Statistical Intervals for Quality Assurance: The Realities! – Mini Paper Taken from the Summer 1994 Newsletter

Abstract: Statistical inference methods, such as interval estimates for quantifying uncertainty, provide important tools for continuous quality improvement. We discuss the assumptions associated with the use of these methods. We also consider how well these assumptions really meet the practical needs of the quality practitioner. We consider common statistical intervals such as confidence intervals, statistical tolerance intervals and prediction intervals. Especially in analytic studies, one is frequently using these intervals to aid in process improvement. In this case, inferences on the future behavior of the process rely considerably on the degree to which the sampled process represents the future process.

Keywords: Confidence intervals - Tolerance intervals - Prediction intervals - Standard deviation - Enumerative study - Analytic study - Random sampling - Convenience sampling - Judgment sampling - Representative sampling