New here! The discussion will prove to be addictive I am sure :-) Great stuff.

One has to be very careful with how data is presented. I agree that the article argues for a correlation effect. This is not causal, but it notes that two events are related without doing a more sophisticated analysis like a regression which does reveal a causal effect between variables. That is to say, if the research presents that one variable regresses on another, then we have a lot more to work with.

So with this in mind if religion correlates with another variable, the next step is to determine the variance of the model. What variables can we pull out that predict other variables to occur under specific conditions? Saying “religion causes violence” is not satisfactory since all religious belief clearly does not. If not all, under what conditions does religion cause violence and is it therefore the conditions in which religion acts on people that is the true cause of the violent behavior? Religion can then act as a secondary cause or even a catalyst, but not the source of violent behavior itself.

Variables and how they are related are vital in reporting research and I think that overall folks get a bit too eager to say “this causes this” and end up misrepresenting a given study. We see it on the Today show expert circuit all the time! :-)

Drew Tatusko