TABLE 4.
Shrinkage in model A and B.
| Linear pool | Logarithmic pool | Fitted normal | Default | |
| MPonMint | 0.67 | −0.00 | 0.54 | > 0.99 |
| MPonAint | 0.89 | 0.04 | 0.89 | > 0.99 |
| APonMint | 0.51 | 0.50 | 0.51 | > 0.99 |
| APonAint | 0.83 | 0.60 | 0.81 | > 0.99 |
| MintonMP | 0.54 | 0.00 | 0.66 | > 0.99 |
| MintonAP | 0.27 | 0.39 | 0.32 | > 0.99 |
| AintonMP | 0.89 | 0.10 | 0.84 | > 0.99 |
| AintonAP | 0.82 | 0.01 | 0.84 | > 0.99 |
| MNonMint | 0.57 | 0.49 | 0.57 | > 0.99 |
| MNonAint | 0.82 | 0.20 | 0.77 | > 0.99 |
| ANonMint | 0.58 | 0.58 | 0.58 | > 0.99 |
| ANonAint | 0.81 | 0.67 | 0.80 | > 0.99 |
| MintonMN | 0.58 | −0.00 | 0.69 | > 0.99 |
| MintonAN | 0.28 | 0.30 | 0.30 | > 0.99 |
| AintonMN | 0.86 | 0.07 | 0.81 | > 0.99 |
| AintonAN | 0.87 | −0.01 | 0.88 | > 0.99 |
M, maternal; A, adolescent; int, internalizing; P, positive interaction behavior; N, negative interaction behavior;on, describes the direction of regression, indirect effects are reported in direction of the association (e.g., MintMPAint describes the indirect effect from maternal to adolescent internalizing symptoms via maternal positive interaction behavior).