Do ethical perceptions influence the use of artificial intelligence as an assistant in academic essay writing by university students? a decision tree analysis
TL;DR
The results indicate that students primarily value the usefulness and fairness of using GenAI, and ethical perceptions are the main determinants of behavioral intention, whereas gender and individual student circumstances exert only a residual influence.
Abstract
Generative Artificial Intelligence (GenAI) is profoundly transforming higher education. Although its use can enhance the quality, efficiency, and equity of learning, it also raises ethical dilemmas related to authorship, academic integrity, and technological dependence. Understanding the ethical factors that influence students’ intention to use GenAI is therefore crucial for designing responsible educational policies. This study examines how different ethical frameworks—justice (JUST), relativism (REL), consequentialism (CON), and deontology (DEO)—affect the behavioral intention to use GenAI as an assistant in the editing of academic essays. To this end, the Multidimensional Ethics Scale was applied to a sample of 151 university students enrolled in social science programs in Spain. Data were analyzed using decision tree models, incorporating three control variables: gender, employment status, and academic performance. The results show that ethical perceptions are the main determinants of behavioral intention, whereas gender and individual student circumstances exert only a residual influence. CON emerges as the most influential factor, followed by JUST, while REL and DEO are the least relevant ethical constructs. Overall, the findings indicate that students primarily value the usefulness and fairness of using GenAI.