Which measure of center is best to use for data that is skewed?

• A. Mean
• B. Median
• C. Mode
• D. Range

Answer: (B)

When data are skewed, you want a center that isn’t pulled toward the tail by extreme values. The median fits that need because it’s the middle value when the data are ordered, so outliers or a long tail don’t shift it much. The mean, by contrast, uses all values, so a single large or small value can pull it toward the tail and give a distorted sense of a typical value. The mode is simply the most frequent value and may not reflect the overall distribution, and the range is about spread, not center.

For example, with data 1, 2, 2, 2, 3, 100, the median is 2, while the mean is about 18.3, showing how the extreme high value pulls the mean upward. Therefore, the median is the best measure of center for skewed data.