Abstract
This dissertation answers several questions on sampling and estimation in information processing systems that are in series with k-steps. Two sampling schemes, sampling each step and sampling only output, are investigated. Criteria are developed to compare the two sampling plans. Sampling each step requires construction of optimal sample size allocation formulas. Optimal allocation means allocation of the sample to the processing steps in order to minimize variance of the estimators being used. Sample size formulas use a priori numerical information. Graphs are exhibited to show the effect of incorrect a priori numerical information on the variance of an estimator. Sample size procedures are compared with an existing procedure. Nonclassical estimation is also adapted to the two sampling plans.
Eakin, Mark Eathan (1980). An examination and comparison of two sampling procedures for estimating the reliability and amount of error in an information processing system. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -647417.