online statistics textbook for biology student
April 7, 2010 9:26 AM   Subscribe

Hello everyone, This is a delurk, so let met start by saying hello. I'm a biology student and I have recently started my first research internship. Now I need to do some statistics, and I have never been any good at it, to the point where I don't even know what exactly a null hypothesis is and why it is important. Please recommend a good basic statistics manual, that will clear up the basics for me. I would prefer it uses R.

posted by Coffer to Science & Nature (15 answers total) 24 users marked this as a favorite
You will need an R-specific book to go with, but you can't get much more basic than the Triola Statistics Series. I am a math numbnuts and I found Elementary Statistics absolutely readable and easy to comprehend.
posted by fairytale of los angeles at 9:29 AM on April 7, 2010

Statistics Without Tears is okay, though the lack of a glossary works against it.
posted by zizzle at 9:30 AM on April 7, 2010

(Ah, shit, didn't see you wanted online. You should be able to score a used copy of a Triola book pretty readily, though, if you want.)
posted by fairytale of los angeles at 9:30 AM on April 7, 2010

Dalgaard's Introductory Statistics with R "This book provides an elementary-level introduction to R, targeting both non-statistician scientists in various fields and students of statistics."
posted by ansate at 9:31 AM on April 7, 2010 [6 favorites]

Seconding the Dalgaard recommendation.
posted by Blazecock Pileon at 9:54 AM on April 7, 2010

Dalgaard's book is a great R reference. If I'm remembering correctly though it might not be the best guide to introductory stats itself, just a guide to running those analyses in R.

What type of biology do you study?
posted by JumpW at 10:12 AM on April 7, 2010

Also, while you should eventually learn R as a language, you might find R Commander to be a helpful GUI while you get your feet wet.
posted by The White Hat at 10:13 AM on April 7, 2010

As somebody who uses R every day, I'm thrilled to see all the previous comments and recommendations. The problem with jumping straight into R, however, is that you may learn how to do lots of statistical tests without understanding what they are or how they work. You don't need to learn all the math behind the statistics, but a conceptual introduction could help you understand, for example, what a null hypothesis really is. Kachigan's Multivariate Statistical Analysis: A Conceptual Introduction is the best book I've encountered on the topic. This isn't the only book you'll need, but it's worth the $18 if you're just starting out.
posted by eisenkr at 10:29 AM on April 7, 2010

Good course websites with good exercises are a good way to learn this sort of thing without having to purchase physical media. I took a geographic data analysis class last Spring that exclusively used R. Although much of the class was dedicated to spatial analysis, there was plenty of statistical stuff as well and some links to free online resources for learning R in general. I remember the exercises and resources being useful refreshers in basic stats while showing me how to actually use R.

Here's the course website:

One particularly helpful item on the site, in terms of actually getting up to speed with the program, was this awesome R command cheat sheet:

Maybe others can suggest and post similar course websites accessible to the public.

posted by Derive the Hamiltonian of... at 10:49 AM on April 7, 2010

You don't indicate your area of study but this book is excellent for analysis of ecological communities.
posted by JackFlash at 11:34 AM on April 7, 2010

Starter: How To Lie With Statistics
Basic/Intermediate: Crawley
Basic/Intermediate: Grafen & Hails.
Basic/intermediate/Advanced: Crawley
Intermediate/Advanced: Bolker
Advanced:Venables & Ripley
Specialist (bendy relationships):Wood
Specialist (randomisation; the swiss army knife of statistics):Manly.

Where starter = null hypothesis? and advanced = doctoral ecology.

I'd also put in a plug for Otto for general maths in biology and Legendre & Legendre (brilliant, brilliant books, although not specifically statistical)
posted by cromagnon at 2:24 PM on April 7, 2010 [1 favorite]

Intuitive Biostatistics is a great reference. It's written by the guy who also wrote GraphPad Prism so you won't find any R references there, but I have had this book for years and still refer to it for stats help.

For that matter, the GraphPad website also has a lot of links to other stats resources, and recommended books for people who want to learn more.
posted by caution live frogs at 2:43 PM on April 7, 2010

Seconding Intuitive Biostatistics.
posted by sevenyearlurk at 8:20 PM on April 7, 2010

Wow, thanks for all the replies. It will definitely be helpful! I'm doing work on bioacoustics. many gigabytes of data that I isolate sounds from and gather variables from. I'm glad to hear suggestions for books or online resources, just whatever is the best for a simple start!
posted by Coffer at 2:33 AM on April 8, 2010

Oh, and in reply to The White Hat, I already know R as a programming language, just don't know statistics. Which is sort of weird, but there you go.
posted by Coffer at 2:35 AM on April 8, 2010

« Older What is the perfect gift to get for a Chicago...   |   Could a rah rah "parody" song spell copyright... Newer »
This thread is closed to new comments.