Dr Edwards' stimulating and provocative book advances the thesis that the appropriate axiomatic basis for inductive inference is not that of probability, with its addition axiom, but rather likelihood - the concept introduced by Fisher as a measure of relative support amongst different hypotheses. Starting from the simplest considerations and assuming no more than a modest acquaintance with probability theory, the author sets out to reconstruct nothing less than a consistent theory of statistical inference in science.
What people are saying - Write a review
We haven't found any reviews in the usual places.
The framework of inference
The concept of likelihood
Bayes Theorem and inverse probability
The Method of Support for several parameters
Expected information and the distribution of evaluates
Application in anomalous cases
adopt alternative hypothesis analogy applied approximation backcross Bayes Bayesian best-supported value binomial boys chance set-up chapter concept constant defined degrees of freedom estimation example expected information expected support function experimental fact fiducial argument figure Fisher formation matrix frequencies gene genotype given ignorance independent inverse probability Jeffreys justified known Laplace Likelihood Axiom likelihood curve likelihood function likelihood ratio linear log-likelihood log-odds m-unit support limits mating maximize Maximum Likelihood mean measure Method of Maximum Method of Support millimetres Newton-Raphson iteration Normal distribution nuisance parameter null hypothesis observed formation matrix observed information matrix obtained population possible posterior probability Principle of Indifference prior distribution prior probability distribution probability density probability model probability statement problem proportion question random relevance sample solution statistical inference statistical model sufficient evaluator sufficient statistics support curve support equations support surface theorem theory tion transformation trial value unbiassed uniform variable variance zero