Studying Socially Acceptable Behavior in Russia: Theory and Methodology

Abstract

This article explores the methodology of randomized responses used to control for socially desirable responses in the sensitive questions. In the first part of this article the basic methodology for the study of socially desirable behaviors is discussed using different variations of techniques, such as the related and unrelated question techniques, the forced response method, the Mangat’s technique, various techniques with cross-cutting questions, the list experiments and the summation techniques with a random variable. The second part provides a comparative analysis of the main polling techniques in terms of their efficiency and anonymity as well as detailed discussion of the practical implementation of these techniques. The third part is focused on empirical analysis of the data collected in January 2014 by VTSIOM (Omnibus), using four proposed techniques (the forced response method, the list experiments and two versions of summation techniques). According to the findings the list experiment and one of the summation techniques despite the relatively large standard errors are both characterized by point estimates in a deflationary direction -- their use is most preferable; as far as the forced response technique is concerned, despite the lowest standard error its estimate is inflated in the sensitive direction, which is explained by the fact of respondents violating their instructions. The substantive findings of this study indicate the presence of socially desirable factor in the current Russian presidential approval ratings, urging for a more complete and comprehensive study of this phenomenon in the future.