If any values in the data set are <u>greater than</u> the upper boundary, then those values are upper outliers.
If any values in the data set are <u>less than</u> the lower boundary, then those values are lower outliers.
The general rule is that the upper boundary is found computing Q3+1.5*IQR, where IQR is the interquartile range. Recall that IQR = Q3-Q1. The Q1 and Q3 are the first and third quartiles respectively.
Another rule is that the lower boundary is found by computing Q1 - 1.5*IQR
On a box and whisker plot, outliers are points that are not connected to the main plot itself. You can think of them as islands far from the mainland.