The epistemology of statistical science
CITATION: Van Aarde, M. 2009. The Epistemology of Statistical Science. Stellenbosch: SUN PReSS. doi:10.18820/9781920338336.
The original publication is available from AFRICAN SUNMeDIA - www.sun-e-shop.co.za
BOOK BLURB: In this book four different schools of inference, namely, frequentist, Bayesian, likelihood and fiducial are introduced and then with appropriate examples and discussions, the gaps and holes in the traditional sense of inferences are exposed. The author goes back deep into the literature, exploring views and fallacies introduced by outstanding scientists and statisticians. The author proposes a coordination test of any model posted for a given (observed) data set. A coordination test is a calculable ordered number triplet, whose three members are statistical coordinates. Various appropriate examples are given in each chapter where the traditional analysis and inference are applied, and then the statistical coordinates are calculated for each example. After each example the shortfalls, gaps and holes in the traditional methods are discussed.