Having left the statistics to my science partners, I now find myself wanting to at least conceptually understand what they are talking about when discussing t-tests, chi squares and power. A quick google search reveals a ton of information.
John C. Pezzullo’s Statpages.org provides an index to 600+ (!) of statistics tools and online textbooks. His home page has dozens of links to other scientific information. Wonderful stuff.
Linked from Dr. Pezzullo’s page, Russ Length’s Java Applets for Power and Sample Size allow you to compute power and needed sample sizes before performing a study. Lots of useful information here to help design a study so that you’ll receive reliable data for analysis.
I’ve also picked up a couple books.
Head First Statistics by Dawn Griffiths. This is part of the Head First series from O’Reilly which attempts to take relatively advanced concepts (Object Oriented Design, for example) and reduce it into entertaining chunks.
Statistics for Dummies by Deborah Rumsey. There is also a companion workbook, and an Intermediate Statistics for Dummies. This book is more descriptive and less interactive than the Head First book above, but may be better for my purposes; to simply learn the lingo.
Statistics Hacks by Bruce Fey is part of the O’Reilly Hacks series. Subtitled “Measuring the World and Beating the Odds”, this book is the only one of the three I had on hand which discussed power analysis, the statistics tool of my immediate interest when we are designing a study.
Still on my bookshelf:
Microsoft Access Data Analysis This book, now updated for Access 2007 doesn’t have hard-core statistics, but it does have lots of ideas of how to take samples and turn these into useful information with charts and reports.
Data Analysis for Politics and Policy by Edward Tufte This is an older book quite technical, but with lots of interesting examples. I believe he wrote this book before he got started with the graphics series…but of course that his is forte now.
Visualizing Data by Ben Fry. Subtitled Exploring and Explaining Data with the Processing Environment. Processing is an open-source programming environment developed by Fry.
All of these books don’t solve my immediate problem, which is trying to learn about power calculations. Instead they deal with data after it has already been gathered.
Don’t forget that you may already have considerable statistical firepower at your fingertips if you have a copy of Microsoft Excel. On the Mac, Numbers has a few functions as well, but in comparison to Excel, Numbers is pretty light.