This paragraph seems to concisely summarize what NYT got wrong in their analysis:
> Age is a common confounder in public health research, and COVID-19 is no exception. The mortality burden of COVID-19 is not randomly distributed across age groups. Indeed, age appears to be the “strongest predictor of mortality” from COVID-19, with one’s risk of death increasing exponentially with age. According to CDC figures, the oldest populations experience a rate of death 570 times higher than that of the youngest populations. This is precisely why older populations were vaccinated first; we knew that prioritizing this population would have the most dramatic effect in curtailing hospitalizations and deaths. Yet the crude county-level analysis reported in The New York Times failed to adjust or account for age at all.
The Times article didn't get anything wrong, COVID is hitting red states harder. You can't just blame the older populations in red states, the vaccination rate is higher in blue states vs similarly aged demographics in red states.
For some, truthful attribution of causality matters.
> The mortality burden of COVID-19 is not randomly distributed across age groups. Indeed, age appears to be the “strongest predictor of mortality” from COVID-19, with one’s risk of death increasing exponentially with age. According to CDC figures, the oldest populations experience a rate of death 570 times higher than that of the youngest populations.
So if 0% of those 65+ in one state have their vaccination compared to 100% in another state, we can still reliably blame this on states with older populations?
> So if 0% of those 65+ in one state have their vaccination compared to 100% in another state
You can prove anything if the antecedent is false and I am not interested in bogus arguments.
1. Mortality from COVID19 (y) is strongly associated with age (z).
2. Voting Republican/Trump (x) is strongly associated with age (z).
3. Therefore, one cannot simply look at the association between x and y without controlling for z.
This is what I spent decades trying to teach in Intro Stats and Econometrics. Ignoring a confounding variable because it is a convenient way to score a political point is in line with propaganda, not a quest for honest inquiry.
You are arguing that a vaccinated 80 year old can be compared to an unvaccinated 80 year old. A 80 year old Republican is less likely to get the vaccination. It's pretty simple.
I am afraid I cannot continue to discuss in good faith because you are making up stuff I did not say.
For the record, the whole point of the substack post is that the NYT article did not do anything close to comparing a vaccinated 80-year olds in a Blue states to unvaccinated 80-year olds in Red states.
In fact, the article makes no attempt to control for age at all.
No, the article assumes there is only one possible factor that might determine whether someone might die from COVID19 ... "Is that person one of us or one of them?"
Journalism is a dumpster fire. If the NYT wants to be considered reputable by more than 20% of the population, it needs DOIs to all its publications, cite-backs, and archive history and tracked changes in a transparent changelog.
As of right now, it's a blog by BAs (at best) with frequent stealth edits.
8 comments
[ 0.17 ms ] story [ 35.3 ms ] thread> Age is a common confounder in public health research, and COVID-19 is no exception. The mortality burden of COVID-19 is not randomly distributed across age groups. Indeed, age appears to be the “strongest predictor of mortality” from COVID-19, with one’s risk of death increasing exponentially with age. According to CDC figures, the oldest populations experience a rate of death 570 times higher than that of the youngest populations. This is precisely why older populations were vaccinated first; we knew that prioritizing this population would have the most dramatic effect in curtailing hospitalizations and deaths. Yet the crude county-level analysis reported in The New York Times failed to adjust or account for age at all.
> The mortality burden of COVID-19 is not randomly distributed across age groups. Indeed, age appears to be the “strongest predictor of mortality” from COVID-19, with one’s risk of death increasing exponentially with age. According to CDC figures, the oldest populations experience a rate of death 570 times higher than that of the youngest populations.
You can prove anything if the antecedent is false and I am not interested in bogus arguments.
1. Mortality from COVID19 (y) is strongly associated with age (z).
2. Voting Republican/Trump (x) is strongly associated with age (z).
3. Therefore, one cannot simply look at the association between x and y without controlling for z.
This is what I spent decades trying to teach in Intro Stats and Econometrics. Ignoring a confounding variable because it is a convenient way to score a political point is in line with propaganda, not a quest for honest inquiry.
For the record, the whole point of the substack post is that the NYT article did not do anything close to comparing a vaccinated 80-year olds in a Blue states to unvaccinated 80-year olds in Red states.
In fact, the article makes no attempt to control for age at all.
No, the article assumes there is only one possible factor that might determine whether someone might die from COVID19 ... "Is that person one of us or one of them?"
Alrighty then.
As of right now, it's a blog by BAs (at best) with frequent stealth edits.