« Discussing "Bias and the Bar" | Main | Discussing "Bias and the Bar" II »

30 April 2009



love to see this discussion! It’s great to see you all working through the issues and also, it’s great to see recommendations for testing. In the end, it’s what your actual users do and prefer that should be your biggest driver in making these decisions.

get degree form home

Male Enhancement

Bias and the Bar are different. Anyway, Nice information that you share on this article. I like to thank you for posting this great article. Good information.

Tracy Lightcap

Coincidentally, I recently used just the kind of strategy Chris is talking about in a project for my research methods class. We are evaluating the new DUI Court in Troup County. Such courts are exemplars of what Henderson, et al., called diagnostic adjudication and we have plenty of data about the probationers themselves. Problem = we have counterfactual difficulties: the defendants we could look at from the old court system had no exposure to the DUI Court and (drat!) nowhere near enough of the DUI Court defendants were carryovers from the old system. We had to predict how likely our DUI Court defendants would have been to recidivate if they hadn't been in DUI Court. Which, unfortunately for us, they had been.

We matched the State Court defendants from before with the DUI Court defendants today using MatchIt and simulated the counterfactual using the logit routines in Zelig. (Well, to be more exact, I did that, but the students actually understood what was going on and followed me when I demonstrated.) Chris is right: it couldn't have been easier from a computational standpoint. The real problem was the students putting together a description of what happened in prose our clients (and the students, of course) could understand. Our final report will be pretty much what they asked for and at virtually no cost to the court.

So, if any of you are apprehensive about the matching processes Chris is talking about, don't be. Just learn a little R (the really hard part; I/O can be tricky) and get to it.


Chris: Excellent post. And great idea to link to actual code.

The comments to this entry are closed.


July 2024

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      

Site Meter

Creative Commons License

  • Creative Commons License
Blog powered by Typepad