Ed Snack (886) Says:
April 24th, 2013 at 8:46 am
Yes, I’m aware of that …. but I was poking fun at the picture and the ‘information’ presented on it and how easy it is to read whatever we want into it.
Scott: Quite right..there are many correlations – weak or strong – between various factors and the plummeting rate of homicide in the US from the early 90’s. All of them of course “warrant further investigation” – except one.
Although there is a strong correlation between more punitive policing and sentencing policies across the US and falling homicide rates, more punitive sentencing had absolutely nothing to do with that. Ask any American liberal academic.
And the non-evidence of causation will be even stronger here in New Zealand, of that you can be certain. Violent crime peaked in this country in 2009-10, and has been falling since. But New York style policing in South Auckland, and “three strikes”, are playing absolutely no part in that fall at all. Ask any liberal New Zealand academic – in fact ask any New Zealand academic; we only have liberal ones.
Ed Snack (887) Says:
April 24th, 2013 at 9:32 am
Yeah, I blame the interwebby thing myself – a lot! (especially when I have wasted time on it and not completed the things in the real world that I should have) Tis a dangerous place the web – hence the name -just ask any fly what a web means to them!
It’s correct that a correlation does not equal causation.
But I think the following point is often missed – a correlation (direct or indirect*) is a necessary ‘precondition’ to establishing causation. If there’s no direct or indirect relationship – there is no causation.
*An indirect relationship is one that is ‘masked’ (or to use the technical terms, mediated, moderated, or suppressed) by one or more other variables.