Servicenavigation

Project Details

Back

Projekt Print View

Sensitive Questions and Social Desirability - Theory and Methods

Subject Area Empirical Social Research
Term from 2017 to 2022
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 387002847
 
Final Report Year 2023

Final Report Abstract

Survey studies are of great importance in the social sciences. They often include sensitive topics (such as self-reported delinquency, voting behavior, health issues, etc.), which are often of great interest to the public and the scientific community. However, it has been shown that many respondents do not answer such questions truthfully, but rather distort their statements in a socially desirable direction ("misreporting") meaning the resulting data are not valid. Against this background, the project addressed the questions of how misreporting can be explained theoretically and which design aspects and questioning techniques can help to increase data validity. Regarding the questioning techniques, a focus was placed on the Item Count Technique (ICT). Instead of individual questions, this technique uses lists of questions with items to be answered simultaneously, thus completely anonymizing the survey situation for the respondents, but still allowing estimates of the prevalence of sensitive characteristics to be generated. Four online surveys with different samples were conducted in the project. The surveys included, among other things, different variants of sensitive questions, survey experiments in which several design aspects and question forms were varied, fictitious items, and response time measurements. In addition, a meta-analysis was conducted in which all existing literature on the effectiveness of ICT in reducing misreporting was reviewed. The project results show that misreporting or invalid data is a massive problem. This concerns, for example, self-reported vaccination status over the course of the Covid-19 pandemic, selfreported sexual behaviors, but also so-called pseudo-opinions, i.e. statements of opinion by respondents about things they are unfamiliar with: Up to 75% make statements about freely invented question items. ICT as a special question technique for sensitive items can certainly mitigate the problem (according to the result of the meta-analysis), but it has issues with reliability. For an alternative special questioning technique, the so-called crosswise model, which also anonymizes the question situation, it is shown that this technique has the opposite effect of the intended increase in validity; namely, it leads to false-positive estimators. A theoretically relevant result of the project is that that invalid answers arise unconsciously/spontaneously rather than through careful consideration/thinking on the part of the respondents. In addition to findings for theoretical and methodological basic research in the field of survey methodology, recommendations for the practice of survey research can also be derived from the project.

Publications

 
 

Additional Information

© 2025 DFG Disclaimer / Copyright / Privacy Policy

Textvergrößerung und Kontrastanpassung