Research

Signals and Artifacts in Health Data

May 18, 2026

Why large datasets make skepticism more important, not less.

Negative Controls and Boring Results

May 7, 2026

The unglamorous checks are often where the science becomes credible.

Why Birth Order is Hard to Study

April 16, 2026

Notes from a phenome-wide scan of 10 million siblings.

What 200 Million Health Records Can Tell Us

April 14, 2026

Some thoughts on working with large-scale claims data.