Professor Edward (Ted) Miguel delivered the keynote lecture at the inaugural Berkeley-Stanford Veridical Data Science Workshop, a pioneering event dedicated to advancing veridical (truthful) data science (VDS). The workshop aimed to promote reproducible and reliable data analysis, foster trustworthy data science, machine learning, and artificial intelligence, and build a community of researchers committed to these principles.

In his lecture, Ted tackled the pervasive issue of publication bias and the challenges of research transparency in the social sciences. Drawing on foundational scientific norms articulated by sociologist Robert Merton—universalism, communality, disinterestedness, and organized skepticism—Ted highlighted the stark gap between these ideals and current practices. He shared insights from his own research in Africa, where pre-analysis plans and open science practices have emerged as critical tools to combat selective reporting.

Ted called for a cultural shift in the scientific community, urging researchers to prioritize truth over career advancement and to embrace transparency as a core value. He introduced the RARE framework (Reporting All Results Efficiently), advocating for full reporting of all hypotheses, even those with null or unpopular results, to combat the “file drawer problem.” Ted’s message was clear: while institutional reforms are essential, lasting change will require a cultural transformation within the scientific community, one that values integrity and truth over sensationalism.