Summarising Complex Longitudinal Data – Sometimes Less is More
University of Edinburgh
Dr Valeria Skafida
Entry (no or almost no prior knowledge)
Research Training Centre, firstname.lastname@example.org
This seminar forms part of the Research Training Centre's Talking Methods seminar series delivered in association with the Scottish Graduate School of Social Science.
For my current research I was interested in exploring how exposure to domestic abuse in the household was related to children's social and emotional developmentary trajectories. I worked with cohort data from the Growing Up in Scotland survey and wanted to summarise trajectories of internalising, externalising and prosocial behaviours over four times points in children's lives using repeated measurements. This led me down a rabbit hole where I learned how to apply group-based trajectory analysis, also known as latent class trajectory modelling, to extract developmental trajectories. In the end, after considering what this complex method had to offer for the data I was using it on, I opted to summarise the same data in a much simpler way, by pooling together standardised scores across time points. This was an interesting exercise of immersion and apprehension of new methods, as well as an exercise in learning to let go of fancy methods in situations where simpler approaches can achieve the same task.
Website and registration:
Longitudinal Data Analysis
Related publications and presentations: