Scientific models establish their validity by working over a wide range of conditions. A model that generates only a few usable values is generally regarded as useless. Here is a simple example of model testing using a model actually advocated by creation scientists! (Index to Creationist Claims: CB620) This model has the advantage that any high-school graduate should have the mathematical skills necessary to do it. This problem is simple enough to do in a spreadsheet program.The Mathematical Details
Consider this simple creationist global population model. P0 is the starting population at the time, t0. The population grows at a constant fractional growth rate, r, defined as the fractional growth per unit time. The population, P, at some later time, t, is then given by
P = P0 (1 + r)(t - t0)
Note that this is the same equation used in calculating earnings in interest-earning bank accounts, where P and P0 are replaced by dollar amounts.
Start in the year of the (claimed) Global Flood (2349 B.C.), t0=-2349. Assign a starting population of eight people, so P0 = 8. Assign a growth rate of 0.5%, so r=0.005. With these parameters and the simple population growth model we get a good match for the t = 2008 global population of about P = 6 billion. (I actually get a better match with 0.47% growth rate but that's a detail).
With only two data points, this equation is guaranteed a solution that matches at two points. From a scientific perspective, the important question in testing a model is if it works for times between these two points (interpolation).Questions
Now for the advanced questions...
I find it interesting that every creationist I've asked to work this problem has tried to evade it. Did they work it and not mention it out of some concern, or were they afraid to even work it? As of this writing (August 10, 2008), this problem is not on AIGs list of Arguments we think creationists should NOT use, but it should be added soon.