Link to the study https://jech.bmj.com/content/jech/early/2023/08/17/jech-2023...
Appears to be a typical “look at dozens of variables and pick the one that turned out to be significant” to my untrained eye. Would love to know if I’m wrong and why!
Appears to be a typical “look at dozens of variables and pick the one that turned out to be significant” to my untrained eye. Would love to know if I’m wrong and why!
The interesting thing is that, with 40 variables the expectation is that they would find 2 spurious correlations. They got unlucky and only got to publish 1.
Also, with about 50 control variables that nobody has any idea how related they are to the outcome, I'm sure I could mess with the numbers until I find some 10 spurious correlations on the set. They quit too early.
Still, those studies are important to lead future research into the actual relevant questions. It's journalists that shouldn't touch them.
The vast majority of epidemiological papers are basically correlations with socio economic status.
There's a huge academic industrial complex generating this type of crap.
>As an observational study, the research was not able to determine what is causing the link between housing tenure and biological ageing, and the DNA samples analysed so far were only from white, European householders.