What is an appropriate step when you find an outlier in data analysis?

• A. Immediately remove it without justification.
• B. Keep it and ignore it in analysis.
• C. Use it as the sole basis for conclusions.
• D. Investigate causes and justify including or excluding it.

Answer: (D)

When you spot an outlier, the important step is to investigate why it occurred and decide, with justification, whether to include or exclude it. Outliers can arise from measurement or data entry errors, data processing mistakes, or they can be genuine extreme values that reflect real variation. By tracing back to how the data were collected and computed, you determine whether the value is erroneous or meaningful. If it’s an error, you have a basis for correcting or removing it; if it’s real, you justify keeping it and possibly use robust methods or show how results change with and without it. Always document the decision and its impact on conclusions, and consider a sensitivity analysis to show how the outlier affects the analysis. The other options would either discard information without justification, ignore potentially informative data, or let a single value drive conclusions, which can bias or mislead.