The hidden question behind social-science measurements
Social-science findings often come with a follow-up question: do they apply beyond the setting studied? MIT political scientist Naoki Egami calls that question external validity. He studies the mathematical and computational methods behind empirical problems, including how AI tools are used in research and how accurate they are. His advice is to combine technical theory with interesting real-world problems: theory alone can detach from useful questions, while problem-first work can become ad hoc.
AI-generated discussion · Different perspectives
Take
AI belongs in the measurement workflow only if researchers test its accuracy and account for its tendencies.
Pushback
A technically elegant method can still miss whether a finding travels from one election or country to another.