How Do You Know When That Psychology Study Won't Replicate?
This is a very good, very readable article about why many psychology studies don't replicate, though most of it is applicable to most any other field of research. The author has four rules to keep in mind, and they are all good. He looks at a number of studies that do not replicate to demonstrate how the rules work. A sample:
This study is basically a p-hacking manual. They’re not even trying to hide it, instead describing in detail how, when a hypothesis failed to yield a p-value below 0.05, they tried more and more things until something publishable popped out by chance.
