[WS:] I did not say it should go unexamined. I said not to use questionable methodologies in that examination and I suggested more suitable ones. As they say, correlation does not mean causation and this applies especially to this case. Correlations (or regression coefficients) are not very useful without a proper causal model and a model that postulates gun ownership as the cause and homicide rate as an effect strikes me as a faulty one. I am inclined to believe that the correlation between these two variables is spurious, meaning that both variables are likely to be caused by another common factor, but there is no direct causal relation between them. It is like the height of the siblings - its is correlated, but the relationship is spurious as both are caused by another factor - height of the parents.
But enough of this already. I do not think much can be added to this discussion. Over and out.
-- Wojtek
"An anarchist is a neoliberal without money."