Abstract
This chapter explores the integration of generative AI in higher education assessment, addressing the inadequacies of traditional methods in meeting the diverse needs of contemporary learners. It highlights the potential of AI technologies, such as natural language processing and computer vision, to offer personalized, scalable, and insightful evaluations. The chapter critically examines both the enhanced capabilities introduced by AI in educational settings and the ethical challenges it poses. Emphasizing the need for a balanced approach, it suggests synergizing AI's analytical strengths with human expertise to ensure equitable and effective assessments. This work aims to guide educators, administrators, and policymakers through the complexities of AI adoption in academic evaluation, focusing on maintaining academic integrity and inclusivity while leveraging the transformative potential of AI in education.
TopChanging Needs For Assessment In Contemporary Higher Education
Higher education has experienced massive growth and diversification of the student body over the past 50 years. Where universities previously served a homogeneous population of society's elite, today's campuses reflect the broad societal spectrum – first-generation students, adult learners, low-income and racial minority populations, non-native language speakers, students with disabilities, and more (Erisman & Looney, 2007). Classes which used to contain a few dozen similar students now often have hundreds of individuals with widely varying backgrounds, skill levels, motivations and needs.
Standardized assessments and one-size-fits-all approaches are clearly inadequate for effectively measuring and supporting learning across this highly heterogeneous group. Unfortunately, such practices still dominate at most institutions. Multiple-choice question assessments remain common due to their efficiency and reliability for statistical analysis but are criticized for their inability to measure higher-order thinking (Newble & Jaeger, 1983). The prevalence of high-stakes final exams over continual formative assessment disincentives risk-taking and causes unnecessary stress. Manual paper grading does not readily permit individualized diagnosis or feedback. Curving grades on a distribution often pits students against each other rather than promoting collective learning.
Key Terms in this Chapter
Adaptive Learning: A personalized educational approach that uses technology to tailor learning experiences to individual needs. It modifies content, presentation style, and pace based on real-time feedback on learner performance.
Machine Learning (ML): A subset of AI, Machine Learning involves algorithms that enable computers to learn and adapt from experience without being explicitly programmed. ML focuses on developing systems that can access and use data to learn for themselves.
Automated Essay Scoring (AES): The use of specialized software to evaluate and score written essays. AES algorithms assess various elements of writing, such as grammar, syntax, coherence, and organization.
Bias in AI: Refers to the tendency of AI systems to make unfair or prejudiced decisions, often due to skewed data or flawed algorithmic design. AI bias can result in discrimination and misrepresentation of certain groups.
Natural Language Processing (NLP): An area of AI that focuses on the interaction between computers and human languages. NLP aims to enable computers to understand, interpret, and respond to human language in a meaningful way.
Intelligent Tutoring Systems (ITS): Computer-based systems that provide personalized instruction or feedback to learners, simulating the methods of a human tutor. ITS adapt to the learner’s pace and understanding, offering tailored guidance and support.
Artificial Intelligence (AI): A field of computer science dedicated to creating machines capable of performing tasks that typically require human intelligence. AI systems are designed to simulate human cognitive functions like learning, problem-solving, and decision-making.