Predictions, explanations and causal effects from longitudinal data

Researchers across the social sciences who are interested in change and development can now draw on a rich array of longitudinal resources – birth cohort studies, panel studies and the like – to help them answer their questions.

The combination of theory, data and modern methods of statistical analysis can be used to describe, to predict and to generate causal explanations.

In his lecture Ian Plewis explores these uses of longitudinal data with illustrations from the recent and more distant past.

These include the early use of statistics to investigate the causes of changes in pauperism, and, more recently, analysis of progress in English using the National Pupil Database.