The landscape of gateway STEM student success and teaching practices at R1 institutions in the United States post-COVID and post-AI

Abstract

Gateway science, technology, engineering, and mathematics (STEM) courses have historically been taught primarily with didactic lectures and high-stakes examinations which can contribute to poor student outcomes. Much has changed in higher education in recent years after two historic disruptions: the COVID-19 pandemic and the arrival of generative artificial intelligence (AI). It is therefore important to know how these courses are currently taught and how students are succeeding post-COVID and post-AI. This study is the first of its kind to survey faculty from five different STEM disciplines (biology, calculus, chemistry, physics, and programming) to widely assess gateway STEM failure rates, course design and teaching strategies, and assessment practices at institutions with very high research activity (R1) post-COVID and post-AI. The results show that while gateway STEM course failure rates have declined, traditional lecturing and assessment practices still dominate while evidence-based instructional practices (EBIPs) are used sparingly. Academic leaders and faculty should act on these data locally to determine the extent of potential impacts at their own institutions to design and teach gateway STEM courses to maximize student success.