Table 1.
Characteristic |
Class of analysis |
|
---|---|---|
Exploratory | Model-driven | |
Level of Parameterization | Low | High |
Reliance on a priori assumptions | Low | High |
Multiple events or population processesa | No | Yes |
Coverage of total ‘scenario space’ | Broad | Narrow |
Permits unexpected discoveries | Yes | Limited |
Accommodates stochasticity | Limited | Yes |
Statistical discrimination among alternatives | Limited | Yes |
Framework for comparisons across species | Qualitative | Quantitative |
NCPA is considered ‘exploratory’ here and is unique in its ability to separate multiple temporally overlying events and processes. Conversely, several model-driven methods explicitly consider temporally sequential events or processes (e.g., IM, simulations within population trees; Sections 4.3 and 4.4).