Getting the most from your data: Using Statistical Process Controls for Data Quality Assurance in Sport Science Data

All levels of sport have seen a surge in data availability, which has allowed for the prospective and retrospective monitoring and tracking of player performance metrics, physical outputs (e.g., distance run, number of accelerations), and injuries. Players, coaches, support staff, governing bodies, and researchers are trying to leverage data to support long-term player welfare and real-time decisions, and this is made possible with advances in data capture, processing, and analysis. Statistical process control (SPC) is a method of quality control designed for understanding, monitoring, and improving process performance over time, historically associated with manufacturing. Visualization of SPC, referred to as a run chart, is accompanied by a mean centerline to show how the underlying process is changing relative to a benchmark. The run chart is visualized with control limits, which are used to determine whether the process is, or is not, operating within statistical control. Deviation of process data outside of the control limits is deemed to be a cause for special variation, highlighting areas that may require investigation. The aims of this methodological report are (a) to provide an example of how SPC can be used in sport and athlete monitoring and (b) provide practical applications for the sports science practitioner. This tutorial provides specific examples from the author's experience in using SPC in the sport field and adjoining simulated data and code to reproduce these results, and more importantly, use as a template for the practitioner's own sport data.