AI and Job-Market signaling

Last registered on September 19, 2025

View Trial History

Pre-Trial

Trial Information

General Information

Title

AI and Job-Market signaling

RCT ID

AEARCTR-0016732

Initial registration date

September 18, 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

September 19, 2025, 10:24 AM EDT

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

Locations

Country

Italy

Region

Padova

Primary Investigator

Name

Moritz Janas

Affiliation

NYU Abu Dhabi

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Antonio Nicolo

PI Affiliation

University of Padua

Contact Investigator

PI Name

Paolo Roberti

PI Affiliation

Free University of Bozen/Bolzano

Contact Investigator

Additional Trial Information

Status

In development

Start date

2025-09-22

End date

2026-03-31

Keywords

Behavior

Additional Keywords

JEL code(s)

Secondary IDs

Prior work

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

Abstract

This study investigates the impact of artificial intelligence (AI) on hiring decisions using a controlled laboratory experiment. We examine whether agents’ access to AI tools influences principals’ ability to identify the most suitable candidates. By randomly varying the AI support available to agents, we assess how algorithmically enhanced applications shape principals’ behavior and selection outcomes. In an additional treatment dimension, we explore how participants self-select into different environments, with or without AI assistance. The experiment is programmed in oTree and conducted in the laboratory of the University of Padua.

External Link(s)

Registration Citation

Citation

Janas, Moritz, Antonio Nicolo and Paolo Roberti. 2025. "AI and Job-Market signaling." AEA RCT Registry. September 19. https://doi.org/10.1257/rct.16732-1.0

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

Intervention (Hidden)

Intervention Start Date

2025-09-22

Intervention End Date

2026-03-31

Primary Outcomes

Primary Outcomes (end points)

Hiring Behavior of Principals (in combination with Logic Score of the participants); Selection into LLM or NoLLM environments; Confidence of participants

Primary Outcomes (explanation)

All outcomes are variables within the laboratory experiment. For details, see the attached pdf.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Participants enter the laboratory and are randomly allocated in different treatments. We will observe whether there are treatment differences in behavior (more details in the field "Experimental Design (Hidden)".

Experimental Design Details

The experiment investigates how the availability of large language models (LLMs) affects job-market signaling. Subjects first complete a logic test and are incentivized based on their performance. In the next stage, depending on treatment, they are either exogenously assigned to an environment with or without LLM access (EXO), or can choose their environment (ENDO, with the choice implemented with 70% probability). Participants then complete a reading comprehension task, with LLMs available in the relevant treatments. They are not directly incentivized for performance in this task. Instead, in a hiring stage, principals select candidates based on reading task performance. Selected participants receive a bonus, while principals are incentivized according to the (unknown) logic test score of the chosen participant.

Randomization Method

the experiment is programmed in oTree. All randomization takes place within the program.

Randomization Unit

Participants are individually randomly allocated to different treatments.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

300 participants

Sample size: planned number of observations

300 participants

Sample size (or number of clusters) by treatment arms

we aim for 300 participants in total. The exact distribution among the treatments will depend on the showup patterns in the laboratory. In general, we aim for 180 in the Exo treatments (90 LLM, 90 NoLLM), and 120 in the Endo treatments (distributed according to the choices of the participants).

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

Supporting Documents and Materials

IRB