TABLE 2.
Model comparison.
| Construct(s) | Model | Description | #Parameters | DIC |
|---|---|---|---|---|
| EmEx | M1 | Unconditional Multilevel AR(1) Model | 5 | 37829.69 |
| M1a | M1 with variance of AR(1) path fixed to a close-to-zero (0.001) value | 4 | 40622.21 | |
| M2 | Residual DSEM (RDSEM) with a linear trend | 9 | 108224.44 | |
| M2a | RDSEM with a linear trend but its variance fixed to a close-to-zero (0.001) value | 8 | 139776.64 | |
| MWB | M1 | Unconditional Multilevel AR(1) Model | 5 | 48175.75 |
| M1a | M1 with variance of AR(1) path fixed to a close-to-zero (0.001) value | 4 | 50666.04 | |
| M2 | Residual DSEM (RDSEM) with a linear trend | 9 | 148709.85 | |
| M2a | RDSEM with a linear trend but its variance fixed to a close-to-zero (0.001) value | 8 | 149756.94 | |
| EmEx & MWB | M3 | VAR(1) model | 22 | 174096.25 |
| M3a | VAR(1) model with “EmEx → MWB” (φ3) level-2 variance (τ44) fixed to a close-to-zero (i.e., 0.001) value | 21 | 174319.45 | |
| M3b | VAR(1) model with “MWB → EmEx” (φ4) level-2 variance (τ55) fixed to a close-to-zero (i.e., 0.001) value | 21 | 174955.32 |
EmEx, emotional exhaustion; MWB, mental well-being; #Parameters, number of estimated parameters; DIC, Deviance Information Criterion. Preferred models are reported in bold.