It has been said that Florence Nightingale was the first to use diagrams for presenting statistical data. This is not true, of course, but she may have been the first to use them for persuading people of the need for change. Minard’s edward tufte the visual display of quantitative information pdf includes a temperature chart which misleadingly suggests that Napoleon’s army froze to death.

Nightingale herself studied this catastrophe, and concluded that Napoleon’s army – like most others – had died of disease. Like Minard’s, Nightingale’s most famous graphics illustrated what she called the ‘loss of an army’ – the British army sent to the Crimea. She published them ten years before Minard’s. Hers also were more topical and conveyed a call to action – they were prescriptive rather than descriptive. She used recent data to persuade the Government to improve army hygiene.

Although she was before Minard, there were others before her. The best-known pioneer of statistical graphics was William Playfair, who published what must be the first ‘pie chart’ in 1801. It was in a graphic showing that, by comparison with other countries, the British paid more tax. Playfair used this graphic to argue for lower taxes. So you could say that, unlike Minard, his graphics are prescriptive. But Playfair’s graphics are merely comparisons.

They do not demonstrate what would happen if you reduced taxes. They look good but make you ask ‘so what’? They do not illustrate cause-and-effect – what Nightingale called a ‘law’. Before going into Nightingale’s graphics, let’s look at the state of statistical science in her day. There was a great revolution in this area in Nightingale’s time. In 1837 the General Registry Office at Somerset House, led by William Farr who later helped Nightingale with her Crimean statistics, began to systematically record births, deaths, and marriages in the UK.

This gave people the opportunity to examine new cause and effect relationships using registration statistics. For example, Florence Nightingale and her sister Parthenope attended the 1847 meeting of the British Association for the Advancement of Science in Oxford. There, they may have seen a report from a Government Actuary, F. Neison, which showed that counties in which people were better educated had a lower crime rate. Neison knew that opponents of his theory would claim that it was prosperity, not education, that reduced the crime rate. Neison estimated the level of education in each county by counting the proportion of people getting married there who were able to write their name on the marriage certificate.