[sradman]

I think the misinterpretation is mainly by the media and the general public. When most of your population is vaccinated and most of the serious disease is in the unvaccinated, you need to report by rate and vaccination status as the Ontario Science Table [1] now does (rather than absolute case numbers).

Simpson's Paradox is more of a data artifact that you have to be aware of. I didn't know about this statistical anomaly before but the takeaway is that if you see a effectiveness percentage decrease from 97% to 77% then you should also check that the value in each age cohort because each individual cohort may surprisingly be above 90%. The Israeli data might be fine but I want to see the "age corrected" range rather than a single effectiveness number.

The bottom line is that we will get good data moving forward from the Israeli 3rd dose program with other quality data sets soon to follow from the U.S., UK, Canada, Singapore, etc.

What we have not yet seen is any good evidence that the vaccinated are contributing to spread, though in the Fauci interview he indicated that the R(t) in the unvaccinated was non-zero. This is an important question, IMO.

[1] https://covid19-sciencetable.ca/ontario-dashboard/#riskbyvac...

[tguvot]

i know. as i said he miscalculated efficiency for age cohorts. israeli data for most age ranges is 45 to 25 percent lower than his calculated 95%-99%+ efficiency. So maybe, as paradox it's nice, but as calculation go, they suck. and now everybody running around with this site as proof that there is no need in booster because efficiency is still 95%

[sradman]

Do you have a link to the full age cohort effectiveness numbers? I didn't think these numbers were published yet, if they are or don't exhibit Simpson's Paradox then the question is what accounts for the discrepancy with the UK and Canadian data and the immunological lab research done by Shane Crotty and Rockefeller [1].

[1] https://www.microbe.tv/twiv/twiv-717/

[tguvot]

There were presentation with numbers stratified by age, but I can't find it now. Unfortunately I still can't figure out how to track down everything that ministry of health releases over there. Closet one that I found is this one https://www.gov.il/BlobFolder/reports/vpb-12082021/he/files_... . Look at slide 7.

Also, if you we are talking about Simpson's Paradox, we need to go deeper. As you can see at same slide, vaccine efficiency going down, the further you get away from second shot. Hence, age cohort effectiveness is useless. You need age/vaccination time frames to judge real efficiency