Outlier detection using modified boxplot

A boxplot, also known as a box and whisker diagram, is a well-known graphing technique that displays centering, spread, and distribution of a continuous data set, perhaps primarily by using quartiles; currently there exist many versions of this exploratory tool, for all of these boxpots the whiskers are usually defined at 1.5 times the difference between upper and lower quartiles, and any data points falling outside the whiskers are considered outliers. In this paper, a modified cut off value for the whiskers using a robust measure of skewness and tail weight of bimodal distributions is explored in a simulation study.