Discrete Choice Experiment (DCE)
A discrete choice experiment (DCE) is a stated-preference method in which respondents choose among hypothetical alternatives described by attributes, and a random utility model estimates each attribute's influence on choice.
이산선택실험(DCE)은 응답자가 속성으로 묘사된 가상의 대안 중 하나를 선택하게 하고 무작위효용모형으로 각 속성이 선택에 미치는 영향을 추정하는 진술선호 방법이다.
How do I measure how people trade off attributes when choosing between options?
사람들이 선택할 때 여러 속성을 어떻게 맞바꾸는지(트레이드오프) 어떻게 측정하나요?
A discrete choice experiment presents repeated choice sets of alternatives with systematically varied attributes and analyzes the choices with random utility models such as conditional logit or mixed logit. It fits valuing attributes of policies, health services, or products, including willingness to pay when a cost attribute is included.
When to use Discrete Choice Experiment (DCE)
- Valuing attributes of health services, policies, or environmental goods
- Estimating willingness to pay with a price or cost attribute
- Choice data in long format with one row per alternative per choice set
When to use something else
- Individual-level utilities for market simulation are the focus: use hierarchical Bayes conjoint
- Only a ranking of many items is needed without attributes: use best-worst scaling (MaxDiff)
Assumptions and requirements
- An efficient experimental design with appropriate attribute levels
- Respondents choose by maximizing utility (random utility theory)
- Independence of irrelevant alternatives for conditional logit, or a mixed logit when it fails
- Sufficient respondents and choice tasks for the number of parameters
What the analysis reports
- Attribute coefficients with standard errors
- Marginal willingness to pay with confidence intervals
- Preference heterogeneity (standard deviations in mixed logit)
- Model fit: log-likelihood, AIC, BIC, pseudo R²
How to report it
In the mixed logit model, respondents preferred [attribute level] (β = [value], p = [value]), corresponding to a marginal willingness to pay of [value] (95% CI [[value], [value]]).
Related methods
- Hierarchical Bayes Conjoint
- Conditional Logit Model
- Mixed Logit Model
Run Discrete Choice Experiment (DCE) on Quorant
- Data you need
- a table (CSV or Excel) with an outcome variable and predictor (feature) columns
- Price
- 250 credits for one analysis up to 10,000 rows (complexity level 2); larger or wider data costs more. Optional methodologist review: 1200 credits.
- Checking
- Python results are cross-checked against R reference outputs in every deployment
- Before you pay
- A free pre-analysis diagnosis checks your data and gives an exact quote; nothing is deducted until you approve it.