Identify outliers in a dataset using Tukey's 1.5×IQR fence method.
Tukey's method flags any data point below Q1 − 1.5×IQR or above Q3 + 1.5×IQR as an outlier. The 1.5 multiplier captures roughly 99.3% of normally-distributed data within the fences.
Lower Fence
LF = Q1 − 1.5 × IQR
Upper Fence
UF = Q3 + 1.5 × IQR
John Tukey chose 1.5 because for a normal distribution it captures about 99.3% of data — extreme enough to be meaningful but not so strict as to flag too many points.
No. Outliers can be legitimate extreme values, data entry errors, or signals of a different process. They should be investigated, not automatically removed.