You can check out the paper and online supplement to learn more about the studies, the interventions, and important limitations and future directions for this work. The data, materials, and scripts are available here: osf.io/m2a9w. The open access preprint is available here: doi.org/10.31234/osf...

For example, providing participants with clear, actionable evaluation strategies seemed to produce strong effects in the context studied. On the contrary, interventions that gave participants relatively generic instructions to avoid biases were not as impactful.

What can we learn from these findings? Many features were found to be shared amongst the interventions that were relatively more or less effective in this large-scale study (30 interventions, N>20,000).

Across four studies, we identified four interventions which reliably reduced overall attractiveness discrimination in our outcome measure, the Judgement Bias Task, in which participants selected candidates who varied in qualifications and physical attractiveness.

We received 30 interventions (thank you to all collaborators!) and tested them against a single control condition. This procedure allowed us to gain knowledge on the relative efficacy of each strategy using the same sample source and our outcome measure of discrimination.