MTH 290: Design of Experiments

Design of Experiments (formerly MTH 248) introduces the statistical thinking and statistical methods needed for scientific research. We discuss planning data collection and data analysis that will provide evidence about a research hypothesis. Planning a data collection includes deciding on what kind of experimental material to use in the study (the experimental units), what conditions to study (the treatments), and what kind of measurements to take (the response variables). We discuss both experimental and observational studies and the issues of confounding, interaction and causal inference.

The data analysis in this course focuses on four basic designs and their extensions: completely randomized factorial designs, randomized block designs, Latin Square designs, and split-plot/repeated measures designs. The statistical methods for analyzing these designs are discussed in detail and we use many real-data examples to illustrate each design. We use many informal and graphical tools to explore data before conducting formal analyses, and we consider whether data transformations may be needed to help the data conform to the assumptions necessary for the analyses. Students learn to use statistical software for data exploration and data analysis.

Each student completes a research project during the course. The student selects a research question, designs a study, carries it out, and presents the results to the class. Students receive feedback throughout the research from peers and from the instructor.