(This is what's known as a relative risk, which means the risk of one thing happening compared to another) 6.5/
-
Show this thread
-
What does this mean? Well, not that much. As I said, this is very crude. We know that the debt notices were probably sent to less healthy than average people (some of them were disability recipients, for example) 7/
2 replies 1 retweet 4 likesShow this thread -
So the fact that the relative risk of death was about 40% higher in the robodebt recipients may not mean anything at all (although it's worth noting that we HAVE - mostly - controlled for age by definition here) 8/
1 reply 1 retweet 6 likesShow this thread -
Ultimately, you'd need to do a much more robust analysis to get a particularly strong answer What we CAN say is that ROBODEBT DEFINITELY DIDN'T CAUSE 776 DEATHS
#notmydebt#robodebt 9/3 replies 0 retweets 4 likesShow this thread -
People who received debt notices *may* be at a higher risk of death than the general population - it's hard to be sure - but the debt notices definitely didn't kill people en masse End 10/
1 reply 0 retweets 4 likesShow this thread -
Worth noting - because it's my eternal bug-bear - that the ABSOLUTE risk of death appears to have been increased by 0.03% by debt notices, which is somewhat less impressive and scary as well
2 replies 1 retweet 5 likesShow this thread -
CORRECTION TO THIS THREAD - I actually miscalculated the relative and absolute risk - I halved the estimate twice accidentally This is why peer review is important thank you
@AndrewHayen1 reply 0 retweets 3 likesShow this thread -
The true relative risk is 3, or a 300% increased risk of death, and the absolute risk should be 0.07 or an 0.07% increased risk of death
@not_my_debt1 reply 0 retweets 2 likesShow this thread -
But still probably worth remembering that this is not very meaningful and we'd have to at the very minimum compare to other Centrelink recipients - not the general population like I did - to get a decent estimate
2 replies 0 retweets 4 likesShow this thread -
Replying to @GidMK
Lots of confounders within an already at risk group. If the data was available pre/post robodebt and/or with/without robodebt notice could be interesting.
1 reply 1 retweet 1 like
Pretty much. This number is ~slightly~ more meaningful than the 776 raw figure, but still not a reasonable comparison
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.