Still, more intelligent than the 'Your Health' report
Unsurprising Neale News item of the day:
Heavily rural province with lots of RCMP has country's highest drunk driving rate
There are many reasons why so much print and TV news is just plain lousy. One is unacknowledged bias; I understand there are a couple of blogs out there devoted to exposing this.
Another is basically laziness; excessive willingness of reporters to accept press releases at face value, so long as it doesn't challenge their biases.
A big reason, though, is that often, reporters just don't know what they're talking about, and worse, they don't know that they don't know what they're talking about.
Look at the story linked above. I wouldn't bet against the accuracy of the headline. In fact, it's entirely believable, for the two reasons noted in my paraphrased headline. But it's not supported by the story - and I mean at all.
Say the Homicide division of the Edmonton Police Service was staffed by lazy wankers, and over the next year, solved no murders and made no arrests. According to the logic of the CBC headline, Edmonton would have Canada's lowest murder rate. (Congratulations, guys!)
The linked CBC story says that Saskatchewan law enforcement makes more per capita drunk driving arrests that the other provinces. That may well be interesting information, but that is all it is. You cannot conclude from that statistic that there's more drunk driving in Saskatchewan than anywhere else. For all we know (from this story, at least), there's the same amount in every province, and Ontario is 3 times lousier at catching and prosecuting offenders than SK.
That said, let's not get too excited about the great job SK police are doing catching impaired drivers. For all we know (again, from this story at least), SK has 10 times as much impaired driving as ON, and ON police are actually 3 times better at catching and prosecuting offenders.
I know it's not sexy (or PC, or brief) but this story and those like it really ought to end with a paragraph like this:
"Interpreting impaired driving statistics requires numerous assumptions that may or may not be true, because the overwhelming majority of 'offenses' are never known or detected, and leave no evidence."
This won't happen, because it reminds the public that in most cases, impaired driving is a victimless crime--and really, who wants to be carrying that banner. (Anyone? I'm sure it's a real friend-winner...)