Participants may give socially desirable rather than candid answers
Aliases: social desirability bias · impression management · public conformity
What it is
Social desirability bias in a focus group occurs when participants adjust public accounts to appear normative, competent, or cooperative. It does not make every statement false, nor does it make a private statement automatically authentic. The more precise question is how audience, identity, and anticipated consequences changed what could be said.
Why it happens
In a group, an answer communicates information and manages an impression. Early speakers cue a local norm; status relations and sensitive topics raise the cost of dissent or disclosure. Participants may therefore agree, soften criticism, or translate experience into institutionally approved language. Public commitments can also induce consistency in later turns despite private reservations.
Studying it
Collect private individual judgments before discussion and compare them with later group accounts. Code who speaks first, when common phrasing emerges, and whether counterexamples appear only after probing. Anonymous surveys, interviews, and groups can be compared when relevant, but mode changes questioning, interaction, and disclosure costs together; a difference is not a pure measure of honesty.
Where it stops holding
When the object of study is a public norm, shared narrative, or negotiation, socially managed talk is evidence rather than noise to remove. People may also revise a judgment sincerely after hearing another experience. Anonymous input reduces some evaluative pressure but does not remove self-image concerns, recall error, or guesses about researcher expectations.
Applying it
- Collect sensitive judgments privately before opening collective discussion.
- Normalize a range of behaviors without marking one answer as correct.
- Probe concrete episodes, actions, and consequences while allowing participants to decline disclosure.
- Report pre-discussion differences and changes instead of treating consensus as independent preferences.