lecture: Science is broken
How much can we trust science in light failed replications, bogus results and widespread questionable research practices?
We're supposed to trust evidence-based information in all areas of life. However disconcerting news from several areas of science must make us ask how much we can trust scientific evidence.
The field of psychology is faced with a crisis where many results that were trusted for decades are called into question. Obviously bogus results like one trying to prove that precognition is real can be created with the existing scientific standards.
In replication attempts in preclinical cancer research more than 90 percent of study results could not be confirmed. Pharmaceutical companies are constantly under attack for questionable research methods. The scientist John Ioannidis asked more than ten years ago "Why most scientific research findings are false".
These aren't just single incidents, they show much deeper problems in the way science is performed today. Scientific results get published if they yield to "positive" results and land in the drawer if the results are "negative", giving an incomplete and often skewed picture. In many fields scientific studies are never replicated. Scientific incentive structures like the Impact Factor prefer sensational results more than rigorous scientific standards.
But there's also some move into the right direction. Trials registers or registered reports can prevent or at least detect many questionable research practices. The replication crisis has led some fields to put more emphasis on repetitions of important results.
Appart from the fact that we get a wrong picture of reality these shortcomings of science also are undeserved munition for those who'd like to reject the scientific principle as a whole.
How broken is science - and what can be done to make more scientific results true?