Disconfirmation Bias

The asymmetry in how people evaluate evidence: evidence that CONFIRMS existing beliefs is accepted quickly and uncritically, while evidence that DISCONFIRMS existing beliefs is scrutinized intensely, picked apart, and dismissed on technicalities. The same study is judged as 'well-designed' when it supports your view and 'methodologically flawed' when it challenges your view. Disconfirmation bias is the mechanism that makes beliefs self-reinforcing: confirming evidence gets waved through while disconfirming evidence gets blocked at the gate. Combined with confirmation bias (seeking confirming evidence), disconfirmation bias makes belief revision extraordinarily difficult.

When to use it

When evidence evaluation seems to produce conveniently one-sided conclusions; when disconfirming evidence is being dismissed on technicalities that confirming evidence wouldn't face; when intellectual honesty requires even-handed evidence evaluation; when Karl Popper's principle of falsifiability is relevant.

How it can help

Apply the SAME level of scrutiny to confirming AND disconfirming evidence. The practice: when evidence supports your position, scrutinize it AS IF it challenged your position—would you accept this methodology, sample size, and conclusions if the results were opposite? Conversely: when evidence challenges your position, evaluate it AS IF it supported your position—would you find these flaws if you agreed with the conclusion? The best thinkers actively seek and steelman disconfirming evidence rather than seeking reasons to dismiss it.