Design of Experiments - (L16) R.A. Bailey



Design of Experiments - (L16) R.A. Bailey

This course is about how to design real experiments, and includes issues about statistical consultancy as well as the necessary mathematics. It includes enough about the analysis of data from an experiment to show what we need to think about when designing the experiment. The following topics will be covered

  • The problem of deciding how to allocate treatment to experimental units.
  • Bias, variance, blocking and randomization.
  • Linear model and analysis of variance.
  • Factorial treatments: main effects and interactions.
  • Complete-block designs, row-column designs, split-plot designs, false replication.
  • General theory of orthogonal designs, including Hasse diagrams for factors, null analysis of variance and skeleton analysis of variance.
  • Incomplete-block designs.
  • Fractional factorial designs.

Pre-requisite Mathematics
Introductory statistics, including estimation, bias and variance. Introductory probability, including the normal, χ2, t- and F-distributions. Introductory linear algebra over the real numbers, including the eigespaces of real symmetric matrices and orthogonal projection onto subspaces. Arithmetic in the integers modulo n.

Literature
Main Text: Design of Comparative Experiments by R. A. Bailey, CUP, 2008.
Other texts: Planning of Experiments by D. R. Cox, Wiley, 1958; Design and Analysis of Experiments by George W. Cobb, Springer, 1998; Experimental Designs by W. G. Cochran and G. M. Cox, Wiley, 1957.

 

 

 

 


Back to top