Kurtosis (Fourth Central Moment)
  • is a measure of whether the data are heavy-tailed or light-tailed relative to a normal distribution
    • data sets with high kurtosis tend to have heavy tails, or outliers
    • data sets with low kurtosis tend to have light tails, or lack of outliers

Formula

for univariate dataset {𝑋1, …, 𝑋𝑛}:

  • sample kurtosis = 𝐄[(𝑋𝑖- 𝑋̅)4] = [𝛴1≤𝑖≤𝑛(𝑋𝑖- 𝑋̅)4] / [𝑠4𝑛]

where:

Excess Kurtosis

kurtosis for a standard normal distribution is 3

  • excess kurtosis = [ [𝛴1≤𝑖≤𝑛(𝑋𝑖- 𝑋̅)4] / [𝑠4𝑛] ] - 3

Kurtosis Types