The reliability and validity of a field hockey skill test.

High test retest reliability is essential in tests used for both scientific research and to monitor athletic performance. Thirty-nine (20 male and 19 female) well-trained university field hockey players volunteered to participate in the study. The reliability of the in house designed test was determined by repeating the test (3 - 14 days later) following full familiarisation. The validity was assessed by comparing coaches ranks of players with ranked performance on the skill test. The mean difference and confidence limits in overall skill test performance was 0.0 +/- 1.0 % and the standard error (confidence limits) was 2.1 % (1.7 to 2.8 %). The mean difference and confidence limits for the "decision making" time was 0.0 +/- 1.0 % and the standard error (confidence limits) was 4.5 % (3.6 to 6.2 %). The validity correlation (Pearson) was r = 0.83 and r = 0.73 for female players and r = 0.61 and r = 0.70 for male players for overall time and "decision making" time respectively. We conclude that the field hockey skill test is a reliable measure of skill performance and that it is valid as a predictor of coach-assessed hockey performance, but the validity is greater for female players.