Overview of Multiple Comparison Methods
In this workshop we will demonstrate when the problem of multiple comparisons occurs, and how inference regarding Type-I error is affected. We will briefly describe the traditional methods for correcting for multiple testing. These methods include the Bonferroni correction, Tukey's HSD, and Dunnett's method. We will then describe methods which are more suitable for large-scale experiments. In particular, we will explain the notion of familywise error rate (FWER), and how to control the false discovery rate (FDR). We will show how different software packages account for multiple testing. This workshop requires basic knowledge in hypothesis testing (Z-test, t-test, ANOVA).
Fee: None to members of the Cornell community, but registration is required.
For times and locations of upcoming workshops, please see the Workshop Schedule.