Field Experiments in Credence Goods
How to cite
Woods, Daniel; Kerschbamer, Rudolf; Kujansuu, Essi; Lorè, Luisa; Neururer, Daniel. 2023. "Field Experiments in Credence Goods." AEA RCT Registry. AEARCTR-0011643. https://doi.org/10.1257/rct.11643-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
We plan to study pricing behavior in the market for cellphone repair, which is a credence environment. Our main variable of interest is the price of repairing a cellphone. We conduct a field experiment in which actual market transactions are conducted in repairing a deliberately damaged cellphone. As a baseline, in all repair stores, we have the customer directly request that the damaged part is replaced. With a different customer in the same store, with the same cellphone model and problem, we have them state they do not know what is wrong with the phone. There are 6 treatments, and the random allocation of these treatments is done at the store level, using the random shuffling of a computer. The treatments are 3x2, with one dimension varying the gender of the customer. The other dimension varies what additional lines the customer is instructed to state (if any). The first is no additional lines, the second is insisting on getting replaced parts back and requiring contact before any repairs are conducted, and the third is stating that they trust the repairer to conduct the necessary repairs without being contacted first. We also elicit beliefs from the ‘mystery shoppers’ we use as customers, and have them play as a FM in an investment game, where the store in question will be the SM. The SM decision will be in a follow-up survey of stores, where other things like their personality traits and perceptions of their customers are asked.
Experimental design
Baseline (in all stores): Customer requests defective charging port is replaced
Treatments:
• Customer not knowing the cause of the problem
• Customer not knowing the cause of the problem, and insists on receiving any replaced parts back, as well as requiring contact before any repairs are conducted
• Customer not knowing the cause of the problem, and states that they trust the repairer to conduct straightforward repairs without being contacted first
Above 3 treatments are crossed 3x2 with gender. Therefore, there are 6 treatment arms.
Each store is visited twice, by 2 different customers of the same gender. One visit is the baseline, and the other visit is one of the treatments.
The target group is cellphone repair stores that are not part of a chain or franchise, and have their business details on Google Maps. The sample size is 180 stores.
The mystery shoppers we use as customers will complete questionnaires and belief elicitations, and there will be a follow-up survey of the stores.
Full details are attached to a pre-registration submission, which will be released when the trial is complete.
Design details
Cited by
No citation data available for this registration yet.