Featured Video
This Week in Quality Digest Live
Statistics Features
Donald J. Wheeler
The key to using any number is to understand its limitations
Mike Richman
Understanding how to estimate requirements and specification limits
Bonnie Stone
Five critical lean tools
Eston Martz
They help quantify probabilities and determine if results are statistically significant
Scott A. Hindle
Entropy will make each and every process unpredictable

More Features

Statistics News
Ask questions, exchange ideas and best practices, share product tips, discuss challenges in quality improvement initiatives
Strategic investment positions EtQ to accelerate innovation efforts and growth strategy
Satisfaction with federal government reaches a four-year high after three years of decline
TVs and video players lead the pack, with internet services at the bottom
Using big data to identify where improvements will have the greatest impact
Includes all the tools to comply with quality standards and reduce variability
A free, systematic comparison of upcoming changes to the ISO 9001:2008 standard

More News

Mike Richman

Statistics

Process Capability Analysis

Understanding how to estimate requirements and specification limits

Published: Monday, December 4, 2017 - 12:01

Sustainable performance improvement is simply impossible without a firm handle on the precepts and tools of statistical process control (SPC). It is for this reason that we cover industrial statistics so frequently here at Quality Digest. After all, as the great Scottish physicist and engineer Lord Kelvin once said, “If you cannot measure something, you cannot improve it.”

With this in mind, I welcome the release of Process Capability Analysis: Estimating Quality (CRC Press/Taylor & Francis, 2018), the forthcoming book from Neil Polhemus. Although not necessarily written with beginners in mind, Polhemus’ work is nevertheless accessible to rank-and-file QA/QC professionals with an abiding interest in the inner workings of process improvement. I myself have no more than a journalist’s basic understanding of SPC, yet found this book’s central premise—i.e., to address “the problem of estimating the probability of nonconformities in a process from the ground up”—to be valid and valuable.

It’s important to note (and in fact Polhemus points this out more than once) that a consideration of process capability is not exclusive to the world of manufacturing. In fact, the statistical tools and techniques described in this book can apply equally well to the capability of a process to produce the proper tolerance or diameter of a given widget, or the optimal provisioning and delivery of a given service.

Replete with dozens of charts, graphs, and formulas, Process Capability Analysis offers guidance on many if not most of the problems that can be solved via industrial statistics. Polhemus’ tone is authoritative and insightful, providing detail on not only how such analyses can be performed, but why as well. With so many statistical tools now available to students of Six Sigma, for example, this sense of discernment over which analysis to select to address a specific problem is more important than ever. Not every problem is a nail; even if it were, not every solution is the appropriate type or size of hammer.

On a practical level, this book addresses many of the questions that one may have when beginning to consider whether one’s process is in control. Chapters cover distinguishing between variable data and attribute data, the statistical unpinning of Six Sigma, the differences between Cp and Cpk (as well as Pp and Ppk), considerations of normal vs. non-normal data distributions (including whether to transform non-normal data), and the role of control charts.

All of these topics, and many more, are held up to the light and examined with perfect logic, and even an experienced reader will learn much about tools from this book. However, the real value of Process Capability Analysis comes from the strategic, big-picture approach that it takes to quality improvement. Using the tools only gets one so far if there’s not an understanding of why one needs to understand capability. Data, properly understood and put into context, has much to teach the quality professional who is willing to study and apply those lessons back to occasionally cranky systems that must be closely monitored if they are to remain in control.

The science of statistical analysis is a core competency of quality, and anyone wishing to do it properly would be well-served by the teachings found in this book. It’s a good overview of the topic, with a nice combination of a high-level approach along with real-world examples. The result is completely relatable to pretty much anyone who wants to understand and control their processes, and thereby reliably sustain long-term and continuous improvement for their organization.

Discuss

About The Author

Mike Richman’s picture

Mike Richman

Mike Richman is Quality Digest’s publisher.

Comments

I would definitely recommend this book

I use StatGraphics frequently, and it has the ability to estimate process performance indices not only for the normal distributions that are far more common in textbooks than in real-world applications (including non-manufacturing such as waiting times for services), but also for non-normal distributions. StatGraphics also performs the goodness of fit tests that should be performed before reporting a process performance index or, for that matter, deploying a control chart on the basis of the assumed statistical distribution. The software can also generate control charts for non-normal distributions, e.g. by setting the control limits at the 0.00135 and 0.99865 quantiles (same as for a 3-sigma Shewhart chart) of the actual distribution.

The book is supposed to become available on Dec. 5 and I would definitely recommend it.