Experimental Design for Process Development and Improvement (R&D)

What you will gain on this course

The course aim is two-fold: firstly to show the benefits of experimental design both in efficiency and clarity of outcomes. You will learn how to design the appropriate experiment to meet your objectives and then how to interpret the statistical and graphical outputs using appropriate software of your choice

Pre-requisite
Knowledge of basic statistical methods (t-test, ANOVA, regression and correlation) will be assumed.

Course Summary
The aim of the course is to introduce attendees to the benefits of statistical experimental design and to use statistical software to both design and analyse experiments. Following the course, attendees will be better able to understand and interpret the graphical and statistical outputs as well as develop a programme of experimentation from screening key variables to process optimism. The course can  include a hands on practical using the ‘Statapult’ to get participants fully engaged with how design issues can affect results.

Software
Access to a general statistics package is required to take part in the course workshops and have a go yourselves at analysing some data to fully understand how to apply the course content. We recommend Minitab but can also advise on the appropriateness of other software packages such as Design Expert, XLSTAT, JMP^® etc. Usually free demo versions can be used for the training if the software is not already available, so no purchases are necessary.

Flexibility
We can customise the course content to meet specific requirements. This course can also be offered as a series of 6 webinar modules, each of 2 hours in length.

Course content

• Statistical modelling refresher
• Factorial experiments – benefits, design issues, interpretation of effects and importance of understanding interactions

• Screening design – to deal with many factors, fractional factorials benefits and dangers
• Optimisation experiments – Experimental designs for process optimization: Box-Behnken, central composite and other options
• Optimisation, identification of viable operating regions when there are many output parameters
• Mixture design.