OK, in the past day or two I've been looking over this study, the one arguing that Hurricane Maria killed about 5,700 Puerto Ricans (the 4,500 number you've seen is unadjusted). Some interesting stuff I found that's a bit too wonky for a blog post. https://www.nejm.org/doi/full/10.1056/NEJMsa1803972 …
-
Show this thread
-
The background is that while the official estimate is 63, the definition of a hurricane death is insanely narrow and recordkeeping is poor. Other estimates put it at around 1,000.
1 reply 0 retweets 4 likesShow this thread -
So these folks did a household-level survey in Puerto Rico, asking if anyone in the household died in 2017.
1 reply 0 retweets 3 likesShow this thread -
One limitation is that the households there in 2017 != the households there now, given many people fled the hurricane. Another is that if an entire household (especially a single-person household) died, by definition they won't be counted in this survey.
1 reply 0 retweets 3 likesShow this thread -
This approach allows two different comparisons. One is checking to see whether deaths went up after the hurricane, within the data collected. ~2/3 of the 2017 deaths took place in the ~1/4 of the year after the hurricane.pic.twitter.com/oZnfZz51bg
1 reply 0 retweets 3 likesShow this thread -
Another possible comparison is between the post-hurricane death rate *in the survey* and the *official* death rate from the same period the year before. The iffy part here is that there are only 38 post-hurricane deaths recorded in the survey. But this produces the headline #.
1 reply 0 retweets 3 likesShow this thread -
The confidence intervals are YUGE. The CIs on "excess deaths" run from 793 to 8498, or, with the adjustment for survivor bias and household-size distribution (to match it to census data), 1506 to 9889.
2 replies 0 retweets 3 likesShow this thread -
I also found this chart of official data really fascinating. September and October deaths were indeed much higher than expected, maybe 1,500 or so from eyeballing? Dunno wazzup with December though, or whether PR authorities could miss, not just misclassify, 1000s of dead bodies.pic.twitter.com/wba40sd390
3 replies 0 retweets 4 likesShow this thread -
Replying to @RAVerBruggen
Did you see
@St_Rev 's thread about this from yesterday?2 replies 0 retweets 0 likes
It wasn't a very good thread, not sure why everyone is plugging it
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.
. Banned in Sweden. SubGenius, Zhuangist, white-hat troll. Defrocked mathematician. Brain problems.