Long page
descriptions

Chapter 3   Basic numerical displays

3.1   Describing centre

3.1.1   Centre and spread

Two important aspects of a distribution are its centre and its spread.

3.1.2   Summaries of Centre

The mean and median are descriptions of the 'centre' of a distribution. The mode is another description of a 'typical' value for discrete data.

3.1.3   Comparison of median and mean

The median is less sensitive to outliers and is easier to interpret than the mean. However there are more 'advanced' uses of means than medians.

3.1.4   Mean of discrete data

The mean of discrete data does not need to be a whole number.

3.1.5   Quartiles, deciles and percentiles

The median splits a data set into two. The quartiles split it into four, deciles into 10, and percentiles into 100.

3.2   Describing spread

3.2.1   Range and interquartile range

The range in a data set is the difference between the maximum and minimum values. A better measure of spread is the difference between the upper and lower quartiles.

3.2.2   Distance from k

The mean absolute error and root mean square error summarise how close a data set is to a target, k.

3.2.3   Measures of spread

The standard deviation describes how close the values are to the centre of the distribution.