Outliers (statistics)
Outliers are data points that fall far outside the overall pattern of a dataset. A common rule flags values more than 1.5 times the interquartile range below the first quartile or above the third quartile.
An outlier is an observation that lies unusually far from the rest of the data. In a list of home prices where most values cluster between $300,000 and $500,000, a $4 million estate is an outlier. Outliers can be legitimate extreme values, data-entry errors, or observations from a different population, so the first step is always to investigate why the point is unusual — not to delete it automatically.
The most common detection method is the 1.5 × IQR rule. Compute the interquartile range (IQR = Q3 − Q1); any value below Q1 − 1.5 × IQR or above Q3 + 1.5 × IQR is flagged as an outlier. If Q1 = 10 and Q3 = 20, the IQR is 10, so the fences sit at −5 and 35, and a value of 42 would be flagged. Boxplots display these flagged points individually beyond the whiskers. Another approach flags points more than 2 or 3 standard deviations from the mean.
Outliers matter because they distort some statistics far more than others. The mean and standard deviation are sensitive to outliers, while the median and IQR are resistant — which is why skewed data is better summarized with the median. In regression, an outlier in the x-direction has high leverage and can be an influential point, single-handedly changing the slope of the least-squares line.
The AP Statistics exam tests the 1.5 × IQR rule, resistant versus non-resistant measures, and outliers' effects on regression, while the Praxis Core math test asks how outliers shift the mean relative to the median. The CMA Part 1 exam touches outliers in its data visualization coverage, where boxplots and scatter plots are used to spot anomalous values.
Key takeaways
- An outlier is a data point far outside the overall pattern of the dataset.
- The 1.5 × IQR rule flags values beyond Q1 − 1.5 × IQR or Q3 + 1.5 × IQR.
- The mean and standard deviation are sensitive to outliers; the median and IQR are resistant.
- In regression, outliers can be high-leverage, influential points that change the fitted line.
- AP Statistics, Praxis Core math, and CMA Part 1 all test recognizing outliers and their effects.