It's xkcd on statistical significance. I need to stop eating all those green jelly beans.
I hear there's also a link between sunshine and lollipops.
If every news article came with a caveat about multiple-testing corrections, we’d be in for some very long news articles indeed.
I actually based a whole lecture on this comic yesterday. My students are just starting to understand p-values so it was perfect!
Can someone help me find the semi-satirical academic paper published debunking an earlier paper finding support for telepathy? It was a hoot, and its dismantling of the original paper was based on precisely this multiple-testing issue.
FlowingData explores how designers, statisticians, and computer scientists are using data to understand ourselves better — mainly through data visualization.
As for me, I'm Nathan Yau, and I have a PhD in statistics, with a background in eating.