Response surface basics – Designs and models for response surface modelling
Robert Collis (Minitab)
Many of the industrial organisations require running experiments in order to optimize (or maximize or minimize) key process output parameters, for example, maximizing the yield in a chemical process. In this case, this will always involve creating and running a response surface design. This presentation will give a practical overview of the main designs that are available, that is Central Composite Designs (Circumscribed, inscribed and face-centred designs) and Box Behnken designs. The advantages and disadvantages of each design will be addressed and a discussion will be provided about the approach that needs to be used to fit a second order polynomial model given the alias patterns that exist between certain terms when constructing such designs.
When the model doesn’t fit - Alternative models and simulation. Steve Ellison (LGC)
Response surface modelling is a valuable tool for exploring the effect of different factors on a process, and is perhaps the standard tool for optimisation using classical experimental design approaches. However, optimisation relies on a relatively smooth and well-behaved response, and software implementations usually assume that the surface can be modelled by a simple polynomial. This presentation will give some examples from optimisation of chemical measurement methods which illustrate some practical problems in fitting surface models, and will show how alternative models and/or simulation can assist in locating optimal settings.
Response Surface Models for Six Sigma – Radouane Oudrhiri
Abstract to follow
Robert Collis (Minitab)
Minitab is a world renowned statistical software application with its HQ in State College, Pennsylvania in the U.S. Robert is Minitab’s most experienced trainer in Europe, bi-lingual in French and English, and delivers Minitab training and consultancy in both languages. He has trained and consulted for companies from diverse sectors such as pharmaceuticals to medical devices to banking for audiences as diverse as R&D and Quality specialists in industry to business professionals in the service sector and all of this not only Europe but also in the Middle East, Africa and the South Pacific.
Robert helps companies who want to improve the statistical culture within their organization to meet industry or quality standards, including those implementing the Six Sigma quality initiative. His cross-sector experience gives him great insights into many different questions and problems including the implementation of Design of Experiments and the analysis of experimental results. Robert, though originally from the UK, is based in France and is a qualified statistician with a degree from University College London.
Steve Ellison (National Measurement Laboratory at LGC, Teddington)
Dr Ellison is a Science Fellow at LGC, Teddington, the UK National Measurement Laboratory for chemical and biological measurement. An RSS Fellow and member of the RSS Quality Improvement Section, his principal interests are in applications of statistics applied to analytical chemistry and measurement science, including experimental design for analytical method development, interlaboratory studies of method performance, proficiency testing, and reference material certification. He contributes to a range of IUPAC, ISO, CEN, BSI and other committees involving applications of statistics, and has contributed to several International Standards and guidelines, co-authored a textbook of statistical methods for analytical chemists, and provides training in statistical methods for measurement science.
Member - free to attend
Non member - £10 entry fee