Generative Artificial Intelligence and Assessment Task Design: Getting Back to Basics through the Lens of the AARDVARC Model

Effective assessments guide student learning, refine teaching practices, ensure curriculum alignment, and foster workforce readiness. However, the emergence of generative artificial intelligence (GenAI) tools, such as ChatGPT, has significantly disrupted traditional assessment processes, raising concerns about academic integrity and necessitating innovative approaches. While higher education institutions are making strides in adapting to this new reality, the foundation of effective assessment remains educators’ assessment literacy. This paper responds to the critical need for improving educators’ assessment literacy by introducing a comprehensive model – the ‘AARDVARC’ framework – that outlines eight key attributes of effective assessment: alignment, authenticity, reliability, developmental appropriateness, validity, accessibility, realism, and constructiveness. By fostering assessment literacy, educators can design innovative, equitable, and discipline-relevant assessments that incorporate GenAI responsibly and meaningfully. The paper further offers actionable recommendations for adapting university assessments to align with institutional goals and meet the evolving demands of the educational landscape. These strategies aim to ensure that assessments continue to promote student engagement, maintain academic standards, and reflect the realities of modern education.