Outlier
Also called: Extreme value · Topic: Economic Data: Census, NSS, Surveys and Statistical Tools · NCERT: Class 11, Ch 5 "Measures of Central Tendency"; Class 11, Ch 6 "Correlation"
Meaning
An outlier is an extreme value that lies far away from the rest of the data. It pulls the arithmetic mean up or down, so the mean stops showing a typical value. The median depends on position, not on size, so an outlier barely affects it. For the same reason, Spearman's rank correlation, which uses ranks rather than actual values, is preferred over Karl Pearson's r when the data have outliers.
Example
In NCERT's series 1, 2, 3000, the mean is 3003/3 = 1001, which is far from most of the values. The median is 2. Income data have the same problem: a few very rich people raise the mean income well above what a typical person earns.
Don't confuse with
- Sampling error: sampling error is the gap between a sample estimate and the true population value. An outlier is a real extreme value within the data.
Related concepts
- Measures of central tendency
- Arithmetic mean
- Weighted arithmetic mean
- Median
- Quartiles
- Decile
- Percentiles
- Mode
- Bimodal and multimodal distribution
- Modal class