When you have (or imagine you have) access to the entire population of interest, perhaps every person in a country or every user of a device, and can calculate the standard deviation (s.d., σ) of some facet of them all; that is √ ( &;Sigma (x_{i}–μ)^{2}))/N – where x_{i} are the individual measurements of the facets, μ is the arithmetic mean (μ) and N the number of people in the entire population.

Used on page 30