15% is a big enough difference that such a coarse methodology is fine to draw preliminary conclusions. Of course you could instrument better, but this is a writ

15% is a big enough difference that such a coarse methodology is fine to draw preliminary conclusions. Of course you could instrument better, but this is a writer at The Guardian, of whom I don't expect the technical prowess to do so, and the article (or what I skimmed of it) doesn't overstate his findings.