Skewness describes how lopsided a distribution is. In a symmetric distribution the mean and the median coincide; in a right-skewed one, a few large values pull the mean above the median, as with household incomes; in a left-skewed one, a few small values pull it below. This calculator estimates the skewness from summary statistics that you often have already: the mean, the median, the mode, the standard deviation and the quartiles.
How the calculation works
- Pearson’s second coefficient = 3 × (mean − median) ÷ standard deviation
- Pearson’s first coefficient = (mean − mode) ÷ standard deviation
- Bowley’s skewness = (Q3 + Q1 − 2 × median) ÷ (Q3 − Q1)
All are zero for a symmetric distribution, positive for a right skew and negative for a left skew.
Example
A data set has a mean of 52, a median of 50, a mode of 47, a standard deviation of 8, and quartiles of 44 and 58. Pearson’s second coefficient is 3 × 2 ÷ 8 = 0.75, a moderate right skew. The first coefficient is 5 ÷ 8 = 0.625 and Bowley’s skewness is (58 + 44 − 100) ÷ 14 = 0.143. All three agree on the direction: the mean sits above the median, so the right tail is longer.
Reading the numbers
As a rough guide, values between −0.5 and 0.5 indicate near symmetry, those between 0.5 and 1 (or −1 and −0.5) a moderate skew, and beyond 1 a strong skew. Bowley’s measure is smaller in scale, since it only looks at the middle half of the data and ignores the tails.
Why skewness matters
Many statistical methods assume roughly symmetric data. With strong skew, the mean can mislead, and the median is a better summary. A log transformation often reduces right skew. In finance, negative skew of returns signals a risk of occasional large losses, while in income and prices, positive skew is the norm.