Table 1.

Linear rates of speed improvement estimated from datasets from model 1. Parameter estimates are from REML models with year fitted as continuous covariate. Inference is by likelihood comparison of full and reduced models fitted by ML (see text for details).

datasetyearsclassesfinishersdistance (furlongs)temporal trend ± s.e. (yards s−1 year−1)Embedded Imagep
1.11850–2012elitewinners5–70.014 ± 5 × 10−4659<0.001
1.21850–2012elitewinners8–120.013 ± 4 × 10−4677<0.001
1.31850–2012elitewinners14–200.011 ± 0.001106<0.001
1.41997–2012elitewinners5–70.020 ± 0.00264.3<0.001
1.51997–2012elitewinners8–120.006 ± 0.0025.80770.016
1.61997–2012elitewinners14–200.007 ± 0.0052.710.100
1.71997–2012eliteall5–70.023 ± 0.001409<0.001
1.81997–2012eliteall8–120.006 ± 0.00126.0<0.001
1.91997–2012eliteall14–200.008 ± 0.00212.3<0.001
1.101997–2012allwinners5–70.014 ± 6 × 10−4466<0.001
1.111997–2012allwinners8–120.006 ± 7 × 10−470.6<0.001
1.121997–2012allwinners14–200.005 ± 0.00211.2<0.001
1.131997–2012allall5–70.018 ± 4 × 10−42212<0.001
1.141997–2012allall8–120.010 ± 4 × 10−4634<0.001
1.151997–2012allall14–200.009 ± 8 × 10−4114<0.001