// learn.shawon.ch / research-methods TRACK
← index

Study Track

Research methods

A working filter for any health, medical, psychological, or social-science claim you meet — in a paper, a podcast, or a supplement ad. It teaches you to place evidence on a ladder, see how each kind of study can fool you, read the numbers honestly, account for the ways the literature itself is skewed, and decide — in proportion to the stakes — whether something is strong enough to act on. Built to be used on real decisions, not just read.

Curriculum · 6 parts, 18 lessons

Part I · Foundations

How findings get distorted on the way to you, how to rank the kind of evidence before you judge its quality, and the single trap — correlation read as cause — under most bad health claims.

Part II · Study designs

What each kind of study can and can't prove, and the threats to each — from passive observation and the biases that fool it, up to the randomized trial that earns the word "caused."

Part III · What they measured, and on whom

A moved marker is not a better life, and a result in one group may not transfer to you. The two questions — was it the right outcome, and does it generalize — that decide whether a finding is even about what you care about.

Part IV · The numbers

Reading the statistics honestly — significance vs size, the base rate a scary risk hides, and why a "significant" result can still be more likely false than true.

Part V · Why the literature is skewed

The systemic reasons the published record overstates what's true — selective reporting, flexible analysis, who paid for it — and how reviews try, and sometimes fail, to correct for it. The part that matters most for supplements and influencers.

Part VI · Putting it to work

Turning all of it into a routine — how to read a paper, a 60-second triage for hype, and a way to decide, in proportion to the stakes, whether to act — including an honest self-experiment.

Learn · Shawon Chowdhury · back to all subjects