Gloomy metastudies

Robin Hanson: "Medical studies are seriously biased by interested funders and by tolerance for sloppy methods. Here are four examples."