An outlier is an observation far from the rest of the data. It may come from a data entry or measurement error, or reflect a rare real case. It strongly influences the mean and standard deviation. You spot it with the 1.5 x IQR rule or a box plot, then examine it before deciding whether to keep it.
Example
Outliers are values outside [Q1 - 1.5 x IQR ; Q3 + 1.5 x IQR].