Bruce Martin told us:


The officer from Traffic Branch who dealt with me at the time of the death said that it was the third fatal accident he had dealt with in six months involving drivers over seventy.

That comment tells us nothing!

First, "involving" doesn't mean "caused by".

We need to know how many fatal accidents in toto he saw during that period. You can't have a meaningful statistical result without a large enough sample. If the sample is too small, the flukes and freak accidents and one-time circumstances can shoot your results all kerflooey. If, for example, he only saw four fatal accidents, then even three which we might assume were caused by drivers over 70 could be a fluke. If he saw sixty, three is not a huge percentage.

Supposing it's a large enough sample to be meaningful, you then need to know how the accident causes were distributed. My guess is that the causes would be overrepresented in the late-childhood (15-18) and early adulthood (19-25) groups.

As I say, the officer's comment, at least as reported, is meaningless.

Dave Oesterreich