A collection of methods for performing random rotations on high-dimensional, normally distributed data (e.g. microarray or RNA-seq data) with batch structure. The random rotation approach allows exact testing of dependent test statistics with linear models following arbitrary batch effect correction methods.