Full-time
psychometrician, using IRT mostly (the OPLM [One Parameter
Logistic Model ] of Eggen & Verhelst is the "home"
model at Cito; my colleague Norman Verhelst and I are modernizing and
'Window-izing' the package), partly in the Dutch National Assessment
Program for primary schools. As I think it good practice to look at data
from different perspectives before delving into IRT, I use several data
analytic tools (Gifi things, Wilks & Gnanadesikan) as preprocessors.
My
interest in scaling started when at Leiden University in the "boom
era" of nonmetric MDS; although I was not part of
the Data Theory group (Gifi), I was thoroughly
influenced by it. As all serious scaling software at the time was
written in Fortran, I became an ardent Fortran user, and still am. I
think Fortran is very good for (heavy)numerical computations; especially
the new Fortran90/95 form . Current production includes
Andrews' Hi-dimensional Scheme, and Homogeneity Analysis.