Tuesday, February 2, 2010
Decorative data are collected and presented not because any useful conclusions can be drawn from them but because they can be used to make a pretty picture for a report. The picture shows an example: all staff at a university by school and gender. (The data are based on real information but the numbers have been randomly adjusted by small amounts.) Yes, I’ve cheated and presented the data in a way that obscures the information content – the numbers are represented by the height of the different coloured sections of the cones but the eye sees the area. However, even if I had presented the data as a boring bar graph, it would still be difficult to draw conclusions from the data as the numbers are for all staff, but the gender balance of academic and research staff varies from discipline to discipline much more than the gender balance of support staff.