Table 2. Dependent variables and covariates in youth ice hockey players at baseline, and magnitude-based inferences for parallel groups Ice1 and Ice2. Test Group (n = 7) Baseline Mean ± SD Observed change Mean ± SD (n = 6) Probability [%; %; %] Ratio (Ice1/Ice2) Mean; ±90% CL Inference Weave agility (s) Ice1 17.9 ± 1.3 -4.9 ± 12.1% [9; 14; 77] -0.5; ± 8.9% unclear effect Ice2 17.8 ± 1.2 -4.4 ± 5.3% [1; 9; 90] Reaction agility (s) Ice1 5.7 ± 0.3 -5.6; ± 5.3% [1; 4; 96] -1.0; ± 5.1% harmful* [47; 31; 22] Ice2 5.8 ± 0.3 -4.7 ± 5.1% [10; 4; 86] CL = confidence limit with 90% confidence interval for the mean. Observed changes are percentages; baseline values are expressed in measurement units. Likelihood that the true effect is substantial: *possibly. Magnitude threshold: 0.20-0.59. Numbers in brackets indicate the % probability of a harmful, trivial and beneficial effect for both probability of group change and group inference. <0.20, trivial; 0.20-0.59, small; 0.60-1.19, moderate; >1.20, large.