Everyday Probability and Statistics. Health, Elections, Gambling and War
Information disseminated by newspapers, radio, television and so on brings a lot of data that we can use as support for our decisions. Since the decisions are made in an environment where uncertainty plays an important role, a scientific approach is also based on probability theory. It is easy to misinterpret given data, which leads to the well-known phrase “lies, damned lies and statistics” (L. H. Courtney). As for me, I prefer another phrase: “It is easy to lie with statistics. It is hard to tell truth without statistics” (A. Dunkels). And the author in the introduction correctly writes: “In the 21st century a cultured man should understand something about statistics otherwise he will be led by the nose by those who know how to manipulate statistics for their own ends”.
The book is a very elementary introduction to probabilistic and statistical thinking. The basic ideas are demonstrated on simple examples from everyday life, e.g. how to bet on a horse. There are also non-intuitive problems like the birthday problem and the problem about switching. As a more practical topic we find an application of probability theory to medicine. The elements of statistics presented in the book concern calculations of the mean and variance and normal and Poisson distributions. The book also contains parts devoted to predicting voting preferences, a sampling technique, some statistical tests and building probabilistic models. The author does not assume that the reader is familiar with mathematics. Because of that, an explanation on how to handle expressions like (22)3 is quite long and, similarly, a description of the definition of the number e is also very detailed. Some places in the book deserve critical remarks. The expectation of the normal distribution is called the mean and denoted by the same symbol as the arithmetical mean – this is confusing (p. 117). I cannot agree with the formulation “Assuming that the lifetimes have a normal distribution ...” (p. 120). The lifetimes are nonnegative and so they cannot have a normal distribution. It would be better to say that a normal distribution can be a good approximation to the unknown and perhaps very complicated true distribution of the lifetimes. The book can be recommended to students who are not specialists in mathematics.