statistics

Pie charts run amok. Tip from Junk Charts. Read more

I knew this, I just didn’t know I knew it:  many people prefer stories to statistics. Tip from the Knowledge Problem. Read more

Cllive Thompson has a nice article in Wired about the need to learn statistics, pity it’s too short, and bit one-sided (heavy snow doesn’t contradict AGW, but a heat wave, that proves AGW!).Tip from Kaiser Fung at Numbers Rule Your World, whose blog post is a better article. Update (3 May).  Discussion at Joanne Jacobs Read more

Steven Strogatz has a go at demonstrating the Bayes’ Box, without the box.  Hundreds of befuddled and off-topic comments ensue.  I teach this thing in about an hour, and all my students can do it. Tip from Andrew Gelman (commenter #17). Update (4 May).  More from Kaiser Fung. Read more

Junk Charts exposes two lemons in graphs of proportions. Read more

Someday I hope to have time do something with this data. Tip from Andrew Gelman’s blog. Read more

Pick the right typeface here. Tip from the Geek Press. Read more

K. O’Rourke has a statistical horror story about a poorly thought-out clinical trial.  It needs more references. Read more

To a small boy with a hammer, everything looks like a nail.  To an experienced statistician, every good industrial experiment is a split-plot design.  I never did really understand these; maybe this paper will help. Tip from the JMP Blog. Read more

Face-detection algorithms can be fooled by random patterns, but only well-designed random patterns. Tip from the Instapundit. Read more