## Definition

The sample can be "normally distributed" or "not normally distributed": a sample has a "normal distribution" when most of the values are aggregated around the mean, and the number of values decrease as you move below or above the mean: the bar graph of frequencies of a "normally distributed" sample will look like a bell curve (above). To understand the concept of "normal distribution", watch this animation which shows the distribution of randomly dropped balls. Of course the information collected in the field will only approximate to the ideal curve: but the larger the sample, the better the approximation.

## Normal and skewed distributions

Some samples do not display a "normal distribution", or show a "kurtosis":