I think perhaps you've missed what happened. As part of a study on covid mortality in schoolchildren, he compared Mar-Jun 2020 child mortality to Mar-Jun 2019 child mortality. He found an increase of 68%. >
-
-
Replying to @DavidSteadson @RickCarlsson and
He wrote to the Anders Tegnell, at the Swedish Public Health Authority - with whom he has a lot of contact - asking for help getting the cause of death. Tegnell refers him to the National Board of Health & Welfare, who handle death certificates This is all fine, even good. >
1 reply 0 retweets 1 like -
Replying to @DavidSteadson @RickCarlsson and
He does however, talk about the effect being mitigated if he includes a younger age group and considers that. This is less good, but just an email. He says he feels once he knows about this data, he should publish it. Great. >
1 reply 0 retweets 1 like -
Replying to @DavidSteadson @RickCarlsson and
The article is published, and it includes mortality data comparing Mar-Jun 2020 to Nov-Feb 2019, which makes no sense at all. There's no significant difference. Zero mention of the concerning year on year comparison at all. >
1 reply 0 retweets 4 likes -
Replying to @DavidSteadson @RickCarlsson and
Since the Science article was published he has claimed he included this comparison *at the suggestion of a reviewer*, which I find frankly ... well, bizarre? What did he submit, the concerning data that showed the 68% increase? And the reviewer suggested changing it?
1 reply 0 retweets 4 likes -
Replying to @DavidSteadson @RickCarlsson and
Or did he just submit the 2020 mortality data with no comparison, and the reviewer suggested comparing it to the immediate 4 months instead of the previous year? None of this seems plausible to me. We *today* now have some data showing the causes of death were likely not covid.
2 replies 0 retweets 1 like -
The problem is lack of transparency in the research and in the editorial process including the peer review. Peer reviewers often request bizarre things. Anyway although data wasn’t strictly available, children dying would be common knowledge in Sweden. >
1 reply 0 retweets 2 likes -
Replying to @RickCarlsson @DavidSteadson and
> even with small numbers that big difference wouldn’t have gone unnoticed here. He should still have verified it before submitting though.
1 reply 0 retweets 1 like -
Replying to @RickCarlsson @DavidSteadson and
I understand your point that there are far worse examples of data manipulation, but I don't think that is a reason to downplay what seems to have happened here. Publishing a big claim while hiding data that may contradict it, even if you think it doesn't, is serious. No?
2 replies 0 retweets 4 likes -
Replying to @NDLoubere @DavidSteadson and
Yes, it should be corrected. That’s very important. Still, the overall method where it’s up to the author to pool ages as they see fit, rather than following some fixed standard set out before analysis, and the poor review process is the real story. >
3 replies 0 retweets 1 like
I reckon the real story is that this sort of shoddy study with selective research is so common. The weirdest part about this whole argument is that it's basically reinforced my initial comment - that it sounds like p hacking - but because p hacking is ubiquitous...
-
-
Replying to @GidMK @RickCarlsson and
Would you still consider it p-hacking if the comparison dates were requested by the reviewer as is claimed? If that claim is true then all of the data for those dates was published from the start
3 replies 0 retweets 3 likes -
Replying to @JacobGudiol @GidMK and
Isn't it a bit much to ask researchers to included data from dates not mentioned in a article just in order for them to not be accused of p-hacking afterwards?
1 reply 0 retweets 2 likes - Show replies
New conversation -
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.