Varios

Advanced Analyses Glossary
This document contains glossary information for the advanced analyses in STATGRAPHICS Plus Quality and Design, and STATGRAPHICS Plus Professional. To facilitate searching for terms while working with a specific procedure, the glossary entries are grouped by analysis in the order they are listed on the Special menu: Analysis Quality Control Design of ExperimentsTime-Series Multivariate Methods Advanced Regression Page 2 9 17 22 28

Introduction

1

Quality Control
Acceptable Quality Level (AQL) The poorest level of quality which the consumer finds acceptable on average. Acceptance Charts A charting method for instances where the specification limits for the process are wide compared to the process performance. Rather than placing the control limits at +3 sigma, the control limits are placed far enough within the specification limits to ensure that a signal is generated if the process is going out of control, but otherwise allows the process to vary around the long-term process mean. Acceptance Sampling Acceptance Sampling is a procedure for determining sample sizes and decision rules for incoming or outgoing product inspection. Alpha Risk Theprobability of a Type I error; determining the probability of declaring a process out-of-control when, in fact, it is not. ARIMA Chart An Autoregressive Individuals Moving Average (ARIMA) Chart is a modification of standard control charts to account for the serial correlation between consecutive periods with the objective of separating long-term changes from short-term dependencies. Attribute Aninherent property usually measured by a finite number or a discrete set of values. Attributes control chart One of four control charts (p, np, c, or u) that allows you to construct control limits based on a particular attribute of a sample. Average A measure of the central tendency of the data. Also known as the mean. Average Outgoing Quality (AOQ) Curve This shows the average percent of defectiveitems shipped as a function of the proportion of nonconforming items in the lot when subjected to the current sampling plan. Average Outgoing Quality Limit (AOQL) Based on the assumption that rejected lots are 100% inspected and that any non-conforming items are replaced by good items, this is the maximum fraction of non-conforming items which are accepted. Beta Risk The probability of a Type IIerror; determining the probability of declaring a process in control when it is not. c Chart An attributes control chart, based on the Poisson distribution, that allows you to construct control limits based on an attribute related to frequency, such as the number of defects in an item. Capability Indices Statistics that indicate the inherent capability of a process; for example, Cp and Cpk.Quality Control

2

Chi-Square Statistic A statistic that lets you compare observed to expected frequencies. If none of the expected frequencies is too small, the statistic follows a chi-square distribution. A large chi-square statistic leads to a conclusion that two distributions are different. A small chi-square statistic leads to a conclusion that two distributions are similar. Also known as aPortmanteau statistic. Confidence Interval An interval that indicates the bounds between which it is likely that the real parameter lies. Confidence Limits The end points of a confidence interval. Common limits are 90, 95, and 99. Control Chart A graph of results from sampling the product of a manufacturing process; consists of a horizontal line that represents the expected mean of somecharacteristic of quality and two lines on either side that represent the extent you are allowed to sample random product deviations. Controlled Process A process for which a control chart shows no points outside the limits and no nonrandom variation within the limits. Control Limits On a control chart, the lines above and below the center horizontal line that designate the area into which a stated...