Should we re-evaluate our assessments?

Dr Rebecca Upsher, Teaching Fellow and Research Associate, Department of Psychology

In the previous step, you considered the general implications of AI for assessment in higher education. In this step, you will learn about key opportunities for re-balancing assessment in the age of generative AI.

Hello! I’m Dr Rebecca Upsher, an educator and researcher from the Psychology Department at King’s College London. I’m deeply passionate about enhancing student wellbeing and learning during their university journey. I co-authored the Education for Mental Health toolkit on this topic. In addition, I run an international online Education for Mental Health special interest group that showcases academic research and practice that aims to support student wellbeing in the curriculum.

Here, I’m eager to share insights, both from my own studies exploring generative AI (GenAI) in higher education and those of fellow researchers, regarding opportunities to improve assessment in the age of GenAI.

The rising prevalence of GenAI use is an opportunity for teaching staff at universities to evaluate their assessment design. For example, thinking about how we can make our assessments meaningful, sustainable and inclusive. Of course, we can do this without the existence of GenAI, however, GenAI brings these issues to the forefront as many existing assessments are vulnerable to GenAI automation. Therefore, we need to design our assessment with GenAI in mind, to make them less vulnerable to GenAI automation, whilst enhancing the learning experience for students.

Let’s go through seven key opportunities here!

1. Authentic assessment

GenAI is not merely a technological advancement; it represents a shift in tasks, eg GenAI is predicted to automate up to 60 % of current tasks in the future workplace [1]. It’s crucial, therefore, to incorporate GenAI into the higher education curriculum to support students in developing skills for their future careers. For example, supporting students to perform real-world tasks that demonstrate meaningful application of essential GenAI knowledge and skills.

2. Enhancing meaningful learning

Research indicates that GenAI excels in lower-order tasks like factual recall. While these tasks showcase foundational knowledge, they do not foster lasting, deep learning. Conversely, GenAI is not currently so good at higher-order tasks, like decision-making, problem-solving and interpersonal skills. So these tasks are becoming vital for the future workforce. These tasks also ensure meaningful, long-term learning for students.

Higher-order thinking activities might include peer feedback to improve critical evaluation skills, peer teaching or seminar debates. Even if AI produces the primary content, supplemental question-and-answer activities ensure genuine comprehension, encourage discussion and provide an opportunity for formative feedback. Students can also use AI for data summarisation, critiquing AI outputs or applying AI in collaborative project management tasks.

3. Sustainability in assessment

Traditionally, assessment has often been performed on students, which lowers intrinsic meaning for students. A better approach is the shift towards ‘assessment for learning’, positioning students at the heart of their own educational journey. This shift is in line with sustainable assessment, which underscores the importance of long-term learning and self-agency. This can be achieved in many ways. Examples include co-creating assessment with students or giving students a choice in assessment (for example which problem to solve, which essay question to answer or in which format to submit the assignment). This aligns with opportunity 1: by additionally considering sustainable elements of assessment, leveraging GenAI tools not only facilitates personalised assessment options but also equips students with skills essential for adaptive, continuous learning in a changing working world.

4. Promoting ethical AI use in formative activities

Educators should integrate GenAI into formative tasks to give students the opportunity to practice ethical use of GenAI tools. This approach not only provides students with practical experience but also maintains academic integrity by highlighting the supportive role of AI in learning, rather than shortcutting results. For example, when writing an essay, students could use GenAI to assist in learning – such as a reading assistant, critiquing writing, providing initial feedback or engaging in a Socratic dialogue to unpick thought processes – rather than automating the entire writing process.

5. Prioritising inclusivity in assessment

Shifting from traditional exams and oral assessments to flexible assessments ensures inclusivity, especially for students with disabilities or social anxiety. Co-creating assessments with these students and offering assessment choices can better address and alleviate potential challenges. GenAI tools offer new opportunities to do this.

6. Commitment to equitable assessment

Educators must be aware of the potential for a digital divide amongst their students. Some students can afford to pay for technologies, whereas others can’t. Students will also vary in their levels of GenAI literacy. To create equitable learning experiences, it’s vital to utilise universally (freely) accessible GenAI tools and to make sure GenAI training is accessible and available for all students.

7. Bringing all these ideas together!

Here are a couple of examples of assessment for you to think about that can bring together the ideas mentioned above.

The PAIR framework: the PAIR framework, as discussed previously in Step 2.4, stands for Problem, AI, Interaction and Reflection. This structured approach empowers students to pinpoint problems, pick suitable AI tools, engage dynamically with them and then partake in reflective learning. This model also promotes higher-order thinking, collaborative learning opportunities and ethical and responsible use of AI. This article, Are your students ready for AI?, which you encountered in Step 2.4 discusses the PAIR framework in more detail.

Processfolios: this technique asks students to not only present a final product, but also to demonstrate the process and learning journey. Students have opportunities to self-reflect on their achievements, obstacles and the methods used to tackle difficulties. In particular, they will reflect on their use of GenAI. This can be used in combination with the PAIR framework.

Concluding thoughts

The rise of GenAI can be daunting for educators in higher education. Yet, when approached and taught with care, GenAI can enhance learning opportunities for students. It also presents an excellent chance for us to reflect on and refine our assessment practices, ensuring optimal support for our students.

Try re-evaluating your own assessment practices from the perspective of the opportunities offered by GenAI. As you do, please consider the opportunities we’ve looked at here. Together, they champion authentic, meaningful, sustainable and inclusive approaches that enhance student engagement and learning.

Now that you have completed this step, you have some ideas of opportunities to re-evaluate your assessment for students. You have some inspiration for developing authentic assessment in light of a world where generative AI exists. In the next step, you will learn about the importance of, and how to develop, AI literacy within a higher education context.

References

1. Chui M, Hazan E, Roberts R, Singla A, Smaje K, Sukharevsky A, Yee L, Zemmel R. The economic potential of generative AI: the next productivity frontier. McKinsey Digital; 2023, Jun 14.

Join the conversation

• Do any of these opportunities resonate with you?
• How will you think about re-evaluating your assessment in the age of generative AI?
• Can you apply the PAIR framework in your assessment?