Table 5.

Average posterior inference on herd variances from ten replicates using the bivariate Student's-t (BSt) and normal (BN) fitted models with different residual degrees of freedom (DF).

		Fitted Model2
		BSt			BN
True Parameters	True Model1	PM ± SE3	95% PPI4	ESS5	PM ± SE	95% PPI	ESS
= 1.5	BSt-4	1.71 ± 0.10	[1.07, 2.56]	12,027	1.74 ± 0.14	[1.01, 2.71]	10,437
	BSt-12	1.82 ± 0.12	[1.19, 2.65]	16,133	1.82 ± 0.12	[1.18, 2.65]	16,802
	BN-∞	1.71 ± 0.08	[1.12, 2.48]	17,543	1.70 ± 0.08	[1.12, 2.47]	17,537
= 6.0	BSt-4	6.33 ± 0.30	[4.47, 8.79]	24,120	6.49 ± 0.28	[4.47, 9.17]	22,810
	BSt-12	6.69 ± 0.28	[4.77, 9.22]	27,704	6.72 ± 0.24	[4.79, 9.28]	27,956
	BN-∞	6.33 ± 0.27	[4.55, 8.71]	29,800	6.33 ± 0.27	[4.55, 8.71]	29,881

¹Used to simulate data

²Used in analysis of simulated data

³Posterior mean ± Standard Error

⁴95% equal-tailed posterior probability interval based on the 2.5^th and 97.5^th percentiles of the posterior density

⁵Effective sample size