Debiasing
Techniques that reduce or correct for cognitive biases in judgment and decisions.
What it means
Debiasing covers the strategies used to counteract systematic errors of judgment, falling into two broad families: modifying the person (training in statistical reasoning, considering the opposite, taking the outside view, slowing down to engage deliberate thought) and modifying the environment (changing how information is presented so the bias is never triggered). Evidence is mixed and effortful person-level debiasing often fades or fails to transfer, which is part of why choice architecture and boosting gained favor — it can be easier to redesign the situation than to retrain the mind. Effective techniques tend to be specific, practiced, and supported by cues at the moment of decision rather than one-off warnings. The deep nuance is that knowing about a bias rarely cures it (the bias blind spot persists). It matters because better decisions sometimes require fixing the judge, not just the choice.
What the evidence shows
The strongest single result comes from a serious game built for US intelligence analysts: one session cut susceptibility to biases like confirmation bias and the bias blind spot by medium-to-large amounts, and the effect still held two months later (Morewedge et al., 2015). A follow-up put graduate students through an unannounced business case and found trained players about 19% less likely to pick the hypothesis-confirming, inferior option, rare evidence that training transfers to a realistic task (Sellier et al., 2019). But a 2025 meta-analysis of 54 trials found only a small pooled effect overall, with some biases, such as the representativeness heuristic, barely budging. The honest reading is that some techniques work, gains are usually modest, and durability and transfer are the exception rather than the rule.
Consider the opposite, and its limits
The most reliable technique is disarmingly simple: before committing to a judgment, ask what would make the opposite true. Lord, Lepper and Preston (1984) showed that explicitly considering the opposite corrected biased assimilation better than merely instructing people to be fair, because the bias springs from a one-sided memory search and the prompt forces retrieval of the evidence that search skipped. It reliably dampens anchoring, overconfidence and hindsight, where a single counter-thought rebalances an estimate. It works less well against motivated reasoning, and asking for many counterarguments can even backfire: when the counterarguments are hard to bring to mind, that difficulty itself gets misread as evidence they do not exist. The same move that helps on a number can harden a person on an identity.
Why the mind resists correction
A recurring finding explains why person-level debiasing disappoints: knowing a bias exists does little to stop it. People readily spot distortions in others' reasoning while judging their own as objective, the bias blind spot, and cling to that self-assessment even after the mechanism is explained to them (Pronin, Lin & Ross, 2002). Biases also operate below awareness and in real time, so a warning delivered in a training room rarely fires at the moment of a live decision. Effects that do appear tend to decay and to stay narrow, tied to the exact task practiced rather than generalizing to new ones. This is why the field increasingly treats fixing the judge as costly and unreliable, and reaches first for changes to the situation.
Two levers, and using them in practice
Debiasing pulls two levers: retrain the person or redesign the decision. Because retraining is effortful and fragile, the durable wins usually come from building the correction into the process so no one has to remember it. Structured protocols do this: fixed scoring rubrics, independent judgments recorded before any discussion, reference-class forecasting for estimates, and pre-mortems that force people to imagine the failure. Where person-level training is used, it works best when it is specific, practiced with feedback, and cued at the point of decision rather than delivered as a one-off lecture. A good practical test is whether the technique still operates once the person has forgotten they were ever trained, which is why environmental fixes, once installed, tend to outlast the memory of the workshop.
Examples
Forecasters trained to 'consider the opposite' and to log predictions for calibration feedback make measurably better-calibrated judgments — a debiasing routine that sticks because it is practiced.
A hiring panel scores every candidate on the same fixed questions, independently, before anyone speaks — redesigning the situation so a strong first impression never gets to steer the room.
A team estimating a new build looks up how long their last ten projects actually took instead of imagining this one: the outside view, applied at the moment the estimate is made.
A radiologist, trained to ask 'what else could this be?' before signing off, runs through the alternative diagnoses that the first, most obvious reading had quietly crowded out.
An investment committee writes its buy thesis and its exit triggers before purchase, so a later run of good news cannot silently rewrite the original case for holding.
First described in Fischhoff (1982); Larrick (2004).
Key references
- Swaryandini, G., Graham, J., Griffith, S., Grilo, V., Ruzzante, F., Zhang, X., ... & Noetel, M. (2025). Systematic review and meta-analysis of educational approaches to reduce cognitive biases among students. Nature Human Behaviour, 9(12), 2510-2538. doi.org/10.1038/s41562-025-02253-y
- Sellier, A.-L., Scopelliti, I., & Morewedge, C. K. (2019). Debiasing training improves decision making in the field. Psychological Science, 30(9), 1371-1379. doi.org/10.1177/0956797619861429
- Morewedge, C. K., Yoon, H., Scopelliti, I., Symborski, C. W., Korris, J. H., & Kassam, K. S. (2015). Debiasing decisions: Improved decision making with a single training intervention. Policy Insights from the Behavioral and Brain Sciences, 2(1), 129-140. doi.org/10.1177/2372732215600886
- Pronin, E., Lin, D. Y., & Ross, L. (2002). The bias blind spot: Perceptions of bias in self versus others. Personality and Social Psychology Bulletin, 28(3), 369-381. doi.org/10.1177/0146167202286008
- Lord, C. G., Lepper, M. R., & Preston, E. (1984). Considering the opposite: A corrective strategy for social judgment. Journal of Personality and Social Psychology, 47(6), 1231-1243. doi.org/10.1037/0022-3514.47.6.1231