Gender and performance on test

Last registered on November 05, 2021

Pre-Trial

Trial Information

General Information

Title
Gender and performance on test
RCT ID
AEARCTR-0008480
Initial registration date
November 01, 2021
Last updated
November 05, 2021, 4:32 PM EDT

Locations

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information

Primary Investigator

Affiliation
University of Groningen

Other Primary Investigator(s)

Additional Trial Information

Status
In development
Start date
2021-11-01
End date
2023-02-28
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This research varies the gender of characters in test questions and examines whether/how this affects the performance of test takers.
External Link(s)

Registration Citation

Citation
Peter, Noemi. 2021. "Gender and performance on test." AEA RCT Registry. November 05. https://doi.org/10.1257/rct.8480-1.0
Experimental Details

Interventions

Intervention(s)
This research varies the gender of characters in test questions and examines whether/how this affects the performance of test takers.
Intervention Start Date
2021-11-01
Intervention End Date
2021-11-02

Primary Outcomes

Primary Outcomes (end points)
Performance (points)
Primary Outcomes (explanation)
Note that boys and girls will be analyzed separately. I will also look into where the effect concentrates.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
This research varies the gender of characters in test questions and examines whether/how this affects the performance of test takers.
Experimental Design Details
Not available
Randomization Method
Stratified randomization using randtreat in Stata
Randomization Unit
Individuals are randomized to treatments
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
No clustering. See next answer for further details.
Sample size: planned number of observations
Individuals are randomized, so there is no clustering. Note that people who were not part of the randomization in Stata will not be part of the analysis sample (e.g. people who did not register for the exam were not part of the Stata randomization, but some of these people may show up to the exam). I cannot simply plan the sample size because the experiment is run at a course, and I have no control over enrollment into the course and participation in tests. Furthermore, note that I will analyze boys and girls separately (randomization was also stratified by gender, among other things), and the number of girls is small (less than a hundred econ&business econ girls were part of the Stata randomization, even with other study programmes the number of girls is below hundred and forty). Therefore, I will have to run the field experiment not only this year but also in at least one later year to gather more observations. Data from this year can be used to formulate a plan on how many more years of data is needed. In this sense this year can be viewed as piloting (but note that for boys, sample is about three times as large as for girls, so it is the female sample that constitutes the bottleneck).
Sample size (or number of clusters) by treatment arms
There are two treatments, so size per treatment is approximately fifty percent
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
IRB

Institutional Review Boards (IRBs)

IRB Name
IRBFEB
IRB Approval Date
2021-09-14
IRB Approval Number
FEB-20210908-13180