Features of Evidence
Evidence ambiguity and conflict can amplify cognitive biases by allowing selective interpretation, while the amount and completeness of evidence can influence confidence and support for hypotheses regardless of diagnostic value.
THE IDEA
Features of Evidence
Evidence Ambiguity and Conflict. Evidence with ambiguous or conflicting implications may be likely to amplify the impact of other features on perceived support. For example, as mentioned earlier in the discussion of influences on subadditivity, Koehler (2000) found that the advantage afforded a focal hypothesis that is pitted against all its alternatives taken together is more pronounced in the face of conflicting evidence than in the face of consistent evidence. In short, unpacking effects may be enhanced by ambiguity or conflict in the evidence.
One possible explanation for this pattern is that evidential ambiguity or conflict allows the judge latitude in evaluating support, making it possible to interpret the evidence in a manner that fits the judge’s desires or expectations. Numerous findings from the social psychological literature are consistent with this view. Lord, Ross, and Lepper (1979), for example, presented mixed evidence regarding the deterrent effectiveness of capital punishment to proponents and opponents of capital punishment. Both groups tended to discount whichever aspect of the evidence did not support their preferred hypothesis. Exposure to the same body of conflicting evidence led each group to believe more strongly in its own position. The conflict in the evidence presumably allowed both groups to be selective about the conclusions they drew from it.
Amount and Completeness of Evidence. The perceived amount of evidence available may also influence judgment, in a manner independent of its diagnostic value. Provision of highly detailed information may convey a sense of confidence, even if the additional detail is not particularly relevant or useful (e.g., Oskamp, 1965). By contrast, knowing that certain pieces of information exist but are not available may inappropriately reduce the support for hypotheses under consideration (Fox & Tversky, 1995). Finally, the amount of information available may also typically be correlated with the level of conflict in the evidence: With enough information, some of it is bound to support each hypothesis under consideration (e.g., Downs & Shafir, 1999).
When to use it
Evaluating scientific studies with conflicting results, such as in nutrition or medicine.
Making business decisions when market data is ambiguous or contradictory.
Assessing legal evidence in a court case where both sides present conflicting testimony.
Reviewing a project proposal with both supporting and opposing data.
FROM THEORY TO EVERYDAY LIFE
In practice
Capital Punishment Debate
Lord, Ross, and Lepper (1979) presented mixed evidence about the deterrent effect of capital punishment to both proponents and opponents. Both groups discounted evidence that contradicted their initial position, and after exposure to the same conflicting evidence, each group became more convinced of their original stance. Ambiguous evidence reinforced each group’s existing bias.
Unpacking Effects with Conflicting Evidence
Koehler (2000) found that when evidence is conflicting, unpacking a focal hypothesis into its alternatives strengthens the perceived difference between the focal hypothesis and its alternatives. This shows how conflict amplifies the unpacking effect.
Detailed Information and Overconfidence
Oskamp (1965) showed that providing clinicians with highly detailed case information increased their confidence in their diagnoses, even though the additional detail was not diagnostically relevant. More information increased certainty without improving the basis for the diagnosis.
A MOMENT TO REFLECT
Put it to work.
Can you recall a time when you encountered mixed or ambiguous evidence? How did your prior beliefs influence your interpretation?
Think of a decision you made where you had a large amount of information. Did the quantity of information affect your confidence, even if some of it was irrelevant?
How might you avoid the pitfalls of biased evidence evaluation when making an important decision?
References
- Subadditivity and conflicting evidence — Koehler (2000)
- Biased assimilation of mixed evidence on capital punishment — Lord, Ross, & Lepper (1979)
- Overconfidence in case-study judgments — Oskamp (1965)
- Availability and support for hypotheses — Fox & Tversky (1995)
- Conflict and information in judgment — Downs & Shafir (1999)
CONTINUE THE CONVERSATION
Source texts
- Heuristics and Biases The Psychology of Intuitive Judgment (Thomas Gilovich, Dale Griffin, Daniel Kahneman)