Anchoring bias in markets

Last registered on March 18, 2021
View Trial History

Pre-Trial

Trial Information

General Information

Title
Anchoring bias in markets
RCT ID
AEARCTR-0003402
Initial registration date
October 16, 2018

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
October 22, 2018, 1:00 AM EDT

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

Last updated
March 18, 2021, 11:08 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country
Netherlands
Region
Europe

Primary Investigator

Name
Konstantinos Ioannidis
Affiliation
University of Amsterdam & University of Birmingham
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Theo Offerman
PI Affiliation
University of Amsterdam
Contact Investigator
PI Name
Randolph Sloof
PI Affiliation
University of Amsterdam
Contact Investigator

Additional Trial Information

Status
Completed
Start date
2018-10-22
End date
2018-12-30
Keywords
Other
Additional Keywords
Anchoring bias, double auction, bilateral bargaining
JEL code(s)
D01, D02, D91
Secondary IDs
Abstract
In a laboratory experiment, we investigate whether the information revealed during market participation reduces the anchoring bias observed in individual decision making. We control the amount of information available in the market using 3 different market settings: a double auction, bilateral bargaining, and a control treatment in which a good different from the anchored one is traded. Anchors are created randomly for each individual subject by rolling a 10-sided die twice; using a median split subjects are divided into a low anchor group and a high anchor group. The main question of interest is whether the difference in the willingness to pay between the low anchor and the high anchor group observed before market participation is eliminated after the market participation.
External Link(s)

Registration Citation

Citation
Ioannidis, Konstantinos, Theo Offerman and Randolph Sloof. 2021. "Anchoring bias in markets." AEA RCT Registry. March 18. https://doi.org/10.1257/rct.3402-2.0
Former Citation
Ioannidis, Konstantinos, Theo Offerman and Randolph Sloof. 2021. "Anchoring bias in markets." AEA RCT Registry. March 18. https://www.socialscienceregistry.org/trials/3402/history/87652
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Anchoring is attempted by the following Yes/No question: Would you sell the good back to the experimenter for X€? X is created randomly by rolling a 10-sided die twice. After the anchoring question, participants interact in a market: either a double auction, bilateral bargaining, or in a double auction for a different good (control treatment).
Intervention Start Date
2018-10-22
Intervention End Date
2018-12-30

Primary Outcomes

Primary Outcomes (end points)
Willingness To Accept (WTA) both before and after the market participation. The low and the high anchor group will be determined by a median split. We will also compare the difference in WTA before and after market experience between the highest quartile and lowest quartile of the anchor distribution.
Primary Outcomes (explanation)
Difference in WTA between the low and high anchor group, both before and after market participation.

Secondary Outcomes

Secondary Outcomes (end points)
Probability of a trade, Speed of a trade, Price of a trade
Secondary Outcomes (explanation)
Probability of a trade = ratio of agreed trades over all possible trades
Speed of a trade = seconds until trade is agreed (Only for realized trades)
Price of a trade = agreed price (Only for realized trades)

Experimental Design

Experimental Design
We anchor subjects with a random price and compare their WTA for a good before and after participating in a market. Three types of market settings are studied: a double auction, bilateral bargaining, and a double auction for a different good.
Experimental Design Details
Our experiment consists of 3 Phases.

In Phase I, we anchor subjects' with a random price and elicit their WTA for a good. The anchoring happens via a Yes/No question of the type: "Are you willing to sell the good to the experimenter for X€? The X is determined by rolling a 10-sided die twice and making a price out of the rolls (eg rolls of 6 and 7 give price 6.7€).

In Phase II, subjects participate in a market. . Three different market settings are studied. In the Double auction treatment, participants are grouped in groups of 8 with 4 buyers and 4 sellers and participate in a double auction (for the same anchored good). In the Bilateral bargaining treatment, participants are grouped in pairs (1 buyer and 1 seller) and negotiate a trade. In the control treatment, participants take part in a double auction for a different good.

In Phase III, we elicit WTA for the good again.

After the first two sessions a manipulation check will be performed to verify successful inducement of the initial anchor.
Randomization Method
For each individual subject, the random anchor is obtained from two (10-sided) die rolls.
Randomization Unit
Subjects are randomly assigned treatment market setting before a session starts. Within each market setting, a median split on the random anchors defines the Low Anchor and the High Anchor groups.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
480 participants from 30 sessions with 16 participants in each.
Sample size: planned number of observations
480 participants
Sample size (or number of clusters) by treatment arms
160 participants in the Double Auction treatment (80 in Low Anchor group and 80 in High Anchor Group)
160 participants in the Bilateral Bargaining treatment (80 in Low Anchor group and 80 in High Anchor Group)
160 participants in the Control treatment (80 in Low Anchor group and 80 in High Anchor Group)

(We may encounter problems with the size of our subject pool. If we cannot manage to have 480 subjects, we will have less in the Control treatment.)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
With our sample and assuming a standard deviation of 3.5 (based on an in-classroom pilot session), the Minimum Detectable Effect size is 1€. Defining the effect size as the difference in means divided by the pooled standard deviation, the percentage is 28%.
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
Ethics Committee Economics and Business (EBEC), University of Amsterdam
IRB Approval Date
2018-09-26
IRB Approval Number
EC 20180926020919
Analysis Plan

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?
Yes
Intervention Completion Date
November 16, 2018, 12:00 +00:00
Data Collection Complete
Yes
Data Collection Completion Date
November 16, 2018, 12:00 +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
98 trading groups (20 Double Auction sessions + 78 Bilateral Bargaining sessions)
Was attrition correlated with treatment status?
No
Final Sample Size: Total Number of Observations
316 subjects
Final Sample Size (or Number of Clusters) by Treatment Arms
160 participants in the Double Auction treatment (80 in Low Anchor group and 80 in High Anchor Group) 156 participants in the Bilateral Bargaining treatment (78 in Low Anchor group and 78 in High Anchor Group) * Given our null result of anchoring, the control treatment was not run.
Data Publication

Data Publication

Is public data available?
Yes
Public Data URL
https://github.com/KonstantinosIoannidis/Anchoring/tree/master/Data

Program Files

Program Files
Yes
Program Files URL
https://github.com/KonstantinosIoannidis/Anchoring/tree/master/Stata
Reports, Papers & Other Materials

Relevant Paper(s)

Abstract
We test whether markets are needed to mitigate the effects of anchoring on peoples' preferences. We anchor subjects by asking them if they are willing to sell a bottle of wine for a transparently uninformative random price. We elicit subjects' Willingness-To-Accept for the bottle before and after the market. Subjects either participate in a small or a large double auction market. The variance in subjects' Willingness-To-Accept shrinks within trading groups. Our evidence supports the idea that markets have the potential to mitigate a bias. However, the market is not needed: our anchoring manipulation failed in a large sample.
Citation
Ioannidis, K., Offerman, T., & Sloof, R. (2020). On the effect of anchoring on valuations when the anchor is transparently uninformative. Journal of the Economic Science Association, 6(1), 77-94.
URL
https://link.springer.com/article/10.1007/s40881-020-00094-1

Reports & Other Materials