<![CDATA[This study examines how the intensity of Artificial Intelligence (AI) use in learning affects students' learning independence. A quantitative approach was applied using a correlational design and simple linear regression. Data were obtained from 63 seventh-grade students at SMP Negeri 2 Saronggi, Sumenep Regency. The instruments measured AI usage intensity through 5 items and learning independence through 6 items, both using a 1–5 Likert scale. Both measures showed high reliability: α=0.889 for the AI intensity scale and α=0.830 for the learning independence scale, and both were categorized as reliable. According to the Shapiro–Wilk test, the AI scores (p=0.156), learning independence scores (p=0.096), and regression residuals (p=0.750) were normally distributed. Pearson correlation indicated a very strong positive relationship between the two variables (r=0.912; p<0.001). Simple linear regression produced the equation KB=6.434+0.866AI (R²=0.832, F=302.819, p<0.001). In the research sample, AI usage intensity accounted for approximately 83.2% of the variation in learning independence. These results require qualification: Harman's single-factor test found significant common method variance, with a single factor explaining 58.427% of the variance. The observed effect may include inflation caused by similarity in measurement methods and should therefore be treated as an upper bound rather than a pure effect estimate. Even with this limitation, the findings support the alternative hypothesis that greater AI usage intensity has a positive and significant effect on students' learning independence. The findings also highlight the need for AI literacy and educator support to ensure that AI strengthens students' self-regulation in learning rather than replacing it.]]>
Copyrights © 2026
