Over-Commitment and gender differences

Last registered on June 23, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

Over-Commitment and gender differences

RCT ID

AEARCTR-0016271

Initial registration date

June 23, 2025

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

June 23, 2025, 3:06 PM EDT

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

Locations

Country

France

Region

Grand-Est

Primary Investigator

Name

Nicolas Jalabert

Affiliation

University of Strasbourg

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Magali Jaoul-Grammare

PI Affiliation

CNRS Université Strasbourg

Contact Investigator

Additional Trial Information

Status

In development

Start date

2023-07-10

End date

2025-12-31

Keywords

Behavior, Gender, Lab

Additional Keywords

Over-commitment, competition, personality

JEL code(s)

C91 D03 J16

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

We remobilize an old and important body of literature on the over-commitment. This bias describes the tendency of an individual or a group to persist in a specific behavior even when facing increasingly negative outcomes related to their actions. In this regard, we aimed to test in a lab setting to elicit over-commitment and determine whether certain individuals (based on gender and personality) might be more exposed to it.

The experiment will be conducted inside the LEES (Laboratoire d'Économie Expérimentale de Strasbourg). Participants will have to chose between two shorts tasks, and peform 15 sessions They will earn money for each level completed, and both the difficulty and the amount earned will increase after each completion. Between each level, they will have the possibility to leave that system to perform the other task with lower difficulty. There will be three different treatments in this experiment. The first will be the standard version described above, where the difficulty is created by a time constraint. The second will introduce competition, where success at each level depends on their performance (time) compared to others. The last treatment will be the competitive setting with the addition of information between each level, where individuals will learn their relative performance.

The idea is that a fully rational individual should optimize their moment of drop-out. They should leave right before their first failure. Given that it might be difficult to estimate the moment they would first fail, we anticipate most of them to drop out after the first failure, especially under competition. However, we are interested in individuals who commit to the task even after that, and what may distinguish them. In treatment 3, giving them information on their relative performance and possible gains should limit the issue of a lack of information and of an under/over-confidence bias. Therefore, individuals persisting in treatment 3 would represent pure over-commitment. The aim of the 15 sessions is to compare, between individuals, how long it takes for them to decide to give up on the initial task.

Our explanatory variables will be the sex of the participants, their personality traits measured with the Big Five, and a set of control variables related to socio-demographic characteristics, liking of the task, and risk aversion. We will also investigate their motivations behind giving up on the task, and whether they had a strategy beforehand (leaving at first failure, after a defined number of attempts) or if it was a decision made in the moment (frustration, aversion to negative feedback).

Our main contributions related to the experiment are to especially center the analysis around gender and integrated personality traits. We also constructed a setting that allows us to test for a longer session of exercises (15).

External Link(s)

Registration Citation

Citation

Jalabert, Nicolas and Magali Jaoul-Grammare. 2025. "Over-Commitment and gender differences." AEA RCT Registry. June 23. https://doi.org/10.1257/rct.16271-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention (Hidden)

Participants will chose between two task and then will perform up to 15 levels of this repetitive task. After each level, they earn a monetary reward that increases with each level completed. The task becomes increasingly difficult, due to a time constraint or competition depending on the treatment. After each level, participants may choose to continue the current task or switch to the other, which will be easier but (to secure a gain, but lower-paying). There are three treatment conditions:

Treatment 1 (Baseline): Participants face increasing time pressure, with no additional elements.
Treatment 2 (Competition): Participants are informed that their success depends on their performance relative to 30 others participants from previous sessions (speed of completion).
Treatment 3 (Competition + Information): Same as Treatment 2, but after each level, participants receive information about their relative performance.

Intervention Start Date

2025-06-26

Intervention End Date

2025-11-19

Primary Outcomes

Primary Outcomes (end points)

The number of missed attempt before dropping out

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

binary/categorical variable of whether the participant left before/at/after the first failure. Number of consecutive failures.

Secondary Outcomes (explanation)

Experimental Design

The experiment follows a between-subject design with random assignment to one of three treatments. Each participant is randomly assigned to one treatment at the beginning of the session. Participants make repeated decisions (up to 15) about whether to continue or stop the main task after each level. The primary outcome of interest is the level at which a participant chooses to stop the task. Additional data will be collected on participant characteristics, including gender, Big Five personality traits, socio-demographic variables, and risk preferences. Each session includes approximately 30-40 participants, but decisions are made individually and independently.

Experimental Design Details

Randomization Method

Randomization done by computer.

Randomization Unit

Randomization at the individual level.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

Not a cluster-randomized design. Randomization is at the individual level

Sample size: planned number of observations

300-350 students

Sample size (or number of clusters) by treatment arms

100-115 participants by treatment. We hope for 50/50 on each of the two tasks.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

We plan a total sample size of 300 participants, with 6 treatment arms of 50 participants each. Our primary outcome is the number offailures before dropping out of a progressive task. Based on our calibration, we anticipate a standard deviation (SD) around 4 on this outcome. Power calculations are based on a two-sided t-test comparing two groups at a time, with: α = 0.05 (significance level) Power = 80% (β = 0.20) n = 50 per group Using these values, the minimum detectable effect size (MDES) in standardized terms is: 0.4

Supporting Documents and Materials

IRB