Half of being a great statistics practitioner is having the right books in which to look things up.
There are many dense statistics
books
out there, and I've read a lot of them. Here I recommend the gems
in the rough--easily readable books that will help you apply
statistics to real
data, not derive anything or compute statistics by hand. These
books are
written for researchers, not statisticians.
I
recommend a few textbooks, mainly as resources. Their intent may
be to teach to
compute statistics by hand, but you can just skip the exercises.
Even in textbooks, I only recommend those that are highly readable.
There are a few in the list
that are pretty dense reading. I've included them only when they
are the definitive book on the topic, and when slogging through them
will be worth it. I have labeled them.
Some of these books are
oriented toward using specific software, but are such great guides to
the statistical technique that they are worthwhile for anyone.
They are described as such.
The following three books are
the exception to the non-technical rule. These are traditional text
books (the kind that make you calculate statistics by hand). But they
are well written, and although I don't recommend them if you are trying
to learn a technique on your own, they are all excellent reference
books.
This is a textbook,
through
and through, but an excellent one. It really covers just about
everything in linear models (regression and anova), and it quite
straightforward. A great reference book.
A fabulous ANOVA
text.
One of the best explanations of simple effects I’ve ever seen. I
used this in an ANOVA class as an undergrad psychology major. It
was challenging then, but is definitely accessible to social
scientists. This is a true textbook and contains lots of
exercises for computing statistics by hand (skip those parts).
But it is a great reference book.
Another excellent ANOVA
text. This one was written for engineering students, so it’s a
bit more theoretical and mathy than Keppel. Definitely a
compute-by-hand textbook, but is another fabulous reference.
This is a textbook,
through
and through, but an excellent one. It really covers just about
everything in linear models (regression and anova), and it quite
straightforward. A great reference book.
The picture will take you
to a previous edition, which will save you about $100 off the latest.
This is a pretty specific topic, but it’s a great book when you need
it. I love all of the books in this series that I’ve read.
Contains enough statistical theory to be helpful, but not lots of
technical jargon or deriving equations. Readers should be quite
familiar with linear regression.
You’ve probably noticed
by now
that I really like Paul Allison’s books. He is one of my favorite
applied statistics authors.
This one is great as well. This book, combined with Scott
Menard’s, are all you need to learn and use Logistic Regression.
A great introduction to
logistic regression, and very reasonably priced. I love all of
the books in this series that I’ve read. Contains enough
statistical theory to be helpful, but not lots of technical jargon or
deriving equations. Readers should be quite familiar with linear
regression.
Very reader-friendly.
One
of “the little green Sage books.” This is an excellent overview, covers
much of what a data analyst needs to know, and very accessible. This is
the book to start with. And
very reasonably priced.
This is a pretty
technical
book, and is not for the statistically feeble. But if you’re
doing mixed models, you’re not statistically feeble. That said,
if you are doing Mixed Modeling in SAS, it’s a must-have
book. Back in the Cornell Statistical Consulting office, we
actually wore out the book.
This is THE book to read
if
you’re doing Factor Analysis. The first chapter alone on
Principal components is amazing. Very well written and
nontechnical. Even if you don’t use SAS, this book explains
Factor Analysis VERY well. Just ignore the SAS code.
This is the book that I used
when I took my first course on SAS during graduate school, although
that was many editions ago. It explains both the logic and the
command language of SAS in a very understandable way. It's one I
refer to often.
This is my favorite “How
to
use SPSS” book. Each chapter covers a single statistical
analysis. There is a brief overview of each statistical method,
but it is not enough information if you’ve never done it before.
Well, of course. It assumes you’ve never
taken or done statistics before, so is at a very basic level. It
shows how to do all analyses using SPSS, rather than by hand. For
someone either just learning or reviewing statistics and learning
SPSS. It is aimed at sociology majors, but even my
mycologist (fungus biologist) neighbor found it helpful when she needed
to review statistics for a job interview. If you’re pretty good
at statistics and just want to learn
SPSS, I recommend Green & Salkind.
If you need to use
Survival
Analysis for the first time (or the 20th), you need this book, even if
you don’t use SAS. I just can’t say enough about how good it
is.
Monthly Tips, Resources, and News.PlusThe Top Resources for Learning 13 Statistical Methods. And our free Webinar Series: The Craft of Statistical Analysis. FREE!
