Session 1:

In the morning session of this workshop two new derivations of item response models were introduced. The first was based on a very influential model in mathematical psychology for decision making: the diffusion model. It explained a wide range of behavioral and neural data. Interestingly, it was possible to derive two variants of the two parameter item response model, one for unipolar traits such as ability and one for bipolar traits as attitudes. Both could be applied to data using recently developed R-packages. The second derivation was based on a prominent model in what is now called network psychometrics: the Ising model. The link between the Ising model and the two parameter item response model gives rise to a new view on the difference between dimensional and typological individual differences. This was explained in the context of general intelligence and attitudes.

Session 2:

The afternoon session focused on a novel application of item response theory. To collect high frequent reliable data on children’s cognitive development we developed a web-based educational platform where children can practice math and language at their own ability level and teachers receive rich information about their students. The platform applies a novel implementation of the Elo rating system, originally developed for chess, including scoring rules for speed of responses. This allows computerized adaptive testing without the (expensive) requirement of pre-testing item banks.
The popularity of this research tool among children and teachers stimulated the start of a university spin-off. This spin-off,, quickly developed multiple platforms (,,,,, now used by over 200.000 pupils of 2000 mainly Dutch primary schools. Schools and parents agree with the use of anonymized data for research. The data (over 1 billion item responses) are used by researchers all over the world, resulting in more than 25 international publications and 3 dissertations in the last 5 years. Examples are papers on the development of logical reasoning, the role of working memory in arithmetic, errors in number recognition, past tense learning in Dutch, but also new contributions to the psychometrics of the Elo rating system.



Han van der Maas (1966) studied Psychology at the University of Amsterdam and received his master (Psychological Methods) in 1989 (Cum Laude) and his Ph.D (Developmental Psychology) in 1993 (Cum Laude) for research on methods for the analysis of phase transitions in cognitive development (advisor: Peter Molenaar). After a five year KNAW fellowship, he joined the faculty of the developmental group of the University of Amsterdam, first as associate professor and in 2003 as full professor. In 2005 he became professor and chair of Psychological Methods group at the University of Amsterdam. Since 2008 he is also director of the Graduate School of Psychology at the University of Amsterdam. In 2009 he founded, a spin-off company of the UvA-holding, selling an unique game-based web-based adaptive child monitoring system. He teaches Psychological Methods, IRT, LCAand R. His research focuses on the formalization and testing of psychological theories in areas such as cognition, expertise, development, attitudes and intelligence.