Statistical muddling at the Herald

Posted by Peter Hemminger in Stranger in the Alps

I don't usually get into politics on this blog, but lazy, misleading statistics are a particular pet peeve of mine. Last Friday's Calgary Herald featured a column on hybrid cars, which pointed to a study on the efficacy of rebate programs in encouraging people to buy hybrids. From the article:

"Researchers at the University of B. C. studied rebate programs over six years for hybrid electric vehicles in B. C., Ontario, Quebec, PEI and Manitoba. Their results, released this week, show two-thirds of those who bought hybrids were going to purchase them anyway. The rebates didn't influence their buying behaviour at all."

Think about that for a second. If two-thirds of those who bought hybrids were going to purchase them anyway, one-third of those who bought hybrids wouldn't have. One third of all hybrid sales, according to that statement, are directly attributable to the rebate programs studied.

Put slightly differently (because it's always fun to play around with the numbers a little), the rebate programs in B.C., Ontario, Quebec, PEI and Manitoba led to a 50% increase in the sale of hybrid cars. How can that be construed as not having an affect on buyer behaviour? If the goal was to increase purchases of hybrids, in what world is a 50% increase considered a failure?


more in News     |     posted Aug 11th, 2009 at 1:30pm     

Comments: 1

brennan wrote:

Great point!

on Aug 13th, 2009 at 2:36pm Report Abuse


Post comment: (Login or Register)


All Content Copyright © Fast Forward Weekly 1995-2012

About Us Contact Us Careers Privacy Policy Terms of Use