How bias leads to entrenched errors

Scientific progress is derailed by two human failings: a desire for positive findings, which leads to publication bias, and a tendency to notice and remember things that agree with our current viewpoint, which affects how we cite prior literature. Either one of these alone is damaging; combined they form a toxic combination that means that many ‘established facts’ in a field are not trustworthy. I will give examples of these processes and discuss how we might overcome them.