Mapping the 21-cm line with the next generation of radio interferometers promises to dramatically advance our knowledge of the process of reionization, and the nature of the first stars and galaxies.

However, even if we can overcome the observational challenges, 21-cm maps will not be straightforward to interpret. For example, the signal will be very non-Gaussian during both reionization, and an earlier phase during which the effects of Lyman-$\alpha$ coupling and X-ray heating dominate the signal. There are also strong degeneracies in the various parameters that can be used to model these physical processes. It is therefore essential that we look beyond the power spectrum, and exploit multiple statistical measures.

 I will discuss the role that statistics like the bispectrum and skewness, both sensitive to non-Gaussianity in maps, will have to play in constraining the properties of the first stars and galaxies. In doing so I will cover a fast method for measuring the bispectrum. I will also discuss the observational prospects for the bispectrum and skewness.