Lesson 7: Experimental Design

Printer-friendly versionPrinter-friendly version
Key Learning Goals for this Lesson:
  • Understand the importance that experimental design has on how we are able to interpret results.
  • Be able to articulate how the treatments are being applied within an experiment to the experimental units.
  • Understand the critical role that randomization plays within an experiment and why this is important as we interpret results.
  • Understanding the role that blocking factors have in restricting randomization within experimental design.

Introduction to Experimental Design

(See Text Section 15.2)

Previously in the course, we have referenced how experiemental design drives our models and not the other way around.  Therefore it is essential that statisticians understand how data are collected.  Experimental design addresses how the experiment was actually conducted. This typically involves physical layout, logistics, etc., and affects the ANOVA. In our discussions of treatment designs we looked at experimental data in which there were multiple observations made for treatment applications. We referred to these loosely as ‘replicates’. In this lesson, we need to work formally with these multiple observations. This brings us to the right hand side of the schematic diagram;

diagram of experimental design

The treatment design addresses the natureof the experimental factors under study.  The randomization design addresses how the treatments are assigned to experimental units.