一,二重混合检查法：大批中剔出缺陷个体的方法

提出了求无需混合的确界的方法,首次提出了比一重混合更节约检查的加倍混合(二重混合)检查方法。按上下界点制作了检察表,对多重混合量平均检查公式提出了一个猜测。

THE METHODS OF TEST FOR SIMPLE AND DOUBLE POOLED SAMPLES

A large population is subjected to a test (e. g. blood test) in order to find out the defective individuals. Each individual can be tested separately, or the test samples taken from k individuals are pooled and and analysed together. On the latter case, the sample is called a simple pooled sample. If the test result is negative, this one test suffices for the kindividuals. If the test result is positive, each of the k individuals must be tested separately. This new technique was proposed by R. Dorfman 1] and is cealled simple pooled sample test.A double pooled sample (or 2-fold pooled sample) is a sample pooled from two simple pooled samples, eachis consisting of kindividuals.Each double sample is analysed together. If the test result is negative, this one test is sufficient. If the test result is positive, the double pooled sample is separated into two constituting simple poole samples and each one is subjected to simple pooled sample test. This procedure is called a double pooled sample test (or 2-fold pooled sample test). N-fold pooled samples and n-fold pooled sample tests are defined similarly.On this paper, it is proved that the double pooled sample test is more saving than simple pooled sample test. A sampling table is given. this table consists of three columns: the first column is titled pooled amount (2xk). the second one is the upper bound and lower bound for p, p being the percentage of defectives (or percentage of positsves members). the last column is the lower bound for the saved amount of inspection. Take a row (2xk, p1-p2, r) as a example, if a p is given, P1, it is best to take a 2xk double pooled test and the inspection amount saved is at least r finally, we also give a conjecture for the formula of expected inspection number of n-fold pooled sample test.