This is a pretty good introduction to statistics, especially for complete beginners. I had to buy it for a course, and at first wasn't too impressed with it, primarily because of the writing style. The author keeps the tone very light and says many jokes and funny things, which never bothered me. What did irritate me was that he also has a tendency to really "dumb-down" everything, to the point where you feel like it has been written for younger students, rather than university students (which I gather it sort of was).
However, I figured out after working my way through the course that the text is actually pretty good: it covers several statistical tests that other texts skip. Tests detailed include: binomial sign test, Chi-square, Wilcoxon matched pairs signed ranks, Mann-Whitney U, Wilcoxon rank sum, t test, Pearson's correlation, Spearman's rho, regression (including multiple), Kruskal-Wallis, Jonkheere trend, Friedman, Page trend, ANOVA (one-way, two-way, more-than-two-way, unrelated and related), MANOVA, ANCOVA. It also covers the design of experiments in detail. And the author really does make an effort to explain everything fully, for readers who have neither a statistical background nor even a strong maths background. It serves as a really good reference, even if reading it is a bit painful.
In short, I doubt this is the best statistics book out there, but it is the best statistics reference book I've seen yet. So I would recommend it if you are starting from scratch. But if you are comfortable with maths or even basic statistics, I wouldn't bother this book, as I'm sure you'll find it as irritating to read as I did.