What is social desirability bias?
Also called: Social desirability response bias, Socially desirable responding
Social desirability bias is the tendency for research participants to answer in ways that make them look good, overreporting behaviour they think is approved of and underreporting behaviour they think is frowned on.
Updated
Where the idea comes from
Allen Edwards named the effect in The Social Desirability Variable in Personality Assessment and Research (1957). He showed that how likely people were to endorse a personality statement tracked how socially desirable that statement was judged to be. Douglas Crowne and David Marlowe published a scale to measure the tendency in 1960, and versions of it are still used in academic studies.
What it looks like in product research
- Customers say they read release notes, back up their files and review permissions regularly. Product data says otherwise.
- B2B interviewees describe their team's process as more organized and documented than it is, especially with a manager on the call.
- Usability participants blame themselves ("I'm sure it's just me") instead of saying the design confused them, particularly when the designer is moderating.
- Survey respondents overstate how carefully they compare vendors before buying.
Why it's stronger in some studies
The pressure rises when the participant knows who is asking and has a relationship with them: their account manager, the company they pay, or the person who built the feature. It is usually stronger in live interviews than in self-completed surveys, because there is a person to impress. Sensitive topics such as security habits, compliance, spending or skill level make it worse.
How to reduce it
- Ask about specific recent instances: "Walk me through the last time you shared a file outside the company" instead of "Do you follow your sharing policy?"
- Normalize the less flattering answer: "Some teams we talk to skip this step entirely. How does it work on yours?"
- Use indirect questions: "How do other people on your team handle this?" People often describe their own habits through others.
- Observe behaviour where you can, in a usability test or a screen share, instead of asking people to report it.
- Keep the builder out of the moderator seat, and consider hiding which company is running the study.
- For surveys, make responses anonymous when you can and say so plainly in the invite.
Go deeper
- User interviews guide
Lists social desirability among the common interview pitfalls.
- How to conduct customer interviews
Interview techniques that keep answers grounded in what happened.
- Survey research guide
Where social desirability sits among the other survey biases.
- Concealed identity studies in Great Question
Run research without revealing your company to participants.
Questions
Is social desirability bias the same as response bias?
It's one type of response bias. Response bias covers any systematic distortion in answers, including acquiescence and question-order effects. Social desirability bias is specifically about people presenting themselves favourably.
Does making a survey anonymous remove social desirability bias?
It reduces it, but doesn't remove it. People also want to see themselves in a good light, and some don't believe a survey is anonymous when it arrives from a company they use. Anonymity helps most on sensitive questions.
How can you tell if social desirability bias affected your results?
Compare what people said with what they did. If interviewees all describe a careful weekly review and your analytics show the review page is rarely opened, the gap is your answer. Academic researchers use dedicated scales, which product teams rarely need.
Run the study, not just the method
Great Question handles recruiting from your own customers, scheduling, incentives, interviews, tests and analysis in one place.
No credit card required.
Related terms
Full glossary- Response bias
Response bias is a systematic pull on what participants say, caused by question wording, who is asking, or how people want to come across, so their answers drift away from what they actually think or do.
- Leading questions
A leading question is worded in a way that suggests the answer the asker expects or wants, so participants drift toward that answer and you learn more about your own assumptions than about their experience.
- Triangulation
Triangulation is checking a research finding against more than one source, method or researcher, so you only act on it when different kinds of evidence point the same way.