1.7 A short history

The statistical design of experiments was created by R. A. Fisher at the Rothamsted agricultural research station in England in the 1920s. Faced with decades of field records from which little could be learned, he introduced randomisation, replication and blocking, the analysis of variance and factorial designs, and set them out in The Design of Experiments (1935). F. Yates extended the work to confounding and incomplete blocks. In the 1950s G. E. P. Box and K. B. Wilson brought the ideas into the chemical industry with response surface methodology, where experiments are run in sequence and the results are available quickly. From the 1980s the work of G. Taguchi on robust design, experiments that make a product insensitive to uncontrollable factors, spread the methods through manufacturing. The principles have not changed since Fisher; the applications now run from agriculture and medicine to engineering and computer simulation.

Questions on this section

Stuck on something here? Ask below and it stays attached to this topic.