SSC Journal Club: MacIntyre On Cloth Masks

Early-COVID journal club on MacIntyre et al (2015), the only RCT of cloth masks (Vietnamese hospitals), which found cloth-mask wearers did WORSE than controls -- raising the possibility cloth masks are actively harmful, just as millions were about to make them. Scott carefully dismantles the inference: the control group wore some masks; the cross-study comparisons are invalid; everything is underpowered, so the 'x not significantly different from y not different from z' chains don't support the conclusion (you could even draw the opposite one). He concludes the prior that cloth masks should be neutral-to-helpful is too high for so weak a study to overturn, and he'd still wear them (especially for blocking OUTGOING particles, which the study ignored) -- while honestly flagging the genuine uncertainty ('the evidence says I'm wrong'). Companion to the Face Masks MMTYWTK.

Quality 71 · Strong. A careful, responsible single-study critique at a high-stakes moment, with sharp statistical reasoning (underpowered-chains; reversible inference) and honest uncertainty -- but narrower than the Face Masks deep-dive. Strong.

Claude’s paradigm shift 42 · Moderate. Moderate. A skeptical methodological critique of one study; the statistical points are sharp but not novel.

Real-world impact 3 · Moderate. A careful, responsible single-study critique at a high-stakes early-COVID moment — dismantling the inference that the one cloth-mask RCT showed masks are harmful (underpowered chains; reversible inference) with honest uncertainty. Real topical relevance to a live public-health question, but a narrow study critique informing the debate — modest RWI.