Skip to main navigation menu Skip to main content Skip to site footer

IMPACT OF ARTIFICIAL INTELLIGENCE TOOLS ON RESEARCH WRITING AMONG UNDERGRADUATE ENGLISH STUENTS IN NORTH-WEST NIGERIA

Abstract

The proliferation of Artificial Intelligence (AI) tools like ChatGPT, Grammarly, and QuillBot has fundamentally transformed academic writing workflows globally. However, empirical understanding of how these technologies alter the writing skills and psychological profiles of undergraduate humanities students in regional African contexts remains limited. This study evaluated the impact of AI tools on the research writing processes of undergraduate English language students in public universities across North-West Nigeria. It investigated the relationships between AI usage frequency, perceived writing quality, writing self-efficacy, task completion, and the presence of institutional guidelines. A descriptive survey design was utilized to sample 123 two hundred-level undergraduate English students across public federal universities in the North-West geo-political zone. Data were gathered using a structured questionnaire measuring AI exposure metrics, writing dependencies, and ethical anxieties. Analysis was performed using Pearson and Spearman correlation coefficients alongside Mann-Whitney U tests. A moderate positive relationship was observed between AI usage frequency and perceived writing quality improvements (r = 0.500, p < 0.001, R^2 = 0.250). Conversely, frequent AI usage was significantly and negatively correlated with writing self-efficacy (r_s = -0.265, p = 0.003$) and perceived task completion efficiency (r = -0.384, p < 0.001). Additionally, a significant negative correlation was identified between general AI reliance and independent writing confidence (r = -0.237, p = 0.008). Finally, students lacking clear institutional frameworks exhibited significantly higher anxieties regarding plagiarism and inaccuracies (U = 2078.0, p = 0.002).While AI tools enhance perceived text quality, they act as a psychological double-edged sword, causing lower writing self-efficacy, slower independent task completion, and structural dependency. Clear institutional policies are urgently needed to mitigate ethical anxieties and preserve core literacy skills.

Keywords

Artificial Intelligence, Research Writing, Writing Self-Efficacy, Undergraduate English Students, Higher Education Policy

PDF