H/T to Sue VanHattum for recently pointing out these two spot-on Tom Siegfried pieces that I totally missed in Science News a month ago:
Siegfried astutely summarizes the statistical problems plaguing science for a long time, that are just recently getting addressed (at one point calling statistics "addictive poison" to science) -- the first article fleshes out the problem, the second offers 10 suggestions.
Siegfried is always worth reading and I need to get into a better habit of checking out his "Context" pieces for Science News.
As long as we're talking statistics and worthwhile reads I'll recommend two current books:
"Standard Deviations" (newly available in paperback) by Gary Smith is one of the most fun math reads I've seen for awhile; a statistical playground (but never too technical) with a fire-hose of examples, one-after-another-after-another, that will enlighten and entertain you (eventually I'll have some sort of review up at MathTango, but don't wait for the review, get this book!).
And I'll also go ahead and suggest, sight unseen, Tyler Vigen's new book, "Spurious Correlations" based on his fun site of the same name.
Wish statistics had been this much fun when I was young!