Triggered by #682777, I ended up reading this extremely well-written (but very long) account of what has become known as the replication crisis in social psychology. Here it is focused on the scientific misconduct that came along with it, but it gives many anecdotes of things you've probably come to believe as true having read them in self-help books or watching Ted talks.
E.g.:
- Putting your hands up in a victory v-shape before giving a talk will help you give a better talk
- Forcing yourself to smile regularly will actually have a positive impact on your mood
- Putting a pledge of honesty at the beginning of a insurance claim will push you to be more honest
- The tendency for people to torture others if it comes from a position of authority (the well-known prison guard experiments)
These studies are mostly garbage. Either due to p-hacking, or due to actual manipulation of the data.
I want to believe that in my field of exact sciences this kind of behavior happens less, but I do not have proof of that. I can probably find comfort that in exact sciences it's easier to replicate and thus confirm if a certain finding is based on fraudulent data. We'll see what the future tells. I tend to think of myself as being able to spot the collaborators that have integrity and the ones that do not have it. I hope I'm not being misled ;)
I've known for a fact that a certain collaborator had copy-pasted his course material from Wikipedia (the charitable explanation would be that he wrote the Wikipedia page for that material himself), but other than that, most people I know I consider them to be clean. This is also probably a grey-zone area of what is considered unethical rather than fraudulent.
If you work in science, have you ever encountered fraudulent behavior from some of your colleagues?