Evaluating the Effectiveness of AI-Driven Applications in Enhancing the Physical Fitness of Tertiary Physical Education Students

Artificial intelligence (AI) has proved its potential in various educational settings. However, its application in exercise-based learning remains unexplored. It is vital as some countries (e.g., the Philippines) have a high prevalence of physical inactivity, and most youth do not meet the recommended levels for maintaining health. Thus, this study investigates how incorporating AI-driven applications into tertiary physical education classes affects students' physical fitness performance. A non-randomized two-group design was utilized to compare pre- and post-test scores. Standardized tests measured cardiovascular endurance (CVE), upper body (UBS), core (CS), and lower body strength (LBS). An 8-week intervention with AI-driven exercise was implemented. Post-test scores showed that the AI-enhanced group (AG) performed significantly better than the traditional group (TG), with CVE at p<0.01 and UBS, CS, and LBS at ≤ 0.001. The TG showed improvement only in CVE and LBS p≤0.01. Conversely, the AG showed significant improvements in CVE, UBS, and LBS (p < 0.001) and CS (p < 0.01). These mean that using AI-driven applications during exercise has a favourable effect on students' physical fitness. Further research is recommended to encompass a variety of populations and educational settings while addressing AI literacy and ethical concerns to maximize its potential in promoting lifelong physical activity participation for overall fitness improvement.