Amazing talk from @ta_broderick on whether dropping a little data can change conclusions of a statistical analysis. The delivery was a masterclass on patient teaching of new and essential concepts that leaves us all with new insights and lines of research.

→ View original post on X — @shakir_za,