrate at the exam. I have no doubt that a chatbot is able to write a first-year assignment that will pass. That is why it is not sufficient to assess the completed assessment alone. We have to be able to assess the process and the student’s own contribution.

That is why I have been working on developing declarations of AI use. If the students use AI, they must document how they use it and what they gain from it. It is not enough simply to state that they have used AI. They have to be able to describe whether AI has been used to generate the structure or the outline, to give feedback or do text revision, for example, and at what stage of the writing process it was used.

I have shown the students examples of declarations of AI use in research articles in which AI has been used to support the writing process, including one of my own publications. The point of the declaration is to make the process transparent, so that we can assess the students’ own contributions to their written work.

I have also written to the external examiners and explained how we have worked with AI and what the requirements are for declaring the use of AI. It is important that the students and the external examiners receive the same information. Otherwise, there may be some uncertainty about whether everyone assesses the students’ use of AI in the same way.

3. How have the students reacted?

The students are curious, but I also get the impression that quite a few of them are a bit reluctant to tell us lecturers about how they use AI. Many of them are concerned that the use of AI may be seen as cheating. That is why my aim has been to start an honest conversation with them.

I am telling them that the use of AI is voluntary. They do not have to use AI to do well in the course. However, if they choose to use AI, they must do so with awareness and document how they have used it.

We have also incorporated dialogue-based exercises into the classes. The students take a stance on statements about the use of AI by moving to a particular spot in the classroom according to how much they agree or disagree. The student teachers in the classes have led the way in this exercise, which highlights the fact that their fellow students may have different experiences and views regarding AI.

It is my impression that first-year students are not really discussing among themselves how they are using AI in their studies. For that reason, it is important to create academic spaces for them to discuss this. It is not enough to teach them what they are and are not allowed to do. They also have to learn to assess when AI can support their learning and when it may end up removing some of the work they have to do themselves.

4. What does AI mean for the students’ learning?

I think it will change the students’ learning processes. Their written work may be of a higher quality, but this also means that we need to pay closer attention to how we maintain their critical reflection.

For example, I have asked the students to use AI as a kind of sounding board to help them in their reflection process. If they give a chatbot a piece of text and ask for feedback, they can tell it to ask reflective questions rather than simply to rewrite the text. If this leads to changes to the text, they must type in the changes themselves, so that they are actively considering each individual change. This is particularly important for first-year students to ensure that they specifically incorporate the reflective process into their academic work early on in their studies.

The point is that AI must not replace academic judgement. The students must learn how to evaluate AI-generated output and understand what it means for their own writing and learning processes.

In addition, there is a specific focus on human interaction on the Psychology programme. Regardless of whether the students go on to become researchers, clinicians or consultants, they have to acquire certain skills with regard to relationships, conversation, analysis, assessment and critical reflection. We must hold on to these skills, despite the current changes in how we work with texts.

5. What should be key principles for the responsible use of AI?

The most important principle is that AI should support learning, not replace it. That is why we should not just ask whether the students are allowed to use AI. We should ask what they are supposed to learn and how AI affects that learning.

The students should know what is expected of them, how they should declare their use of AI and how this will be taken into account in the assessment of their written work.

We need more discussions about AI on the study programmes and in the departments. We need to discuss which reflections the students should engage in at which stages of the programmes. We must make sure that all of this does not mean that AI will overshadow the subject-specific content.

I hope that, in the long run, AI will help us to change our priorities. If the technology can be used for planning and organisation purposes and for certain types of training, there may be more time for in-depth reading, in-depth learning and the uniquely human elements in the field of psychology. However, this requires us to maintain our academic judgement and critical thinking at all times.