Build a decision habit and check its effect

A useful next step is a small decision habit that can be observed. Write the purpose and criteria before comparing options. Identify which evidence supports each judgment. When a description rests on familiarity, ask what behavior or result it refers to. Seek a plausible alternative explanation before treating a first impression as settled. These practices do not eliminate all bias, but they make reasoning more available for review.

Choose a decision you can influence without exceeding your role. A meeting agenda, a review of draft materials, or a choice about who receives information can provide a manageable starting point. Formal employment, procurement, or eligibility decisions may have required processes and responsible offices; this learning exercise does not replace them. The same principle applies: relevant standards should be clear, evidence should be checked consistently, and departures should have a reason.

Evaluate the habit using the decision itself. Did the group identify missing information? Did it apply the criteria consistently? Did an initially overlooked option receive a fair examination? Do not use a later IAT result as a certificate that the process improved. Reflection and action are related, but they are not the same evidence. Keep the review focused on decisions and access rather than building a file of individual attitudes. A modest, well-examined change is more credible than a declaration that everyone has completed bias training and the issue is solved.

Fictional case: A communications team chooses which draft to use for a public notice. The lead usually selects the version from the most familiar writer. For the next review, the team agrees on accuracy, readability, and accessibility before comparing drafts. It identifies a strong section in a previously overlooked version and combines the best-supported elements. The review notes describe which criteria changed the decision. They contain no IAT scores and make no claim that the team’s biases have disappeared.

Write a one-page decision record with purpose, criteria, evidence, missing information, and a reasoned conclusion. Test it on the fictional notice or a non-sensitive decision you are authorized to review. At the next use, check whether the record exposed an unsupported assumption. Revise the tool if it adds work without improving the reasoning.

Which repeated decision in your work could become more transparent? What result would convince you that the check helped rather than merely added a form?