Tuesday 14th July
Wednesday 15th July
Thursday 16th July
Friday 17th July
Wednesday 15th July, 16:00 - 17:30 Room: O-101
Social Desirability and Non-Reactive Methods in Survey Research: Improving Theory and Data Collection 1
|| Dr Ivar
Krumpal (University of Leipzig )
|Coordinator 1||Professor Roger Berger (University of Leipzig)|
|Coordinator 2||Professor Mark Trappmann (IAB Nürnberg)|
Social desirability bias is a problem in surveys collecting data on private or norm-violating issues (e.g. sexual behavior, health related issues, voting preferences, income, or unsocial opinions) as soon as the respondent’s true score differs from his or her perception of the social desirable score. Due to the respondents’ strive for social approval and keeping a favourable self-image as well as data protection concerns, collecting valid data on private or norm-violating issues is a challenging task. More specifically, respondents may engage in impression management or self-deception or edit their answer before reporting it. Non-reactive data collection methods could improve measurements and data quality in surveys where social desirability bias is a potential problem. Therefore, the possibilities and limits of non-reactive methods (e.g. record linkage approaches, surveys without questions, biomarkers, field experiments or administrative data usage) will be critically discussed and compared to methods which are based on self-reports.
This session has four main goals: (1) discussion of the theoretical foundation of the research on social desirability bias in the context of a general theory of human psychology and social behavior. A clearer understanding of the social interactions between the actors that are involved in the data collection process could provide empirical researchers with a substantiated basis for optimizing their survey design and data collection to achieve high quality data; (2) presentation of current empirical research focusing on non-reactive methods of data collection in connection with the problem of social desirability; (3) discussion of new designs combining or contrasting non-reactive methods with standard ‘question-and-answer’ survey measurement in innovative ways; (4) exploration of possibilities of integrating such new and innovative approaches in well-established, large-scale population surveys taking into account problems of research ethics and data protection.
Paper Details1. Surveys in Legislative Research: Methodological Challenges and Opportunities
Mr Pirmin Bundi
(University of Zurich)
Professor Frédéric Varone (University of Geneva)
Professor Thomas Widmer (University of Zurich)
In this article, we illustrate the methodological challenges of surveys in legislative research and demonstrate their effects on the analytical power of the sample. By using a methodological triangulation, we are able to identify the effects of non-response, self-reporting, and social desirability. In doing so, we compare responses of parliamentarians in a survey with their real behavior in the parliament. Our article contributes to the legislative research by illustrating methodological challenges in parliamentary surveys and by pointing out factors influencing a biased sample by social desirability.
2. Application of Paradata to Identify Social Desirability in Surveys
Mr Fabian Schüßler
(University of Kaiserslautern)
Professor Jochen Mayerl (University of Kaiserslautern)
Sensitive questions result in misrepresented responses and identifying those questions is crucial to get straight answers to sensitive topics. By tracking response and page change in addition to item non-response and response latencies, the researcher has a powerful paradata tool to identify social desirable response patterns in surveys.
3. Social Desirability and Experimental Vignettes in Survey Research. A Validation Study Using Non-Reactive Data
We examine 30 questions of personality traits of a student tablet survey and data of two additional surveys with a multi-level structural equation modelling approach. Sensitive questions and response behavior should be associated with response latencies and higher rates of event tracking in interaction with individual need for social approval.
Dr Knut Petzold
(Department of Sociology, CU Eichstätt-Ingolstadt)
Due to the more detailed description of decision situations in experimental vignettes it is generally believed that they provide less room for interpretation and evoke therefore less normative and socially desirable answers. However, systematic studies on the validity of the measurement of action intentions by vignettes are rare and inconsistent. In the present study the results of vignettes are compared with the results of non-reactive methods in tow situations, fundamentally different with regard to the consequences of the decisions: horn-honking and hiring. The analysis suggests that the validity of vignettes is possibly moderated by feature of the situation.
4. External Validation of a Factorial Survey with Longitudinal and Administrative Data
Mr Konstantin Mozer
(Goethe-Universität Frankfurt am Main)
Do stated preferences measured with factorial surveys and revealed preferences diverge? If yes, for what reasons? To gain better knowledge on the external validity of FSs and their possibility to reduce social desirability bias, we discuss first theoretical reasons for stated and revealed preferences to diverge. Second, results from two validation studies concerning sensitive questions are presented. The studies use for the first time panel data and additionally offer the possibility to draw on non-reactive register-data.