The goal of any scientific experiment is to maximize information gain, while simultaneously minimizing the cost of the experiment. A well-designed experiment can help the researcher obtain this goal and increase the chance of obtaining significant results. This short course will discuss the three fundamental principles of experimental design (randomization, replication, and local control of error) and their relative importance in designing a useful experiment. To help illustrate these principles, we will work with the Paint Hardness data set, which can be found on the Minitab website. We will also discuss some commonly designed experiments, such as randomized complete block and split-plot designs, as well as describe some more advanced designs which may be of interest. Any and all who have little to no background in basic statistical techniques and are interested in acquiring a short introduction to experimental design are invited to attend this lecture session.
LISA Short Course: Designing Experiments from LISA on Vimeo.