Red light stats unclear

Statistics are dangerous things, and so is ignoring what statistics actually are.

A Jan. 26 headline in The Spokesman-Review said red-light cameras make Spokane’s streets safer – “statics suggest,” it declared!

After further emphasizing how convincing the numbers are, the article eventually mentions that collisions at the original three red-light camera intersections are down 45 percent, 33 percent and 10.5 percent. But wait. During that same time period collisions at all Spokane intersections are down 25 percent. So does this mean the cameras caused the number of collisions to go down at two intersections and up at a third? Of course not.

Scientific statistical analysis can be – should be – used to compare the decline in collisions at camera intersections versus the amount of decline witnessed at no-camera intersections with sample size taken into account so it can be determined if the difference is significant or random. Council member Mike Fagan says, “There’s no way to refute the data.”

Really? What if I flip a nickel and quarter each 10 times and the quarter comes up heads six times and the nickel comes up heads four times? Will the Spokesman and Fagan proclaim we have irrefutable statistics showing quarters come up heads more than nickels?

Don Larsen

Spokane