The author is the first to acknowledge the benefits of having as little complicated mathematics as possible between raw experimental results and conclusions, in terms of the ability of complicated calculations to conceal mistakes. Therefore, some justification seems to be required for the intervention of a sophisticated statistical algorithm. The justification will have two parts: the first will explain why any statistical inference process is needed at all, and the second will explain why a new Markov Chain Monte Carlo algorithm was coded, rather than using computer codes already in the public domain.