Statistical modelling of longitudinal psychological problems

One of the areas of interest is the application of structural equation modelling (SEM) to longitudinal data from large scale birth cohorts.

This involves, for example, longitudinal path models to establish the predictive value of early risk factors and personality features (e.g. difficult temperament) in relation to a later diagnosis of ADHD or conduct problems in childhood. Another example for the application of SEM analysis is path analysis in the area of cognitive neuropsychology aiming at disentangling reciprocal effects, for instance between different types of attention as shown in the figure below.

The team are also interested in robustness research, with a current project comparing the performance of the Parallel Method for factor extraction to standard methods such as the eigenvalue criterion or the Scree plot.

Finally, they are also involved as statistical consultants in external projects, for example the psychometric evaluation of the Movement ABC Checklist and a new handwriting speed test DASH (Dr. A. Barnett, Oxford Brooks University).