The inability to self-diagnose bias

Litigants are guaranteed the right to an impartial jury-one that bases its judgment only on the evidence presented in the courtroom. The Supreme Court, as recently as in Skilling v. United States, has instructed courts on how to screen for potentially impartial jurors: simply ask them. The empirical presumption behind this directive is that jurors can self-diagnosis bias with sufficient accuracy. Across two initial experiments, we exposed 248 mock jurors to news articles that were either prejudicial to the defendant or irrelevant. We then gave jurors the admonitions and questions endorsed by the Supreme Court for identifying bias, prior to all of them watching a thirty-two minute condensed video of a civil trial. Results showed that jurors were unable to diagnose their own biases. Excluding those jurors who identified as either prejudiced against the defendant or simply unsure of their impartiality, jurors exposed to prejudicial news remained significantly more likely to rule against the defendant (odds ratio = 2.4, p = .004), and those that did so awarded eight-times larger median damages (p = .01) than those exposed to irrelevant news. Two subsequent experiments, with 751 mock jurors, replicated this finding. These experiments also explored the viability of a remedy suggested by the literature on the "bias blind spot": jurors who were asked whether the article they read would bias others-rather than themselves-were, indeed, significantly more likely to say yes. Yet, they were too eager: whereas individuals using self-diagnosis failed to sufficiently screen for their own biases, they too often identified others as being biased.