Implementation of Gemini AI in the Teaching at the Right Level (TaRL) Approach to Improve Critical Thinking and Problem-Solving Skills in Hydrodynamics Courses Rosane Medriati, Desy Hanisa Putri, Netriani Veminsyah Ahda, Tiara Hardyanti Utama
Physics Education, Bengkulu University, Jl. WR. Supratman, Kandang Limun, Kec. Muara Bangka Hulu, Sumatera, Bengkulu, Indonesia
Abstract
This study investigates the effectiveness of implementing Artificial Intelligence (AI) Gemini through the TaRL approach in hydrodynamics courses to enhance students critical thinking and problem-solving abilities. Using a quantitative descriptive method with a pretest-posttest design, the study involved students from the Physics Education Study Program at Bengkulu University. The research was conducted in two stages: a limited test and an extensive test. In the limited test, the average critical thinking score increased from 27.9 (SD = 14.5) to 55.5 (SD = 15.5), while problem-solving skills rose from 33.7 (SD = 11.5) to 56.3 (SD = 16.7). This stage served to refine the Gemini AI module to improve adaptability. In the extensive test, critical thinking improved from 51.73 (SD = 25.86) to 70.10 (SD = 16.28), and problem-solving from 53.85 (SD = 25.12) to 70.58 (SD = 22.50). The effect size values were high (d = 0.849 for critical thinking and d = 0.701 for problem solving), demonstrating significant learning gains. The higher initial standard deviation indicated varied learning outcomes, while the reduced standard deviation in the extensive test showed more uniform performance. Overall, the AI Gemini-TaRL integration effectively enhanced student achievement, adaptability, and learning consistency in physics education.
Keywords: AI Gemini, Critical Thinking, Hydrodynamics, Problem Solving, TaRL
