When does the difference between mean and median matter?

We're not trying to sound mean, but you should pay attention to what your maths teacher tells you at school.

4th November 2015
When does the difference between mean and median matter?

Asked by: Lisa Cooper, Brighton

We all remember those boring maths lessons where the teacher droned on about the difference between the mean of a set of data, and the median. Yet the difference can be vital to understanding some controversial issues. For example, the latest official statistics show that men working full-time in the UK get paid on average around 17 per cent more than women. But that figure masks the impact of the relatively small proportion of men who get paid colossal amounts.

Whenever data is seriously skewed like this, the median becomes far more representative of what’s ‘typical’ than the mean, as it’s the value which splits the data exactly in two, with 50 per cent being above the median and 50 per cent below.

In the case of pay, taking the median shrinks the gender gap among those in full-time work by around one-third, but its effect on part-time pay statistics is even more dramatic. While the mean says that men working part-time get around 5 per cent more than women, the median figures reverses this, showing that men typically get paid 5 per cent less than female counterparts.

 


Get more fascinating Q&As from BBC Focus magazine by following @sciencefocusQA

You are currently reading: When does the difference between mean and median matter? - 4th November