Artificial Intelligence in Population Research Involving Families with Young Children: A Scoping Review

Joyce Lee, The Ohio State University College of Social Work
Hunmin Cha, The Ohio State University College of Social Work
Amy Xu, The Ohio State University College of Social Work
Yujeong Chang, The Ohio State University College of Social Work

Population research involving families is increasingly exploring the application of artificial intelligence (AI). However, there is limited understanding of how AI is being employed to advance the field. There is a need to synthesize relevant evidence to inform demographers, as well as family policymakers. This scoping review systematically examined the literature to identify AI use in studying families, especially those with young children. Nine databases were searched, with the following eligibility criteria: (1) is a peer-reviewed journal article; (2) published 2014–2024; (3) written in English; (4) used AI methodology; (5) included families raising children 0–5 years; and (5) was quantitative in analysis. Of the 10,022 studies identified, 21 met the eligibility criteria. Results from synthesizing evidence showed that most studies focused on maternal and child health (MCH) outcomes in low- and middle-income countries (LIMCs) (60%). Majority of the studies used AI to identify the most important predictors of MCH, with random forest performing as the best model. Only one study directly mentioned the ethical use of AI. Although population research using AI can inform family policies aimed at promoting MCH, thoughtful consideration of AI ethics and fairness is needed to prevent the negative social impacts of AI on marginalized populations.

Keywords: Data and Methods, Children, Adolescents, and Youth, Health and Morbidity, Families, Unions and Households

See extended abstract.