We use mean and MAD as the measure of centre and the measure of variability respectively for a skewed shaped data distribution.