Don’t Add Centre Points to Screening Designs when Designing Experiments

Don’t Add Centre Points to Screening Designs when Designing Experiments

Some experimenters believe that you must always include centre points when designing experiments. This is not the case, and in particular, it doesn’t make sense to include centre points when designing screening-type experiments, with large number of factors.

Centre points have an important role in testing for curvature in the appropriate circumstances, but will not provide an aid in the search for factors with significant main and/or interaction effects , which is the main objective when screening.

Centre points can be added to the design, if desired, when we have completed our screening work, and we are contemplating proceeding to optimisation using Response Surface Methodology (RSM).

I sometimes see people including a single centre point when designing screening experiments. This makes no sense. A single centre point is completely useless and is simply a wasted experimental run.

There is a view among some experimenters that adding centre points gives you a three-level design. This is not the case. Centre points are located at the centre of the design space and have no role in determining effects of the factors at the middle value of the factor levels.

Learn more about the correct use of centre points by attending our Design of Experiments courses. We present two Design of Experiments courses; a three day Design of Experiments on both a public and in-house basis and a six day Design of Experiments for R&D on an in-house basis.