Integrating AI in Assessments (part 2): Lessons from My Action-Based Research

Completing my action-based research was an eye-opening experience, carried out over two cycles within the constraints of a business school operating on a carousel model—where each unit runs for just six weeks. Given this short timeframe, I could only conduct two cycles, but the results were transformative.

Cycle 1: The Struggles of AI Adoption in the Classroom

In the first cycle, students faced significant challenges using AI to complete their tasks. They were restricted to using ChatGPT as their sole generative AI tool and had to apply specific decision-making theories from the unit. Many struggled to generate useful outputs, with difficulties in creating effective prompts and verifying AI-generated content. Issues with made-up references and a lack of understanding of AI’s reliability were prevalent.

This initial struggle highlighted an important gap: Students needed guidance on how to work with AI effectively. Rather than seeing AI as an academic shortcut, they needed to learn how to use it as a tool—critically, ethically, and efficiently. The results from their first assignment and the survey questionnaire revealed that.

Cycle 2: A Shift in Learning and Engagement

For the second cycle, I made key changes to how AI was integrated into the assessment process:

• Students were explicitly guided in developing prompts to refine their queries better and get more relevant information.

• While the restriction to decision-making theories remained, they were now encouraged to use multiple generative AI tools—beyond ChatGPT—to generate text, images, and even their final presentation.

• Students had creative freedom in presenting their work through role play, games, quizzes, and other innovative formats.

• A critical component was verifying AI-generated academic citations to ensure accuracy rather than relying on fabricated references.

• I emphasised that I was less concerned about grammatical perfection and more interested in their understanding and application of concepts in real life.

  
  

  Student Presentations: Creativity Meets Critical Thinking

  This cycle saw four diverse and engaging student presentations:

  1. ESG and Multinational Corporations – This group creatively used role play to demonstrate ESG (Environmental, Social, and Governance) issues in a multinational corporation (MNC). Protestors in the presentation challenged the company’s top management, leading to in-depth discussions on decision-making theories.

  2. ESG and Global Impact – Another group took a similar approach but expanded their focus to the effects of ESG in another MNC in various countries. They also incorporated protestors questioning corporate decisions, making the presentation dynamic and immersive.

  3. Big Data and Social Media Breaches – This group tackled data privacy and security. A short role play on a data breach in social media set the stage, followed by a more traditional presentation format. Internal leadership issues within the group were reflected in their presentation style, showing real-world dynamics of teamwork and adaptability.

  4. The Influence of Indian Cuisine on the Global Market – The most outstanding presentation, this group explored how Indian food became a global phenomenon, using a real Indian restaurant in London as a case study. They creatively combined music, dance, role play, PowerPoint, and live demonstrations to showcase business expansion decisions and strategies. They accurately applied the theories studied in the unit to their case study.

     
     

     Key Reflections and Student Takeaways

     After the presentations, students participated in focus group discussions on their experiences using AI in assessments. Some key insights emerged:

     • Initially, they were unsure how to use AI effectively, but structured classroom time and guidance helped them develop ethical and strategic AI usage skills.

     • They learned that AI is often unreliable and that different AI tools produce varied results—reinforcing the need for human verification.

     • They appreciated being given the freedom to think in their native languages before formulating their thoughts in English. My multilingual skills played a role in helping them bridge this gap.

     • They valued the support and motivation provided throughout the process, which encouraged them to engage deeply with AI tools rather than just using them superficially.

     The Future of AI in Higher Education

     This research confirmed my belief that AI can be successfully integrated into assessments, but it requires careful planning and classroom implementation. Educators need training in identifying AI-generated content, assessing AI-integrated work, and applying ethical guidelines to uphold academic integrity. More importantly, successfully integrating AI into education requires a collaborative effort between students, teachers, and institutions.

     With the right approach, AI can enhance learning rather than hinder it. But for this to happen, higher education must shift its mindset—moving away from fearing AI and instead focusing on using it as a tool for critical thinking, creativity, and ethical engagement.

     
     

     What are your thoughts on AI in higher education? Have you integrated it into your teaching or learning process? Let’s start a conversation.