Multiple Sampling with Constant Probability

Idle thoughts of a ‘well‐calibrated’ Bayesian in clinical drug development

“…Predating Wald's work, the earliest sequential test procedure in which, in contrast to statistical tests developed in agriculture, the number of observations is not fixed in advance -goes as far back as Dodge and Romig [52] who developed a double sampling procedure. The advantage of these schemes, recognised by Dodge and Romig and by Bartky [53], is that on average, they require fewer observations than traditional single sampling schemes. The development of similar, sequential ideas in medicine in the 1950s is due to Peter Armitage in the UK and Irwin Bross in the US [54][55][56][57][58].…”

A Destructive Sampling Method Designed for Outsourcing Situations Involving High Quality Production Processes

“…An extension of double sampling is "multiple sampling, which involves taking a first sample and making a decision based on the evidence of the first sample (Bartky, 1943). If the sample is not good or bad enough, then just like in double sampling, a second sample is taken and combined with the first sample in order to attempt to make a decision.…”

Statistics in Medical Research

“…This is more efficient than the traditional single-sample procedure in reducing the number of items to be inspected (Ghosh, 1988). Walter Bartky (1943) extended the scheme to include more than two steps and Harold Hotelling (1941) mentioned chain or large-scale experiments designed to include successive stages. Abraham Wald (1947) regarded this latter idea as preliminary to sequential analysis.…”