Debiasing Decision Making Through Observational Learning

We find that observational learning, indirectly acquiring attitudes, norms, and skills by observing others, can be an effective debiasing intervention. Participants in a three-hour laboratory experiment (Experiment 1; N = 277) who received an observational learning-based training intervention exhibited a large reduction (overall d = 1.86) from pretest to posttest in their propensity to exhibit three cognitive biases––anchoring, overgeneralization of the representativeness heuristic, and social projection––more than did controls and participants who received an information-based intervention. Participants in an online experiment (Experiment 2; N = 494) who received an observational learning-based training intervention exhibited greater improvements in judgment and advice taking (d = .24) than did controls, and no less than did participants who received an information-based intervention. The combination of both observational learning and information-based debiasing interventions had a larger debiasing effect than either intervention alone, suggesting that observational learning can have unique debiasing effects.