Automation of learning outcome generation using a Large Language Model

Abstract

Learning Outcomes are clear and specific statements about what students should know, understand, and be able to do at the end of a learning process. They are a fundamental pillar in curriculum design and competency assessment, ensuring coherence between teaching, learning, and evaluation. Their proper formulation is key to constructive alignment and educational planning. Traditionally developed through manual writing, learning outcomes present challenges such as structural variability, the lack of standardized guidelines, and the workload for educators. To address these difficulties, methodologies based on competency matrices, cognitive taxonomies, and tools such as analytical rubrics have been developed, facilitating the connection between competencies, teaching strategies, and assessment. However, the increasing need for standardization and efficiency in their formulation highlights the necessity of automating their generation. This article analyzes the writing process of learning outcomes and the feasibility of automation to improve coherence and applicability in higher education.