Maybe they should just write some papers about their priors and not mess around with actual data!

This came up in a blog discussion a few years ago. As I wrote at the time:

Yeah, people always think I’m being sarcastic or patronizing, but I’m dead serious here. I see this so much, that researchers have strong beliefs and, at least from their perspective, deep qualitative understanding, but instead they write really weak quantitative papers based on bad data and sloppy analysis. No need for them to focus on their weaknesses.