Last month, Imperial College’s React study claimed that new cases of Covid were static or even rising slightly. This contradicted the figures for confirmed new cases, obtained through the Test and Trace system, which had shown a sharp fall in new cases from the second week of January onwards. Given that React tests a randomised sample of the population to arrive at an estimate for prevalence of the disease — and is therefore not capable of being skewed by the number of tests being performed — some were more inclined to believe that this, and not the number of confirmed...