I think John Kelso made a good point about unintentional biases we sometimes induce while interpreting statistical data. It reminded me of Andrew lang's funny quote:In case you haven't been following along at City Hall, some City Council members want to make it illegal for panhandlers to do their thing on roads and sidewalks, and within 1,000 feet of schools.
When you stop and think about it, aren't toll roads a high-tech form of panhandling?
Besides, panhandlers working near schools is a better educational tool than Career Day. Just having the panhandlers out there looking ugly says, "This could happen to you if you don't finish your algebra homework."
Anyway, Council Members Brewster McCracken and Jennifer Kim point to statistics that show there are lots of traffic accidents at intersections where panhandlers commonly hang out.
Of course there are lots of accidents at intersections where panhandlers hang out. This is because they are busy intersections, which is why panhandlers pick them. Where there are cars, there is more loose change. This is why you rarely see panhandlers in the Sam Houston National Forest.
An unsophisticated forecaster uses statistics as a drunken man uses lamp-posts - for support rather than for illumination.
No comments:
Post a Comment