Table 1.
Comparison of two methods based on 100 simulation runs. Mean (SE) of the integrated absolute errors of estimating the mean functions and predicting the unit and sub-unit level random effects. Numbers shown are the actual numbers multiplied by 10. “Reduced rank” refers to our method; “Bayesian” refers to the Bayesian method of Baladandayuthapani, et al. (2008).
| Setup | Method | Mean | Unit | Sub-Unit | |||
|---|---|---|---|---|---|---|---|
| 1 | Reduced rank | 1.332 | (0.061) | 1.361 | (0.072) | 0.861 | (0.051) |
| Bayesian | 1.945 | (0.041) | 3.512 | (0.046) | 1.637 | (0.030) | |
| 2 | Reduced rank | 1.633 | (0.057) | 2.152 | (0.097) | 1.628 | (0.100) |
| Bayesian | 2.161 | (0.046) | 3.797 | (0.052) | 2.051 | (0.023) | |
| 3 | Reduced rank | 1.151 | (0.040) | 1.487 | (0.059) | 1.117 | (0.062) |
| Bayesian | 1.886 | (0.023) | 2.681 | (0.036) | 1.502 | (0.016) | |
| 4 | Reduced rank | 1.492 | (0.056) | 1.517 | (0.055) | 0.543 | (0.014) |
| Bayesian | 2.049 | (0.034) | 3.366 | (0.046) | 1.055 | (0.011) | |
| 5 | Reduced rank | 4.170 | (0.128) | 5.571 | (0.094) | 4.063 | (0.039) |
| Bayesian | 4.086 | (0.146) | 5.518 | (0.108) | 4.104 | (0.046) | |