Please turn on JavaScript

Brooks Wilson's Economics Blog: Bad Experimental Design

Sunday, August 8, 2010

Bad Experimental Design

From the right, a newly released CBS news video feed from September 2008 showing Katie Couric lambasting  the names of Sarah Palin's children, Trig and Track ("Where the hell do they get these [names]?"), is being used by some conservative groups to demonstrate Katie Couric's bias against all things conservative ("Couric's Palin vid leaked").  What is lacking is experimental control, a film of Couric reaction to the unusual names of a liberal's kids, perhaps Calc and Econ.  Even then the number of observations would be so small to make a strong conclusion. 

From the left, James Harrington, a pastor, set up a mock fight between two black men to test the reaction of an invited speaker, a white police officer ("Police: Officer Baited In Altercation That Injured Him").  Again there is no experimental control.  How would the same cop react to a fight between two white men?

We can't always make decisions on controlled experimentation, but we should be careful about the conclusions we draw without it. 

No comments:

Post a Comment