Trial-to-Trial Repeatability and Curve-Level Reliability of Rotational-Progress-Aligned Kinematic Waveforms During the Primary 360° Rotation of a Rumba Turning Sequence.

Abstract

The rumba turning sequence is technically demanding, but the repeatability of continuous kinematic waveforms during repeated performance remains unclear. This study examined trial-to-trial repeatability and curve-level reliability of rotational-progress-aligned kinematic waveforms during the primary 360° rotation in a standardized rumba turning sequence, corresponding to the initial spiral turn. Fifteen high-level adolescent female Latin dancers completed 5 valid trials at 76 beats per minute. Kinematic data were collected using a 10-camera Vicon 3-dimensional motion capture system (100 Hz). Support-foot progression angle, pelvic obliquity, and thorax-pelvis relative frontal-plane angle were analyzed across 101 rotational-progress points. Whole-waveform similarity, point-by-point SD, integrated intraclass correlation coefficient(A,1), integrated standard error of measurement, and key-event reliability were calculated. All variables showed high whole-waveform similarity (r = .932-.944). Integrated intraclass correlation coefficient(A,1) ranged from .545 to .675. Integrated standard error of measurement was smallest for pelvic obliquity (3.98°) and largest for support-foot progression angle (10.55°). Variability hotspots occurred in specific rotational-progress regions, and magnitude-related key events generally showed smaller absolute error than timing-related events. These findings indicate that continuous kinematic variables differ in reliability and measurement error and should be selected cautiously for repeated technical assessment.