Predicting ‘Who Voted for Brexit’ Through Ecological Analysis – An Example of the Problem of Confounding, and its Resolution

10 Pages Posted: 9 Feb 2018

See all articles by Ron Johnston

Ron Johnston

University of Bristol - School of Geographical Sciences

Kelvyn Jones

University of Bristol - School of Geographical Sciences

David Manley

University of Bristol

Date Written: January 31, 2018

Abstract

In a recent paper Becker et al. have analysed the pattern of voting at the UK’s 2016 Referendum on membership of the EU, testing four hypotheses with a wide range of variables. Re-analysis of their data identifies a number of problems of confounding resulting from collinearity among the independent variables which create difficulties in interpretation of the results. This brief comment suggests an alternative analytical procedure that removes the confounding problems and generates regression results that are readily interpreted.

Keywords: Confounding, UK, Voting, Brexit

JEL Classification: C18, C36, D72

Suggested Citation

Johnston, Ron and Jones, Kelvyn and Manley, David, Predicting ‘Who Voted for Brexit’ Through Ecological Analysis – An Example of the Problem of Confounding, and its Resolution (January 31, 2018). Available at SSRN: https://ssrn.com/abstract=3114055 or http://dx.doi.org/10.2139/ssrn.3114055

Ron Johnston (Contact Author)

University of Bristol - School of Geographical Sciences ( email )

University Road
Bristol BS8 1SS
United Kingdom

Kelvyn Jones

University of Bristol - School of Geographical Sciences ( email )

University Road
Bristol BS8 1SS
United Kingdom

David Manley

University of Bristol ( email )

University of Bristol,
Senate House, Tyndall Avenue
Bristol, Avon BS8 ITH
United Kingdom

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
138
Abstract Views
627
Rank
376,949
PlumX Metrics