Relating the choice of measures of center and variability to the shape of the data distribution and the context in which the data were gathered.