Ignorance-based quantitative models and their practical implications
Published Web Locationhttps://doi.org/10.1177/0951692899011003008
Some basic models in science are not grounded in what we know but in careful specifying of the limits of our ignorance. Boundary conditions on the conceptually possible range determine whether a distribution is likely to be closer to normal or lognormal and whether arithmetic or geometric means should be used. If finite and non-zero boundary values can be defined, then most actual cases tend to be around the center of the possible range. This approach makes it possible to estimate the number of parties in a national assembly within a factor of two, using only assembly size and electoral district magnitude. Deviations from ignorance-based base line are a signal to look for specifically political causes. This approach may be of use in many aspects of political science.