A good, vintage introduction to applied statistics.
I occasionally teach mathematics at this level and some of the ordering of topics surprise me, such as Poisson distribution before normal distribution. However, I can see a point to discrete distributions ahead of continuous sample space. But, multinomial distribution before binomial?
Regardless, I really like the content on control charts, Dodge Romig Sampling Plans, and general sampling theory for uncovering defects.
You would need a pen, paper and a calculator to follow this author. This might seem to be a simple book but it does get pretty number-heavy towards the middle. That being said, the treatment of each concept is comprehensive and a patient reading shall definitely give a lot of clarity. My personal favorite chapters are Safety in sampling and Ranking methods
I was recommended this by the late authors of P-Stat, Roald and Shirrell Buhler, when I was a young programmer. It's by far the best book on statistics for the numerate beginner. The manual calculations have been superseded, as another reviewer said, but still valuable because you get to see where all these numbers come from. Moroney's wry humour and explanations are priceless. Even now, I still refer to it for clarification or technique.
I skimmed this book about 20 years ago, seemed very good (and was highly recommended by a friend). Should get back to it someday, specially if I need to delve deep into its subject matter (statistics).