Reliability and Validity of a Global Positioning System for Measuring Movement Patterns during Field Hockey

Abstract

The scarcity of research into the physiological demands of women's field hockey has inhibited the development of a systematic and scientific approach to preparation for competition at any level of field hockey. Global positioning systems (GPS) have the potential to further our understanding of the physiological demands of competitive sport through the process of tracking player movement patterns. PURPOSE: To assess the reliability and validity of GPS technology for measuring player movement patterns in field hockey. METHODS: Nine subjects volunteered for the study. A predetermined circuit, painted onto an Astroturf surface, was designed to replicate the movement patterns and time spent in each motion category (stand, walk, jog, cruise, sprint), with that performed during a game of field hockey. Each subject completed 14 laps of the circuit to replicate the average distance covered in a match (6818 m). Four timing gates were positioned at various stages of each lap to accurately measure time over a known distance. A calibrated trundle wheel pedometer was used to measure the exact length of the circuit. Players were tracked throughout the duration of the circuit using a GPS athlete-tracking device (SPI Elite, GPSports, Canberra, Australia). The data was edited to establish GPS and actual time, speed and distance at each timing gate, lap and whole circuit. Triplicate repeat trials were conducted at 9am, 1pm and 4pm to assess satellite positioning on reliability of the system. RESULTS: The mean (± SD) total distance covered was 6818.7 ± 5.2m, 6818.7 ± 4.3m and 6819.0 ± 5.4m during the 9am, 1pm and 4pm trial respectively. Repeated measures ANOVA revealed no significant difference (p>0.05) between time of day and the ICC was significant (r=0.809) for total distance covered. Paired t-tests indicated no systematic bias relative to random error. The Pearson correlation for the total distance covered and speed was r = 1 (p<0.01) and r = 0.998 (p<0.001) respectively. A typical error of 4.77 m for total distance covered and 0.01 m/s for speed was observed. Bland and Altman plots for total distance covered and speed shows a mean difference and limits of agreement of 2.2 ± 13.2m and 0.0 ± 0.9 m/s respectively. CONCLUSIONS: In conclusion GPS is a reliable and valid measurement tool for assessing the movement patterns of field hockey players.