Table 2. Summary of the Wilcoxon signed-ranks and Shimodaira-Hasegawa tests (SH) test results for topologies generated under the different search methods.
| Method | Templeton | SH | ||
|---|---|---|---|---|
| Length | P | -LogL | P | |
| Mu-UMP | 445 | 0.9857–1.0000 | 2336.973–2353.912 | 0.5070–0.9890 |
| Co-UMP | 446–447 | 0.6374–1.0000 | 2343.877–2352.36 | 0.5730–0.8800 |
| IWMP | 449 | 0.4142 | 2367.288 | 0.2370 |
| ML | 447 | 0.7327 | 2334.732 | 1.0000 |
| Bayesian inference | ||||
| Exp+GA | 521 | <0.0001* | 2441.756 | 0.0000* |
| Exp+LN | 511 | <0.0001* | 2425.579 | 0.0000* |
| Uni+GA | 525 | <0.0001* | 2448.811 | 0.0000* |
| Uni+LN | 505 | <0.0001* | 2416.245 | 0.0000* |
Results are summarized for all trees obtained with each method and presented as a range of values (for the Bayesian analyses, results are shown for each different combination of parameters). Abbreviations: Co-UMP, unweighted parsimony analysis with contingent character coding; IWMP, maximum parsimony under implied weighting; ML–maximum likelihood analysis; Mu-UMP, traditional unweighted parsimony analysis with multistate character coding.
* Statistically significant values indicating tree topologies that are not equally well-explained by the data and significantly differ from the other analysed topologies.