ABSTRACT
The median does not rely on all values in its calculation, so it is not affected by extreme values. This makes it important when describing the central value in skewed distributions. If the data are skewed, the mean and median will be different values, and the difference between them indicates how skewed the data are. When the median is reported as the central value, its parallel measure of dispersion or variability is the range. It measures how much participants vary or differ from one another by simply taking the highest score minus the lowest score. To get around this problem, researchers often use a modified version of the range, which is called the interquartile range (IQR). The formal definition of the interquartile range is the range of the middle 50% of the participants. To calculate the IQR, line up all numbers from smallest to largest, similar to calculating the range or median.
