About Smoothing

SpaceStat allows you to smooth your data. Smoothing reduces the variability in your data, essentially reducing the extreme values more towards the mean of the dataset. There are several smoothing algorithms in the literature. SpaceStat implements the empirical Bayesian smoother which smooths values from low-population areas towards the mean (Clayton and Kaldor 1987, cited in Waller and Gotway 2004).

Why smooth?

Particularly with health data, we may have different levels of confidence in values at different locations. For example, in cancer mortality data, we may have more confidence that the mortality rate in a large population is representative of that population or is related to the underlying risk. Whereas the mortality rate derived from few deaths in a small population is much more subject to noise. Small numbers can create unstable rates.