Appendix A - Design of the 2010 British election study

Summary

The design of surveys conducted in the 2010 British Election Study is displayed in Figure A.1. The figure shows there are three distinct components: the in-person pre-campaign and post-election panel survey, the multiwave rolling campaign panel study (RCPS) and the monthly Continuous Monitoring Surveys (CMS). The first of these surveys repeated the classic in-person probability sample which has been the staple of the BES since the first study was conducted by Butler and Stokes in 1963. Although the BES originally included inter-election panel components, this design was replaced in the 1980s by a single post-election survey with no pre-campaign component. The latter design was changed in 2001 when the Essex team assumed responsibility for the study. For the first time the in-person survey consisted of a pre-campaign and a post-election panel survey.

In 2010, fieldwork for the in-person survey was performed by BMRB, under supervision of Study Director, Nick Howat. As Figure A.1 shows, the 2010 pre-campaign survey consisted of slightly less than 2000 respondents. These respondents were approached again immediately after the election, and the post-election sample was increased by a top-up survey giving a total of 3075 respondents. The top-up component was included to increase the representative quality of the survey given that a panel design was employed.