|M.Sc Thesis||Department of Quality Assurance and Reliability|
|Supervisors:||Dr. Yefim Haim Michlin|
|Assoc. Prof. Ingman Dov|
|Full Thesis text|
The Sequential Probability Ratio Test (SPRT) ensures high authenticity at the smallest sample number (SN) and is the most common acceptance test in the field of reliability and quality control. Truncation compensates for the absence of a limit on the test duration, with the relevant parameters chosen on the basis of their influence on the supplier- and user risks (α and β error probabilities of I- and II- kind, respectively) and on the mean SN. The thesis deals with planning of an optimal truncated binomial SPRT with equal α and b, the acceptance test being a check of the hypothesis that the probability of nonconforming items does not exceed a specified value. One of the proposed measures of the test quality and an optimality criterion is the increase in the Average Sample Number caused by the test truncation. It facilitates solution of the problems in automatic planning of the test. Also given are a method and formulas for determining parameters of the test boundaries depending on the required test characteristics. A user's algorithm for the planner is also included.