Scaling Developmental Science with PsyNet

Building tools and paradigms to run high-powered, cross-cultural developmental experiments online.

Developmental science faces a fundamental paradox: infants are among the most important populations to study yet also one of the most challenging to test. Early cognitive development shapes lifelong learning, individual differences, and the cultural transmission of knowledge across generations — making it essential for building theories of human cognition. Yet infants cannot verbalize their thoughts, require substantial caregiver coordination, have fleeting attention spans, and mature rapidly, severely constraining re-testing opportunities. These inherent difficulties have created methodological bottlenecks that limit what we can learn about the developing mind.

In a joint grant, we are extending PsyNet — our scalable online experimentation framework — to meet the specific needs of developmental research. The goal is to achieve greater granularity, efficiency, and global reach in studies with infants and children: running experiments that manipulate core cognitive parameters to improve the interpretability of results, including direct extensions of recent large-replication reports.

PsyNet architecture for developmental research
PsyNet for developmental research. (A) Modular Python components provide reusable building blocks for experiment design. (B) One-click deployment to cloud servers with secure data management. (C) Integration with citizen science platforms, lab recruiters, and crowdsourcing with automated payment. (D) Quality control through video recording and AI verification for child participants.

What we do

  • Extend PsyNet with developmental workflows: caregiver coordination, child-friendly interfaces, video recording, AI-based age and identity verification, and adaptive protocols.
  • Enable high-powered studies deployable across multiple languages, cultures, and populations, reaching populations historically excluded from developmental research.
  • Run empirical studies targeting core cognitive parameters — including infant music perception, statistical learning, and child-directed speech — to sharpen theoretical interpretability.
  • Extend and challenge existing findings from large replication studies, contributing to a more robust and generalizable developmental science.
Infant musical preference mapping paradigm
Infant musical preference mapping. (A) Touchscreen swipe paradigm: infants control music playback by swiping; swipe latency measures preference. (B) Music sampling via Spotify API across 30 cultures, mid-popularity range, AI-classified and native-speaker validated. (C) Infants prefer children's songs over adult songs; adults show the opposite pattern. (D) AI-based feature analysis of preferred music identifies relevant acoustic dimensions.

Why it matters

The questions developmental science asks — how minds form, how culture is transmitted, why individuals differ — are among the most fundamental in cognitive science. Yet the field’s methodological toolkit has lagged behind its ambitions. By building infrastructure that enables large-scale, cross-cultural, and theory-driven experimentation with infants and children, this project aims to transform how developmental science is conducted: generating more inclusive datasets, sharper statistical inference, and genuinely generalizable theories of early human cognition.