The Most Efficient Way to Develop a Method is…

Martin Perkins

9th May 2018

Bioscience, Biotech, Camilla Liscio, Design of Experiments, DoE, Newsletter,


Design of Experiments (DoE) is the most efficient way to develop analytical methods but has required a grasp of statistics that has put it beyond most of us.  Nowadays, anyone can use DoE, because software does the difficult work.

In analytical laboratories, the most common approach to method optimisation is to change one experimental variable at a time, while keeping the other variables fixed.  This is known as one-variable-at-a-time approach (OVAT).

This approach often leads analysts into exploring only part of the relevant experimental space, especially when multi-parameter processes are involved.

On the other hand, a multivariate approach (MVAT) changes multiple variables simultaneously and reveals the presence of interactions between variables.

This approach will provide information on a larger region of the experimental space.

To summarise, you get to:

  • Fully explore the problem
  • Uncover the best solution
  • Take the shortest possible route

Our lab is a busy one and we have a lot of work to get through.

My own experience of using DoE is very positive.  It has shortened my method development times by 50% or more and I love the certainty of knowing, that I have found the best possible combination of parameters.

The software package I use is called JMP and I am eternally grateful to Tom Howard, University of Newcastle, for teaching me how to use it.

If, like me, you do a lot of analytical method development, then my advice is to invest a little time to learn about DoE.  It will be a good use of your time!