AI Tutors or Distractions? How AI Affects University Students' Learning Habits In Lower Dir

Abstract

Adoption of artificial intelligence (AI) tools, such as ChatGPT, Grammarly, adaptive learning tools and others have swiftly spread in higher education. However, their effectiveness in influencing the learning process and outcome of the students’ needs more exploration in developing regions particularly in Lower Dir, Pakistan. This study was designed to describe the frequency and purpose of using AI tools in university students in Lower Dir, to evaluate the relationship between AI uses and students' self-reported learning behaviours, and to investigate if the excessive use of AI is associated with learning outcomes. A cross-sectional survey was conducted of 260 students (130 males, 130 females) from four higher education institutions in Lower Dir, Pakistan. The participants were selected via stratified random sampling. The questionnaire was used as a data collection tool in which the use of AI tools and frequency, learning practices, and self-reported average in grades were assessed. Descriptive statistics and chi-square were used for the analysis.The findings revealed that 78.5% of students use AI tools frequently, with ChatGPT the most often used (68.1%). The most common uses were grammar checking (71.2) and learning hard concepts (64.6). Nevertheless, 42.3 percent of students have consented that they occasionally depend on AI rather than find their own solutions to problems. The frequency of AI usage was positively correlated with self-reported grades, although weakly and significantly (r = 0.24, p < 0.01). The use of AI was also less by female students compared to male students (mean 2.9 vs. 3.5 on a 5-point scale, p < 0.05). AI applications are extensively utilized but not always on in-depth work. Although it is modestly related to improved grades, over-reliance is a problem. Universities are to include AI literacy training and provide equal opportunities to female students.