Precursors of young adults' world beliefs across cultures: A machine learning approach.

Abstract

Primal world beliefs ("primals") capture individuals' basic understanding of what sort of world this is and are strongly associated with a wide range of behaviors and outcomes, yet we have little understanding of how primals come to be. This study used a data-driven machine learning approach to examine what individual, parenting, family, and cultural factors in childhood best predict young adults' beliefs that the world is Abundant, Alive, Enticing, Good, Hierarchical, Progressing, and Safe, contributing a long-term longitudinal perspective to the nascent work in developmental science on primal world beliefs ("primals"). Participants included 770 young adults from eight countries (Colombia, Italy, Jordan, Kenya, Philippines, Sweden, Thailand, United States). During childhood, participants and parents reported on 76 factors available as potential predictors of primals. Factors at individual, parenting, family, and cultural levels all had some predictive value in relation to specific primals, but no single factor or cluster of factors was predictive of all primals. Developmental pathways to perceiving the world as Abundant, Alive, Enticing, Good, Hierarchical, Progressing, and Safe are not uniform. The current data-driven approach successfully unearthed several promising leads for developmentalists to probe in further research.