Social Desirability Bias as Substance and Not Nuisance

• In Event: New Approaches to the Study of European Elections

Abstract

{\em
Social scientists have long debated the question of how threatening social desirability biases are for what scientists can learn from survey respondents. Until today such biases are treated as a nuisance at best. In contrast, we argue that social desirability biases are key to understand for social scientists as they are important proxies for social norms and the need for conformity within societies. We compare six different designs to measure social desirability bias in relation to support for democracy in three European countries. Besides a list experiment, we rely on various approaches that trigger respondents' need to provide socially acceptable answers. We show that triggering social desirability is outperforming list experiments and observational measures for various reasons. Most importantly it allows researchers to test whether social desirability is actually manipulated. It also ensures that social desirability is not treated as a nuisance but as a matter of substantive research interest.

Authors

• Daniel Bischof, Aarhus University

• Tim Lars Allinger, Aarhus University

• Morgan Le Corre Juratic, Aarhus University

• Kristian Vrede Skaaning Frederiksen, Aarhus University