This was used in one of my stat classes to show the difference between causation and correlation.
A higher percent of older women in neighborhoods correlated with fewer songbirds. Old ladies didnt kill songbirds, but they had increased cat ownership over the general population, and the cats were the cause.
Correlation is not causation. Older folks in the neighborhoods also suggest, for example, older suburbs closer to the urban ring.
(Data also only as good as its gathering...)