Statistical Methods have been in widespread use for many years.
Their purpose is:

• to identify which process variables are critical
• to quantify the effect the critical process variables have on the process output
• to optimise the process output through controlling the critical process variables
• to allow the process to be managed and improved on an ongoing basis

Equally applicable to transactional and manufacturing processes, the courses listed below are taught using a fun, hands-on approach, using real customer data and examples where possible. Participants learn where each of the tools fits within a continuous improvement programme, how to apply the tools and how to use statistical software which ensures the focus is on improving the process rather than on detailed statistical theory. Implementation can also be approached as a hands-on project or a combination of classroom and project-based learning.  

Course List

• Process Measurement and Graphical Analysis
• Measurement Systems Analysis (Gauge R & R) - Manufacturing
• Statistical Process Control and Process Capability
• Design of Experiments
• Response Surface Methodology
• Hypothesis Testing
• Regression Analysis
• Data Collection and Measurement Systems Analysis for Transactional Processes

Process Measurement and Graphical Analysis

Although knowledge and experience are vital to management of the process and making improvements, continued progress can be achieved only by making and analysing measurements. This course uses a process simulation game where the participants learn to decide what to measure, how to measure it and then how to analyse and interpret the data using graphical tools. They then present their recommendations to “Management”.

Topics include: Identifying and Selecting Process Measures, Types of Data, Process Sampling, Data Collection Plan, Histograms, Scatter Plots, Box Plots, Time Series Plots, Basic Statistics.
(return to top)


Measurement Systems Analysis (Gauge R&R) - Manufacturing

In order to be able to control and improve a process, it is essential to obtain confidence in the measurement system.

This highly-practical, one-day course provides participants with a solid grounding in Measurement Systems Analysis (MSA) techniques. The morning session covers the theory – how to recognise when an MSA study is required, planning the study, and how to analyse and interpret the results using Graphical Techniques. The afternoon session gives the participants the opportunity to practice these new techniques on their own measurement systems and to consider what actions might be appropriate to improve incapable systems.

Advanced training can also be provided, to extend the technique to Attribute and Destructive Gauge R&R Studies. Application to complex
(return to top)


Statistical Process Control and Process Capability

The SPC Chart was developed to give managers the information they need to make appropriate day to day process decisions and bring about improvement over time. This course explains the basis of SPC Charts, how to go about implementing them and how to interpret and react to them. It also introduces Process Capability and the Cp/CpK indices. A game is used to illustrate the concepts of SPC and generate data which the participants use to set up and interpret an SPC chart. We encourage clients to provide some of their own process data for discussion in the class. 

Topics covered – Understanding Common and Special Cause Variation, Data Collection, Chart Selection and Construction, Interpreting Patterns in Control Charts, Out-of-Control Action Plans, Introduction to Process Capability and calculation and interpretation of common indices (Cp, Cpk, Pp, Ppk), Process Improvement using SPC and Process Capability. Chart types covered are Individuals Charts, Xbar-R, p, np, c, u, EWMA..
(return to top)

Design of Experiments (DOE)

Often the root causes of problems can be exposed using quite simple tools such as process maps and graphing data. On other occasions, however, the causes remain hidden. Where this is the case, Design of Experiments is the technique used to force them into the open. A common example of its use is to establish which machine settings are critical to control product quality but it is increasingly being used in more imaginative, non-industrial applications too – what has the biggest effect on number of data entry errors - operator gender, age, experience or grade, time of day, style of keyboard etc? 

The course covers: when to use designed experiments, identifying the right experimental design, planning and running the experiment, using graphical and statistical tools to understand the results. Full Factorial, Fractional Factorial and Screening Designs are included.

Again the teaching approach uses class exercises where the participants plan, run and analyse their own experiments. Software is used to assist in the design and analysis.
(return to top)

Response Surface Methodology (RSM)

Design of Experiments establishes which factors are most important in controlling process output. Response Surface Methodology goes on to determine the optimum settings for the factors.  This is an advanced topic suited to participants who have some DOE experience.
(return to top)


Hypothesis Testing
When evaluating new processes, it is important to determine whether apparent differences between the old and the new are truly present or simply due to random variation in the process. Reaching the wrong conclusion can lead to unnecessary costs or missed opportunities. This course teaches the use of the most commonly used hypothesis tests to determine whether differences seen are statistically significant or not. Statistical software is used which relieves the need for deep theoretical understanding and manual calculations but allows the focus to be placed on interpretation.

Topics include: 1 and 2 sample t-tests, ANOVA, Chi-Squared test, Comparing Variances, Assumptions for tests, sample size calculation, Type I and II errors.
(return to top)


Regression Analysis
Problem solving and process improvements often require an understanding of whether a relationship exists between 2 or more variables. Participants will learn how to use regression analysis techniques to determine whether these relationships are present and produce a prediction model which can be used to control and improve the process. Statistical software is used which relieves the need for deep theoretical understanding and manual calculations but allows the focus to be placed on interpretation.

Topics include: Simple Linear Regression, Multiple Regression, Curvilinear Regression
(return to top)

Data Collection and Measurement Systems Analysis for Transactional Processes

Operators in transactional processes continually have to make decisions – but are those decisions always right? If the same purchase order was entered into the system a number of times, would the same operator always assign the same budget code? Would another operator assign the same code? Are they both always right? Wrong decisions can have a dramatic impact on productivity, customer satisfaction and cost.

This course will teach participants how to collect good data and use simple Agreement Analysis techniques to improve the quality of decisions made in service processes. 
Topics include:  Operational Definitions, Process Sampling, Data Collection, Attribute Agreement Analysis
(return to top)