Description
Describe two options on one shared attribute and one attribute unique to each, and test whether the shared attribute gets more weight in choice than in single-option ratings.
When people compare two options, an attribute both options share is easy to compare and tends to dominate, while attributes present for only one option are discounted. Slovic and MacPhillamy showed this with pairs of students described by one common test score and one score unique to each: the common score carried more weight in predictions, respondents said afterwards they had not intended this, and choices diverged from separate ratings of the same profiles. The test makes the bias measurable for any comparison interface. Because respondents are unaware of the bias, self-reported importance cannot detect it; only the contrast between common and unique presentations, or between choice and separate rating, does.
Attributes
- Discipline
- Behavioral Science & Economics
- Loss type
- Unmigrated
- Stage
- Design, Evaluation
- Data
- Quantitative
- Effort
- Days
Steps
- Build option pairs each described on two attributes: one present for both options and one different for each.
- Ask respondents to choose or predict between the options in each pair.
- Separately collect ratings of each option on its own, from the same or a matched group.
- Estimate each attribute's weight when common and when unique, and count choice-rating reversals.
- Ask afterwards whether respondents intended to favour the common attribute, to document awareness.
Sources
- Decision Making: Descriptive, Normative, and Prescriptive Interactions (Bell et al., 1988)
Also known as
- Dimensional commensurability test
- Compatibility bias test (common versus unique cues)
Categories
- Analytics & Behaviour
- Behavioral Science & Economics