I agree: it is wrong to infer causal stories for the individual based on population level statistics.