God and the true ideals of the American Dream are unchanging
I don’t think there is a strong partisan pattern in the overall level of Covid deaths per capita. However there does appear to a strong partisan pattern when it comes to vaccinations which is plausibly caused by politics. And that in turn is likely to lead to a strong partisan pattern of current and future deaths until the red states are able to reach somewhere near herd immunity.
If I were peer reviewing your study, it would be immediately rejected due to poor statistics. Correlation in this case demonstrates nothing. Are you controlling for e.g. rurality, SES? Comparing NJ and WY on speed of access is a fool’s errand. Not to say that R state governments/people are more resistant to vaccines, but I’ll bet the R^2 is small.
Run the stats. I’d guess they are statistically significant, when comparing the two populations (High 10 covid v. low 10 covid.) R2 is going to be pretty good, I’d say.
I’m not comparing "speed of access’.
The proposal is simply "States that voted in higher numbers for Trump will tend to have higher current covid numbers (7 day average).
I think the numbers bear this out. Run the stats. Show your work.
Just to go on some more ('cause my wife says this is one of my “strong points”^)
My claim was not that there should be a direct and pure correlation between the strength of Trump support and the worst current covid numbers.
My claim was that Trump-supporting states are doing poorly with covid containment compared to Biden supporting states. I think we can agree on which states are the top 10 and bottom 10 for recent 7 day average covid cases/million.
I think we can agree which states are Republican leaning v. Democrat leaning.
It’s that simple.
^I don’t think this is a compliment
Political lean and number of cases are both things that it’s pretty easy to put a quantitative number on. Surely someone has made a graph of the one vs. the other?
Easy to do… but I’m rather busy at the moment. Will try to do a more detailed analysis later if there is any interest. Or anyone else?
I’ve thrown together some numbers that I’m interested in, based on total numbers, rather than a recent snapshot.
I didn’t bother running a regression, because just eye-balling it shows difference to be rather minimal.
Rolling bar graph, 1 June 2020 until yesterday:
Again, total numbers are misleading because several blue states got hit hard at the beginning of the pandemic when there were probably tons of unknown covid cases out there. We also didn’t know how to treat the disease, so the case fatality rates were super high.
Recent cases/hospitalizations are showing a clear trend because vaccines are now widely available. Someone posted it in another thread. The trend is not perfectly strong because, within cities, a lot of Black people are still afraid of getting the vaccine due to historical reasons. Minorities were also having issues with access but that appears to be improving.
I like this. It starts after the first wave of the pandemic. That’s the point when we had much more information on how to deal with the virus.
I am not sure I understand your graph. Does each state have two dots (one for Biden and one for Trump)? What’s the hypothesis that’s driving this graph?
I think that if you wanted to test whether states which had a higher percentage of trump voters had a similar death rate to states with a lower percentage of trump votes, it would be clearer if you mapped trump voters per capita - biden votes per capita by deaths per capita.
It’s not just that. In the rolling bargraph @Skywatcher posted, the sorting was already nearly perfect in December 2020, before the vaccines. Presumably, that’s due to masking, social distancing, and all the other things we were doing before vaccines were available.
A few days ago, I made up a linear regression graph using an Apple Numbers template and data from all 50 states. I was trying to correlate each states vote percentage for Biden with its vaccination rate. The graph looks like a pretty fair correlation, but I can’t find a way to post it here. The R-squared value is 0.754.
My only training is statistics is from a high school class I took about 45 years ago. Can someone tell me just how good a correlation that is? Better still, if you can explain how I can post a .png chart, I’ll do that.
The error message reads “Sorry, you can’t embed media items in a post.”
I’d love to see this done with the most recent 7 day mean, rather than the total numbers. I’d already eyeballed the total numbers, and agree that there did not seem to be much of a pattern.
As I’ve said, I’m more interested in what is happening now, with variable vaccination rates and attention to public health recommendations.
Wow, that’s one awesome graphic!
I would say the total numbers are not misleading because several blue states got hit hard at the beginning of the pandemic when we also didn’t know how to treat the disease and the case fatality rates were super high. I want to know how many died, period.
I have no hypothesis. I want to see the relation between Trump, Biden votes and COVID deaths, both absolute and relative. That is what I plotted, two dots per state so I could see how COVID deaths vary independently of votes.
Of course there are other data one can look at and I’m happy to see what others do.
Yes, that would be interesting, but not what I wanted to do. Snapshots let you see the evolution, but you need the totals to know the cumulative effect.
You also looked at deaths, rather than cases, which was in my original post.
I’m more interested in what is going to happen in the near - mid future, and the most recent data is more important in giving me that information. For example, if New York had a lot of deaths in 2020, but few cases in the most recent 7 day average, then that tells me that New York will probably not have a lot of cases next month. I can’t tell that from your presentation of the data.
Sorry to keep harping on this but I don’t see what you’re trying to do.
The OP proposes that political ideology effects the approach to covid on both individual and governmental level. If you want to see the outcome of different political ideologies, you cannot include the first wave. The correlation may be weak after the first wave (however, two other posts suggests it’s not bad), but you’ll never see a correlation if you include the first wave.
Cumulative effect of what? I don’t see what you’re trying to look at.