DISCOVERING PATTERNS OF GENERATIVE AI USE AMONG UNIVERSITY STUDENTS: EVIDENCE FROM AN ONLINE SURVEY AND ASSOCIATION RULE MINING

Abstract

Generative artificial intelligence tools are gradually reshaping how university students read, write, and prepare for academic tasks. However, within Nigerian universities, there is still limited empirical evidence on how students actually use these tools and how such use relates to verification practices, institutional policy awareness, and learning experiences. This study examined patterns of generative AI use among university students in Nigeria using an online survey and association rule mining. A total of 700 valid responses collected between 18 December 2025 and 18 January 2026 were analyzed. Descriptive statistics were used to examine the extent to which AI tools are used, how frequently students rely on them, which tools they prefer, and the academic activities for which they are used. Association rule mining was then employed to identify recurring relationships among AI use behaviour, checking and editing practices, perceived policy communication, policy understanding, and students’ learning experiences. The findings show that generative AI has already become part of the everyday academic routine for many students. Most respondents reported using AI to clarify difficult concepts, summarize learning materials, organize assignment ideas, and prepare for examinations. The analysis also indicates that students who perceived clearer communication about institutional AI policies were more likely to understand acceptable AI use. At the same time, frequent use sometimes appeared alongside signs of reliance that may affect deeper engagement with learning. These results highlight the need for clearer guidance and practical AI literacy support within Nigerian universities.