Table 2.

Maximum likelihood estimation of parameters of six models.

Model	$C_a\mathrm{ }$ (95% CI)	$C_e$ (95% CI)	$C_0\mathrm{ }$ (95% CI)	Maximum log likelihood	AIC	$\Delta$ AICa
(M6) $Ns=C_{a}a.\mathit{wilcox}+C_{e}e+C_0$	0.126 (0.073, 0.185)	0.572 (0.265, 0.922)	0.115 (-0.012, 0.243)	–243.881	493.761	0
(M5) $Ns=C_{a}a.\mathit{diff}+C_{e}e+C_0$	0.032 (0.018, 0.047)	0.521 (0.206, 0.870)	0.141 (0.016, 0.265)	–244.149	494.298	0.537
(M2) $Ns=C_{a}a.\mathit{diff}+C_0$	0.031 (0.018, 0.045)	−b	0.228 (0.115, 0.329)	–249.672	503.345	9.584
(M3) $Ns=C_{a}a.\mathit{wilcox}+C_0$	0.119 (0.065, 0.175)	–	0.217 (0.102, 0.333)	–250.655	505.310	11.549
(M4) $Ns=C_{e}e+C_0$	–	0.526 (0.219, 0.867)	0.225 (0.109, 0.343)	–254.443	512.886	19.125
(M1) $Ns=C_0$	–	–	0.313 (0.207, 0.421)	–260.441	522.882	29.121

^a $\Delta$ AIC is calculated by subtracting AIC of M6 from the AIC of the model.

^bThe hyphens indicate the model does not use that parameter.