Table 2.

Ability of the applied approximate Bayesian computing (ABC) methods to discriminate among the four hypothetical scenarios, measured as rates of Type I and Type II errors. The latter are shown for all alternative scenarios combined (mean), as well as separately for each suboptimal scenario

Error	Method
	Direct estimate	Logistic approach
Type I	0.144	0.182
Type II (mean)	0.116	0.055
Type II (Scenario NCB–NRB)	0.010	0.006
Type II (Scenario CB–NRB)	0.296	0.134
Type II (Scenario NCB–RB)	0.042	0.026

NCB–NRB, No Colonization Bottleneck – No Recent Bottleneck.