User manual SAS JMP QUALITY AND RELIABILITY METHODS

[. . . ] Quality and Reliability Methods Get the Most from JMP® Whether you are a first-time or a long-time user, there is always something to learn about JMP. Visit JMP. com and to find the following: · · · · · · live and recorded Webcasts about how to get started with JMP video demos and Webcasts of new features and advanced techniques schedules for seminars being held in your area success stories showing how others use JMP a blog with tips, tricks, and stories from JMP staff a forum to discuss JMP with other users http://www. jmp. com/getstarted/ Release 9 Quality and Reliability Methods "The real voyage of discovery consists not in seeking new landscapes, but in having new eyes. " Marcel Proust JMP, A Business Unit of SAS SAS Campus Drive Cary, NC 27513 The correct bibliographic citation for this manual is as follows: SAS Institute Inc. JMP® 9 Quality and Reliability Methods. JMP® 9 Quality and Reliability Methods Copyright © 2010, SAS Institute Inc. , Cary, NC, USA ISBN 978-1-60764-594-8 All rights reserved. [. . . ] The logNormal column in the table has the formula: Chapter 6 Reliability and Survival Analysis II Nonlinear Parametric Survival Models 157 The lognormal loss function can be very sensitive to starting values for its parameters. Because the lognormal is similar to the normal distribution, you can create a new variable that is the log10 of Time and use Distribution to find the mean and standard deviation of this column. Then, use those values as starting values for the Nonlinear platform. In this example the mean of log10 of Time is 2. 05 and the standard deviation is 0. 15. Run this example as described in the previous examples. In the Nonlinear Fit Control Panel give Mu and Sigma the starting values 2. 05 and 0. 15 and click Go. After the Solution is found, you can click Confidence Limits, on the Control Panel and see the table shown here. Figure 6. 26 Solution Report Note: Remember to Save Estimates before requesting confidence limits. The maximum likelihood estimates of the lognormal parameters are 2. 2223 for Mu and 0. 3064 for Sigma (in base 10 logs). The corresponding estimate of the median of the lognormal distribution is the antilog of 2. 2223 (102. 2223), which is approximately 167. This represents the typical life for a locomotive engine. 158 Reliability and Survival Analysis II Nonlinear Parametric Survival Models Chapter 6 Chapter 7 Variability Charts Variability Chart and Gauge R&R Analysis A variability chart plots the mean for each level of a second factor, with all plots side by side. Along with the data, you can view the mean, range, and standard deviation of the data in each category, seeing how they change across the categories. The analysis options assume that the primary interest is how the mean and variance change across the categories. · · · A traditional name for this chart is a multivar chart, but because that name is not well known, we use the more generic term variability chart. A variability chart shows data side-by-side like the Oneway platform, but it has been generalized to handle more than one grouping column. A common use of a variability chart is for measurement systems analysis such as gauge R&R, which analyzes how much of the variability is due to operator variation (reproducibility) and measurement variation (repeatability). Gauge R&R is available for many combinations of crossed and nested models, regardless of whether the model is balanced. Just as a control chart shows variation across time in a process, a variability chart shows the same kind of variation across categories such as parts, operators, repetitions, and instruments. The Variability Chart platform can compute variance components. Several models of crossed and nested factors of purely random models are available. Attribute (multi-level) data can also be analyzed with this platform. · · · Figure 7. 1 Example of a Variability Chart Contents Variability Charts . 184 Chapter 7 Variability Charts Variability Charts 161 Variability Charts A variability chart is built to study how a measurement varies across categories. Along with the data, you can view the mean, range, and standard deviation of the data in each category. [. . . ] (2000), "Change-point detection of mean vector or covariance matrix shifts using multivariate individual observations, " IIE Transactions, 32, 537-549. Tan, Charles Y. , and Iglewicz, Boris (1999), "Measurement-methods Comparisons and Linear Statistical Relationship, " Technometrics, 41:3, 192­201. (1976), "An Introduction to Quality Control, " Nagoya, Japan: Central Japan Quality Control Association. (1958), "Estimation of Relationships for Limited Dependent Variables, " Econometrica, 26 24­36. [. . . ]