In the initially published piece, the authors did not post the raw number of seropositive cases in their study, preventing anyone from computing standard statistical analyses of their results. Many experts also objected to the number of statistical regressions run to pull signals out of their very complex intervention.
Recently, the authors finally released their code and data.
Following the release of their data, statistical analyses were able to be conducted. There were 1,106 symptomatic individuals confirmed seropositive in the control group and 1,086 such individuals in the treatment group. The total difference between the two groups was miniscule: only 20 cases out of over 340,000 individuals over a span of 8 weeks.
In a post by one author who ran an analysis following the release of the data, he described several other issues with the study:
“The intervention was highly complex and included a mask promotion campaign and education about other mitigation measures including social distancing. Moreover, individuals were only added to the study if they consented to allow the researchers to visit and survey their households.
“There was a large differential between the control and treatment groups here, with 95% consenting in the treatment group but only 92% consenting in control. This differential alone could wash away the difference in observed cases. Finally, symptomatic seropositivity is a crude measure of covid as the individuals could have been infected before the trial began.”...<<<Read More>>>...
