Integrating Feasibility and Realism: Examining Hybrid Incentive Structures in Discrete Choice Experiments

Last registered on December 03, 2024
View Trial History

Pre-Trial

Trial Information

General Information

Title
Integrating Feasibility and Realism: Examining Hybrid Incentive Structures in Discrete Choice Experiments
RCT ID
AEARCTR-0014816
Initial registration date
November 25, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
December 03, 2024, 1:25 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country
United States of America
Region

Primary Investigator

Name
Bachir Kassas
Affiliation
University of Florida
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Jerrod Penn
PI Affiliation
Louisiana State University
Contact Investigator
PI Name
Wuyang Hu
PI Affiliation
The Ohio State University
Contact Investigator

Additional Trial Information

Status
In development
Start date
2024-12-02
End date
2025-01-06
Keywords
Behavior
Additional Keywords
Discrete Choice Experiments, Hypothetical Bias, Incentive Compatibility
JEL code(s)
C90, D11
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
Hypothetical bias (HB) stands as a major drawback in the literature using discrete choice experiments (DCEs) to estimate consumer preferences and willingness-to-pay (WTP) for market goods and attributes. While a fully incentivized DCE - where one choice set is randomly selected as binding and the subject's choice in that choice set is implemented for real purchase and payment - overcomes the issue with HB, this approach is significantly limited by several logistical hurdles. First, many DCE studies are interested in learning about consumer preferences for products that don't yet exist on the market, rendering a fully incentivized DCE infeasible. Second, even if all products are available, DCEs usually involve multiple choice sets with multiple alternatives each, which would require the researcher to go through the cumbersome task of obtaining a large number of product alternatives to have available for implementing the real purchase. In this light, hybrid DCE approaches, where some of the choice sets are real and others are hypothetical, offer a promising alternatives to mitigate HB while at the same time addressing potential infeasibility and logistical challenges of the fully incentivized DCE. This study aims at examining different hybrid DCE designs that vary the hypothetical portion of the DCE across choice sets vs. alternatives within choice sets, while also exogenously varying the information disclosure between full vs. ambiguous. We compare the different hybrid DCE designs to a hypothetical DCE and a fully incentive compatible DCE to determine their relative effectiveness in addressing HB. Our results can provide useful insights and guidance to researchers on how to mitigate HB in DCEs while keeping costs manageable and maintaining feasibility of their study designs.
External Link(s)

Registration Citation

Citation
Hu, Wuyang, Bachir Kassas and Jerrod Penn. 2024. "Integrating Feasibility and Realism: Examining Hybrid Incentive Structures in Discrete Choice Experiments." AEA RCT Registry. December 03. https://doi.org/10.1257/rct.14816-1.0
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Intervention (Hidden)
Each subject will complete 6 choice sets in their DCE, with each choice set containing two alternatives and a no-purchase option. Subjects will be randomly assigned to one of 6 treatment groups. Subjects in the first two groups will complete a hypothetical and fully incentive compatible DCE, respectively. In the remaining four groups, subjects will complete a hybrid DCE. This study will examine four different hybrid DCE designs, which will be compared to the hypothetical DCE and a fully incentive compatible DCE. Below is a description of the treatments.

1. Hypothetical DCE: all 6 choice sets will be hypothetical (i.e., no purchase and no payment will be made for subjects in this group).
2. Fully Incentive Compatible: all 6 choice sets will be real. One of the 6 choice sets will be randomly selected as binding and subjects will purchase and pay for the item they chose in this binding choice set.
3. Hybrid DCE - CS Disclosure: subjects will be told that 2 out of the 6 choice sets are hypothetical and the remaining 4 are real, but they will not know which choice sets are hypothetical or real. One of the 4 real choice sets will be randomly selected as binding and subjects will purchase and pay for the item they chose in this binding choice set.
4. Hybrid DCE - CS Ambiguous: subjects will be told that some choice sets are hypothetical and some choice sets are real, without knowing how many or which choice sets are hypothetical and real. Again, 2 choice sets will be hypothetical and 4 will be real (similar to treatment 3). One of the four real choice sets will be randomly selected as binding and subjects will purchase and pay for the item they chose in this binding choice set.
5. Hybrid DCE - Alt Disclosure: subjects will be told that one item in two choice sets is hypothetical and all other items are real, but they will not know which items are hypothetical and to which choice sets they belong. One of the 6 choice sets will be randomly selected as binding. Subjects will purchase and pay for the item they chose in the binding choice set if the item is real. If the item they chose turns out to be one of the two hypothetical items, they will instead receive their full participation compensation (i.e., no purchase or payment will be made in this case).
6. Hybrid DCE - Alt Ambiguous: subjects will be told that some alternatives in some choice sets are hypothetical, without knowing how many alternatives or which alternatives are hypothetical, nor in which choice sets they belong. Again, 2 alternatives will be hypothetical (similar to treatment 5). One of the 6 choice sets will be randomly selected as binding. Subjects will purchase and pay for the item they chose in the binding choice set if the item is real. If the item they chose turns out to be one of the two hypothetical items, they will instead receive their full participation compensation (i.e., no purchase or payment will be made in this case).
Intervention Start Date
2024-12-02
Intervention End Date
2025-01-06

Primary Outcomes

Primary Outcomes (end points)
Subjects' choices in the DCE will be used to estimate their WTP for products and product attributes. The WTP estimates will be compared between the four hybrid DCEs vs. the hypothetical and fully incentive compatible DCE to examine the relative effectiveness of each hybrid DCE design in addressing hypothetical bias.
Primary Outcomes (explanation)
A random parameter logit model will be estimated to obtain preference estimates for the different product attributes. WTP will be calculated by taking the negative ratio of the coefficient of an attribute to the coefficient of the price attribute. Krinsky and Rob bootstrapping procedure will be used to generate mean and confidence interval estimates for WTPs, and the complete combinatorial test will be used to compare WTP across treatment groups. Individual WTP estimates from the random parameter logit model will also be used as an outcome variable and regressed against the treatment variables, while controlling for individual factors and sociodemographic variables.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
Each subject will complete 6 choice sets in their DCE, with each choice set containing two alternatives and a no-purchase option. Subjects will be randomly assigned to one of 6 treatment groups. Subjects in the first two groups will complete a hypothetical and fully incentive compatible DCE, respectively. In the remaining four groups, subjects will complete a hybrid DCE. This study will examine four different hybrid DCE designs that vary the hypothetical portion of the DCE across choice sets vs. alternatives within choice sets, while also exogenously varying the information disclosure between full vs. ambiguous. The hybrid DCEs will be compared to the hypothetical DCE and a fully incentive compatible DCE.
Experimental Design Details
This study will be conducted online using Qualtrics platform to program the experimental instructions. Subjects will answer a few questions about their shopping habits and preferences, complete the DCE based on their treatment, and fill out some sociodemographic information. A representative sample of primary shoppers in the US will be obtained and subjects will be assigned to one of the 6 treatment conditions in a manner that ensures balance in subject characteristics across treatments. Below is a description of the 6 DCE treatments.

1. Hypothetical DCE: all 6 choice sets will be hypothetical (i.e., no purchase and no payment will be made for subjects in this group).
2. Fully Incentive Compatible: all 6 choice sets will be real. One of the 6 choice sets will be randomly selected as binding and subjects will purchase and pay for the item they chose in this binding choice set.
3. Hybrid DCE - CS Disclosure: subjects will be told that 2 out of the 6 choice sets are hypothetical and the remaining 4 are real, but they will not know which choice sets are hypothetical or real. One of the 4 real choice sets will be randomly selected as binding and subjects will purchase and pay for the item they chose in this binding choice set.
4. Hybrid DCE - CS Ambiguous: subjects will be told that some choice sets are hypothetical and some choice sets are real, without knowing how many or which choice sets are hypothetical and real. Again, 2 choice sets will be hypothetical and 4 will be real (similar to treatment 3). One of the four real choice sets will be randomly selected as binding and subjects will purchase and pay for the item they chose in this binding choice set.
5. Hybrid DCE - Alt Disclosure: subjects will be told that one item in two choice sets is hypothetical and all other items are real, but they will not know which items are hypothetical and to which choice sets they belong. One of the 6 choice sets will be randomly selected as binding. Subjects will purchase and pay for the item they chose in the binding choice set if the item is real. If the item they chose turns out to be one of the two hypothetical items, they will instead receive their full participation compensation (i.e., no purchase or payment will be made in this case).
6. Hybrid DCE - Alt Ambiguous: subjects will be told that some alternatives in some choice sets are hypothetical, without knowing how many alternatives or which alternatives are hypothetical, nor in which choice sets they belong. Again, 2 alternatives will be hypothetical (similar to treatment 5). One of the 6 choice sets will be randomly selected as binding. Subjects will purchase and pay for the item they chose in the binding choice set if the item is real. If the item they chose turns out to be one of the two hypothetical items, they will instead receive their full participation compensation (i.e., no purchase or payment will be made in this case).
Randomization Method
Subjects will be randomized across the 6 treatment groups using the randomizer feature in Qualtrics. This feature allows complete randomization of subjects evenly across the treatments to ensure comparable sample sizes and subject characteristics across treatments.
Randomization Unit
Randomization will be conducted at the individual level.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
Each of the 6 treatments will consist of 150 subjects, for a total of 900 subjects in the study.
Sample size: planned number of observations
Each subject will complete 6 choices in the DCE. With a total of 900 subjects, the total number of observations from the DCE will be 5,400.
Sample size (or number of clusters) by treatment arms
Each of the 6 treatments will consist of 150 subjects.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
LSU Institutional Review Board
IRB Approval Date
2024-10-11
IRB Approval Number
IRBAG-24-0062
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials