Like, "Poor people live in darker homes, and the photosensitive ganglion cells are connected to the part of the brain that controls melanin production. And this triggers a sleep response more often, which causes the immune system to behave in X, Y, and Z".
How does a statistician make a decision about what to change given only statistical data? It seems very mysterious to me. I understand how to interpret a scientific result(give people brighter homes), but not a statistical one.
(See also for example Judea Pearl's work http://causality.cs.ucla.edu/blog/ )
@Michael, biology is messy compared to machines. To judge a measure of certainty in the methodology and results, read the peer-reviewed papers submitted to scientists working in this and other fields ~ http://dunedinstudy.otago.ac.nz/publications
Already done.
I refer to the "Dunedin Study", (Dunedin Multidisciplinary Health and Development Study). [0] Read the summaries of how this was/is being conducted. The most revealing insight into the links between poverty, disease and mortality. People born into poverty and who increase their wealth considerably in later life, do not get healthier.
Disease (damage to the immune system) is directly related to poverty.
[0] https://en.wikipedia.org/wiki/Dunedin_Multidisciplinary_Heal...
[1] http://dunedinstudy.otago.ac.nz and http://dunedinstudy.otago.ac.nz/publications