Voice AI and Job Interviews

Last registered on August 15, 2025
View Trial History

Pre-Trial

Trial Information

General Information

Title
Voice AI and Job Interviews
RCT ID
AEARCTR-0015385
Initial registration date
March 06, 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
March 10, 2025, 9:26 AM EDT

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

Last updated
August 15, 2025, 2:03 PM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country
Philippines
Region
Country
United States of America
Region

Primary Investigator

Name
Brian Jabarian
Affiliation
University of Chicago Booth Business School
Contact Primary Investigator

Other Primary Investigator(s)

PI Name
Luca Henkel
PI Affiliation
Erasmus University Rotterdam
Contact Investigator

Additional Trial Information

Status
In development
Start date
2025-03-08
End date
2026-05-08
Keywords
Behavior, Labor
Additional Keywords
JEL code(s)
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
See the attached analysis plan.
External Link(s)

Registration Citation

Citation
Jabarian, Brian and Luca Henkel. 2025. "Voice AI and Job Interviews." AEA RCT Registry. August 15. https://doi.org/10.1257/rct.15385-2.1
Sponsors & Partners
Experimental Details

Interventions

Intervention(s)
Intervention (Hidden)
We partner with a global recruitment solutions provider that recruits workers for large US-based firms. Our treatment variation concerns the interview stage, specifically who conducts the interview. Applicants who successfully pass the initial screening are subsequently randomized into one of three experimental conditions: Human interviewer, AI interviewer, and Choice of interviewer.

In the Human interviewer condition, applicants are interviewed by a human recruiter for their job interview. Human recruiters in the organization follow a structured interview script designed to ensure a standardized interview process. The script includes essential questions tailored to the candidate's background, such as gaps in employment or transitions between jobs, while also allowing flexibility for recruiters to adapt their approach as needed.

In the AI interviewer condition, applicants are instead interviewed by a voice AI agent based on a proprietary large language model. The voice AI agent follows a scripted conversation that mimics the same questions and follow-up strategies used by human interviewers.

Lastly, in the \textit{Choice of interviewer} condition, applicants can choose whether a human or a voice AI agent will interview them.

For further details, see the attached analysis plan.
Intervention Start Date
2025-03-08
Intervention End Date
2025-06-08

Primary Outcomes

Primary Outcomes (end points)
See the attached analysis plan.
Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
See the attached analysis plan.
Experimental Design Details
The jobs advertised online by the company are entry-level jobs, such as a customer service representative position. Applicants submit their materials through the company's portal. Then, the company's recruiters conduct an initial screening. If the recruiter judges an applicant suitable, they are invited by text message for an interview. There are two modes of interviews depending on the job and recruitment center: either the interview is conducted remotely, or in one of the company's recruitment centers. Complete interviews typically last between 10 and 30 minutes and serve to assess the candidate's background, experience, and fit for the job. After the interview, a recruiter reviews both the transcript and audio recording of the interview and decides whether to extend a job offer to the applicant or not. If an applicant receives and accepts an offer, there are additional background and medical checks. Once passed, the applicant receives job training and starts their job with the respective firm. For details, see the attached analysis plan.
Randomization Method
Randomization is done by a computer. For details, see the attached analysis plan.
Randomization Unit
Unit of application is the individual. For details, see the attached analysis plan.
Was the treatment clustered?
No

Experiment Characteristics

Sample size: planned number of clusters
See the attached analysis plan.
Sample size: planned number of observations
See the attached analysis plan.
Sample size (or number of clusters) by treatment arms
See the attached analysis plan.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Supporting Documents and Materials
IRB

Institutional Review Boards (IRBs)

IRB Name
University of Chicago Social & Behavioral Sciences Institutional Review Board
IRB Approval Date
2025-01-07
IRB Approval Number
IRB24-1894
Analysis Plan

Analysis Plan Documents

Preregistration_main_experiment.pdf

MD5: c66260810443d3913784e8c7dd81ed9d

SHA1: 2ad34625f42695d3f1aefc865a28d70209e2a165

Uploaded At: March 07, 2025

Preregistration_main_experiment_corrected.pdf

MD5: e788f14a4e444d7c1ebb8a5b1a0186cc

SHA1: 118e762d8b681e98caf3f098ca4d13ae16bdd46b

Uploaded At: March 25, 2025

Post-Trial

Post Trial Information

Study Withdrawal

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

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials