Great new study about superspreading of #SARSCoV2 confirming our results from in @Eurosurveillanc from January that the dispersion parameter k ~ 0.5 is higher compared to SARS, MERS and Ebola (where k ~ 0.1 - 0.2) but lower than for influenza (where k ~ 1) https://www.eurosurveillance.org/content/10.2807/1560-7917.ES.2020.25.4.2000058 …https://twitter.com/msy_lau/status/1296579558062383105 …
-
I agree that your earlier study was important in estimating R0, but isn't it too much to say you got k~0.5 from your study when 90% range is 0.014–6.95? I guess you could get a different median with a different prior.
I was referring to medians or point estimates (same for other infections). That’s why these additional studies are so important to better understand how much overdispersion we can truly expect from #SARSCoV2 and by what factors it is influenced.
-
-
My point is, point estimates are not useful when uncertain range is too huge to draw a conclusion. Eg. your flat prior (0.01-10) on logk has a median of 0.3, but this tells nothing. And from the posterior very close to prior I'd conclude the data didn't have much info on k.
-
Agree we need more studies from various approaches. Currently I think most studies suggest it's almost certainly <1, likely to be 0.05-0.5, but we're not yet sure how different settings/study designs could account for variations in these estimates.
Koniec rozmowy
Nowa rozmowa -
Wydaje się, że ładowanie zajmuje dużo czasu.
Twitter jest przeciążony lub wystąpił chwilowy problem. Spróbuj ponownie lub sprawdź status Twittera, aby uzyskać więcej informacji.