We all know the percentages for standard deviation if you have a normal distribution: approximately 2/3 of the data within one standard deviation of the mean, approximately 95% within two standard deviations, etc. But what if the data are skewed: How do we interpret the standard deviation with skewed data?