TABLE 5.

Posterior means of ${\beta }_k$ and $\mu$ based on the Detroit data. For the SSVS methods, we report the mean of ${\beta }_k\mid I_k=1$ if the posterior mean of $I_k$ exceeds 0.5 and 0 otherwise.

		$p\left(I_k=1\right)=0.5$		Informative $p\left(I_k\right)$ $g_1={0.2}^2$
	Laplace prior	$g_1={\tau }^2/100$	$g_1={0.2}^2$		Subset
CBCL Social Problem	−0.090	0.000	0.000	0.000
CBCL Attention Problem	−0.230	−0.274	−0.268	−0.339	−0.475
CBCL Internalizing	0.083	0.000	0.000	0.000
CBCL Externalizing	−0.195	0.000	0.000	−0.324
TRF Social Problem	−0.486	−0.404	−0.412	−0.460	−0.612
TRF Attention Problem	−0.464	−0.396	−0.404	−0.451	−0.602
TRF Internalizing	0.064	0.000	0.000	0.000
TRF Externalizing	−0.499	−0.407	−0.417	−0.464	−0.617
Verbal IQ	−0.112	0.000	0.000	0.000
Performance IQ	−0.141	0.000	0.000	0.000
Freedom from distractibility	−0.561	−0.447	−0.468	−0.505	−0.622
Verbal fluency	−0.327	−0.332	−0.330	−0.388	−0.525
Digit span backwards	−0.453	−0.389	−0.400	−0.446	−0.574
Story memory	−0.061	0.000	0.000	0.000
μ		−0.324	−0.303	−0.398	−0.572
τ		0.219	0.233	0.187	0.172