What is it about?

Various polling companies study the opinions and intentions of the voters in the time before a referendum takes place. In this article, we propose a statistical method that predicts the outcome of the referendum based on data gathered from multiple opinion polls from various polling companies over time before the referendum. We illustrate the method with two important referendums in the UK: 2014 Scottish Independence referendum and 2016 UK membership in the European Union (or Brexit) referendum.

Featured Image

Why is it important?

Opinions change over time and polling companies deploy varying strategies for collecting information from the population. Thus, it is important to take into account changes of opinions over time as well as differences between pollsters when preparing forecasts of the referendum outcomes.

Read the Original

This page is a summary of: Hierarchical model for forecasting the outcomes of binary referenda, Computational Statistics & Data Analysis, May 2019, Elsevier,
DOI: 10.1016/j.csda.2018.09.007.
You can read the full text:




The following have contributed to this page