Will Artificial Intelligence get in the way of gender equality?

How to cite

Carvajal, Daniel; Franco, Catalina; Isaksson, Siri. 2023. "Will Artificial Intelligence get in the way of gender equality?." AEA RCT Registry. AEARCTR-0012452. https://doi.org/10.1257/rct.12452-2.2

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Daniel Carvajal* , Catalina Franco , Siri Isaksson

* Primary investigator

Registry ID
AEARCTR-0012452
DOI
https://doi.org/10.1257/rct.12452-2.2
First registered
Last updated
Status
In development
Trial dates
2023-11-07 – 2024-12-31
Countries
Private
JEL codes
D91, I20, J16, O33
Registry record
http://www.socialscienceregistry.org/trials/12452

Abstract

This research project aims to investigate the existence of gender differences in the adoption and use of AI technologies, specifically ChatGPT. Previous studies have highlighted a "Digital Divide," showing disparities in internet usage between men and women. This project seeks to extend our understanding of this divide to the realm of AI technologies. We aim to (i) identify whether a gender gap exists in the adoption of ChatGPT, (ii) explore the underlying mechanisms driving any observed disparities, (iii) assess the impact of this gap on productivity, and (iv) evaluate the efficacy of interventions aimed at reducing the gender gap in AI adoption.

Experimental design

Study 1 will mostly collect survey measures on ChatGPT use. The only randomization in Study 1 concerns intended ChatGPT use under different hypothetical scenarios as follows. Participants will be exposed to a series of two hypothetical scenarios. In each scenario they are given a description of a course they are hypothetically signed up for. In the description, they are indicated whether the professor explicitly allows or forbid the use of ChatGPT in class. Students will be exposed to both cases, one where a professor allows ChatGPT and one where a professor forbids ChatGPT. The randomization will correspond to which scenario is presented first to a participant.

Design details

Randomization method
Randomization done using Qualtrics.
Randomization unit
Randomization will be done at the individual level, by gender of the participant.
Clusters
Around 900 participants.
Observations
Around 900 participants.
Arms
450 students in each treatment arm. Around 40% women and 60% men.
Primary outcomes
Willingness to use ChatGPT, and self-reported measures of ChatGPT use.

Cited by

No citation data available for this registration yet.

Keywords