« Mix-and-Match Lawmaking | Main | Favorite Magazine Articles for 2006 »

Thursday, December 14, 2006

One Way to Conduct Empirical Legal Scholarship

Peter Hook has an interesting piece on SSRN, in which he maps, using various statistical techniques, the "topic space" of Supreme Court opinions--i.e., mapping the commonalities in the topical categories assigned to each case by West.  It's a nice article, but I also like it in part because it gives me an entree to talking about methodology in empirical legal analysis, which I'd like to offer a couple posts on--if only to remind empirical legal scholars that multiple regression isn't the be-all and end-all....

The first is multidimensional scaling (MDS), which is one of the approaches Hook uses. As I mention in a recent paper, "MDS is a procedure that helps researchers uncover 'hidden structures' in existing data by graphically plotting respondents’ perceptions of perceived similarities (or dissimilarities) among various stimuli.  When these stimuli are located on a plot based on such perceptions, underlying dimensions that respondents may have used (consciously or not) can be inferred."

So, for instance, a bunch of stimuli are identified -- crimes, Supreme Court opinions -- and every pairwise combination is created. Each pair is rated on how similar or close they are on some scale -- usually unarticulated.  Those similarity ratings are then plotted to give the best fit to the data, yielding a "map" of the ratings that also reflects the perceived structural relationship of all those stimuli.  Finally, by examining the map, one might be able to identify the dimensions that, implicitly or explicitly, underlay the ratings (and also look at individual differences, e.g., do men and women come up with different maps?).

Hook used MDS to map Court opinions; I used it to map people's perceptions of the similarities of crimes.  But I think it's a more broadly useful methodology, particularly for looking at lay intuitions about legal institutions, punishments, or cases; or to assess whether different punishments may have been proportionate; or any context in which (a) similarities among stimuli may be interesting or (b) we are interested in understanding what the basis might be for people's judgments or perceptions.

I'm painting with a broad brush here, but there are useful resources on how to conduct MDS (it's pretty easy and, as Hook's paper shows, gives pretty pictures), and it's another useful part of an empirical legal scholar's toolbox.

One more methodology post to come...

Posted by jeremy_blumenthal on December 14, 2006 at 10:02 PM in Criminal Law, Current Affairs, Legal Theory | Permalink

TrackBack

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8341c6a7953ef00d83461462b69e2

Listed below are links to weblogs that reference One Way to Conduct Empirical Legal Scholarship:

» Round-Up from SCOTUSblog
At ACSBlog, Martin Magnusson discusses last week's argument in the Rockwell case here. In this post at PrawfsBlawg, Jeremy Blumenthal highlights this new paper by Peter Hook called "Visualizing the Topic Space of the United States Supreme Court." Deepa... [Read More]

Tracked on Dec 18, 2006 10:16:11 AM

Comments

We are offering such scholarships so that you can search. http://www.freetoapply.com is a scholarships, fellowships and grants website talior to everyone.

Posted by: College Person | Jun 13, 2008 3:33:33 PM

The comments to this entry are closed.