Estimating excess deaths in the UK, methodology changes: February 2024
2710 sats \ 5 comments \ @erict875 22 Feb charts
The UK has changed the methodology used for calculating the excess deaths. Instead of a "simple five year average," it now uses a multivariate statistical model that takes into account factors like age, sex, geographical location etc.
What is fascinating is that this new sophisticated model now produces the expected outcome of post-pandemic negative excess deaths!
WOW! I could be wrong but it feels like we're literally sweeping bodies under the carpet? I'm really looking forward to reading more about this from medical and maths experts!
Source: https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/causesofdeath/articles/estimatingexcessdeathsintheukmethodologychanges/february2024
See previous post on subject for more context: #432311
write
preview
related posts
100 sats \ 0 replies \ @erict875 OP 22 Feb
The ending remark in this brilliant short video is precisely why this new methodology caught my attention.
reply
31 sats \ 0 replies \ @DudeJLebowski 22 Feb
They've CPI'd the excess deaths statistic!
reply
60 sats \ 1 reply \ @Undisciplined 22 Feb
One of the benefits of simple metrics is that they are simple to understand.
I'm all for using more sophisticated methods to add some nuance, but it's hard to overlook how often revised methods generate more convenient results for those in charge.
reply
123 sats \ 0 replies \ @erict875 OP 22 Feb
I couldn't agree more! I am not a trained statistician, but I actually like the motivations for the new model. I'm nevertheless also suspicious, for obvious reasons. Anyway, I'm sure it's going to be scrutinised by other experts and we'll surely hear more about it soon.
reply
0 sats \ 0 replies \ @erict875 OP 23 Feb
Dr Campbell opines on the subject. Interesting is that the ONS' new methodology produces different numbers than the methodology used by the OECD, and the difference is quite substantial. What a mess!!
reply