Table 1.

Estimated growth parameters, standard errors, and model selection criteria from fixed- and mixed-effects growth models for mixed-sex New Hampshire and Brown Leghorn heritage birds.

Model 1	Model [1]			Model [2]			Model [3]			Model [4]
Parameter 2	Estimate	SEM	Pr > |t|	Estimate	SEM	Pr > |t|	Estimate	SEM	Pr > |t|	Estimate	SEM	Pr > |t|
Wm	1.98	0.0505	<0.001	1.97	0.097	<0.001	2.94	0.029	<0.001	1.99	0.105	<0.001
B	0.027	0.0009	<0.001	0.027	0.0001	<0.001	0.017	0.0006	<0.001	0.027	0.0004	<0.001
σ2e	0.053	0.0013	<0.001	0.0019	4.8E-5	<0.001	0.0049	0.0001	<0.001	0.001	2.5E-5	<0.001
σ2Wm	-	-	-	0.2985	0.075	0.004	-	-	-	0.34	0.087	0.004
σ2b	-	-	-	-	-	-	1.2E-5	4.6E-6	0.013	5.8E-6	2.34E-6	0.10
σWmb	-	-	-	-	-	-	-	-	-	−0.00042	0.00029	0.15
Model fitting statistics3
−2log-likelihood		−263			−10,733			−7,714			−12,679
AIC		−255			−10,723			−7,704			−12,665
AICC		−255			−10,723			−7,704			−12,665
BIC		−230.7			−10,716			−7,697			−12,655
Cross-validation statistics4
MAE		0.1565			0.0304			0.0565			0.0183
MSE		0.0548			0.0019			0.0049			0.0010
RMSE		0.2342			0.0439			0.0705			0.0317
R-square		0.824			0.993			0.984			0.996

¹Model [1]: Fixed-effects model $BW=Wm\times ex{p}^{-ex{p}^{-b(t-t_{inf})}}$. Model [2]: Mixed-effect model including a random effect (Wm_u) for Wm. Model [3]: Mixed-effect model including a random effect (b_u) for b. Model [4]: Mixed-effects model including random effects (Wm_u and b_u) for both Wm and b.

²Wm = mature BW; b = rate of maturing; σ²_e = residual BW variance; σ²_Wm = the individual variance in mature BW; σ²_b = the individual variance in rate of maturing; σ_Wmb = individual covariance between random effects.

³AIC = Akaike information criterion; BIC = Bayesian information criterion.

⁴MAE = Mean absolute error; MSE = Mean square error; RMSE = Root mean square error; R-square = Coefficient of determination of observed BW with predicted BW by the testing model in a k-fold cross validation.