Exploring the use of AI in mathematics and statistics assessments

Abstract

The mathematical sciences and operational research (MSOR) community in higher education is still largely unprepared to adapt to the rapid rise of generative artificial intelligence (genAI) and its impact on assessment strategies. Whilst in-person exams remain an essential assessment mode for MSOR, take-home assignments are also an integral assessment tool. This work investigates concerns that current assignments are not robust against genAI and the way students use genAI. In this work, we address the following questions: 1) How well can genAI perform in current assignments? 2) To what extent do students currently use AI in take-home assignments? 3) How should assessment strategies evolve given the rapid improvement of genAI? Our research involves an investigation of genAI’s performance in a range of MSOR assignments. We also conducted surveys and discussions with mathematics and statistics students and staff at the University of Warwick. We make recommendation and conclude that genAI represents a catalyst for innovation and assignments, perhaps adapted, should remain a core assessment in MSOR.