Statistics refers to a series of mathematical methods used to analyse events which are either experienced through the senses or purposefully observed. These are referred to as empirical data.

Why is it necessary to undertake an analysis of data? This short clip will not make a statistician out of you, but two simple examples will show you how important statistics is.

Imagine a large city with a population of 1 million. It is the year 1880, in which many people are dying of tuberculosis. Years pass and mayors come and go, until the year 1980.
The mayor who is currently in office is a very concerned man who wants to know whether medical care in his city has improved. He selects the number of deaths from tuberculosis as an indicator and commissions a study to find out the number of all deaths from TB from 1860 to 1980.

His employees get to work and soon provide some results – the mayor is satisfied, pats his councillors on the back and shows them the study – in his city the number of deaths due to TB steadily decreased from 1880 to 1980, the mayor is satisfied with the medical organisations, who are obviously doing good work, and a new and expensive hospital is built.

What he does not realize is that the statistical analysis of this data is false, since the number of deaths was shown in decimals, which distorts the graph. If the analysis is performed differently, and the number of deaths is displayed logarithmically, it becomes obvious that the death rate decreased dramatically after 1946. Knowing that a terrible war had just come to an end and that people once again had enough to eat and a roof over their heads, then there was no point in building a new hospital. It may have made more sense to build apartment buildings in order to provide housing for the homeless, and also use the services of a statistician for the study.

Mistakes like this do happen, but this lack of knowledge can also be abused, and even doctors are not immune to this. In 1990 a large pharmaceutical company sent a letter to many doctors playing down the side effects of Prozac – an SSRI blocker. In the letter the pharmaceutical company claimed that this medicine had been tested on 11.000 individuals. The doctors were satisfied and kept on prescribing the medicine, not knowing that the numbers had been very skillfully faked. Actually, the medicine had only been tested on 286 persons instead of 11.000. Only 63 patients were tested for over 2 years – the rest had to quit because of side effects or because the medicine failed to work. To this day there is no study proving the efficacy of this medicine, but it continues to be prescribed and the severity of the side effects has only resulted in a longer package insert.

So you can see how important it is to know something about statistics and studies – like we do!