Quote:
Originally posted by Larry_OHF:
Simple logic. Although The color of the car and accident rate is positively correlated...the color of the car is not the causal factor in the matter.
|
Well maybe
. You are inferring causality (and lack of) on no more information than they are. It
may be that the colour of the car affects visibility which causes more accidents, it
may be that people who buy certain coloured cars are more prone to accidents, it
may be that there are more of some coloured cars which will result in more total accidents even if the percentage of accidents is the same for each car colour. You simply can't tell from the data. Correlation does not automatically imply causality, but it does not necessarily preclude causality either [img]smile.gif[/img]
Though I will comment that I believe that white is the most common coloured car there is. Presumably there are a fair number of younger people with white cars as well as older, more experienced drivers. More research collecting information of this sort would be able help determine which of, if any, the above hypotheses are correct, because the effects of contributing variables like type of driver and type of car would be able to be measured and removed from the effect of car colour on number of accidents.
Also (to everyone) it's not really fair to bag a study based on a four paragraph article about it. It's amazing how often I read about a study that has been done, of which I have read the original published paper, where the findings have been misrepresented or even got completely wrong in some cases, or the type of controls not discussed, or the interpretative caution expressed by the study's authors ignored and the results presented as fact. It's usually best to go directly to the original paper, where issues such as relevance, alternative explanations, and methodology are addressed.