Department of Economics and Business Economics

Uncovering Voter Preference Structures Using a Best-Worst Scaling Procedure: Method and Empirical Example in the British General Election of 2010

Research output: Contribution to conferencePaperResearchpeer-review

Standard

Uncovering Voter Preference Structures Using a Best-Worst Scaling Procedure: Method and Empirical Example in the British General Election of 2010. / Ormrod, Robert P.; Savigny, Heather.

2011. Paper presented at PSA Conference 2011, London, United Kingdom.

Research output: Contribution to conferencePaperResearchpeer-review

Harvard

Ormrod, RP & Savigny, H 2011, 'Uncovering Voter Preference Structures Using a Best-Worst Scaling Procedure: Method and Empirical Example in the British General Election of 2010', Paper presented at PSA Conference 2011, London, United Kingdom, 18/04/2011 - 21/04/2011.

APA

CBE

MLA

Vancouver

Author

Bibtex

@conference{7b65b83248034134bb9513eb5196c2e5,
title = "Uncovering Voter Preference Structures Using a Best-Worst Scaling Procedure: Method and Empirical Example in the British General Election of 2010",
abstract = "Best-Worst scaling (BWS) is a method that can provide insights into the preference structures of voters. By asking voters to select the ‘best’ and ‘worst’ option (‘most important’ and ‘least important’ media in our investigation) from a short list of alternatives it is possible to uncover the relative importance of each of the options. Using a Balanced Incomplete Block research design we reduce the number of comparisons to the number of media, thus reducing fatigue and complexity problems associated with the standard paired-preference scale from which BWS is developed. Scale variables can be calculated to conduct statistical procedures such as multiple regression and MANOVA. We demonstrate the utility of the method for analysing events in the political sphere using data collected from 282 voters immediately after the British General Election of 2010 on voter preferences regarding the relative amount of information provided by seven alternative media. Results indicate that voters in our sample gained the most information about the election campaign through the broadcast and print media and the leadership debate, whilst Party-controlled media such as party websites and election posters provided the least information. We furthermore investigate group differences using an ANOVA procedure to demonstrate how contextual variables can enrich our empirical investigations using the BWS method.",
keywords = "Best-worst scaling, research method, 2010 British General Election, voter preferences, media",
author = "Ormrod, {Robert P.} and Heather Savigny",
year = "2011",
month = "4",
day = "18",
language = "English",
note = "PSA Conference 2011 ; Conference date: 18-04-2011 Through 21-04-2011",

}

RIS

TY - CONF

T1 - Uncovering Voter Preference Structures Using a Best-Worst Scaling Procedure: Method and Empirical Example in the British General Election of 2010

AU - Ormrod, Robert P.

AU - Savigny, Heather

PY - 2011/4/18

Y1 - 2011/4/18

N2 - Best-Worst scaling (BWS) is a method that can provide insights into the preference structures of voters. By asking voters to select the ‘best’ and ‘worst’ option (‘most important’ and ‘least important’ media in our investigation) from a short list of alternatives it is possible to uncover the relative importance of each of the options. Using a Balanced Incomplete Block research design we reduce the number of comparisons to the number of media, thus reducing fatigue and complexity problems associated with the standard paired-preference scale from which BWS is developed. Scale variables can be calculated to conduct statistical procedures such as multiple regression and MANOVA. We demonstrate the utility of the method for analysing events in the political sphere using data collected from 282 voters immediately after the British General Election of 2010 on voter preferences regarding the relative amount of information provided by seven alternative media. Results indicate that voters in our sample gained the most information about the election campaign through the broadcast and print media and the leadership debate, whilst Party-controlled media such as party websites and election posters provided the least information. We furthermore investigate group differences using an ANOVA procedure to demonstrate how contextual variables can enrich our empirical investigations using the BWS method.

AB - Best-Worst scaling (BWS) is a method that can provide insights into the preference structures of voters. By asking voters to select the ‘best’ and ‘worst’ option (‘most important’ and ‘least important’ media in our investigation) from a short list of alternatives it is possible to uncover the relative importance of each of the options. Using a Balanced Incomplete Block research design we reduce the number of comparisons to the number of media, thus reducing fatigue and complexity problems associated with the standard paired-preference scale from which BWS is developed. Scale variables can be calculated to conduct statistical procedures such as multiple regression and MANOVA. We demonstrate the utility of the method for analysing events in the political sphere using data collected from 282 voters immediately after the British General Election of 2010 on voter preferences regarding the relative amount of information provided by seven alternative media. Results indicate that voters in our sample gained the most information about the election campaign through the broadcast and print media and the leadership debate, whilst Party-controlled media such as party websites and election posters provided the least information. We furthermore investigate group differences using an ANOVA procedure to demonstrate how contextual variables can enrich our empirical investigations using the BWS method.

KW - Best-worst scaling, research method, 2010 British General Election, voter preferences, media

M3 - Paper

ER -