3 ms·

This looks like someone's picked out a variable and missed the bigger picture. If people from Dunedin do have a positive correlation between lead in their bloo

This looks like someone's picked out a variable and missed the bigger picture. If people from Dunedin do have a positive correlation between lead in their blood and criminal convictions, then I'd be asking why the convicts have more lead in their system in the first place. From a quick google it looks like lead paint can cause higher blood lead levels. Perhaps the use of something like lead paint could be highly correlated to poverty. I know my gran has wallpaper on her walls instead of using paint - perhaps people in order to save money on their properties used lead paint instead of wallpaper. Perhaps state houses have more lead in them than richer family houses. The point is without knowing much about this area, and from a quick scan of this article, it seems that they're trying to find evidence to support their current position of thinking around lead poisoning and crime. I don't know if what I've mentioned above around housing is accurate at all, but it's a quick thought off the top of my head to demonstrate a potential for missing the forest for the trees, or an attempt to find 'evidence' for your preconceptions.

Gibbon1 9y ago
As I understand it lead paint in the US was a scourge in poor neighborhoods. Old lead paint tends to flake and small children love the taste of lead paint chips because it tastes sweet. 1950-1970's lead paint was a huge issue in poor neighborhoods. Point I'd make is, often you see people trying or use epidemiological studies to try and prove some cause and effect. And this doesn't actually work very well. The world is filled with associations that have nothing to do with each other. However with lead the neurotoxic effects are fairly well understood in detail. As well as population exposure rates. So in this case epidemiological studies are confirming what one might well expect. Lead exposure ->
xxgreg 9y ago
Agreed. If crime is correlated with child poverty, and cheap rentals are old houses which haven't been renovated in a long time, and tend to have more lead paint. Then on average criminals will have higher blood lead levels. This doesn't provide any evidence that lead had any causal affect in crime.
mr_toad 9y ago
This study does record details such as income and economic status over the lifetime of the participants, so in theory you might be able to control for those effects. But the sample size is not huge, and there is a risk of overfitting.

Gibbon1 9y ago

As I understand it lead paint in the US was a scourge in poor neighborhoods. Old lead paint tends to flake and small children love the taste of lead paint chips because it tastes sweet. 1950-1970's lead paint was a huge issue in poor neighborhoods. Point I'd make is, often you see people trying or use epidemiological studies to try and prove some cause and effect. And this doesn't actually work very well. The world is filled with associations that have nothing to do with each other. However with lead the neurotoxic effects are fairly well understood in detail. As well as population exposure rates. So in this case epidemiological studies are confirming what one might well expect. Lead exposure ->

xxgreg 9y ago

Agreed. If crime is correlated with child poverty, and cheap rentals are old houses which haven't been renovated in a long time, and tend to have more lead paint. Then on average criminals will have higher blood lead levels. This doesn't provide any evidence that lead had any causal affect in crime.

mr_toad 9y ago

This study does record details such as income and economic status over the lifetime of the participants, so in theory you might be able to control for those effects. But the sample size is not huge, and there is a risk of overfitting.