Multiple Regression, Longitudinal Data and Welfare in the Nineteenth Century: Reflections on Yule (1899)ISER External Seminars

In 1899 G. Udny Yule read a paper to the Royal Statistical Society on the causes of changes in pauperism which is a landmark in social statistics. It was the first time that the multiple regression method had been applied to social science data and Yule applied it in a way that will be familiar to today’s quantitative social scientists. Another remarkable feature of Yule’s 1899 paper is that it is an analysis of longitudinal data. He also linked Census data to administrative data in a way that seems quite modern. These data have been reconstructed from Poor Law returns etc. and then reanalysed. The presentation considers the social context of Yule’s paper and its relevance to current methodological and substantive debates about the analysis of observational data and about the operation of the welfare system. I then outline the strengths and weaknesses of Yule’s analysis before going on to show how twenty first century social statisticians might use random and fixed effects approaches to answer Yule’s main research question in the light of the data available to him in the 1890s. These reanalyses indicate that Yule’s main statistical finding about pauperism stands up but that the effects are much more heterogeneous than he was able to elucidate, and the interpretation of those effects much less straightforward.

Presented by:

Ian Plewis (University of Manchester)

Date & time:

5 Oct 2015 14:00 pm - 5 Oct 2015 15:30 pm



External seminars home


Latest findings, new research

Publications search

Search all research by subject and author


Researchers discuss their findings and what they mean for society


Background and context, methods and data, aims and outputs


Conferences, seminars and workshops

Survey methodology

Specialist research, practice and study

Taking the long view

ISER's annual report


Key research themes and areas of interest