Default Bias and Preference for Randomization

How to cite

Zhang, Xing; Zhong, Songfa. 2024. "Default Bias and Preference for Randomization." AEA RCT Registry. AEARCTR-0013290. https://doi.org/10.1257/rct.13290-1.1

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

Xing Zhang* , Songfa Zhong

* Primary investigator

Registry ID
AEARCTR-0013290
DOI
https://doi.org/10.1257/rct.13290-1.1
First registered
Last updated
Status
In development
Trial dates
2024-04-08 – 2024-08-01
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/13290

Abstract

Using an online experiment, we investigate when people are likely to commit default bias in decision-making and whether offering a coin-flipping option can mitigate the default bias. The participants engage in a series of decision-making tasks, choosing between two lotteries. One lottery is provided as the default option and they must decide whether to keep the default option or trade for another lottery. The participants are randomly assigned to a control condition and a treatment condition, where they can toss a coin to make the decision. We further vary the statistical relationship between the two lotteries and measure participants’ attentiveness. We seek to answer the following questions: 1) Do the statistical relationships between the two lotteries affect the tendency to commit the default bias and the tendency of choosing coin-flipping? 2) Does participants’ attentiveness affect committing the default bias and the tendency of choosing coin-flipping? 3) When coin-flipping is more likely to mitigate the default bias?

Experimental design

The experiment will be conducted on the online survey platform Qualtrics. Participants will be recruited on Prolific. They will receive a fixed show-up fee and be given an experiment stock as a bonus. The return for stock is determined as follows: A computer will randomly draw one card from a set of 4 cards, which includes 1 red card, 1 white card, 1 black card, and 1 blue card. Depending on which card is drawn, the stock will pay a certain amount of money. For example, suppose the following stock (namely, Stock A) is given to the participant:

• If the red card is drawn, the participant will receive a return of $15.
• If the white card is drawn, the participant will receive a return of $2.
• If the black card is drawn, the participant will receive a return of $0.
• If the blue card is drawn, the participant will receive a return of $8.

The participant can keep Stock A, i.e., choose the default, or she can trade the stock for another stock (namely, Stock B). The experiment follows a mixed design including a between-subject factor (Coin vs. No Coin) and a within-subject factor (Stock A is the default vs. Stock B is the default). When a participant chooses to keep Stock A when Stock A is the default but chooses to keep Stock B when Stock B is the default, we consider this participant commits default bias facing Stock A and B.

At the end of the experiment, we select 5 choice tasks they did, and ask them to recall their decisions. Each decision that was successfully recalled would receive US$0.2.

The following information will also be collected from the participants:
1. Response time
2. Demographic information (gender, age, education, employment, and religion)

Design details

Randomization method
Treatment group assignment will be randomized by the Qualtrics online software.
Randomization unit
Individual participant.
Clusters
Around 240 participants.
Observations
Around 240 X 60 = 14,400 (number of participants X number of tasks).
Arms
Around 120 participants.
Primary outcomes
The default bias by experimental conditions, attentiveness, and statistical relationship of the two lotteries; The choice of coin-flipping option by attentiveness and statistical relationship of the two lotteries.
Secondary outcomes
Correlation of default bias and the choice of coin-flipping by the same individual (attentive vs. inattentive) across different statistical relationships.

Cited by

No citation data available for this registration yet.

Keywords