Group attribution bias

Group attribution bias occurs when an outcome is only true for some individuals but when that outcome is generalized for an entire group.

Example: A model that selects the candidates only with a master’s degree.

How to prevent group attribution bias

To prevent group attribution bias, make sure that you have an understanding of

  • The outcomes of your model on different groups.

  • Under which circumstances you can and cannot generalize your model.

Was this page helpful?
Yes No