This textbook on statistics is intended for students of the behavioural sciences. I have been co-teaching a module on statistics for a Masters programme in Human-Centred Interactive Technologies, and this is one of the recommended books. I used it as one of the texts to reference as I prepared my lectures.
Since it is a textbook, I have not read it from cover to cover. However, I have read considerably more of it than I originally intended. Each time I looked up a particular technique, and started reading, I just seemed to keep on reading. The book is very well written and extremely engaging.
It also has the right depth for students who are technically literate, but not mathematicians. Many statistics books can be either all theory, targeted at mathematicians, or all ‘magic’, targeted at those who want recipes rather than understanding. This book hits the sweet spot of formalism combined with pragmatism with great verve: providing background and context so that the workings of the various tests make sense, with sufficient technicalities to inform rather than confuse.
Some of the included vignettes demonstrate the care that is needed. There is one tale of a renowned statistician who had invented various tests, yet still got the number of degrees of freedom wrong on one of them. These precision tools are hard to invent (so probably best not to invent your own!), but once they have been, they can be taught and used with relative ease. This is how we progress.