Distinguishing patterns of charge order: Stripes or checkerboards

In two dimensions, quenched disorder always rounds transitions involving the breaking of spatial symmetries so, in practice, it can often be difficult to infer what form the symmetry breaking would take in the ideal, zero disorder limit. We discuss methods of data analysis which can be useful for making such inferences, and apply them to the problem of determining whether the preferred order in the cuprates is stripes or checkerboards. In many cases we show that the experiments clearly indicate stripe order, while in others (where the observed correlation length is short), the answer is presently uncertain.