Identify unusually extreme observations using the rule specified in a question.
Flag unusual values before deciding what to do with them
An outlier is an observation unusually far from the general pattern of the data. It may be a genuine extreme event or a recording error; identifying it does not establish its cause.
Use the definition in the question. A common rule flags values below Q1−kIQR or above Q3+kIQR, where IQR=Q3−Q1 and often k=1.5. Another rule uses values outside xˉ±kσ, often with k=2.
On a box plot, mark outliers separately, usually with crosses. When the observations are available, whiskers end at the most extreme non-outliers. If those values are unavailable, use the stated plotting convention; a question may use the outlier boundaries.