TILS655 Design of Experiment (5 cr)

In the course, the students will familiarize themselves with the general principles of designing scientific experiments, which are the same in all fields of science. A poorly designed experiment wastes resources and may not succeed in producing new knowledge at all. First, the course will cover random variation, randomization, and sample size calculations through examples. After this, factorial setups for linear models are discussed and optimality criteria based on the information matrix, especially D-optimality, are presented. Finally, adaptive experimental designs and experimental design for nonlinear models are discussed.

A student

• knows the general principles of experimental design,
• knows the most commonly used optimality criteria,
• can determine the optimal experimental design with the help of software,
• can apply the general principles of experimental design in practical situations.

The course is held every two years or less frequently.

Statistical inference 1 & 2, Generalized linear models 1 & 2, Linea algebra and geometry 1 & 2, Bayesian statistics 1

Lecture notes