Detecting lying on the individual level? Revisiting the die-under-the-cup paradigm.
Last registered on December 06, 2017

Pre-Trial

Trial Information
General Information
Title
Detecting lying on the individual level? Revisiting the die-under-the-cup paradigm.
RCT ID
AEARCTR-0002607
Initial registration date
December 04, 2017
Last updated
December 06, 2017 11:40 AM EST
Location(s)

This section is unavailable to the public. Use the button below to request access to this information.

Request Information
Primary Investigator
Affiliation
University of Bern, Institute for Organization and Human Resources
Other Primary Investigator(s)
PI Affiliation
University of Bern, Institute for Organization and Human Resources
Additional Trial Information
Status
In development
Start date
2017-12-05
End date
2018-12-31
Secondary IDs
Abstract
Fischbacher and Foellmi-Heusi (2013) introduced the die-rolling task to investigate lying behavior in the laboratory. Subjects are asked to roll a 6-sided die in private and report the number they rolled. The higher the number they report, the more money they gain, unless it is a 6, which leads to a payoff of zero. Since the actual result of the die roll cannot be observed by the experimenter, there is an incentive to cheat by reporting a higher number and thereby getting a higher payoff. An important aspect of this approach is that lying can only be detected on the aggregate, but not on the individual level by comparing the observed distribution with the uniform distribution of a die roll. The advantage is that potential demand effects are reduced and that real lying behavior can be observed. The downside is that it makes the measurement of lying behavior noisy and does not allow to track a single subject. In this experiment, we want to investigate the effect of using a digital die. Subjects learn that every result of the digital die roll is stored, but not checked by the experimenter for the purpose of the payment. If using the digital die leads to the same lying pattern as in Fischbacher and Foellmi-Heusi (2013), this would be a strong motive to use the digital die since it keeps the demand effect low, is easy to implement and offers the additional advantage of allowing to observe individual behavior.
External Link(s)
Registration Citation
Citation
Bieberstein, Frauke and Ann-Kathrin Crede. 2017. "Detecting lying on the individual level? Revisiting the die-under-the-cup paradigm.." AEA RCT Registry. December 06. https://www.socialscienceregistry.org/trials/2607/history/23742
Experimental Details
Interventions
Intervention(s)
Intervention Start Date
2017-12-05
Intervention End Date
2018-06-30
Primary Outcomes
Primary Outcomes (end points)
Reported result of a 6-sided die roll
Primary Outcomes (explanation)
Secondary Outcomes
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
Experimental Design
The design and procedures of the control treatment are exactly the same as in Fischbacher and Foellmi-Heusi (2013), in the experimental treatment the conventional die is replaced with a digital die: This experiment will be added to the end of other experiments since it only takes 10 minutes to conduct. After participants finished the first experiment, they are told that a second, very short experiment will follow and that this experiment has nothing to do with the experiment before. The task consists of completing a short questionnaire and afterwards determining the payoff for it by a die roll. Subjects are asked to report the first number they rolled, which determines the payoff: Reporting a 1 leads to a payoff of 1 euro, reporting a 2 leads to a payoff of 2 euros and so on until reporting a 5, while reporting a 6 leads to a payoff of 0.

There are two treatments, which are one-shot and designed in a between subject-design:

Control treatment: Conventional die
In the control treatment, subjects find a cup and a die on their table and are asked to use this die to determine their payoff. They are asked to remember and report the first number they rolled, but are told to roll the die a couple of times to check whether the die is fair.

Experimental treatment: Digital die
The experimental treatment is identical to the control treatment, but instead of a die on the table, subjects find a digital die on their screen which randomly shows a number between 1 and 6 after clicking on a button. In addition, subjects are told that every result of the digital die roll is stored, but not checked by the experimenter for the purpose of the payment.

Experimental Design Details
Not available
Randomization Method
Randomization done by computer (Laboratory experiment)
Randomization Unit
individual
Was the treatment clustered?
No
Experiment Characteristics
Sample size: planned number of clusters
-
Sample size: planned number of observations
Based on power calculations with different specifications, we plan to collect data from 270 subjects. Since we are interested in the null result, it is crucial for our data analysis to consider beta, i.e. the probability to accept H0 when it is false. Our power calculations are based on a high power (1-beta) of 0.95.
Sample size (or number of clusters) by treatment arms
We plan to collect 135 observations per treatment.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Name
IRB Approval Date
IRB Approval Number