Doing science is hard. Like, really hard. Because even if you do your best and follow all the best practices, there is still plenty of chance you results will be “wrong” – because of type I errors.
But “worst” than that, incentives to get published can deter researchers from following these best practices. If like me you’re an economist, you won’t be surprised by that – because humans respond to incentives. Social psychologist Brian Nosek summarised things well:
There is no cost to getting things wrong. The cost is not getting them published.
This 12 minutes video by Veritasium is one of the clearest explanation I know about type I errors and how incentives actually shape scientific discoveries. A must watch!