Self-reporting is a frequently used method to measure various constructs in many areas of social science research. Literature holds abundant evidence that social desirability bias (SDB), which is a special kind of response bias, can severely plague the validity and accuracy of the self-report survey measurements. However, in many areas of behavioral research, there is little or no alternative to self-report surveys for collecting data about specific constructs that only the respondents may have the information about. Thus, researchers need to detect or minimize SDB to improve the quality of overall data and their deductions drawn from them. Literature provides a number of techniques for minimizing SDB during survey procedure and statistical measurement methods to detect and minimize the validity-destructive impact of SDB. This study aims to explicate the classical and new techniques for mitigating the SDB and to provide a guideline for the researchers, especially for those who focus on socially sensitive constructs.
TopIntroduction
Surveys are a widespread method of data collection for quantitative research. In this method, self-reported measures are often and sometimes inevitably used to measure participants’ attitudes, beliefs, feelings, values, intentions, behaviors, personalities, and many other directly unobservable constructs within the context of social science research. However, the vulnerability of these kinds of measures to response bias is still a matter of debate today. Response biases are non-relevant sources of systematic error and need to be controlled or registered by researchers before, during or after the application of self-report surveys. Among these biases, social desirability bias (SDB hereafter) is considered as the most complicated and pervasive bias especially while measuring socially sensitive constructs.
Social desirability generally refers to the “tendency for an individual to present him/her, in a way that makes the person look positive with regard to culturally derived norms and standards in test-taking situations” (Ganster, Hennessey and Luthans, 1983). It subsists especially in sensitive questions which are designed to explore potentially embarrassing and invasive matters and yields a high level of non-response rates and intense underreporting (Tourangeau & Smith, 1996). More specifically, SDB is considered as an evident threat to the validity of research which involves measurement of self-report scales, since it may “(a) produce spurious results; (b) hide real results (suppression); and (c) moderate relationships” (Ganster et al., 1983). Due to its not only validity-destructive impact and but also obscure and intricate nature (King & Bruner, 2000), SDB has received the attention of personality researchers and behavioral scientists since the 1930s (Paulhus, 2002). But it was first Edwards (1957) who enunciated the plight as “social desirability bias” and provided the very first scale to the literature. Since then, many methods and scales have been suggested to mitigate social desirability bias and its effects. Today, many researchers utilize these measurement or attenuation methods to identify and manage social desirability bias in order to improve the overall quality of their data-set and the deductions drawn from them.
Because of the extensive cluster of contentious subjects that have accumulated throughout the years, this study aims to explicate the extant literature, provide a clear guideline for young researchers and bring upon an invigorated impetus to current research by introducing some classical and new techniques for mitigating the social desirability bias. In this context, the current chapter is composed of the following sections;
- •
SDB and its relation with sensitivity,
- •
impacts of SDB on the accuracy of measurements,
- •
techniques for attenuating the SDB,
- •
SDB measurement methods,
- •
and finally conclusion and further research.