that estimates the mean with less dependence
on the outlying values
of a statistical distribution
producing a more robust
estimation for the mean. By replacing outlying, high
and low values with values that are closer to the mean, the resulting estimation is likely to be biased towards the mean.