> I wonder if you can even reliably measure a 0.4 IQ point loss. Which is why one never ever trusts an article or paper with a bare number as claim. Confidence

> I wonder if you can even reliably measure a 0.4 IQ point loss. Which is why one never ever trusts an article or paper with a bare number as claim. Confidence intervals, p-values, standard deviation, whatever it is, there are ways to tell whether it's significant or what the odds are that it's noise. I haven't read the article btw so I don't know if it does, but since we're talking about a bare number, the answer to your question is that it indeed by itself doesn't tell us anything about measurability even if it had been five points.