Statistics Manual

Rating:
Summary: An excellent refference and tutorial
Review: If you need to be spoon fed, this book is not for you. If you want the latest advanced theories with lots of proofs, this is not for you. If you, like me, are in between, and need a guide basic and intermediate statistics, the gives it too you straight, this book *IS* for you. If you want to know all the commonly used statistics, with consise explanation of why they work and how to use them, I doubt you could find a better source.

Rating:
Summary: An excellent refference and tutorial
Review: If you need to be spoon fed, this book is not for you. If you want the latest advanced theories with lots of proofs, this is not for you. If you, like me, are in between, and need a guide basic and intermediate statistics, the gives it too you straight, this book *IS* for you. If you want to know all the commonly used statistics, with consise explanation of why they work and how to use them, I doubt you could find a better source.

Rating:
Summary: Complete 'How To'
Review: This is a superb 'How To' book for all basic statisticsl tests, methods. etc. Simple explanantions, lots of examples, step-by-step how-to do many statistical analyses. I used it during grad school and for 30+ years thereafter. Very useful and useable. If you are trying to buy a book to explain, use, practice statistics, then definitely buy this one!

Rating:
Summary: Very helpful, very practical
Review: This is an outstandingly clear handbook of basic, traditional statistics. There's no theory here, but there are large numbers of practical procedures, including significance testing, regressions, and simple experiment design. Each chapter cover a wide variety of circumstances, and forces a reader to think very hard about just what information is available and just what answers are expected of it.

If you're used to books that list the tests, the procedures, and the tables of p-values, this will probably offer more choices and more detailed discussion than usual. It won't help a lot if you have an unusual circumstance that requires customized analysis, though, and doesn't talk much about the principles under the procedures. Also, it makes almost no mention of non-parameteric techniques.

The one warning I have is that it dates back to pre-computer days. It's full of slick tricks for pencil&paper calculation, with approximations that let you skip a step or two. Back then, computational tips were very welcome. Today, though, they just put a false face in front of the actual operations being performed. The insight available in the basic statement of a computation is often lost.

If you just want basic answers to basic problems, this is about the most useful book I've seen. It won't make you a statistician, but will put you well ahead of most stats users.