This is a measure of variation calculated as the difference between the upper quartile and lower quartile. That is, half of the data items will fall within the interquartile range. It is typically a far more robust measure than the more common standard deviation (s.d., σ), but less tractable for calculations.