Results of the model comparison, including the likelihood achieved by the best-fitting model as well as the WAIC, and PSIS-LOO (± standard error), which penalize for model complexity.
| Model . | Variant . | Likelihood . | WAIC . | PSIS-LOO . |
|---|---|---|---|---|
| A&A | Homogeneous | −28.1 | 57.7 ± 9.9 | 28.8 ± 9.9 |
| MAS | Homogeneous | 8.2 | −13.3 ± 9.6 | −6.6 ± 9.6 |
| Heterogeneous | 8.2 | −11.3 ± 9.5 | −5.6 ± 9.5 | |
| RSA | Homogeneous | 8.1 | −13.3 ± 9.5 | −6.7 ± 9.5 |
| Heterogeneous | 8.1 | −10.5 ± 9.3 | −5.2 ± 9.3 | |
| Speaker-dependent | 12.0 | −16.4 ± 9.1 | −9.2 ± 9.1 |
| Model . | Variant . | Likelihood . | WAIC . | PSIS-LOO . |
|---|---|---|---|---|
| A&A | Homogeneous | −28.1 | 57.7 ± 9.9 | 28.8 ± 9.9 |
| MAS | Homogeneous | 8.2 | −13.3 ± 9.6 | −6.6 ± 9.6 |
| Heterogeneous | 8.2 | −11.3 ± 9.5 | −5.6 ± 9.5 | |
| RSA | Homogeneous | 8.1 | −13.3 ± 9.5 | −6.7 ± 9.5 |
| Heterogeneous | 8.1 | −10.5 ± 9.3 | −5.2 ± 9.3 | |
| Speaker-dependent | 12.0 | −16.4 ± 9.1 | −9.2 ± 9.1 |