On a classification of the problems of statistical inference
成果类型：
Article
署名作者：
Deming, WE
刊物名称：
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
ISSN/ISSBN：
0162-1459
DOI：
10.2307/2279212
发表日期：
1942
页码：
173-185
关键词：
摘要：
Two types of problems confront the statistician in his job of making predictions in type A, action is based on a prediction regarding future measurements of a product already in existence; in type B, the product is not yet subject to measurement, perhaps not even produced yet, the evidence coming from other product originating from the same or similar processes. The fundamental problem of mathematical statistics (type A) is to set fixed limits within which percentages of other terms of a random sequence will fall. When the criterion for randomness is not met (type B) the interpretation of the results of the survey or expt. must lean heavily upon the statistician''s knowledge of the subject matter, and his ability to cooperate with experts. Attaining control of quality in industry is a type B problem. Similarly, frequent small sample surveys of population and agriculture would constitute an advance in social and economic planning.