Accurate Quality Elicitation in a Multi-Attribute Choice Setting

TL;DR

This paper compares two methods for accurately eliciting multi-attribute quality, demonstrating that the choice of method significantly affects the quality measures obtained.

Contribution

It introduces and compares two MPL methods, showing that m-MPL requires fewer assumptions and that elicitation method impacts quality measurement.

Findings

m-MPL requires fewer assumptions for accuracy

Switch points differ between the two MPLs in experiments

Quality measures are sensitive to the elicitation method

Abstract

This paper studies how to accurately elicit quality for alternatives with multiple attributes. Two multiple price lists (MPLs) are considered: (i) m-MPL which asks subjects to compare an alternative to money, and (ii) p-MPL where subjects are endowed with money and asked whether they would like to buy an alternative or not. Theoretical results show that m-MPL requires fewer assumptions for accurate quality elicitation compared to p-MPL. Experimental evidence from a within-subject experiment using consumer products shows that switch points between the two MPLs are different, which suggests that quality measures are sensitive to the elicitation method.