Gender and Task Allocation in Group Work

Last registered on June 25, 2024

View Trial History

Pre-Trial

Trial Information

General Information

Title

Gender and Task Allocation in Group Work

RCT ID

AEARCTR-0013849

Initial registration date

June 19, 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

June 25, 2024, 10:43 AM EDT

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

Locations

Country

United Kingdom of Great Britain and Northern Ireland

Region

essex

Primary Investigator

Name

Friederike Mengel

Affiliation

Uni Essex

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Lingqing Jiang

PI Affiliation

PI Name

Laura Mangold

PI Affiliation

Additional Trial Information

Status

In development

Start date

2024-06-19

End date

2025-06-18

Keywords

Behavior, Gender

Additional Keywords

JEL code(s)

Secondary IDs

Prior work

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

Abstract

In organizations, routine tasks are often associated with low promotability and considered as a barrier to completing tasks with higher priority. A common observation is that routine (non-routine) tasks are more often done by women (men) either because the routine tasks are more often assigned to women (demand driven) or women are more likely to volunteer to do such tasks (supply driven). We conduct an online experiment to study the demand side of task allocation by investigating managers’ decision in assigning routine and non-routine tasks and eliciting managers’ belief about workers’ performance.

External Link(s)

Registration Citation

Citation

Jiang, Lingqing, Laura Mangold and Friederike Mengel. 2024. "Gender and Task Allocation in Group Work." AEA RCT Registry. June 25. https://doi.org/10.1257/rct.13849-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention (Hidden)

Intervention Start Date

2024-06-19

Intervention End Date

2024-06-21

Primary Outcomes

Primary Outcomes (end points)

Manager’s selection decisions, specifically the share of women (men) chosen for each of the two tasks.

Primary Outcomes (explanation)

Our primary question is whether female (male) candidates are relatively more likely to be chosen to perform the routine (non-routine) task and how this pattern changes across two treatment dimensions: gender quota and task description.

Secondary Outcomes

Secondary Outcomes (end points)

Information requests: the pieces of performance information requested by the manager
Guesses: guesses about the overall performance of candidates selected by another manager

Secondary Outcomes (explanation)

We are interested in comparing the number of pieces of information requested by managers depending on the gender of the candidate and the sequence of realizations of prior signals.

2) The belief about the performance is measured by the range, i.e., the highest and lowest possible score, guessed by the manager. We are interested in both the midpoint and the size of the range.

Experimental Design

Our experiment consists of two stages in random order. All participants are assigned the role of a “manager”.
In stage 1, managers select desirable candidates to perform a group task. The group task consists of two subtasks, a routine task and a non-routine task. Managers can request up to two pieces of information about candidates’ past performance (information requests) before they make the selection decision.
There are two treatment dimensions. First, we either describe the two tasks in terms of features of the task per se (task description) or in terms of the personnel required to conduct the task (personnel description). Second, we either impose a gender quota in which case one woman and one man have to be chosen or no gender quota.
In stage 2, managers are asked to guess the performance of a randomly selected candidate from a different candidate pool. In particular, they are asked to guess the range of the performance, i.e., the highest and the lowest possible performance.
They are told that another manager (in a pre-experiment data collection) has selected this candidate for either the routine or non-routine task. They are also informed about whether that manager was subject to a gender quota.
Participants in our experiment are randomly assigned to one of the treatments for each treatment dimension, respectively.

Experimental Design Details

We will test for our main outcomes using OLS regressions (Linear Probability Model).
We regress a gender dummy indicating whether the candidate chosen is female on (i) a dummy identifying the non-routine task; (ii) a dummy indicating gender quota; (iii) the interaction of the two. This first regression will aggregate data from the person-based and task-based conditions.
We regress a gender dummy indicating whether the candidate chosen is female on (i) a dummy identifying the non-routine task; (ii) a dummy indicating whether a person-based description was used; (iii) the interaction of the two. This second regression will be run separately on data from the non-quota and quota conditions.
Our main hypothesis is that the coefficient on (i) is negative. A secondary hypothesis is that also (iii) is negative.
Our baseline specifications do not include any controls, but we will include specifications with participant demographics as well as candidate characteristics.
Additional tests: We will study our secondary outcomes and conduct heterogeneity analysis.

Randomization Method

online by a computer

Randomization Unit

individual

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

1000 individuals

Sample size: planned number of observations

1000 individuals

Sample size (or number of clusters) by treatment arms

250 individuals

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

We conducted a small pre-test with fourty participants to get an idea of effect sizes. In that pre-test 64% of the candidates chosen for the routine task were women as opposed to 47% in the non-routine task. After the introduction of a gender quota the ratios were 50% for both tasks. In the condition with person-based description of the task 61% of the candidates chosen for the routine task were women as opposed to 50% in the non-routine task. In the condition with a task-based description the ratio was 52% for the routine task and 48% for the non-routine-task. Our chosen sample size allows us to detect the basic effect size (aggregated non-quota condition) at the 5% level with over 80% power. In fact a smaller sample size would be sufficient, but because of our secondary outcomes as well as heterogeneity analysis we commit to this higher sample size.

Supporting Documents and Materials

IRB