Pretend play is a ubiquitous learning tool in early childhood, enabling children to explore possibilities outside of their current reality. Here we demonstrate how pretend play can be leveraged to empower girls in scientific domains. Four- to seven-year-old children (N = 240) played a challenging science activity in one of three conditions. Children in the Exposure condition heard about a successful gender-matched scientist; children in the Roleplay condition pretended to be that scientist; children in the Baseline condition did not receive information about the scientist. Girls in the Roleplay condition, but not in the Exposure condition, persisted longer in the science activity than girls in the Baseline condition. Furthermore, pretending to be the scientist equated girls’ persistence to that of boys. These findings suggest that pretend play of role models is an effective strategy to motivate young girls in science and may help reduce gender gaps from their roots.

Over-engaged parenting is on the rise: Parents in the U.S. are increasingly completing developmentally appropriate tasks for young children, which relates to reduced child persistence. What can help parents to step back? Across two pre-registered studies, we test the novel hypothesis that parents step back when they prioritize their child’s learning. In line with this hypothesis, parents of 4-5-year-olds (N = 77) reported taking over less on tasks from which they think their child learns more, which was more often the case for academic (versus non-academic) tasks (Study 1). Study 2 found causal evidence for this hypothesis: Emphasizing children’s learning in a novel, non-academic context reduced parents' (N = 60) taking over actions by half compared to a control condition. This work shows that parents take over less when they think that their young child is learning and opens new avenues for interventions to reduce over-engaged parenting.