Math 567 Statistical Design and Analysis of Experiments General Course Description
Math 567 examines basic principles of statistical design and analysis of experiments. An introductory course in statistics is required.
Basic Principles of Experimental Design
We will look at the role of randomization and replication in designing experiments. We start with a single variable and move on to factorial designs for more than one variable. We
How do we test the hypotheses and generate the estimates that were the driving force for the experiment in the first place? The statistical analysis of experiments is closely tied to the experimental design. We examine some basic principles of statistical analysis including, ANOVA tables, confidence intervals and F-tests. We also will investigate how analysis is dependent on design.
More Experimental Design Principles
Factorial Designs are not always possible or desirable. Frequently other aspects of the experimental process which are not of direct interest in the study, but may contribute to variability of response may be present. Also, it is possible that one can not completely measure the experimental material of interest. The principles of blocking and nesting examine these design effects.
When there are many variables in an experiment, a full factorial design can quickly require a large number of runs. Of course the more runs needed the more expensive and time consuming the experiment. Can we get the needed information out of the experiment without carrying out a full factorial design? In most cases, the answer is yes. We will explore fractional factorial designs and see how to choose the design so that you can test the hypothesis of interest and minizmize the number of runs.
Math 567 Home page
Department of Mathematics and Statistics Home page
Statistics Area Group Home Page
Western Michigan University Homepage