Friday, February 04, 2011

Benetech Statistician Megan Price talks to local ASA chapter

Guest Beneblog by Megan Price

I recently had the opportunity to present several of HRP’s projects to the local San Francisco chapter of the American Statistical Association (SFASA). Despite an audience of fellow statisticians, I chose to focus my talk more on the research questions and challenges posed by our work in human rights and less on the nuts and bolts of our statistical methods (though I did include a few equations and Greek letters!).

Specifically, I presented the audience with the following questions:

  1. Were acts of genocide committed against the Mayan people in Guatemala?
  2. How many Kosovars were killed between March and June 1999?
  3. How much did Hissene Habré know about political killings during his presidency?
  4. Did high-ranking officials within the Guatemalan National Police know about Edgar Fernando García’s disappearance?

As I told the audience, for those who like to skip to the last page of novels, the answers are 1) yes, 2) approximately 10,000, 3) a lot, and 4) we’re not sure. For details on how we arrived at these answers, you can find a copy of my presentation on the SFASA webpage.

The audience was a diverse group of statisticians, and the Q&A featured lively discussion. In particular, we commiserated over the challenge of presenting statistical results as an expert witness and wondered if perhaps a version of the Daubert Standard could be created specifically for statistical analyses. There was much excitement and interest in recruiting Stephen Pierson, the American Statistical Association’s Director of Science Policy, to consider such an initiative.

In general the local chapter members were very excited about, and more than a little envious of, our work and the potential it has to impact human rights research. Quite a few are hoping to volunteer for us in the future!

No comments: