Over the past three years, the landscape of academic assessments has been transformed by advancements in artificial intelligence (AI) technology. As generative AI tools become more prevalent, universities worldwide have taken a spectrum of approaches. Some institutions have outright banned the use of these technologies, while others permit limited usage or embrace them completely. This growing presence of AI in education has led to a swirling mix of confusion and anxiety among both students and faculty, particularly surrounding what constitutes “appropriate use” of these tools. At the heart of the matter lies a pressing concern: is AI enabling a rise in cheating?
The conversation extends beyond academic integrity, prompting fundamental questions about the value of a university degree in an era where AI can significantly influence student assessment outcomes. In light of these complexities, our latest journal article probes current approaches to AI in academic settings, posing a critical question: How should universities assess students in the age of AI?
Read more:
Researchers created a chatbot to help teach a university law class – but the AI kept messing up.
Understanding Assessment Validity in AI Contexts
The rise of generative AI has led universities to develop various policies aimed at clarifying what actions are permissible. An innovative example comes from the University of Leeds, which has introduced a “traffic light” framework to guide students on AI usage during assessments. A “red” light indicates no use of AI, while “orange” allows limited applications such as brainstorming or idea generation. A “green” light encourages unrestricted use of AI tools. Through this model, institutions aim to maintain assessment validity, ensuring that evaluations genuinely reflect students’ capabilities rather than their proficiency in utilizing AI.
Despite these measures, merely establishing clear guidelines may not be enough to sustain assessment validity. It’s essential to reflect on whether assessments are genuinely measuring students’ learning or merely gauging their skills in navigating and deploying AI tools.
The Distinction Between Discursive and Structural Changes
In our recent research, we explore an important distinction: discursive changes and structural changes in relation to AI and assessments. Discursive changes involve modifications to existing instructions or guidelines around assessments but rely heavily on students understanding and adhering to new rules. For instance, directing students to “use AI only for editing” exemplifies a discursive change.
In contrast, structural changes entail modifications of the assessment tasks themselves. These changes are designed to either constrain or enable student behavior inherently, rather than relying solely on directives. A notable example would be switching from traditional essay formats to in-class writing exercises, where educators can observe a student’s developmental progress over time. This approach eliminates ambiguity and promotes accountability, making it far more challenging for students to engage in dishonest practices.
The Limitations of Current Strategies
The majority of university responses to the challenges posed by generative AI have focused on discursive changes. Initiatives like traffic light frameworks and mandatory AI usage declarations have only tweaked the rules surrounding assessments. They have not fundamentally restructured the assessments themselves, leaving the door open for potential dishonesty.
Our findings suggest that protecting assessment validity necessitates a shift towards structural changes. As AI continues to evolve, traditional forms of assessment are becoming increasingly susceptible to undetectable rule-breaking. Therefore, merely informing students of what they can or cannot do with AI tools is insufficient.
Embedding Validity Through Structural Change
To ensure that assessments are valid and equitable in this new AI-dominated landscape, universities must embrace structural changes in their evaluation methods. This means designing tasks where the integrity of the assessment is built directly into the assignment. Such a nuanced approach will not be uniform across all disciplines; instead, adjustments will be required to suit the specific context of each field of study.
Moving towards structural changes may necessitate a complete rethinking of how we assess students. This transition won’t be simple, but it is crucial if we aim to preserve the value of educational qualifications in the face of rapidly advancing technology.
Inspired by: Source
