Matching Versus Pricing: Compatibility and Prominence Effects
A study by Tversky et al. demonstrates that the method used to elicit preferences—choice, pricing, probability matching, or payoff matching—affects the relative weight given to attributes, leading to preference reversals due to scale compatibility and the prominence of probability over money.
THE IDEA
Study 5: Matching versus Pricing
This study examines how different elicitation procedures—choice, pricing, probability matching, and payoff matching—affect the relative weight given to probability and payoff attributes in evaluating bets.
Compatibility Hypothesis
The attribute on which a match is made (probability or payoff) is overweighted relative to the other attribute. Therefore, probability matching should favor high-probability (H) bets, while payoff matching should favor low-probability (L) bets.
Prominence Effect
Probability is more prominent than money in choice tasks, leading to a stronger preference for H bets in choice compared to matching tasks.
Results
The results (Table 12.1) show that the tendency to favor H bets was highest in choice (76%) and probability matching (73%), and substantially smaller in payoff matching (49%) and pricing (37%). This demonstrates both choice–pricing and probability matching–payoff matching preference reversals.
Combined Effects
The data reflect both compatibility and prominence effects, with choice and probability matching yielding similar results due to the two effects having roughly equal impact.
When to use it
When designing market research surveys, be aware that asking for willingness to pay (pricing) vs. direct choice can yield different product preferences.
In public policy valuation, the method of eliciting values (choice vs. matching) can affect the perceived importance of environmental benefits.
In experimental psychology, researchers must consider that the mode of response can influence the weighting of attributes and thus the observed preferences.
FROM THEORY TO EVERYDAY LIFE
In practice
Probability Matching
Subjects were asked to set the missing probability in a low-probability bet to make it equally attractive as a high-probability bet. For example, given H bet (33/36, $50) and L bet (18/36, $125), subjects set the probability in L such that the bets were indifferent, revealing a preference that overweighted probability.
Payoff Matching
Subjects set the missing payoff in a bet to equate two options. This procedure overweighted the monetary dimension, leading to more preferences for the L bet.
A MOMENT TO REFLECT
Put it to work.
How might the compatibility effect influence the results of surveys that use different response formats?
In what real-world situations might the prominence of an attribute lead to a discrepancy between what people choose and what they state they would pay?
Can you design a study to isolate the prominence effect from the compatibility effect?
References
- Tversky, Sattath, and Slovic (1988) — Tversky, A., Sattath, S., & Slovic, P. (1988)
- Goldstein and Einhorn (1987) — Goldstein, W. M., & Einhorn, H. J. (1987)
- Slovic and Lichtenstein (1968) — Slovic, P., & Lichtenstein, S. (1968)
CONTINUE THE CONVERSATION
Source texts
- Heuristics and Biases The Psychology of Intuitive Judgment (Thomas Gilovich, Dale Griffin, Daniel Kahneman)