A method of statistical inference proposed by Sir Ronald Fisher in 1930. Fisher's aim was to prescribe an entirely objective procedure that avoided prior assumptions or hypotheses. Fisher's approach is not easy to understand and continues to be the subject of discussion.