Which statistic is least affected by extreme values?

Which statistic is least affected by extreme values?

• A. Mean
• B. Median
• C. Standard deviation
• D. Interquartile range

Answer: (D)

Extreme values tug on statistics that rely on every data point. The interquartile range is different because it measures spread using only the central portion of the data—the distance between the 25th and 75th percentiles. Since these quartiles summarize the middle half, values at the ends (the outliers) have little influence on where those cutoffs fall, especially in larger samples. In contrast, the mean and standard deviation depend on all observations and are pulled toward outliers, making them more responsive to extreme values. The median is also resistant to outliers for location, but the interquartile range is the measure of spread that best minimizes the impact of extremes, so it is least affected by extreme values.

Extreme values tug on statistics that rely on every data point. The interquartile range is different because it measures spread using only the central portion of the data—the distance between the 25th and 75th percentiles. Since these quartiles summarize the middle half, values at the ends (the outliers) have little influence on where those cutoffs fall, especially in larger samples. In contrast, the mean and standard deviation depend on all observations and are pulled toward outliers, making them more responsive to extreme values. The median is also resistant to outliers for location, but the interquartile range is the measure of spread that best minimizes the impact of extremes, so it is least affected by extreme values.