4.25.2006

The projects

Everything we do is a randomized trial. That is to say, we identify some effect we want to test then create a natural experiment where half the subjects are randomly assigned to receive treatment and half serve as a control. This design eliminates the selection and nonresponse bias I spent basically the last three years of my life dealing with. Poof! Gone.

Some areas we're testing in:

-local newspapers' bias's influence on national political races
-various framings of the intelligent design/strict evolution debate
-media and perceptions of Americans abroad - in tea houses in North Africa!
-best methods of nonprofit fundraising.

This was going to be really long, but it's late and I'm gonna run before work at 8 tomorrow. Maybe if there's begging for it.

2 Comments:

At 29/4/06 00:40, Blogger grammar lammer said...

for some good media criticism on religion, check out www.therevealer.org. i know you guys are doing studies rather than scattershot finger-pointing, but the revealer (run by my prof at nyu) does some good picking apart of skewed (and occasionally absurd) coverage of religion in the media. it might serve as a source for ideas, or casual barometer, at the very least.

 
At 2/5/06 06:13, Anonymous Greg HARRIS said...

I'd like to know more about what you're testing in the Philippines. I beg. I assume it has to do with the media/elections, since they've had such heatedness with their president.

 

Post a Comment

<< Home