Review: There is a familiar cry that goes up from many students when the idea of statistics is mentioned. The need for such work is not denied but the issue seems to be how we can get students (or staff) to recognise the importance of statistical literacy with minimum effort and maximum effect. This text aims to do both using the devise of the title to lure readers. If not terrified at least confused and so we are looking for a text that understands the requirements of the beginner whilst still being true to key principles.

The text works by gradually building upon a few simple ideas. It starts with few basic assumptions and then shows how, by adding layers of complexity, statistics can be properly used and interpreted. We start with a brief run through of the books components where icons help highlight key ideas. An extremely brief introduction outlining the nature of statistics leads us into the first idea - that of variation. There are a limited number of categories in statistics and this is one of the most fundamental means, standard deviation and variance. Chapter four starts to look at ideas in more detail adding a quicker way around basic calculations and demonstrating how it works. The next move is to the normal distribution, described in chapter five and expanded upon in the next two chapters. Chapter 8 moves to the t-test - a more common way of testing variation. At this stage in the text the reader is moving towards more complex (and statistically useful) calculations and so the earlier ideas are assumed rather than fleshed out as happened before. Chapter 9 looks at the need for careful examination of the probabilities of extreme measures - usually referred to as one or two tailed tests. It highlights the care that must be taken when comparing results especially if a small error can lead to the rejection (or not) or an hypothesis. Following this the reader moves to chapters describing the uses of and calculations for analyses of variance. Another level of complexity is added when factorial experiments are discussed. These are studies where two or more variables are studied concurrently leading to the need to separate out which changes, if any, are related to which combination of factors. The importance of these tests are shown by the fact that four chapters are devoted to it. Starting with an overview of the technique, we move to 2 and multi-factor tests and then to split plot variations. Chapter 16 returns to t-test and variance but this times considers some of the more subtle variations needed to distinguish between differences in many means. Given that the general movement is towards more complex testing, chapters 17 and 18 might seem out of place. The former gives a robust account of correlation noting how and when it is appropriate. The latter is the familiar chi-squared test comparing "observed" with "expected" results. Finally there is a discussion of non-parametric tests with some examples to guide the reader.

This is an interesting way to cover statistical methods. The title conveys the idea that it will calm the nerves. Certainly there is every attempt to reduce problems e.g. summaries, memory devices and lots of examples but most of these suggest the need for more calculations. This is not a text stating simply which test can be used, when. It's more of an exploration of both how and why certain statistical tests work (or don't) and what issues are involved as well as giving the test details. In this way it seems to fall between two stools lacking the simple approach of test finding whilst giving less detail for more complete statistical exploration. A useful text for statically education but one that may well best be used in conjunction with other approaches.