What is an outlier and how should you treat it in analysis?

• A. A value far from the rest; investigate causes and justify including or excluding it.
• B. A value that is always excluded from analysis.
• C. A value that is exactly the mean of the dataset.
• D. A value that should be replaced with the median automatically.

Answer: (B)

An outlier is a value that lies far from the rest of the data. When you spot one, don’t decide immediately what to do; focus on why it appears. It could be a measurement error, a data-entry mistake, or it might be a genuine observation that reflects real variation in the process or population. The appropriate action is to investigate the cause and then justify whether to include or exclude the value in the analysis, or to use methods that are robust to outliers if the value is legitimate but extreme. Outliers aren’t automatically removed, they aren’t defined by being equal to the mean, and they aren’t automatically replaced with the median. The key is understanding the reason for the anomaly and making a justified, transparent decision.