Industrial Statistics with Minitab

Pere Grima Cintas, Lluis Marco Almagro, Xavier Tort-Martorell Llabres

• Publisher: Wiley
• Year: 2012
• Print ISBN: 9780470972755
• eBook ISBN: 9781118383780
• Edition: 1
• Topic: Mathematics
• Subtopic: Probability & Statistics
• Index: Mathematics

Industrial Statistics with MINITAB demonstrates the use of MINITAB as a tool for performing statistical analysis in an industrial context. This book covers introductory industrial statistics, exploring the most commonly used techniques alongside those that serve to give an overview of more complex issues. A plethora of examples in MINITAB are featured along with case studies for each of the statistical techniques presented.

Industrial Statistics with MINITAB:

• Provides comprehensive coverage of user-friendly practical guidance to the essential statistical methods applied in industry.
• Explores statistical techniques and how they can be used effectively with the help of MINITAB 16.
• Contains extensive illustrative examples and case studies throughout and assumes no previous statistical knowledge.
• Emphasises data graphics and visualization, and the most used industrial statistical tools, such as Statistical Process Control and Design of Experiments.
• Is supported by an accompanying website featuring case studies and the corresponding datasets.

Six Sigma Green Belts and Black Belts will find explanations and examples of the most relevant techniques in DMAIC projects. The book can also be used as quick reference enabling the reader to be confident enough to explore other MINITAB capabilities.

Part One

INTRODUCTION AND GRAPHICAL TECHNIQUES

As much as 95% of all problems within a company can be solved using the seven basic tools.

K. Ishikawa

A good way to start learning how to use statistical software is to give an overview: see what it looks like, consider the different possibilities offered, enter some data, develop an idea on how to handle it and start doing some simple analysis. This is what we do with Minitab in Chapter 1.

The remaining chapters of Part 1 are devoted to graphical techniques. One might think that this is an easy subject, but when we are faced with some data – usually in a format and a structure that are not best suited for our purposes – it is not always easy to decide upon the best graph (or graphics) to convey the desired information. To make a good, informative graphic requires patience: producing different types and different versions of the most promising ones, subject knowledge, and mastering the software to make them. The examples presented here are aimed at presenting and practising Minitab using its many graphical capabilities, while developing practice in identifying the graphic type most suited to different situations.

The graphs presented are organized into three groups: to analyze variability, to split the data into categories to compare frequencies and to show the relationship between variables.

To analyze the variability of a data set we can use:

• Histograms. This is the graph of choice for visualizing dispersion. It is very useful to compare process output with product tolerances. And also very useful to stratify data – by machine, shift, operator, etc. – and compare it.
• Dotplots. This is similar to histograms and is especially recommended when you have few data. They are also very useful for comparing different situations by placing one diagram above another.
• Boxplots. These are very powerful and useful graphs. They have a box-shape showing the measures of position (quartiles), in addition they highlight the outliers. These plots are very intuitive and specially suited to compare a large number of groups.

To split the data into categories and show frequencies, there are:

• Bar Charts. These are similar to histograms but with a qualitative variable (gender, process, zone, etc.) or discrete (number of defects, age, etc.) on the horizontal axis instead of a continuous variable (weight, moisture, density, etc.). The bars can also be stratified.
• Pie Charts. These are widely used in commercial environments to show the size of the different parts of a whole. In general, their use is not recommended in technical environments.
• Pareto Diagram. This diagram is one of the most useful and widely used charts in quality management. They are a special case of bar charts and serve to show the parts of a problem, hoping to find the so-called Pareto principle: the vital few and the trivial many. As available resources are not usually sufficient to address all the causes at the same time, efforts should start and concentrate on the most important ones.

Cause-and-effect diagrams are a special case. Although they are not ...

Table of contents

• Cover
• Title Page
• Copyright
• Preface
• Part One: Introduction And Graphical Techniques
• Part Two: Hypothesis Testing. Comparison Of Treatments
• Part Three: Measurement Systems Studies And Capability Studies
• Part Four: Multi-Vari Charts And Statistical Process Control
• Part Five: Regression And Multivariate Analysis
• Part Six: Experimental Design And Reliability
• Appendices
• Index