Accounting for Social Desirability Bias in Online Research
February 15, 2012
An article posted a year ago in Research Design Review – “13 Factors Impacting the Quality of Qualitative Research” – delineated three broad areas and 13 specific components of qualitative research design that can influence the quality of research outcomes. One factor, under the broad category of “The Environment,” is the “presence of observers/interviewers as well as other participants.” In other words, how does the inclusion of other people – whether it be client observers, interviewers, fellow participants, videographers, or note takers – affect the attitudes, behaviors, and, responses we gain from our research efforts? Does research, almost by definition, create an artificial social context where participants/respondents seek others’ approval leading to a false understanding of their realities?
Social desirability bias is not a new concern in research design and its influence on the ultimate usefulness of our qualitative and quantitative research has been the focus of attention for quite some time. Tourangeau, Rips, and Rasinski (2000) discuss social desirability in the context of sensitive questions:
“[The] notion of sensitive questions presupposes that respondents believe there are norms defining desirable attitudes and behaviors, and that they are concerned enough about these norms to distort their answers to avoid presenting themselves in an unfavorable light.”
Nancarrow and Brace – in their article “Saying the ‘right thing’: Coping with social desirability in marketing research” (2000) – address the under- and over-reporting associated with social desirability bias and outline numerous techniques that have been used to deal with the problem – e.g., emphasizing the need for honesty, promises of confidentiality, and question manipulation by softening the suggestion that the respondent should know the answer to a particular question or behave in certain way.
Online technology and the ever-growing online research designs that are emerging – within social media, mobile, bulletin boards, communities, and survey research – have allayed social-desirability concerns. The belief among some researchers is that one of the beauties of the virtual world is that inhabitants basically live in solitude, away from the influences of a social reality. Paul Rubenstein of Accelerant Research has actually stated that a key advantage to online qualitative research is the obliteration of social desirability bias and hence the heightened validity of online vs. offline designs:
“[The] Internet affords a stronger sense of anonymity [compared to in-person interviewing] among study participants and typical response biases such as social desirability and other “faking strategies” are virtually eliminated online. Qualitative data collected online tend to be “brutally honest” in nature as respondents feel wrapped in a cocoon of privacy and facelessness and have no apprehension about telling a moderator anything.”
The idea that researchers who design online studies can ignore potential bias due to social desirability seems misguided. In fact, a good case can be made that the Internet and online technology have unleashed a dynamic capacity for posturing and the need for approval. Popularity and even celebrity – so elusive to the everyday person in earlier times – have become preoccupations. You only need to witness the apparent race for Facebook friends, LinkedIn connections, Twitter followers, and YouTube or blog views – as well as the “vanity” online self-publishing craze – to gain some insight into the potential competitiveness – i.e., pursuit of social stature – fueled by the realm of online. In this way, the virtual social environment has encouraged a look-at-me way of thinking and behaving.
So, how real are those at-the-moment snippets transmitted by mobile research participants (which may be meant to impress the researcher more than inform)? How honest are those product reviews or blog comments? What is the extent of bravado being exhibited in our online communities, bulletin boards, and social network exchanges? The answer is we don’t know, and yet it doesn’t take a great leap of faith to acknowledge that the individual attitudes and behavior we capture online are potentially distorted by an underlying need for social approval.
To paraphrase Mark Twain, the reports of the death of social desirability bias in online research are greatly exaggerated; and, to the contrary, social needs have blossomed in the online world. More than ever, people are asking, “Do you like me?” and, in doing so, presenting the researcher with a critical design issue that impacts the quality of our outcomes.
Nancarrow, C., & Brace, I. (2000). Saying the “right thing”: Coping with social desirability bias in marketing research. Bristol Business School Teaching and Research Review, 3(11).
Tourangeau, R., Rips, L., & Rasinski, K. 2000. The Psychology of Survey Response. Cambridge University Press.