. 2017 Jul 11;34(11):2982–2995. doi: 10.1093/molbev/msx195

Table 1.

Precision (measured by the root mean squared deviation), Bias, and Coverage (% of simulations in which the true value is found in the 95% highest posterior density intervals) of Parameter Estimates When Fitting Models to Either Epidemiological, Phylogenetic, or Both Types of Data.

Data	$R_{0}$ : RMSD	$R_{0}$ : Bias	$R_{0}$ : in HPD	$k$ : RMSD	$k$ : Bias	$k$ : in HPD
Both	0.1507	−0.0756	100%	15.0737	0.0103	75%
Epi	0.1695	−0.0387	100%	673.9374	93.4255	100%
Phy	0.1863	−0.0855	100%	31.7680	0.5461	100%
Data	$T_{0}$ : RMSD	$T_{0}$ : Bias	$T_{0}$ : in HPD	$ρ$ : RMSD	$ρ$ : Bias	$ρ$ : in HPD
Both	11.0638	0.1813	100%	0.1183	−0.0766	75.00%
Epi	11.4962	0.3933	100%	0.1267	−0.1002	83.33%
Phy	11.9352	−1.2768	100%	NA	NA	NA

Note.—These statistics were evaluated across all simulations presented in figure 1. The estimates of epidemic start dates $T_{0}$ were converted to the number of days after an arbitrary date. For bias and precision, we normalized the statistics by the true parameter value.