Lancaster-Warwick-Stirling node

Modelling of individual measures such as inhibitory control taking into account serial effects, temporal and age effects and separating out within-subject age effects from between subject age effects (Diggle et al, 2002), . This has identified important age-specific interactions and learning effects.

 

 

 Led by Andrea Towse, this sub-project explores the idea that collections of measures produce clearly defined developmental sequences of the underlying skills that are of interest. Latent transition analysis is being used to identify the nature of developmental change in a range of cognitive skills tested over time in the same children. We use software from existing statistical packages to identify and model these transitions and then explore which other skills predict the developmental changes that occur within individual children.

 

[An example here?]

 

 

These are a useful tool for the representation and assessment of such relationships (Whittaker, 1990). This sub-strand, conducted by Ivonne Solis-Trapala follow the approach by Wermuth and Cox (2004) to enable a likelihood-based assessment of competing causal models in developmental psychology. She is building likelihood-based inference for graphical models that involve multivariate binary data after marginalising over the latent structure. Her theoretical concern centres on recent ideas about the mechanisms of cognitive development and how these can be represented.