<![CDATA[Accurate passing is a fundamental skill in soccer, yet delivering individualized, objective, and timely feedback in large physical education classes remains challenging. This quasi-experimental study evaluated the effectiveness of AI-based video feedback on improving passing accuracy among 64 physical education students enrolled in a soccer practicum course. Participants completed pretest and posttest measurements using an adapted Loughborough Soccer Passing Test (LSPT), with six intervention sessions providing personalized visual-numeric feedback generated from smartphone-based pose estimation. Results showed statistically significant improvement from pretest (M = 58.3, SD = 12.7) to posttest (M = 71.8, SD = 11.4), t(63) = 9.84, p < .001, with a large effect size (Cohen's d = 1.23). Instrument reliability was high (ICC = 0.89), and 81.3% of participants achieved minimal detectable change. Students with lower initial scores demonstrated greater absolute gains. Technology acceptance measures indicated high perceived usefulness (M = 4.3/5.0) and ease of use (M = 4.1/5.0). Findings suggest that affordable, smartphone-based AI feedback is feasible and effective for enhancing motor skill acquisition in resource-limited educational settings, though future randomized controlled trials with retention measures are needed to confirm causal effects and long-term sustainability.]]>
Copyrights © 2025
