Mathestate Logo

 

The Histogram

When we view the actual data in histogram form we note that (a) the distribution is not symmetric and (b) that the distribution has a long right tail.

SFevrHist

To deal with this shape we need two new terms . "Skewness" describes the extent to which the distribution is non symmetrical. "Kurtosis Excess" measures the fatness of the tails. A check of the distribution shows it is not normal because normal distributions are symmetrical (skewness = 0) and have "skinny" tails (kurtosis excess = 0). Neither is the case for this data.

Skewness 0.221308
Kurtosis Excess 1.07313

NextStep