SSN Memo

Employing Artificial Intelligence to Measure and Improve Representation in Children’s Books

Policy field

Connect with the author

University of Chicago

The below is an excerpt from a memo written by Anjali Adukia.

Research shows that the educational resources instructors choose can greatly affect how children perceive their world. In particular, the way that people are represented within books can contribute to children’s understanding about what social and professional roles they and others can or cannot inhabit. Given persistent racial and gender inequality in society and the importance of identity and representation in driving beliefs, aspirations, and academic effort and outcomes, these materials offer a key means to either address, perpetuate, or entrench core societal inequalities.

Given their impact, it is critical for those responsible for publishing and sharing these works to understand how their content stacks up. New software used for systematic analysis of images can be applied alongside established text analysis methods to analyze children’s books, and gain a deeper understanding of what children currently have access to and what changes are needed going forward.