PsychologicalDUAL-USE

Social Desirability Bias

What it is

People answer questions and make public choices in the way that will be approved of rather than the way that is true — over-reporting virtue and under-reporting vice — and anyone who controls the question or the audience can steer the answer.

How it works

Crowne and Marlowe built the first widely used scale in 1960 to measure how strongly a respondent needs approval, and Paulhus showed in 1984 that the tendency has two parts: self-deceptive enhancement, in which people sincerely believe the flattering answer, and impression management, in which they know better and answer for the audience. The lever is reputational: most questions carry a visible “right” answer, the cost of giving it is zero, and the cost of the honest answer is a small loss of face. Tourangeau and Yan's 2007 review found the effect largest for sensitive topics, in interviewer-administered modes, and when answers are not anonymous. Two consequences matter for persuasion. First, self-report figures are inflated: claimed church attendance, voting, hand-washing, charitable giving and reading all exceed observed rates. Second, the answer can be engineered: ask in public, in front of a peer, or with the virtuous option already named, and the socially desirable response becomes the sale, the pledge, or the poll number. Gerber, Green and Larimer demonstrated the benign version in 2008, raising turnout substantially by telling households that their voting record would be shared with neighbours.

Real-world examples

  • •Polls overstated Tom Bradley's support in the 1982 California governor's race and understated the Conservative vote before the 1992 British general election; both were partly attributed to respondents giving interviewers the expected answer. The American Association for Public Opinion Research's post-mortem of the 2016 U.S. polls, by contrast, found little evidence of a comparable “shy Trump” effect and pointed instead to late deciders and weighting by education.
  • •Charity “round-up” prompts at a staffed checkout and pledge drives conducted in front of colleagues convert a private preference into a public one, and giving rises accordingly.
  • •Customer-satisfaction surveys delivered by the salesperson who made the sale, or by the clinician who delivered the care, return far higher scores than the same questions asked anonymously.
  • •Validated turnout studies routinely find that more survey respondents report having voted than the official records show actually did.

Ethical guidelines

Where the line is

Asking people what they think in ways that let them answer honestly, and using public accountability only for behaviour they already endorse and with disclosure, is legitimate; staging the question in front of an audience or wording it so that the approved answer is the one you profit from, then reporting that answer as sincere preference, is manipulation.

  • ●Design questions so that the honest answer is as easy to give as the approved one: neutral wording, anonymous modes for sensitive topics, and no interviewer present when the topic is one people lie about.
  • ●Do not report self-described virtue — intention to buy, willingness to recommend, claimed use — as if it were measured behaviour.
  • ●Using public accountability to encourage behaviour people already endorse, such as voting, is defensible when disclosed; engineering a public moment so that saying no costs face is coercion.

How to defend against it

  • ►When a figure is about virtue or vice — how many people recycle, would recommend, never text while driving — ask how it was collected; if it was self-report to a person, discount it heavily and look for observed or administrative data.
  • ►For your own decisions, refuse to answer commitment questions in public: “let me get back to you” moves the choice out of the audience's sight and takes the reputational thumb off the scale.
  • ►When a survey or petition names the approved answer in the question, treat the result as a measure of what people think they are supposed to say, not of what they believe.
  • ►Read poll methodology notes for mode (live interviewer, online, automated) before comparing numbers across polls; mode differences on sensitive questions can exceed the gap being reported.

References

  1. Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24(4), 349-354
    The original need-for-approval scale and the construct of socially desirable responding.
  2. Paulhus, D. L. (1984). Two-component models of socially desirable responding. Journal of Personality and Social Psychology, 46(3), 598-609
    The distinction between self-deceptive enhancement and impression management.
  3. Tourangeau, R., & Yan, T. (2007). Sensitive questions in surveys. Psychological Bulletin, 133(5), 859-883
    Review showing misreporting is largest for sensitive topics, in interviewer-administered modes, and without anonymity.
  4. Gerber, A. S., Green, D. P., & Larimer, C. W. (2008). Social pressure and voter turnout: Evidence from a large-scale field experiment. American Political Science Review, 102(1), 33-48
    The neighbours mailer field experiment: disclosed social pressure raised turnout substantially.
  5. Kennedy, C., Blumenthal, M., Clement, S., Clinton, J. D., Durand, C., Franklin, C., McGeeney, K., Miringoff, L., Olson, K., Rivers, D., Saad, L., Witt, G. E., & Wlezien, C. (2018). An evaluation of the 2016 election polls in the United States. Public Opinion Quarterly, 82(1), 1-33
    The AAPOR post-mortem finding little evidence of a shy-Trump effect and attributing polling error to late deciders and education weighting.
Last reviewed
Suggest a correction

Detect Social Desirability Bias in any text

Paste any message, email, or article into our free Manipulation Detector to see if Social Desirability Bias or other techniques are being used on you.

Try the Detector FreeCreate Free Account