An investigation of the reliability, validity, and impact of operator expertise on assessing vertical jump height in collegiate badminton athletes with My Jump Lab

TL;DR

This study found that My Jump Lab reliably measures vertical jump heights in badminton athletes, with minimal impact from operator experience.

Contribution

The study demonstrates that My Jump Lab is a reliable and valid tool for jump height assessment, even with varying operator experience.

Findings

Operator experience had no significant effect on My Jump Lab measurements.

My Jump Lab showed excellent agreement with the gold-standard OptoJump system.

Proportional bias was observed for CMJAM but not for CMJ or SJ.

Abstract

This study aimed to evaluate the influence of operator experience on vertical jump height measurements in university badminton athletes using My Jump Lab, while concurrently assessing the tool's reliability and validity against the gold-standard OptoJump system.

Seventy-six university badminton athletes (32 females, 44 males) participated in the study. Three vertical jump modalities—countermovement jump (CMJ), countermovement jump with arm swing (CMJAM), and squat jump (SJ)—were simultaneously measured using My Jump Lab and OptoJump. My Jump Lab data were processed by two operators with substantially different levels of experience. A mixed-design repeated-measures analysis of variance (ANOVA) was conducted to examine the effect of operator experience on measurement outcomes. Bland-Altman analysis, complemented by linear regression of differences vs. means, was employed to evaluate measurement agreement and detect proportional bias.

Intraclass correlation coefficients (ICC₁,₁) for both experienced and inexperienced operators were ≥0.92, with coefficients of variation (CV) < 5%, indicating excellent intra-operator reliability. Inter-operator reliability was exceptionally high, with ICC₂,₁ values of 0.992 (CMJ), 0.992 (CMJAM), and 0.983 (SJ), all accompanied by CV < 5%. Repeated-measures ANOVA confirmed that the main effect of operator experience was statistically non-significant [F(1, 150) = 0.338, p = 0.562, ηₚ2 = 0.002], as were its interaction effects with jump type and number of jumps (all p > 0.89). Validity analysis demonstrated excellent agreement between My Jump Lab and OptoJump: ICC₂,₁ ≥ 0.990, mean differences <1 cm, 95% limits of agreement (LOA) ranging from −3.26 to 2.28 cm, and high linear fit (R2 > 0.98) via Ordinary Least Products Regression. My Jump Lab exhibited a slight systematic overestimation across all jump types. Proportional bias testing revealed no significant bias for CMJ (slope = 0.0118, 95% CI [−0.0047, 0.0283], p = 0.1594) or SJ (slope = 0.0177, 95% CI [−0.0010, 0.0364], p = 0.0639), whereas significant proportional bias was observed for CMJAM (slope = 0.0398, 95% CI [0.0216, 0.0581], p < 0.001), with differences between the two systems increasing proportionally with jump height.

Operator experience exerts minimal practical impact on My Jump Lab derived vertical jump height measurements in university badminton athletes. The tool demonstrates robust reliability and validity, making it suitable for routine training monitoring by operators with varying experience levels in grassroots sports settings. While significant proportional bias was identified for CMJAM, this does not compromise My Jump Lab's utility for relative performance monitoring (e.g., longitudinal training progress tracking), and absolute measurement accuracy can be enhanced via calibration if required. Strict standardization of movement protocols and post-test video review remain essential for SJ testing to mitigate inaccuracies associated with unintended countermovement.

Figures