Dr Temesgen Kifle writes for Times Higher Education.
As artificial intelligence (AI) becomes increasingly prevalent in education, there are growing concerns about how it impacts fairness and authenticity in student assessments. With AI tools potentially compromising the integrity of evaluations, educators must innovate assessment methods that accurately gauge students’ true understanding and capabilities. Here are eight strategies for designing AI-resistant assessments that uphold fairness and authenticity:
In-Person Written Assessments: Traditional written exams conducted in controlled environments minimize AI assistance. These assessments allow for direct observation of students’ skills and understanding. However, they require meticulous planning to manage logistics and ensure a fair testing environment.
Oral Presentations and Q&As: Oral presentations offer a dynamic way to assess students' communication skills and their ability to engage with an audience. By asking questions in real time, instructors can evaluate students' ability to articulate ideas and respond to queries, which is challenging for AI to replicate.
Complex Problem-Solving Tasks: Tasks that involve real-world problems and require innovative solutions assess students’ ability to apply their knowledge practically. For instance, developing a policy to address unemployment involves analyzing various factors and proposing solutions—skills that AI struggles to mimic comprehensively.
Reflection Essays: Reflection essays require students to analyze their personal experiences and learning processes. These essays emphasize critical thinking and personal growth, demanding introspective and subjective insights that AI cannot easily generate. Effective grading of these essays involves clear rubrics and selective feedback to manage the time-consuming nature of marking.
Debates and Discussions: Structured debates and discussions test students' abilities to present and counter arguments on complex issues. These activities involve real-time engagement and critical thinking, areas where AI's contributions are limited due to the need for nuanced and context-specific analysis.
Project-Based Assessments: Projects assess students’ comprehensive skills from inception to completion. For example, creating a business plan requires practical application of knowledge and strategic thinking. Unique, context-specific projects combined with AI detection tools can ensure the authenticity of student work.
Case Studies: Analyzing detailed case studies involves applying theoretical knowledge to real-life contexts. For instance, studying the integration of international students at a university requires nuanced understanding of social interactions and cultural adjustment—areas where AI lacks depth and subjective judgement.
Progressive Assessments and Defense: Ongoing assessments that track students’ progress over time help ensure that their final work reflects their personal efforts. By requiring regular submissions and defenses of their work, this method maintains academic integrity and verifies students’ engagement throughout the assignment process.
Focusing on Learning Outcomes
While these methods can mitigate AI’s impact on assessments, the primary goal should remain on measuring student learning and achievement of course objectives. Balancing the need to prevent AI misuse with the need to accurately assess educational outcomes is essential. Effective assessments should thus integrate AI-resistant methods while prioritizing the evaluation of learning outcomes.
By adopting these strategies, educators can foster a fair and equitable academic environment that genuinely reflects students’ understanding and abilities, even in an era where AI is prevalent.
