Table 5.

Posterior means of variation parameters for case study 2 (multilevel model).

Cycle	Method${}^1$	Level-1 SD	Level-2 Var.-Intercept	Level-2 Covar.	Level-2 Var.-FEMALE
2003	BLR non inf	77.12	1069.02	$-$35.54	10.40
2006	BLR non inf	76.42	1185.44	$-$16.86	7.98
2009	BLR non inf	73.83	1358.10	3.57	10.67
2012	BLR non inf	74.35	1221.41	$-$11.83	4.43
2015	BLR non inf	78.41	1152.04	$-$55.28	18.66
2018	BLR non inf	79.21	949.58	$-$39.44	37.80
	BLR inf	79.22	951.49	$-$36.50	34.35
	BLR pooling	76.56	1255.06	$-$73.36	16.36
	BDB IG(1,1) W2	76.48	972.51	$-$42.84	10.16
	BDB IG(1,.1) W2	76.48	974.05	$-$40.97	9.71
	BDB IG(1,.001) W2	76.48	971.30	$-$42.69	9.82
	BDB IG(1,1) W20	76.47	1016.67	$-$57.70	15.22
	BDB IG(1,.1) W20	76.47	1015.61	$-$59.58	15.66
	BDB IG(1,.001) W20	76.46	1011.85	$-$57.25	14.68
	PP (.25)	79.88	618.45	$-$14.61	6.33
	PP (.5)	77.99	974.91	$-$26.88	5.70
	PP (.75)	77.08	1151.31	$-$49.57	9.43

${}^1$BLR non inf: Bayesian linear regression with non-informative priors; BLR inf: Bayesian linear regression with informative priors; BDB: Bayesian dynamic borrowing; IG: inverse-gamma prior for level-1 variance of the joint prior distribution, which determines the degree of level-1 borrowing; W2: Wishart prior with weak borrowing for level-2 precision matrix (results were converted back the covariance matrix); W20: Wishart prior with strong borrowing for level-2 precision matrix (results were converted back the covariance matrix); PP: power priors.