How do you read me? LII launches a new research collaboration.
“It should be borne in mind that all the time we were studying bats, bats had a unique opportunity to study us.”
— old researcher’s folk saying, quoted by John Brunner in “The Skynappers”.
The LII is a place where people — a lot of people — come to find and understand the law. They come to us looking for information. But what if we looked for information from them? It would help us understand what they don’t understand, what we need to organize and present more clearly, and how they make use of what we have. We already do that through reader surveys, e-mail followups with our donors, and of course an extensive set of web analytics provided by Google and our own monitoring. But what if we really asked the 170,000 people who visit the site on an average weekday to tell us what they find difficult to understand?
At last September’s Law via the Internet conference, we attended the presentation of a paper by Michael Curtotti and Eric McCreath of the Australian National University. Later published in the new Journal of Open Access to Law, the paper describes a series of methods for assessing the readability of legislation. This is actually much harder than it looks — most readability measures are designed specifically to sort reading materials into grade-level bins for use in the teaching of reading, and don’t help much if the problem is figuring out how hard a particular statute or regulation is to understand. Linguists have, in the past, done fairly deep research on comprehension in criminal-trial settings, particularly jury instructions. But Curtotti and McCreath are the first to do much with legislation, and the first to build an extensive online system for testing particular pieces of text.
We were fascinated. What if we could use their methods to test readability across the really large audience that is drawn to the LII? A rough plan was hatched (in the hotel bar, of course). Over the next few months it was refined, and the Australians began working on software that would integrate the testing with some carefully-chosen parts of the LII site. We are now finishing up the technical integration of the tests at our end; a particular concern is that we not slow the site down for readers who are not working on the tests. The actual tests should roll out within a month.
The work is exciting in itself, but it is just the tip of the iceberg. The LII assembles a huge audience that is doing something that is interesting in its own right. The ways that experts and laypeople use to find the law, and figure out how or if it applies to them, remain a little mysterious — and we hope that this will be the first of many occasions for outside researchers to help us dispel some of the mysteries.
We look forward to reporting some results in the next newsletter. In the meantime, you might want to take a look at the paper, and experiment with Michael and Eric’s system using text of your own (municipal regulations are a particular favorite of ours).