The Validity and Reliability of the My Jump Lab Artificial Intelligence Application

Abstract

Jump height (JH) achieved in a countermovement jump (CMJ) has been suggested to allow for the monitoring of neuromuscular fatigue (NMF) and assessment of lower body power. Although force platforms (FP) are considered the gold standard for measuring CMJ height, they are expensive compared to mobile apps such as My Jump Lab (MJL). Therefore, this study aimed to assess the concurrent validity and agreement of the MJL app compared to a FP (ForceDecks [FD]) system and to determine its test-rest reliability. A convenience sample of 26 (n = 11 females and n = 15 males) recreationally active university sport students and staff (mean ± SD; age: 23.08 ± 6.33 years; mass: 72.85 ± 9.93 kg; stature: 176.63 ± 10.18 cm) participated in the study. Participants attended the laboratory for testing on two separate occasions, separated by one week. After a standardised warm-up, they completed three CMJs on each occasion, with CMJ height simultaneously assessed by the FD and MJL app. The MJL Artificial Intelligence mode showed a mean bias of 4.32 cm [95% CI: 3.4, 5.26] overestimation with 95% limits of agreement ranging from -3.33 cm [95% CI: -4.96, -0.85] to 11.98 cm [95% CI: 10.13, 13.41]. Both methods demonstrated minimal mean bias (FD = 0.61 cm [95% CI: -0.31, 1.37] and MJL = 0.25 cm [95% CI = -0.48, 0.98]) between sessions, and both showed a similar width to their limits of agreement, ranging ~7 cm about the mean bias. In summary, the MLJ overestimated CMJ height in this sample compared to the FD system, but both methods were reliable. Given the significant differences in cost for these two methods, teams on a budget may interested in trialling the MJL app.