COST-UTILITY AS A MEASURE OF THE EFFICIENCY OF A TEST
成果类型：
Article
署名作者：
BERKSON, J
刊物名称：
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
ISSN/ISSBN：
0162-1459
DOI：
10.2307/2280654
发表日期：
1947
页码：
246-255
关键词：
摘要：
Testing procedures for separating populations into 2 mutually exclu-sive categories must be judged in terms of 2 numbers rather than a single index. A test has utility or sensitivity in proportion to its efficiency for correctly selecting from the population all members of category A. The cost or specificity of a test is proportional to its yield of false-positive results; non-A individuals classified as A. For normally distributed test scores a mean-cost-rating index (M.C.R.) may be derived, thus avoiding theoretical objections inherent in applications of biserial or tetrachoric correlation to such problems.