The Impact of Information on Valuation in Experimental Auctions: A Comparison of Between and Within Subject Designs

Abstract

Experimental auctions are an important technique for measuring preferences for products, product attributes, and the impact of information. While these techniques are widely used, there is a paucity of evidence about an important design decision available to guide researchers: the choice of a between-subject vs. within-subject design. Within-subject designs offer clear value in terms of providing multiple observations per participant, which increases statistical power, but there are long-standing concerns about properties that could decrease the external validity of results generated in within-subject experiments. In this paper, we examine the impact of information on the economically motivated mislabeling of extra virgin olive oil (EVOO) on consumer valuation of EVOOs produced in the country that has experienced mislabeling scandals, along with EVOOs from two unimplicated countries, in between-subject and within-subject designs. Our findings show that the significance and relative impacts on valuation are identical between the two conditions. In fact, the valuation of the implicated EVOO differs only by a few cents ($3.53 vs. $3.60) after participants received information about mislabeling. There were larger differences in valuation of the two unimplicated EVOOs post-information, though the relative preferences implied by the results and the statistical significance did not differ between the conditions. The impacts of information on the product “targeted” by the information are measured consistently in both between-subjects and within-subjects designs, while we observe more variation in off-target products, suggesting that the researchers who are interested in informational spillovers may need to be more careful in design choice than those who want to estimate the impact of information on the target products.