Week 11 Round: Reading the mortality data

A Round at a desk. You take the famous claim apart, find where each number came from, and write the version you would defend.

13 May 2027

Bring a laptop. This Round has no site.

Find the claim in the wild. A news article, a government campaign, a charity page, a minister’s speech, a LinkedIn post. Something that cites the loneliness figure at the public. Save the exact wording.

Trace it back. Follow it to the paper and find the actual number. You will usually find one of three things: the figure is quoted correctly, the effect size has been converted into something snappier along the way, or the chain of citation does not reach a paper at all.

Write the two comparisons side by side. What Holt-Lunstad and colleagues compared, and what the headline compared. The gap between those two sentences is the exercise.

Three confounds. Name them, and then go back to the paper and find out which ones the authors addressed and how. They are not naive about this; the discussion does real work, and the honest criticism is narrower than the lazy one.

Then write the defensible version, in no more than four sentences, as if you had to say it to someone who would push back. Not a hedge that says nothing, and not the headline. This is the exact skill your Place Audit needed and the one your week 12 proposal will need, because you are about to argue that somewhere should exist and someone will ask you what the evidence is.