Skewness
A statistic summarizing distributional asymmetry using the standardized third central moment.
Skewness measures left–right asymmetry around a distribution's mean. The common population coefficient based on the third central moment is:
γ1 = E[(X−μ)^3] / σ^3μ is the mean and σ>0 the standard deviation. Positive skewness is commonly associated with a longer right tail and negative skewness with a longer left tail. Zero third-moment skewness does not prove complete symmetry: some asymmetric distributions also have a vanishing third central moment.
Sample coefficients differ
One sample definition is g1 = m3/m2^(3/2), where mr = (1/n)Σ(x_i−x̄)^r. Packages may apply a small-sample correction; where defined, an adjusted form is G1 = [√(n(n−1))/(n−2)]×g1. These estimators need not give the same number. Samples with n≤2 or zero variance require separate handling.
Why skewness enters a binning rule
Doane's rule adds a skewness-dependent term to the Sturges estimate of histogram bin count. Asymmetry may require different resolution from a roughly symmetric sample of the same size. A Q–Q plot shows quantile behavior graphically, whereas the coefficient compresses one aspect of shape into a number.
Moment skewness is sensitive to extreme observations and does not independently measure multimodality or data quality. A box plot, quantile summaries and the actual observations provide complementary information.