Introduction

Imagine you’re analyzing a dataset of monthly incomes in a neighborhood. Most people earn between ₹20,000 and ₹50,000, but there are a few individuals who make ₹5,00,000 or more. Suddenly, your average (mean) income shoots up to ₹75,000 — far from the reality most residents experience. This is a classic case where outliers distort the central tendency. So what should you do?

What to Do Instead:
To better represent the central tendency in such situations:

Consider the Mode: In certain cases (like survey responses), the mode (most frequent value) might also be relevant.

Use the Median: The median gives a more accurate picture of the “typical” value in skewed distributions. In our income example, if the median is ₹30,000, that reflects the earnings of the average resident more realistically than the inflated mean.

Report Both Mean and Median: Sometimes, reporting both measures helps readers understand the nature of the data — and see that it’s skewed.




Visualize the Data: Box plots, histograms, and distribution curves make it easier to spot skewness and outliers.

Trim or Winsorize the Data (With Caution): For specific statistical purposes, you can remove or adjust extreme values — but only when justified and transparently explained.

Conclusion

When analyzing datasets with extreme values, it’s crucial to recognize how outliers can mislead interpretation. The median, not the mean, often tells the truer story of central tendency. Being thoughtful about which measure to use helps ensure your analysis reflects the real-world patterns your data represents.