Back from the 17th annual ISOQOL conference, in London, where I had the occasion to present my ongoing work on the detection of DIF on an health-related quality of life questionnaire. The slides (PDF) are available.

The advantages of ISOQOL over ISPOR conferences are that: (a) they tend to focus more on patients reported outcomes studies than cost-effectiveness studies, and (b) the pharmaceutical sponsoring/lobbying is far less glaring.

I didn’t attend any workshop at ISPOR this year, but I did go to a four-hour workshop on IRT modelling during ISOQOL. I was initially interested by this workshop because the basic idea was to show how to carry out psychometric analyses with R and WinBUGS. It turned out that the organizers mainly compared output from BUGS and R for a Graded Response Model. But, at the end, there was a nice application on a short CAT with the ltm package. The supporting material is available online: The Use of R and WinBUGS in Fitting Item Response Theory Models.

In the meantime, I just discovered that there is now an R package for CAT simulation on CRAN website, catR. It looks promising, and I should definitively have a look at it.