The study investigates the social desirability of four characteristics often asked as self-descriptive questions in surveys. These are the intensity of watching television, the consumption of alcohol, the frequency of past blood donations and the degree of happiness in life. We compare the validity of three differently elaborated operationalizations of the magnitude and direction of the perceived incentives for social desirable response behavior. The standard method only accounts for the social desirability of a strong occurrence of the respective attribute, however, the second procedure takes differences between strong and weak occurrences into account, and the third implementation considers additionally the evaluation of a medium occurrence. Using data from a local survey based on a random probability sample, first, we find a substantial proportion of the respondents reporting u-shaped desirability profiles for the consumption of alcohol and intensity of watching television. For these topics only the third and most sophisticated operationalization is an appropriate measure of the potential distortions caused by social desirability. Secondly, monotonic desirability profiles are observed for past blood donations and happiness in life. Here, the additional effort necessary for the global difference scores as the second most elaborated measure, improves considerably the magnitude of incentives from social desirability which can be identified. Contemporary research has almost exclusively applied the standard method; however, our investigation suggests that this is under no conditions an adequate measure of the strength and the direction of potential social desirability bias.
Additional information:
Das Dokument wird vom Publikationsserver der Universitätsbibliothek Mannheim bereitgestellt.
![[img]](https://madoc.bib.uni-mannheim.de/style/images/fileicons/application_pdf.png)
Search Authors in