"Center for Discourse Control" 😂 Brilliant! (No, I didn't see what you did until you brought my attention to it. More proof that we look for patterns, which even though efficient, is often to our detriment.)
Whenever people see the word "statistics", they should immediately recall that well-known axiom about them.
Here is one question the corrupt CDC's statistics cannot answer:
"Are you aware that some countries (Sweden, Denmark, Norway, Romania, etc) and some States in USA - and some universities - don't have compulsory vaccination policies? Are those said jurisdictions care less about their people than others who do?"
The CDC study strikes me as nonsense. I went back to the original to check, and their sampling is faithfully represented in the numbers above.
Sample: patients hospitalised with COVID-like symptoms.
Grouping: previous infection vs previous vaccination
Measure: ratio that tested positive
Kulldorf briefly touches on the confounds, but let me point them out in more detail.
(1) There are actually six groups, but only two are reported. Previous COVID (90-179 days), other), previous vax (90-179 days, other), both COVID & vax, neither.
(2) The sample sizes are quite different between the two evaluated groups: 1020 vs 6328.
What has actually been tested is:
Given people hospitalised for COVID-like symptoms, what proportion of them actually had COVID, by past treatment status (if exactly one in last 90-179 days)?
And the correct conclusion is "people who have been vaccinated are half as likely to actually have COVID when hospitalised for COVID-like symptoms". And the researchers have then proceeded to draw conclusions about vax efficiency when they should be asking why vax subjects have twice as many "false positives".
As Kulldorf points out, answering this would require a lot more information about the sampling mechanisms, given that the groups differ in both quantity & demographic.
If they actually wanted to examine vax efficiency, there was a far more useful question they could have asked using the same source data: "Given patients who were COVID-positive (i.e. ignore all the noise about COVID-like symptoms), what was their past COVID-history? e.g. No vaccination or COVID, one or the other, or both.". Again, you'd need information about population sampling to make sense of the data, but it's at least asking the right question, rather than using "COVID-like symptoms" (i.e. false positives), which could well be a dependent variable, as a proxy.
MSM media Germany reports that This covid wave will be the last one. What relief:)
I was thinking if Israel keeps getting nailed by these waves and they have a small-ish population. How come they haven’t reached herd immunity yet?
"Center for Discourse Control" 😂 Brilliant! (No, I didn't see what you did until you brought my attention to it. More proof that we look for patterns, which even though efficient, is often to our detriment.)
Whenever people see the word "statistics", they should immediately recall that well-known axiom about them.
Here is one question the corrupt CDC's statistics cannot answer:
"Are you aware that some countries (Sweden, Denmark, Norway, Romania, etc) and some States in USA - and some universities - don't have compulsory vaccination policies? Are those said jurisdictions care less about their people than others who do?"
The word "autopsy" had me scrambling for the mouse-click. Alas, autopsy of a study, not a body.
The CDC study strikes me as nonsense. I went back to the original to check, and their sampling is faithfully represented in the numbers above.
Sample: patients hospitalised with COVID-like symptoms.
Grouping: previous infection vs previous vaccination
Measure: ratio that tested positive
Kulldorf briefly touches on the confounds, but let me point them out in more detail.
(1) There are actually six groups, but only two are reported. Previous COVID (90-179 days), other), previous vax (90-179 days, other), both COVID & vax, neither.
(2) The sample sizes are quite different between the two evaluated groups: 1020 vs 6328.
What has actually been tested is:
Given people hospitalised for COVID-like symptoms, what proportion of them actually had COVID, by past treatment status (if exactly one in last 90-179 days)?
And the correct conclusion is "people who have been vaccinated are half as likely to actually have COVID when hospitalised for COVID-like symptoms". And the researchers have then proceeded to draw conclusions about vax efficiency when they should be asking why vax subjects have twice as many "false positives".
As Kulldorf points out, answering this would require a lot more information about the sampling mechanisms, given that the groups differ in both quantity & demographic.
If they actually wanted to examine vax efficiency, there was a far more useful question they could have asked using the same source data: "Given patients who were COVID-positive (i.e. ignore all the noise about COVID-like symptoms), what was their past COVID-history? e.g. No vaccination or COVID, one or the other, or both.". Again, you'd need information about population sampling to make sense of the data, but it's at least asking the right question, rather than using "COVID-like symptoms" (i.e. false positives), which could well be a dependent variable, as a proxy.
Flipping a weighted coin?
You might be on to something here.