As a teaching assistant in my grad school days, I saw a number of books
used for intro statistics and biostatistics classes. Freedman et al. was
one, but I must say that I never saw the attraction. A classic case of YMMV,
as I know many others think it's great.
Which brings me to my main point: What works for me, or for you, may not
be what works best for the original asker. If possible, I would suggest that the
woman in question peruse through books in a college book store, college library,
or even a well-stocked general book store, along with what one can see of the insides
of books on Amazon, to see what strikes her as a book at the right level.
That said, I would add Zar's Biostatistical Analysis into the mix. The current edition is the
5th, but I'm still happy with my 3rd edition, and it looks like the 4th edition can be picked up
used for cheap on Amazon, anyway:
http://www.amazon.com/Biostatistical-Analysis-5th-Edition-Jerrold/dp/0131008463/ref=pd_sim_b_3?ie=UTF8&refRID=195BD7SRQ0SP970BVFGK http://www.amazon.com/gp/offer-listing/013081542X/ref=dp_olp_used?ie=UTF8&condition=used I second the suggestion of perusing the offerings on Coursera or edX. You can also find lots of
intro material online, often created as material for college intro statistics courses.
At a slightly more advanced level, I would also recommend taking a look at Rupert Miller's "Beyond ANOVA."
This book looks at the problem first (e.g., two-sample comparison), then addresses the various possible
methods, along with potential problems (e.g., non-normality) and what one might do about them.
Besides teaching a data analyst about potential pitfalls in her analysis, it helps clue her in on what
the consulting statistician is nattering on about :-)!
http://www.amazon.com/Beyond-ANOVA-Applied-Statistics-Statistical/dp/0412070111 (Unfortunately, I don't see any used copies of earlier editions on sale at Amazon at the moment.)
I'll also pass along a piece of advice I got years ago: Often, the second book you read on a subject
is the one that you'll turn back to over the years. Not necessarily because it's the best book, but
because the first book you read laid the ground for you to understand the material the second time
you saw it presented slightly differently. So maybe pick a book and an online course, or two books.
>>Kathy
-------------------------------------------
Katherine Godfrey
-------------------------------------------
Original Message:
Sent: 06-02-2014 13:58
From: James Garrett
Subject: Resources for learning statistics
I find that one of the best resources for providing an understanding of the ideas undergirding a lot of standard statistics is the book _Statistics_ by Freedman, Pisani, and Purves. This is an unconventional answer because this book actually provides relatively little in terms of standard analysis methods. I find that a lot of introductory statistics material is of the recipe-book variety: in this situation do this, in that situation do that, without conveying a coherent foundation that someone totally new to statistics can understand. But if someone actually reads the book (which will take some time) one would actually gain a decent understand of the underlying theory, which would put one in a good situation to learn more about specific methods.
-Jim
-------------------------------------------
James Garrett
Sr. Assoc. Dir. of Biostatistics
Novartis
-------------------------------------------