The Gold Standard of Anthropology | Vignette Interviews, Hypothetical Scenarios and Social Desirability

1. Start with the inference the vignette must support

Back to contents

A vignette is not simply an engaging introduction. Decide whether the study needs perceived norms, moral reasoning, likely choices, explanations of institutional practice or language for a sensitive experience. These are different claims. “What should this character do?” does not establish what the respondent has done, and “what would most people do?” measures a different social frame again. State the unit of inference before writing the story.

2. Build a situation that is incomplete in useful ways

Back to contents

Give enough detail for participants to enter the scene while leaving room for interpretation. Age, gender, class, kinship, place and institutional role can change judgement, so include each feature only for a reason. Excess detail can force a single reading; too little produces guesses about missing context. Pilot the vignette with people who understand the setting and ask which cues they noticed, supplied or ignored.

3. Distinguish fixed comparison from open exploration

Back to contents

A standardised vignette holds wording and presentation constant so responses can be compared. A branching vignette adds information or consequences in stages to show when judgement changes. An open vignette invites participants to complete the story in their own terms. Each design answers a different question. Record whether follow-up prompts were fixed, responsive or improvised, because interviewer additions can become part of the stimulus.

4. Use distance without pretending it is neutral

Back to contents

A hypothetical character may let someone discuss stigma, danger or contested practice without immediate self-disclosure. That distance can reduce threat, but it may also encourage performance of an approved public position. Participants can speak about “people like that” while protecting their own identity, or use the character to communicate personal experience indirectly. Treat shifts between third person and first person as evidence, not as a cue to demand confession.

5. Probe the reasoning, not only the verdict

Back to contents

After an initial answer, ask what information mattered, what remained uncertain, who would agree, what might change the response and how the character’s options are constrained. Compare descriptive norms—what people commonly do—with injunctive norms—what people say ought to be done. Preserve the sequence of probes. A later explanation may be a considered revision rather than the reason that produced the first judgement.

6. Worked example: reporting an unsafe workplace

Back to contents

A vignette describes a temporary worker who sees a supervisor bypass a safety check. Most participants initially say the worker should report it. When the interviewer adds that the worker’s visa renewal depends on the employer, several answers change to seeking advice anonymously or waiting for corroboration. The finding is not that participants are dishonest. The staged story reveals how security, power and perceived retaliation condition a public norm. The report separates the initial principle, the added constraint and the revised action.

7. Track social desirability as a response process

Back to contents

Social desirability can enter through the moral charge of the topic, the interviewer’s identity, the research sponsor, the setting and the participant’s guess about the preferred answer. Neutral delivery helps but does not erase these forces. Compare direct, third-person and “most people” questions; note laughter, hedging and rapid moral certainty; and ask for plausible alternatives. Divergence across frames can show how answers are managed, although it cannot identify one hidden true response.

8. Changed-context transfer check: a school bystander scenario

Back to contents

Move the method to a student interview about witnessing exclusion in a group chat. A named age-matched character may increase realism, but it can also resemble an identifiable classmate. The transferable design principle is to vary only the cue being studied and to protect participants from feeling accused. The context adds safeguarding duties and peer consequences, so researchers must plan what happens if a response becomes a real disclosure rather than treating every statement as hypothetical.

9. Know the common misconceptions

Back to contents

Vignettes do not automatically eliminate social desirability, predict behaviour or make samples representative. A realistic story is not necessarily interpreted consistently, and identical words do not guarantee identical imagined scenes. Differences attributed to one manipulated detail may actually arise from changed plausibility or emotional intensity. Participants who reject the premise are not failed respondents; their refusal may reveal that the scenario violates local knowledge.

10. Audit the story from construction to interpretation

Back to contents

Archive every version, translation, visual or spoken presentation, pilot note, prompt and branching rule. Code both the substantive response and how the respondent positioned the character, the interviewer and the scenario. Compare interpretations across relevant groups and rerun conclusions after excluding cases that understood the premise differently. Report which details were held constant and where conversational flexibility was intentionally allowed.

11. Frequently asked questions

Back to contents

Should a vignette be realistic? It should be recognisable enough for the intended inference, not packed with documentary detail. Can it measure what someone would really do? It can elicit anticipated action and reasoning, not guarantee behaviour. Should characters be anonymous? Usually fictional and non-identifiable, while still culturally intelligible. Can several versions be compared? Yes, if the manipulated difference is isolated and assignment, order and analysis are planned.

12. The gold standard keeps the hypothetical claim modest

Back to contents

Vignettes are powerful when they make norms, constraints and interpretive frames discussable without claiming direct access to private truth. The gold standard defines the intended inference, pilots how the scene is imagined, separates verdict from reasoning and treats social desirability as part of the encounter. A hypothetical answer becomes useful evidence when the path from story to claim stays visible.