All of the CONNX Statistical Functions are aggregate functions. They can be divided roughly into three classes: calculation of central tendency, calculation of dispersion, and calculation of shape.

Calculation of central tendency means finding out what observations are likely.

Calculation of dispersion shows how scattered the data is. One example of dispersion is range, which is the difference between the biggest and smallest value in the set.

Calculation of shape defines the shape of the curve of our observations. Unusually shaped populations can be detected through examination of skew and kurtosis. Skew shows whether there is an abundance of small observations or an abundance of large observations. Kurtosis shows whether distribution is flat with tiny tails, or sharply peaked with large tails. The skewness and kurtosis of the normal distribution are zero.

Many of the statistical functions come in two flavors:

• For the population (the result is a parameter)

• For the sample (the result is a statistic)

Important: Use the population-specific function only if your data set contains a measurement for each and every member of the complete population of objects and all of them are included in the query.