Determining whether a data point is an "outlier" - a result that does not fit, is too high or too low, is extreme or discordant - is difficult when using small data sets, such as the data from three, ...