Over 400,000 people visited Farnam Street last month to learn how to make better decisions, create new ideas, and avoid stupid errors. With more than 100,000 subscribers to our popular weekly digest, we've become an online intellectual hub. To learn more about we what do, start here.

Confirmation Bias in Criminal Investigations

False convictions are rarely intentional. Even in the worst cases, when police or laboratory technicians deliberately tamper with evidence or commit perjury, they do not intend to frame an innocent person; they believe that they have the right person and that they are aiding the pursuit of justice. Motivation to catch the culprit can also lead investigators to become overly committed to a theory of the case.

Cops are just like the rest of us. The problem is once we are committed to a theory, we interpret ambiguous or even contrary evidence in a way consistent with that theory.  At some point investigations change from figuring out what happened to proving it. This commitment affects not only interpretations of evidence but also the rest of the investigation. One psychological process by which investigators (and most of us) reach the wrong conclusion is confirmation bias.

Of particular relevance to decision making in criminal investigations is the effect of simply forming a hypothesis. The hypothesis becomes a framework that guides the decision maker to interpret and search for new information in a way that makes it fit the hypothesis.

Hypothesis driven analysis leads investigators to unconsciously assume that the hypothesis in question is true, and search for evidence accordingly. They are not completely indifferent to contrary information, but assuming the truth of their hypothesis causes them to undervalue conflicting evidence or to be less likely to notice it in the first place. This can undermine accuracy by leading the hypothesis tester to overlook evidence that a hypothesis is false or that an alternative is plausible.

The research below brings into question hypothesis based analysis in disciplines where answers are not always easily distinguishable as right or wrong and information is materially incomplete. “Articulating a hypothesis early affected not only how participants thought about the case, but also what course of action they advocated.”

According to the research below, you can mitigate the effect of confirmation bias in hypothesis based analysis by “taking the additional step of articulating why it might be wrong” (Participants who listed the evidence both for and against their hypothesis did not differ on any measure from participants who articulated no hypothesis at all.) Contrary to what most people think, considering alternative hypotheses did not not work to reduce confirmation bias.

Abstract:

Confirmation bias is the tendency to bolster a hypothesis by seeking consistent evidence while disregarding inconsistent evidence. In criminal investigations, preference for hypothesis-consistent information could contribute to false convictions by leading investigators to disregard evidence that challenges their theory of a case. Two studies examine factors that influence confirmation bias in criminal investigations. In study 1, participants (N = 108) who stated hypotheses early in their review of a mock police file showed bias in seeking and interpreting evidence. In study 2 (N = 109), asking participants to consider why a hypothesis might be wrong remedied bias, but asking them to generate additional hypotheses did not. Implications for improving accuracy of investigations and suggestions for future research are discussed.

Source: Confirmation Bias in Criminal Investigations (link)