Consumer Decision-making When Shopping for Medical Services

How to cite

Chan, Alex. 2020. "Consumer Decision-making When Shopping for Medical Services." AEA RCT Registry. AEARCTR-0006815. https://doi.org/10.1257/rct.6815-2.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Alex Chan*

* Primary investigator

Registry ID
AEARCTR-0006815
DOI
https://doi.org/10.1257/rct.6815-2.0
First registered
Last updated
Status
Completed
Trial dates
2020-11-28 – 2021-03-31
Countries
United States of America ()
JEL codes
I11, I10, M31, J7, D91, D83
Registry record
http://www.socialscienceregistry.org/trials/6815

Abstract

We designed a field experiment in the United States to evaluate how the price sensitivity and sensitivity to provider personal characteristics of consumers of medical care change under an environment with richer information on provider quality compared to an environment without such information. Our motivation stems from the facts that evidence on whether healthcare consumers are price sensitive remain largely equivocal, while the evidence on the nature of consumer racial and gender discrimination of healthcare provider remain scant. More broadly, our study will shed light on the little known aspects of consumer behavior and preferences of the 30M+ uninsured Americans. We will launch a large-scale choice-based conjoint survey experiment with real stakes with actual healthcare consumers. Finally, this experiment will contribute to the important topic if healthcare consumerism.

Intervention

We compare consumer behavior with incentivized choice-based conjoint surveys. In the control group, the incentivized choice-based conjoint surveys does not include a provider quality score ("Comprehensive Quality Score") as a feature in the conjoint profiles. In the treatment group, the incentivized choice-based conjoint surveys includes a provider quality score ("Comprehensive Quality Score") as a feature in the conjoint profiles.

Experimental design

We conduct hypothetical discrete choice experiments (choice-based conjoint) to measure the baseline effects of price, distance, provider characteristics (e.g. gender and race) among subjects. A choice-based conjoint in a survey design in which respondents are offered the choice between sets of options represented by combinations of attributes and features. In our design, the choice-based conjoint experiment is incentivized by making the list of 10 actual provider options (out of hundreds) made available to the subject for actual purchasing contingent on the subject's responses to the choice-based conjoint. We conduct information treatments (quality information) to see the effects of differential treatments on consumer preferences overall and the consumer sensitivity of provider features (price, gender/race, distance). Treatment heterogeneity will be investigated along subject characteristics like demographics, socioeconomic status, location, subjective interpretation of the quality metric, subjective interpretation of the relationship between quality and travel distance and procedure type (all variables we will gather as part of the experiment). An actual shopping action could be executed by the consumer and this actual transaction data will be used to estimate consumer preferences, validate the choice model estimated using the hypothetical profiles-based discrete choice experiment and evaluate incentive compatibility properties.

Design details

Randomization method
Randomization done by Qualtrics' Randomizer element to assign respondents to either a control block of questions or an experimental block of questions
Randomization unit
Individual
Clusters
Approx. 300-600 patients shopping for one of the 3 medical procedures between December 2020 to February 2021
Observations
Approx. 300-600 patients shopping for one of the 3 medical procedures between December 2020 to February 2021
Arms
50% of 300-600 patients shopping for one of the 3 medical procedures between December 2020 to February 2021 will be in the treatment arm and the other 50% in the control arm
Primary outcomes
Patient preferences for providers of medical services measured using discrete choice experiments. A key "outcome" variable is the probability a patient prefers a provider profile.

Cited by

No citation data available for this registration yet.

Keywords