Vignette Method and Factorial Survey

Description

Presents systematically varied short scenarios and asks for judgements, revealing which factors drive decisions and attributions without asking people to explain their own reasoning.

People are poor at reporting which factors drive their judgements, but they judge concrete scenarios easily. The factorial survey exploits this: the researcher defines the judgement of interest and a set of dimensions (for example price, social proof, data requested, brand), generates vignettes by crossing their levels, and has each participant rate a random subset of 10 to 20. Because dimensions vary independently, multilevel regression estimates the causal weight of each factor on the judgement, controlling for respondent effects. Fractional designs keep the vignette universe manageable while preserving key effects. A short qualitative follow-up explores reasoning. Rigour comes from orthogonal design, randomisation of vignettes and order, plausible combinations, and analysis that respects the nested data structure.

Attributes

Loss type
Unmigrated
Effort
Weeks

Steps

  1. Define the target judgment: What decision, evaluation, or attribution are you investigating?
  2. Identify experimental dimensions: What scenario characteristics do you want to vary? Each dimension should have 2–4 levels.
  3. Generate vignettes by factorial combination. Use fractional factorial design for large parameter spaces.
  4. Randomised presentation: Each participant sees 10–20 vignettes randomly selected. Rate or respond to each.
  5. Analysis: Multilevel regression with rating as outcome and vignette characteristics as predictors. Coefficients reveal which characteristics significantly influence judgments.
  6. Qualitative follow-up: After the rating task, interview participants about their reasoning on selected vignettes.

Sources

  • Cambridge Handbook of Research Methods
  • Research Methods for Everyday Life
  • Sage Handbook of Qualitative Data Collection (Flick)
  • Auspurg, K. & Hinz, T. (2015) Factorial Survey Experiments

Also known as

  • Factorial survey experiment
  • Vignette experiment
  • Policy capturing

Categories

  • Analytics & Behaviour
  • Survey & Psychometrics