Everyone needs to know a bit about the interpretation of statistical data. I would go so far as to say that some understanding of statistics is an essential part of democracy. The average citizen, faced with statements like “kills 99 per cent of all known germs” or “contains 25 per cent less fat”, needs to know what they actually mean. This understanding is rendered difficult by the fact that many ordinary citizens have an aversion to numbers, and most especially to numbers accompanied by Greek squiggles. People who would be ashamed to admit to any degree of illiteracy will laughingly announce that their eyes glaze over when they see numbers on a page. There is therefore a need for general books and other guides to help. The classic of these is, of course How to Lie with Statistics (Huff, 1954) – still a useful tool sixty years on.
Producing and understanding statistical data is an important part of all scientific, and especially, social scientific research. Though qualitative approaches have made an edge in the social sciences in recent years, it is by and large true that “if you cannot count it, it does not count.” Research workers therefore need detailed guides to statistical methods. We have recommended numerous reference tools in this journal over the years, notably the American Psychological Association APA Dictionary of Statistics and Research Methods (Zedeck, 2014) (RR 2014/163); Dictionary of Statistics (Upton and Cook, 2014) (RR 2014/264); and, my first choice among such dictionaries, The Cambridge Dictionary of Statistics (Everitt and Skrondal, 2011) (RR 2011/224).
Many scientific researchers speedily get a grip on statistical methods. The two disciplines which most depend on statistics however, are psychology and sociology, and unfortunately these disciplines tend to attract students who have specialised in humanities-based subjects at school, and are more averse to numbers than most scientists. Students in these disciplines therefore need extra guidance.
This book seems to me to be aimed at scientific researchers rather than at the general public or at social science students. It consists of 75 articles, arranged in alphabetical order. Each has a first page of brief definitions, followed by three or four pages of detailed analytic discussion. These take statistical methods up to a very advanced level – the author claims that it is possible to pass Six Sigma Black Belt using it (I must admit that I had never heard of Six Sigma before, until I looked it up - www.bing.com/search?q=six+sigma+explained&form=PRGBEN&pc=EUPP_UP97&httpsmsn=1&refig=5125c51690b14906bee1fdff3f22d9ba&pq=six+sigma&sc=8-9&sp=2&qs=AS&sk=AS1 but, having done so, even the depth of statistical skills needed for the yellow belt level looked terrifying to me, so black belt must be pretty impressive).
The potential market for this book, lying somewhere between a specialised dictionary and an advanced textbook, is therefore more limited than its title might suggest. It would be too off-putting to be of use to general readers, so I would not recommend it for public library purposes. Mathematics undergraduates would find it a very useful tool for getting into the practical applications of their subject, but it would probably not suit undergraduates in non-mathematical subjects. Even among taught postgraduates, looking around my own institution I can see that many of the neuroscience students would be able to use this as a quick reference tool, though a lot of them probably already know enough about statistical methods for their own purposes. Many of the clinical psychologists would need something more basic to start with, though they might work up to this level. The social work masters students need some understanding of statistics but would find this completely beyond them.
All libraries should try to acquire as many basic texts on the interpretation of statistics as possible. Most libraries would benefit from acquiring reference tools such as The Cambridge Dictionary of Statistics. Academic libraries catering for mathematicians and for postgraduate scientific researchers would find this book a useful quick reference tool, supplementing the alphabetic dictionaries on one hand and more organized textbooks of statistics on the other.
