Buenas prácticas que integran la Inteligencia artificial en el proceso de evaluación en la Educación Superior

Abstract

Artificial Intelligence (AI) is causing a significant transformation in higher education, reshaping both teaching and learning methods, as well as the way we evaluate them (Fischer, et al., 2020). In this regard, academic assessment holds a crucial role as as it provides feedback to students, assesses their comprehension, and facilitates continuous learning improvement. Traditional assessment techniques, however, have their limitations in terms of personalization, efficiency, and their ability to adapt to the unique needs of individual students (Huang, et al., 2021). Therefore, the aim of this study is to analyze the scientific production regarding assessment in artificial intelligence in higher education. To this end, the study adhered to the guidelines outlined in the PRISMA Declaration (Urrútia & Bonfill, 2010). Out of the 110 articles pulled from the Web of Science database, 56 empirical studies were selected for examination. The possibilities offered by each of the developed systems are diverse, and all conclude that their implementation leads to an improvement in the academic performance of higher education students. In the various machine learning practices, we can find: the adjustment of question difficulty and content based on each student’s skill level and prior knowledge, automated and personalized feedback on tasks completed, or the analysis of data sources that provide relevant information about learning preferences, interaction patterns, or the prediction of possible study abandonment. All of this allows for a more precise and specific assessment, optimizing the time dedicated to assessment and providing a more meaningful assessment experience (Luckin, 2017). In conclusion, this review provides us with a range of empirical studies in which the integration of Artificial Intelligence in higher education assessment shows promising results, positioning this possibility as a new path toward improving the learning assessment process